An Assessment Of Primary Health Care Providers In Chhattisgarh, India

Item

Title
An Assessment Of Primary Health Care Providers In Chhattisgarh,
India
extracted text
REPORT

Which Doctor For Primary Health Care?
An Assessment Of Primary Health Care Providers In Chhattisgarh,
India

June 2010

PUBLIC
HEALTH
FOUNDATION
OF INDIA

I
I

±1

o

LJbJS& J

111

LU
\\
•«TS$fhr lIF’fr0!

^)10-

J ■

CD

Fl

Which Doctor For Primary Health Care?
An Assessment Of Primary Health Care Providers In
Chhattisgarh, India

Authors
Public Health
Foundation Of India

National Health
Systems Resource
Center

State Health
Resource Center,
Chhattisgarh

Krishna D. Rao
(Principal Investigator)

Dr. Garima Gupta

Dr. Kamlesh Jain

Aarushi Bhatnagar

Dr. T. Sundararaman

Puni Kokho

Neha Kumra

Dr. K.R. Antony

Dr. Saujanya Khanna
Prof. Peter Berman

Prof. Srinath Reddy

This report was prepared by a v

Rao,

Dr. Saujanya Khanna, Neha Ki

and

Dr. Garima Gupta. The views e>

f the

authors and not of their inst
Krishna D. Rao (kd.rao@phfi.oi

SOCHARA
Community Health
Library and Information Centre (CLIC)

Centre for Public Health and Equity
No. 27,1st Floor, 6th Cross, 1st Main,
1st Block, Koramangala, Bengaluru - 34

Tel : 080 - 41280009
email: clic@sochara.org / cphe@sochara.org
www.sochara.org

e to

Acknowledgement

This study was made possible because of the contribution of many people. The authors would

like to thank the Alliance for Health Systems and Policy Research and the WHO for
providing technical support, as well as the Global Health Workforce Alliance for their
financial support. They would also like to thank the National Health Systems Resource
Center and the State Health Resource Center (SHRC), Chhattisgarh for additional funding

and technical support in the design and conduct of the study. Thanks also to the anonymous
reviewers for their extensive comments on the study proposal and design. Finally, the authors

would like to thank Dr. Pascal Zurn and Laura Stormont of WHO for their valuable support

throughout the study.

f-1

Several people in Chhattisgarh provided valuable technical support in the execution of this

study. At SHRC, Chhattisgarh - Virendra Kumar, Premshankar Verma, Samir Garg and

Neelam Nag. Several specialists in Chhattisgarh made important technical contributions to
this study. These include Dr. R.C. Ram (Professor, Raipur Medical College), Dr. Khemraj

Sonwani, Dr. Smit Srivastava (Assistant Professor, Raipur Medical College), Dr. Raka
Sheohare, Dr. Sumanth Panigrahi (Assistant Professor, Raipur Medical College), Dr. S.K.
Mandric, Dr. Pawan Jain, Dr. Sonali Jain, Dr. Sheela Tiwari, Dr. Chandrashekhar Shrivastava
(Assistant Professor, Raipur Medical College). In addition, the authors would like to thank
Dr. Shinjini Bhatnagar of AIIMS, Delhi for her technical contribution.

The study benefited invaluably from a team of dedicated interviewers. These include Niketa
Pawar, Sanjeev Kumar Sharma, Shubhra Mishra, Ashish Mishra, Saurabh Bafna, Rajendra

Kumar Verma, Rajesh Kumar Mishra, Arvind Kaushik, Balkrishana Chandrakar, Ravi Kant
Jatwar, Kapoor Chand Dewangan, Vijeta Masih, Prabhu Datt Nand, Kalyani Tamrakar,
Ashish Pandey, Gauri Bole, Umesh Pandey, Ritesh Kumar Tamboly, Bhuwan Lal Sahu,
Abhinav Chandrakar and Christopher Francis. The authors wish them the very best for their

future work.

ii

Recently, in an article in a leading national daily3, India’s former central health minster

articulated several popular criticisms against a rural cadre: introducing this cadre would be an
act of “discrimination against rural folk, who are taken for second-grade citizens deserving

medical care by a brigade of ‘qualified quacks’.” Further, it will dissuade physicians from

serving in rural areas. Other clinical care providers like AYUSH doctors and nurse­

practitioners were mooted as being better alternatives to a rural cadre of clinical care
providers with shorter training. Finally, it was also argued that, producing more physicians
coupled with compulsory rural service or recruitment of students from rural areas will
eliminate the rural physician shortage.

The debate on the ability of non-physician clinicians is, unfortunately, not based on any
empirical assessment of their performance. This study attempts to fill this important gap by
taking advantage of a “natural” experiment in the state of Chhattisgarh where, because the
public sector could not adequately staff PHCs with Medical Officers, non-physician

clinicians like AYUSH doctors and RMAs provide clinical services. Quite frequently

paramedical staff (e.g. nurses, medical assistants, pharmacists), with little or no clinical
training, provide clinical services at PHCs because no one else was available. The
functioning, either by design or circumstance, of non-physician clinicians and paramedical

staff (pharmacists and nurses) as the main providers of clinical services at PHCs represents a
set of alternatives to Medical Officers. Yet, there is little known about how well they perform

the duties expected of Medical Officers at PHCs.

Study objectives

This study provides a comparative assessment of the performance of different types of

clinical care providers working at the primary care level - Medical Officers and non­
physician clinicians i.e. RMAs, AYUSH physicians and paramedical staff (nurses and
pharmacists) - in their capacity as the main providers of clinical services at PHCs. Their

performance is examined on several dimensions. First, provider competence (‘how much they

know’) to manage the following conditions: malaria, diarrhea, pneumonia, TB, preeclampsia

3 Ramadoos A. The Wrong Way For Rural Doctors. The Hindu. 27th February, 2010. Downloaded on March 31,
2010 from www.editorialjunction.com/7pM061

vii

and diabetes.4 Second, how patients and the communities they work in view them in terms of

satisfaction with services and perceptions about the quality of care received. Third, how
much the PHCs they work in are used by ill community members. Finally, it examines their

attitudes towards rural service and levels of job satisfaction. Assessing the performance of
these different types of clinical care providers on a variety of dimensions enables a

comprehensive understanding of their suitability as primary health care providers. The study

was conducted in the state of Chhattisgarh in central India between July and September 2009.

Provider competence5

Findings from this study suggest that Medical Officers and RMAs are equally competent to
manage conditions commonly seen in primary care settings. AYUSH Medical Officers are

less competent than Medical Officers (and RMAs) and Paramedicals6 are the least competent.
This was observed for infectious, chronic and maternal health conditions and for a range of

patient types - infants, children and adult men and women. Further, these results hold even
after controlling for various individual, facility and location characteristics. This relative
performance is consistently found in all aspects of the outpatient clinical care - history taking,

examinations, investigation, diagnosis, prescription and home recommendations. An
important implication of this is that, in terms of clinical competence for primary health care,

clinical care providers with short duration of training appear to be a viable alternative to
physicians.
Most states in India have been posting AYUSH doctors to fill vacancies of Medical Officers
at PHCs and to mainstream Indian systems of medicine. This study casts doubt on the

appropriateness of this practice when it results in the AYUSH doctor becoming the primary

clinical care provider in the PHC. The overall competence of AYUSH doctors and their
performance on the different aspects of the consultation (except for prescriptions) is below

4 For the TB and preeclampsia case, clinical care providers were not evaluated on their prescription
practices because PHCs are expected to refer such cases.
5 The term ‘competence’ is used because in this study provider knowledge (‘what they know’) is measured and
not practice (‘what they do’). However, the two are related since the competence measure can be thought of as
the maximum attainable performance in practice. Further, studies have shown that measures of competence and
practice are correlated.
6 Paramedicals in this study - pharmacists and nurses - are not trained nor are they expected to diagnose
and treat sick people. However, in PHCs where no other clinical care provider is present, Paramedicals
perform these functions.
viii

that of Medical Officers and RMAs. Without additional clinical training, particularly in
primary health care, AYUSH doctors, at current levels of training, do not seem to be the best

alternative to Medical Officers at PHCs. However, they certainly are a better alternative to
having paramedical staff provide clinical care. Moreover, their competence with additional
allopathic training can be quite different and it is important to explore the potential of this.

The low competence of paramedics is both expected and disturbing. They do not receive any
formal training in clinical care nor are they meant to perform such activities. Yet, because
there is no physician or competent alternative, these paramedics continue to treat patients in
numerous PHCs across Chhattisgarh and in the rest of the country. Their functioning as

clinical providers is clearly a danger to their patients and undermines trust in the public health
system. However, in other countries paramedical staff like nurses, have been found to be as
effective as physicians in providing primary health services. Paramedics in the study sample
are, however, dominated by pharmacists. Nevertheless, the potential of paramedical staff with

appropriate training to serve in PHCs is an important area of future research.

Prescription practices

In every clinical care provider group there was a substantial proportion of prescriptions that

were ineffective in treating the patient. The majority of prescriptions written by Medical
Officers and RMAs were effective in treating the presented condition. Notably, almost half
the prescriptions written by AYUSH Medical Officers and the majority written by

Paramedicals were ineffective. In terms of rational drug use, Medical Officers wrote the most

rational prescriptions followed by AYUSH Medical Officers, RMAs and Paramedicals.

All clinical care providers did better at prescribing effective treatments for the malaria and

diabetes cases. However, they did poorly in treating diarrhea and pneumonia - all diseases
contributing substantially to the burden of disease and to mortality. Malaria is the focus of an

established national disease control program and the ability of clinical care providers to
prescribe effectively reflects the success this vertical program has had on frontline health

workers.

ix

Patient and community satisfaction and quality perceptions

Patients were equally satisfied with Medical Officers, AYUSH Medical Officers and RMAs.
However, patients of Paramedicals were less satisfied with their visit. The ‘medical advice’

and ‘physician behavior’ dimensions of patient perceived quality directly refer to the clinical
care provider in the local PHC. Medical Officers and RMAs received the highest scores
followed by AYUSH Medical Officers on both these dimensions; however, there was no

significant difference between their average scores. This suggests that patients perceive the

medical advice and behavior of Medical Officer, AYUSH Medical Officers and RMAs
similarly. Paramedicals consistently received the lowest scores and these were generally

significantly lower than the other clinical care provider categories.

Community satisfaction with PHC services was moderately high for all provider types,
except Paramedicals. Households were equally satisfied with local PHCs where the clinical

care provider was a Medical Officer, AYUSH Medical Officer or RMA. Household
perceptions of the technical skills of the clinical care provider at the local PHC indicate that

they saw no difference in the ability of Medical Officers, AYUSH Medical Officers and
RMAs to treat common illnesses. However, Paramedicals were perceived as having
significantly lower ability to treat common conditions. In contrast, for the treatment of
serious conditions, all the other clinical care providers were perceived to have significantly

lower ability compared to Medical Officers. If we assume that the household definition of
‘common’ or ‘serious’ conditions is similar across groups, then there is a clear vote that

Medical Officers, AYUSH Medical Officers and RMAs are equally able to treat common
conditions but are less able than Medical Officers to manage serious illnesses. There is little
ambiguity that communities view Paramedicals as having little ability to treat either common

or serious conditions. The discerning ability of the community, in what different providers
can or cannot do, was interesting and matches with both the study findings and theoretical

expectations.
Service utilization and equity

Perceptions of service quality are important drivers of where people choose to go for

treatment. PHCs headed by AYUSH Medical Officers received the largest share (60%) of

total visits, followed by RMAs (35%), Medical Officers (29%) and Paramedicals (20%).

x

Where there is no qualified clinical care provider, as in the Paramedicals case, the local PHC

is hardly used for treatment. Indeed, the share of visits to private providers is largest for this

group. The pattern of visits to the local PHC suggest that, all else being the same, there is at
least as much public trust in AYUSH Medical Officers and RMAs as there is in Medical
Officers.

The majority of visits by households located in the vicinity of a PHC were not to the local
PHC - the most assessable government provided care - but to private providers irrespective
of the type of clinical care provider present in the local PHC. The only exception to this case

was areas where AYUSH providers headed PHCs, but this appears to be due to their
exclusively tribal area location where there are few alternatives to the local PHC.
Importantly, the presence of Medical Officers, a rarity in rural India, did not seem to

influence this pattern.

Poorer people depend more on the local PHC compared to those who are better-off. When no
qualified clinical care provider present (i.e. Paramedical PHCs), use of the local PHC is low,

even by the poor, because people seek care elsewhere. When private alternatives are not
available, as in the AYUSH Medical Officer case because of their tribal area location, the
local PHC becomes very important and is heavily used. This highlights the importance of

investing in primary health services. However, the continued use of private fee-for-service

providers, often unqualified, even by the poorest and even when qualified clinical care
providers are present at PHCs, indicates the need to look beyond the package of services
available and examine issues of access and the nature of social relationships in primary care

settings.

Attitudes towards rural service andjob satisfaction

The majority of the sampled clinical care providers intended to transfer from their current

PHC posting at some point. Medical Officers, RMAs, AYUSH Medical Officers were
equally likely to seek a transfer. However, Paramedicals have a slightly lower inclination to
transfer. Among those seeking an immediate transfer, the largest proportion was from

Paramedicals, followed by Medical Officers, AYUSH Medical Officers and RMAs. No
significant differences in the intention to transfer were found between groups.

xi

The sampled clinical providers had moderate levels of job satisfaction. AYUSH Medical

Officers reported the highest levels of job satisfaction, followed by Medical Officers and
Paramedicals. RMAs had the lowest job satisfaction. Similar patterns are seen across the job
satisfaction dimensions of ‘Family life’ ‘Job performance’ and ‘Professional development’;

AYUSH Medical Officers have the highest job satisfaction, followed either by Medical
Officers or Paramedicals, and then RMAs.

Conclusions

The acute shortage of health workers in rural areas, particularly physicians, severely

constrains the delivery of clinical services to rural communities. Findings from this study
support the claim that in primary health care settings in India, clinical care providers with

shorter duration of training are a competent alternative to physicians. This endorses the
introduction of rural cadres like the RMAs of Chhattisgarh or the BRHC course by the health

ministry. Indeed, limiting BRHC graduates to serve only at sub-centers, as is currently

proposed, is under utilizing their potential in a rural environment of physician shortages.5
AYUSH doctors and paramedical staff like nurses also, as demonstrated in other countries,

have the potential to be competent primary care providers but would require substantial
further training.

Introducing a rural cadre requires careful planning; there are many pitfalls on this road but
much can be learnt from the experience of states like Chhattisgarh.9'10 For one, it is important

that the medical establishment supports the creation and sustenance of this cadre. In this

regard, the government has done well to involve the Medical Council of India in developing
BRHC program.6'7 Secondly, local recruitment, as proposed for the BRHC program, can
increase rural retention though it remains to be seen if quality candidates and instructors will
be locally available everywhere. Third, a rural cadre needs a clear career path - where they

can work, what title they can or cannot use, and what functions they can or cannot perform

need to be clearly defined beforehand. Fourth, it is important that graduates be allowed, after
some years of service, to become fully qualified medical graduates either through a ‘bridge

course’ or through preferential admission to medical schools.10 This will improve the prestige

and status of the rural cadre, increase the legitimacy of program graduates among their

superiors and colleagues, and provide avenues for professional advancement within the
health system.
xii

The finding that even PHCs with Medical Officers receive such few visits from people living
in close proximity strongly suggests that improving primary health care in India requires
efforts which go beyond simply placing a qualified clinician in a PHC.11 The quality of care

is important - this study found that the quality of clinical providers at PHCs was poor

irrespective of who was providing clinical services. It is alarming that even after decades of
emphasis, clinical providers find it difficult to correctly diagnose and treat conditions like
diarrhea, pneumonia and preeclampsia.

Clearly, there is an urgent need for improving

standards. Further, planned action is needed on the organizational environment, adequacy of
supplies, management at different levels, incentives, and even the social context of health

care.

A related point is that successful primary health care is built on the trust and rapport between

physician and the communities they serve. Clearly, as this study shows, the mere presence of
a qualified clinical provider is not adequate to make a PHC ‘successful’. The importance of

community trust in the clinical provider cannot be over emphasized and can only be achieved
through the provider’s continued engagement with the community and by being part of them.
This casts doubt on the effectiveness, from the perspective of successful primary health care,

of human resource policies that involve placing clinical providers in PHCs for a short

duration (e.g. compulsory rural service for a few years) or which allow providers to live away

from the communities they serve.

In some ways, the debate over whether non-physician clinicians are a reasonable substitute

for physicians misses the point because the correct comparator is not the physician but the

situation where no physician is present. Non-physician clinicians offer a substantial
improvement over the latter. Placing qualified clinical care providers in rural health facilities
is an important first step in the process of expanding quality health services. However, it is
presumptuous to believe that the simple act of placing a qualified provider in a PHC will
automatically result in increased service utilization and better health. Successful primary

health care results from the interplay of many factors; placing qualified providers is just one
of many necessary, but by no means sufficient, conditions required for a ‘successful’ PHC.

xiii

Chapter 1

The Challenge of Placing Health Workers in Rural Areas

Health systems planning in India has always aspired to universal and affordable health care.

Since Independence, considerable government effort and resources have been devoted to
establishing a wide network of health facilities through which qualified health workers

deliver affordable health services (Box 1.1). Yet, the aspirations of this massive undertaking
have remained unfulfilled. As recent surveys show, substantial socioeconomic and
geographic inequities in health outcomes and health service use mark the Indian landscape

(NFHS, NSSO). Critically, many Indians, particularly those living in rural areas and the

urban poor, do not receive health care from qualified providers.12,13
This chapter provides an overview of the human resource situation in India. It draws attention

to the national and rural shortage in qualified physicians and other health workers. The
challenges involved in addressing this rural scarcity are examined, particularly, the many
state level experiments in recruiting and retaining health workers. The chapter concludes by

exploring the feasibility of and experience with non-physician clinicians to deliver primary
health care.

1.1 India’s health workforce - a situation analysis

India’s health workforce is characterized by a diversity of health workers offering health
services in several systems of medicine. According to the National Occupation Classification
(NOC) providers of allopathic health services broadly include doctors (general and

specialists),

dentists,

nurses,

midwives,

pharmacists,

technicians,

optometrists,

physiotherapists, nutritionists, sanitarians and a range of administrative and support staff. 14
Physicians and surgeons trained in Indian systems of medicine - Ayurveda, Yoga, Unani,

Sidha - and Homeopathy, collectively known as AYUSH, are also important health care

providers. In addition, there are community health workers and practitioners of traditional
medicine and faith healers. Certain states have introduced state specific cadres. For example,
the state of Chhattisgarh has deployed Rural Medical Assistant (RMA) at PHCs and female

RMAs in CHCs without lady doctors. RMAs receive three years of training in allopathic
medicine followed by a year’s internship before being certified.

A large number of informal medical practitioners, commonly called RMPs (Registered
Medical Practitioners), constitute a substantial part of the health workforce.121 RMPs are often
the first point of contact for medical care for the rural population and the urban poor. They

typically practice allopathic medicine, but have no formal qualification or license to do so.

While it is difficult to estimate their numbers, one study estimates that 25% (42% in rural and
15% in urban) of the individuals classified as allopathic doctors reported no medical
training.4 Another study conducted in Udaipur district of Rajasthan in 2003 found that 41%

of private practitioners who called themselves doctors had no medical degree, 18% had no
medical training at all and 17% had not even graduated from high school. 12

Estimates based on the 2001 Census suggest that there were close to 2.2 million health

workers in 2005, which translates into a density of approximately 20 health workers per
10,000 population (Figure 1.1). The estimated density of allopathic doctors is 6.1, nurses &
midwives is 5.8 and AYUSH practitioners is 1.8 per 10,000 population. Allopathic doctors

comprise 31% of the workforce, followed by nurses & midwives (30%), pharmacists (11%),
AYUSH practitioners (9%) and others.4 Census and NSSO estimates are based on self­

reported occupation which is susceptible to unqualified providers being counted as qualified
ones. Census estimates adjusted for educational qualifications are revealing; the density of
qualified health workers reduces to a little over 8, of allopathic physician to 3.8 and of nurse

& nurse-midwives to 2.4 per 10,000 population.4 Overall health workforce estimates do not
include the substantial number of community health workers introduced under the National

Rural Health Mission (NRHM) after 2005.

2

Figure 1.1

Health Worker Density (Per 10,000 population)-All India, 2005

Allopathic Physician
Nurse
Midwife & Related
AYUSH
Dentist
Pharmacist
Others
Other Traditional
All

8

2
Census

10

12

14

16

18

20

Qualified Practitioners*

Estimates based on self-reported occupation in NSSO

Source: Rao K, Bhatnagar A, Berman P (2009)
Others = Dietician & Nutritionist, Opticians, Dental Assistant, Physiotherapist,
Medical Assistant & Technician and Other Hospital Staff
Other Traditional = Traditional Medicine Practitioner, Faith Healer

Estimates based on the Census suggest that the combined density of allopathic doctors,
nurses and midwifes (11.9) is about half of the WHO benchmark of 25.4 workers in these

categories per 10,000 population for achieving 80% attended deliveries by skilled personnel
in cross-country comparisons.16 When adjusted for qualification, the density falls to around

one fourth of the WHO benchmark.4 The Census estimates also indicate that India has a
skewed mix of nurses and allopathic doctors. There is approximately one nurse and nurse­

midwife per allopathic doctor and the qualification adjusted ratio falls to 0.6 nurses per

doctor. Although there is no gold standard for a nurse-doctor ratio, a higher ratio is desirable
because nurses can deliver basic clinical care and public health services at a lower cost than

doctors. The only states in India which have two or more nurses per doctor are the north­
eastern states, Kerala, and Orissa.4

3

Box 1.1

India’s Health System: An Overview

India’s health system is characterized by a
large public and larger private sector. The
public sector consists of a hierarchy of health
facilities comprising of sub-centers, primary
health centers (PHC), community health
centers (CHC). district hospitals and
specialty/research hospitals (see figure). The
private sector is largely unregulated and
heterogeneous: it comprises of super­
specialist hospitals, nursing homes, clinics,
unqualified allopathic practitioners, trained
practitioners of indigenous systems of
medicine and traditional health care providers.

Tertiary
Hospital
District
Hospital

• V. :.■! -t nospitaS

Community
Health Center

Primary Health Center

• SO/XKJ 120,000
an
• 50+ beiii
.n:<! spec-^iu -suiget.'
(H’dUh-'riiin, jCvnsixikjg'St
•Jix tois

• ? ptiysKian?. .*
• 2 « BCiJri

.jnn

Sub Health Center level
Community Health Workers

Despite the presence of an extensive network of public sector health facilities, the majority of
inpatient and outpatient care is provided by the private sector and this share has gradually increased
over time. Latest estimates indicate that 80% of all ambulatory and 50% of in-patient treatment occurs
in the private sector. However, the public sector continues to be the major provider of preventive
services. One of the unfortunate consequences of India’s highly privatized health system is that,
coupled with insurance covering only a small percentage of the population; nearly 80% of the total
health expenditure is paid out-of-pocket. For many Indians, especially those who are poor, health care
payments place an enormous burden leading to people falling into poverty, experiencing catastrophic
health care payments or undertaking distress financing to pay for health services.
Since health is a state subject in India’s federal system, the respective state governments are
responsible for administering and funding the public sector. Common norms guide the states resulting
in similar public sector structures across the country. The Central government, however, is also an
important financier of health care. This is primarily done through centrally sponsored schemes
through which health initiatives of national importance receive direct funding from the center.
Examples of these programs include, all the national disease control programs, the family planning
program, the reproductive and child health program and. most recently, the National Rural Health
Mission (NRHM). These programs, depending on the situation, have their own cadre of workers or
fill vacancies in the public sector by hiring workers on contract or make use of the state level health
workforce.

The National Rural Health Mission (NRHM), which was launched in 2005, is a key recent health
system initiative launched by the central government. It aims to bring about an ‘architectural
correction to the health system’ through a variety of strategies, such as, substantial increases in
government funding for health, integrating vertical health & family welfare programs, providing a
female health activist in each village, de-centralized health planning, communitization of health
services, strengthening of rural hospitals, providing untied funds to health facilities and
mainstreaming traditional medicine systems into the public health system (NRHM Mission
document). It covers the entire country, with special focus on 18 states, which have relatively poor
infrastructure and demographic indicators. One of the core strategies of NRHM is to integrate into the
general health system the different national programs, including the disease control programs - the
only exception to this is the HIV/AIDS program.

4

Doctor Density (per 10,000 population), 2005

Figure 1.2

f JSAW J 41 ASfftIP

y

MAHARASKrW

□3-4 LAhS^
□ 4-6

□ 6-8
■ 8-23



Source: Rao K. Bhatnagar A, Berman P (2009)

Health workforce distribution

Health workers in India are unevenly distributed across the country (Figure 1.2). In general,
the north-central states, which are among the poorest in terms of both their economy and
health, have low health worker densities. The distribution of health workers is highly skewed

in favour of urban areas with around 60% of the health workers present there (Figure 1.3).4
This mal-distribution is substantially exacerbated when adjusted for the larger share of the

population residing in rural areas. The density of health workers per 10.000 population in

urban (42) is nearly four times that of rural (10.8) areas. The density of allopathic doctors is
four times larger in urban (13.3) compared to rural (3.3) areas and for nurses and midwifes
the difference is three times as large (15.9 urban, 4.1 rural) 4 If these estimates are adjusted
for health worker qualification, then the density of allopathic physicians in urban and rural

areas reduces to 11.3 and 1.9 per 10,000 population, respectively, reflecting the higher

proportion of physicians in rural areas reporting insufficient qualifications 4 AYUSH doctors
also have a stronger presence in urban (3.6) compared to rural (1.0) areas.4

5

Figure 1.3

Rural-Urban Distribution of Health Workers in India, 2005
Note: Numbers on the Bars Indicate Density (Per 10,000 Population)

I

Allopathic Physicianl

3.28

Nurse & Midwife I

4.13

r

L
AYUSH I
L
Dentist I

r
Pharmacist I

13.34

I

1.04

3.64

0.0‘S

0.59

___________

1.33

4.28

0.66

3 37

Other Traditional!

0.28

0.92

L
Alli

10.78

0

20

r
Others I
r

40

100

80

60
Percentage

BB Rural

B

■ Urban

Source Rao K, Bhatnagar A, Berman P (2009)

AYUSH = Ayurvedic, Yoga, Unani; Others = Dietician & Nutritionist, Opticians, Dental Assistant. Physiotherapist, Medical Assistant &
Technician and Other Hospital Staff; Other Traditional = Traditional Medicine Practitioner Faith Healer

Figure 1.4

Distribution of Health Workforce by Sector, 2005
Rural

Urban

Allopathic Physician

Allopathic Physician

Nurse & Midwife

Nurse & Midwife

AYUSH

AYUSH

Dentist

Dentist

Pharmacist

Pharmacist

Others

Others

Other Traditional

Other Traditional

All Health Workers

All Health Workers
0

20

40

60

80

100

Iffll
0

Percentage

20

40
60
80
Percentage

100

JBNon-Government MII<-’0VernrTient

'UBNon-Government ^Bcovernment

Source: Rao K, Bhatnagar A, Berman P (2009)
AYUSH = Ayurvedic, Yoga, Unani; Others = Dietician & Nutritionist, Opticians, Dental Assistant, Physiotherapist, Medical Assistant & Technician
and Other Hospital Staff; Other Traditional = Traditional Medicine Practitioner, Faith Healer

6

The majority (70%) of health workers in India are employed in the private sector (Figure
1.4). Significantly, the vast majority of allopathic doctors (80%), AYUSH physicians (80%)

and dentists (90%) are employed in the private sector. In contrast, only about half the nurses
and midwifes are employed in the private sector.4 This pattern holds for rural and urban areas
and also after adjusting for health worker qualification. However, the distinction between
public and private sector is not always clear since, for instance, public sector doctors often

indulge in private practice.

1.2 The scarcity of qualified health workers in rural areas

Many rural and poor urban Indians receive curative care from unqualified providers due to
the scarcity of qualified physicians in these areas.1217 This

scarcity is due to both the

disinclination of qualified private physicians to work in underserved areas and the inability of
the public sector to adequately staff rural health facilities. Latest government estimates
indicate that currently 18% of the PHCs are without a doctor, about 38% were without a lab

technician and 16% lacked a pharmacist.18 Specialist allopathic doctors are particularly in
short supply in the public sector with 52% of the sanctioned posts of specialists at CHCs

vacant. This includes vacancies in 55% of surgeon, 48% of obstetricians & gynecologist,
55% of physician and about 47% of pediatrician posts. Nurse vacancies are also high - 18%

of the posts for staff nurses/nurse-midwives at PHCs and CHCs are vacant.

The actual

number of PHCs and CHCs without adequate staff will be considerably higher given high
health worker absenteeism.19

There are several reasons for the scarcity of qualified health workers in rural areas. The

opportunity to earn a better income, to utilize skills, good living conditions, education

opportunities for children and safe working and living environments are other important job

attributes which tilt the balance in favor of urban location.4 Of particular concern is the
inability of the public sector to place adequate doctors in rural areas. For many medical
graduates the desire for post-graduate specilization dissuades them from entering the job

market and thereby the possibility of rural posting in the public sector.20 Once they have

specialized, government employment and rural service is not attractive. Nurses are more
amenable to public sector enployment than doctors - around half the nurses in India work in
government jobs.4 However, poor service and living conditions for their families makes
7

urban employment preferable.2122 Public sector efforts to recruit and retain health workers to
rural postings is also compromised by institutional issues such as changes in service rules,

recruitment delays, the lack of transparancy in identifying vacancies, promotions, transfers

and the numerous related court cases faced by the state health Directorates.2'"

Historical perspectives

Several policies that affected human resources for health in India were adopted in the early

years of Independence. The Indian Medical Service, an all India medical service, which was

primarily concerned with the health of the colonial Army, was abolished. The colonial policy
of sharing of powers with states was retained; health and the workforce became principally a
responsibility of the states leaving the central government with a limited role in these areas,

except through financing centrally sponsored programs. Key health workers in undivided

India were few with 1.6 doctors and 0.23 nurses per 10,000 population.24 The majority (70%)

of physicians were private practitioners and mostly in urban areas.24 There were two classes
of allopathic physicians present; doctors who underwent a five and a half year course and
licentiates (LMPs) who underwent a three-to-four-year course. Nearly two-thirds of the

qualified medical practitioners were Licentiates who naostly served in rural areas.2'3

In

addition, there was a substantial presence of Indian system of medicine practitioners and

traditional healers.2

The adoption of the Bhore (1946) and subsequent committee reports directed government

attention and resources towards establishing a publicly funded and managed health system.
The focus was on bringing primary health care to rural areas and central to this plan was the
production of a ‘basic’ doctor schooled in clinical skills and public health. The licentiate

physician did not find favor with the Bhore Committee leading to, despite much dissent, this
cadre being abolished. India was to produce only one type of allopathic physician, the five
year MBBS graduate, and this physician would be on par with western doctors.3 Further,

there was no role for physicians schooled in Indian systems of medicine. The result of these

policy decisions is starkly visible today. Qualified doctors are scarce in rural areas while the

country exports a substantial number to developed countries.25’26 At primary health centers,
AYUSH doctors and paramedical staff fill this vacuum. Interestingly, two states in India

8

(Chhattisgarh and Assam) have recently reincarnated the licentiate by creating a new cadre of

allopathic physician having three years of medical training for serving in rural areas.

Box 1.2

Medical Education

Medical education in India consists of a basic undergraduate degree (MDBS) and then the
option to specialize leading to a post-graduate degree. Undergraduate medical education
consists of one and a half years' preclinical and three years' clinical teaching, after which the
MBBS degree is awarded. The graduate then undertakes one year of compulsory internship in
different hospital departments. The relevance of this medical education to India’s context and
healthcare problems has been questioned. One study noted that students do not learn enough
about the common infectious diseases or problems of child and maternal health.27 Further, the
Government itself has noted, most medical graduates are not adequately trained to perform in
the primary health care setting.28’3” Students themselves do not seem confident to treat
patients on the completion of their degree.2” 31 Further, the strong desire to specialize, and
there is intense competition for the limited post-graduate seats, leads students to utilize the
compulsory internship period, which is meant to strengthen clinical skills, for studying for
post-graduate entrance examinations.2”32 Once students specialize there is little incentive or
inclination to serve in a primary care setting or in a rural area.

Task shifting and non-physician clinicians

Clinical care providers with shorter duration of medical training are now seen as a cost-

effective means of delivering primary care services/3 Where qualified physicians are scarce,
these non-physician clinicians offer an important way to continue services by performing

many of the clinical and non-clinical functions of physicians. This task shifting is common in
many developing and developed countries. For instance, nurse-practitioners in the United

States provide basic clinical services. In developing countries, a variety of non-physician
clinicians have been deployed in response to rural shortages of health workers. Depending on

their location and function, these health workers have been known as feldshers, medical

assistants, hospital assistants, health officers, rural health technicians, health post aides,

village/community health workers, health officers, clinical officers, physician assistants,
nurse practitioners, or nurse clinicians.134

The training and functions of these non-physician clinicians vary substantially. Community

health workers (CHW) and similar cadres undergo a short period of training, typically lasting
a few months or less, and serve to treat simple ailments, provide health education and connect

9

households with health facilities.35 At the other end of the spectrum are non-physician
clinicians with upto four years of clinical training. They seem to be particularly popular in
sub-Saharan Africa with one study reporting their presence in 25 of 47 countries there.1 In

several African countries they form the backbone of the health system being the main
providers of clinical care and, in some instances, surgical procedures.1,36

Few evaluations have been conducted on the quality of care produced by non-physician

clinicians. A systematic review of primary care provided by nurses (practice nurses, nurse
practitioners, clinical nurse specialists, or advanced practice nurses) found that, in general, no
appreciable differences were found between them and doctors (general practitioners, family
physicians, pediatricians, general internists or geriatricians) in health outcomes for patients,

process of care, resource utilisation or cost. Further, patients treated by nurses reported higher
satisfaction level. '7 An assessment of primary health facility surveys in Bangladesh, Brazil,
Uganda and Tanzania compared the clinical performance of health workers with longer

duration of pre-service training (those with >4 years of post-secondary education in Brazil or

>3 years in the other three countries) and shorter duration trained health workers providing

clinical care. Perfomiance was assessed in terms of assessment, classification, and
management of sick children according to.lMCI guidelines. The authors concluded that IMC1
training is associated with much the same quality of child care across different health worker

categories, irrespective of the duration and level of pre-service training. In Tanzania,
Assistant Medical Officers, who receive three years of training as clinical officers and then an
additional two years (including surgery and obstetrics) to become Assistant Medical Officers,

provide emergency obstetric surgery. An assessment based on records of patients admitted

for complicated deliveries at fourteen district hospitals found that that there were no

significant differences between the Assistant Medical Officers and Medical Officers in
patient outcomes or quality.36

In India, the focus on doctors to provide primary care ignored experiments taking place in
other parts of Asia (e.g. barefoot doctors in China) and elsewhere with non-physicians

providing clinical services. Though several small area experiments with CHWs took place in
different parts of India starting in 1970,38,39 only in 1978 was the first national Community

Health Volunteer program launched. These early attempts were not successful, though later,
programs like the Mitanin programme (2002) in Chhattisgarh, which is the forerunner of the

10

national ASHA program under NRHM, have fared much better.39,40 With the coming of the

NRHM in 2005, CHWs are once again being introduced in a major way onto the health
landscape.

A notable development in India is the provision of primary care services by AYUSH doctors
and clinical care providers with a short duration of training. In keeping with the National
Health Policy 2002 and NRHM guidelines to mainstream Indian Systems of Medicine into

primary care, AYUSH doctors (see Box 1.3) are being deployed at PHCs as a second medical
officer, the first one being the MBBS doctor. AYUSH physicians are recruited on contract

and have either an under-graduate or post-graduate degree in a particular Indian system of

medicine or homeopathy. AYUSH practitioners in government service undergo some training
in allopathic medicine, particularly in the control and treatment of diseases covered in the
national disease control programs (e.g. TB, malaria, leprosy, and blindness) and vaccinations.

AYUSH physicians posted at health facilities prescribe both AYUSH and allopathic
treatment.

In Chhattisgarh and Assam, a significant strategy to address rural doctor shortages has
involved creating a clinical care provider with three years of training to serve in rural areas.

In Chhattisgarh, these physicians receive a degree of Practitioner of Modern and Holistic

Medicine at the end of their three year course (Box 1.4). They are recruited to government
service to serve at PHCs as RMAs, though lady RMAs are posted in CHCs where there are
no lady doctors.

11

Box 1.3

AYUSH Doctors

Indian medicine systems such as ayurveda, unani, siddha and yoga, together with
homeopathy, have had a long history of existing alongside allopathic systems. These systems
of medicine are collectively known by the acronym of AYUSH. Ayurveda and siddha are
ancient Indian systems of medicine, yoga is the ancient Indian system of exercise postures,
and unani has its roots in ancient Graeco-Arab medicine.
The Government has supported the development of AYUSH by establishing teaching
institutions and a dedicated network of hospitals and clinics. After the National Health Policy
(1983), integration of AYUSH and allopathic medical systems was promoted. This led to the
establishment of the Department of Indian Systems of Medicine and Homeopathy in 1995
within the Ministry of Health and Family Welfare. In 2005 the National Rural Health Mission
made mainstreaming AYUSH within the public health system one of the central features of
the program.

The Central Council of Indian Medicine (CCIM), a statutory body constituted by the
Government of India under the Indian Medicine Central Council Act. 1970, prescribes
minimum standards of education, syllabi and course, codes of practice and advises the
Government in matters relating to recognition of medical qualification. Currently there are
254 ayurveda. 41 unani and 9 siddha medical colleges offering undergraduate and post­
graduate degree courses. 41
The first degree in ayurVeda (BAMS), unani (BUMS), siddha (BSMS) and homeopathy
(BUMS) lasts approximately five and a half years and is comparable in duration with a first
degree in allopathic medicine (MBBS). The course comprises of approximately four and a
half years of in-class work and a one year internship period typically at a District Hospital.
While the BAMS (i.e. Bachelor of Ayurvedic Medicine and Surgery) syllabus mainly focuses
on training in ayurvedic medicine, during their course work and internship students also
receive training in allopathy. They study anatomy, modern physiology and basic pathology.
They learn to conduct and evaluate blood, urine and stool examinations. Students are trained
to conduct minor surgical procedures like incision, abscess drainage, suturing, wound
management, and management of surgical emergencies. They also learn how to conduct anteand post-natal care and the management of minor and major pregnancy disorders.
The Supreme Court of India has permitted ayurveda, siddha, unani and homoeopathy
practitioners to prescribe allopathic medicines under Rule 2(ee) of the Drugs and Cosmetics
Rules, 1995 in those states where they are authorized to do so by a general or special order
made by the concerned state government. Practitioners of Indian medicine holding degrees in
integrated courses can also prescribe allopathic medicines if any state Act in the state in
which they are practicing recognizes their qualification as sufficient for registration in the
State Medical Register.

12

Box 1.4

The Rural Medical Assistant (RMA)

Efforts to expand rural health services in the state of Chhattisgarh, like in most states in India,
are constrained by a lack of physicians willing to serve in rural areas. The state governments
responded to this problem by introducing a new cadre of rural clinical care provider, the
Rural Medical Assistant (RMA). RMAs undergo a shorter duration of training in allopathic
medicine relative to physicians and can only serve in PHCs.
Six colleges to train RMAs opened in Chhattisgarh between October 2001 and the end of
2002. admitting upto 150 students per year. However, as the course faced opposition from the
Indian Medical Association, its name underwent several changes and ultimately came to be
known as 'Diploma in Modern and Holistic Medicine*. The training students receive includes
three years of course work in allopathic medicine followed by one year of internship. They
study the same texts as students in a MBBS degree program. The internship is focused on
rural health; it includes one month training at a sub-centre, three months at a PHC, 4 months
at a Community Health Center (CHC) and 4 months rotational posting at a District Hospital.

The address the challenge of providing health services in rural areas and to create
employment opportunities for these students, the state government created the post of Rural
Medical Assistant (RMA) in the state health service. These posts were sanctioned at PHCs
classified as remote or in tribal districts. RMAs are hired on contracts. In non-tribal areas
they receive Rs. 8,000 per month (significantly less than the salary of a MBBS trained
physician) and those in tribal areas receive Rs. 9,000 per month. The state recently increased
the total number of RMA posts to 858 and introduced one RMA post at all PHCs and an
additional post for Lady RMAs at the CHC level in all the 18 districts (difficult, rural or
tribal) of Chhattisgarh.9 In principle, RMAs are always posted under the supervision of a
Medical Officer. Private independent practice is not permitted.
RMAs are permitted to carry out a variety of functions, including:









Assist in implementation of all national and state level health programs
Provide preventive health care
Provide primary health care services and referral
Provide basic maternal and child health care, conduct deliveries, basic management of
complications of pregnancy and childbirth, suturing of first degree perineal tears
Perform simple operative procedures: repair of small wounds by stitching, drainage of
abscess, burn dressing, applications of splints in fracture cases, application of
tourniquet in case of severe bleeding wound in a limb injury
Procedures not to be performed: Medico-legal cases, post-mortem

13

Chapter 2

Study Objectives
The shortage of physicians in rural areas of India is one of the biggest constraints to

achieving universal health care. This scarcity is due to both the disinclination of qualified
private physicians to work in underserved areas and the inability of the public sector to
adequately staff rural health facilities. One consequence of this public sector shortage is that

clinical services in primary health care settings are being provided by a variety of non­
physician ' clinicians. In several states AYUSH doctors, who were meant to serve as the
second Medical Officer in a PHC, have become the main providers of clinical services. Some
states (e.g. Chhattisgarh) have introduced a new cadre of rural clinical care provider, the

RMA, who undergo a shorter duration course in allopathic medicine and can serve only in
PHCs when they join government service. Quite frequently paramedical staff (e.g. nurses,
medical assistants, pharmacists), with little or no clinical training, provide clinical services at

PHCs because no one else is available.

The functioning, either by design or circumstance, of clinical care providers with shorter

training duration or AYUSH doctors and paramedical staff (pharmacists and nurses) as the

main providers of clinical services at PHCs represents a set of alternatives to physicians. Yet,
there is little known about how well these non-physician clinicians perform duties expected
of Medical Officers at PHCs.

This study provides a comparative assessment of the performance of different types of

primary health care providers - Medical Officers, clinical care providers with short training
duration i.e. RMAs, AYUSH Medical Officers and paramedical staff (nurses and

pharmacists) - in their capacity as the main providers of clinical services at PHCs. The

performance of these health workers is examined on several dimensions. Their ability to

manage common illnesses seen at PHCs was assessed in terms of their competence (‘how

7 Throughout this report we use the term physician to refer to a medical doctor or a Medical Officer i.e.
those with a MBBS degree or higher qualification. We use the term ‘non-physician clinician’ to denote
clinical care providers who have undergone shorter duration training in modem medicine relative to physicians.
In the context of this study non-physician clinicians include AYUSH doctors (BAMS degree or higher). Rural
Medical Assistants (RMA) and paramedical health workers (pharmacists and nurses).

14

much they know’). The perception of their patients was assessed in terms of satisfaction with

services and perceptions about the quality of care they received. The use of the local PHC by

ill community members and community perceptions of the local PHC was also examined.
Assessing the performance of these different types of health workers on a variety of
dimensions enables a comprehensive understanding of their suitability as primary health care

providers. The study was conducted in the state of Chhattisgarh in central India in 2009.

The rest of this report is structured as follows. Chapter 3 describes the state of Chhattisgarh,
the location of this study. Chapter 4 provides a description of the sampling design and
Chapter 5 describes the characteristics of the sample. Chapter 6 discusses results from the

assessment of the clinical competence of the different clinical care providers sampled in this
study. Chapter 7 examines the prescription practices of these clinical care providers. Chapter

8 examines results from the assessment of patient satisfaction and perceived quality. Chapter
9 discusses results from the community survey, including, care seeking and community

perceptions of the local PHC. Chapter 10 discusses findings from job satisfaction assessment
of the sampled clinical care providers.

This study was primarily funded by a grant from the Alliance for Health Systems and Policy

Research, WHO to the Public Health Foundation of India (PHFI). Additional funding for this

study was from the National Health Systems Resource Center (NHSRC), Ministry of Health
and Family Welfare, Government of India and the State Health Resource Center (SHRC),

Chhattisgarh. NHSRC and SHRC also provided technical support for the design and conduct
of field survey. Department of Health and Family welfare, Government of Chhattisgarh

facilitated the fieldwork. The views in this report are solely that of the authors. Ethical
clearance for this study was received from the PHFI Institutional Ethics Committee and the
WHO Research Ethics Review Committee. Informed consent was taken from all those who
participated in the study.

15

Chapter 3

Chhattisgarh State
Chhattisgarh is one of the youngest states in India. It was carved out of the state of Madhya

Pradesh in November 2000 with Raipur as the state capital. Chhattisgarh is the 10th largest
state of India and is spread across an approximate area of 135,194 sq km. Nearly half the
state is forested and accounts for 12% of India’s forests.

Chhattisgarh shares borders with the states of Madhya Pradesh and Maharashtra in the west,

Orissa in the east, Andhra Pradesh in the south, Jharkhand and Uttar Pradesh in the north.
Eighteen districts make up the state (Figure 3.1). According to the 2001 Census, Chhattisgarh
has a population of 20.7 million people and a population density of 154 persons per square

kilometer. This compares favorably with the national average of 324 persons per square
kilometer.

Chhattisgarh is a predominantly rural state with one fifth of the population living in urban

areas (Table 4.1). It also has a large tribal population; 30% of the population is tribal (Table
4.1). It is home to many of the primitive tribes of India and has a high concentration of Gonds
who inhabit the hilly region of the state. Literacy levels are low, particularly for females.

Remarkably, there are twice as many literate males than females (Table 4.1).

Insurgency

Large areas of Chhattisgarh state are currently experiencing armed conflict. Seven districts

(Bastar, Narayanpur, Dantewada, Bijapur, Kanker, Sarguja and Rajnandgaon) are particularly

affected by violence between the government and violent left wing groups (Maoists and
Naxalites). The Government is finding it difficult to implement public programs in these
areas. Out of these seven districts, five are in the Bastar region in southern Chhattisgarh.

These districts lack basic amenities including transportation and communication facilities,

electricity, water supply and skilled human resources. Since the start of the conflict in 2006
there have been a lot of killings and social unrest in the northern and southern parts of the
state. Frequently, the violence spills into other parts of the state as well.

16

Figure 3.1

Districts of Chhattisgarh

INDIA
**'

T'"
Koriya

Surguja
Jashpur

"V
Korba
Bila^pur

JanfgirCh^mpa

l< a birdham

RAIRUR

Rajnandgaon

Raigarh

Mahasan' und

Durg

Dham
tart
Kanker

Ba'.tar

Dan tew a da

J

Economy

Chhattisgarh is rich in natural resources. The state’s dense forests offer much potential for

forest products. It is also one of the richest states in mineral resources. The state has mega
industries in steel, aluminum and cement. Agriculture and allied activities forms the base of
the state’s economy and provides livelihood to 80% of the rural population.42 The central

plains of the state are known as the ‘rice bowl’ of central India. The major crops produced
include sugarcane, pulses, banana, wheat, rice and pulses. The rural economy also contains a
thriving non-farm economy.

17

Table 4.1 Social and Demographic Characteristics of Chhattisgarh
Indicator
Demographic

Caste and tribe

Religion

Education

Health

Economy

Population (millions)
Male (%)
Urban (%)
Sex ratio

India
1,028.7
52
28
933

Chhattisgarh
20.7
50
20
990

8

30

Tribal (%)
Schedule caste (%)
OBC/Other

19

14

72

57

Hindu (%)
Muslim (%)
Christian (%)
Sikh (%)
Other (%)

82
13
3
2
2

95
3

No education (%)
Infant mortality rate
Stunting in children (%)
Children fully immunized
Institutional deliveries
Total fertility rate
Life expectancy at birth

42
57
48
43
39
2.68
64

Households with electricity (%)
Per capita net state domestic product (Rs)
Population below poverty line (%)

68
37,490
28

1
0.1

0.7

47
70.8
53
49
14

2.62
58
60
34,483
41

Sources: NFHS (2005-06), Census (2001), SRS (2007), Economic Survey of India 2009-10

Despite all this natural wealth, Chhattisgarh is one of the poorest states in India with 41% of
the population below the poverty line. Nationally, 28% of the population is below this line

(Planning Commission of India 2004-05). The state’s per capita income is slightly below the
national average (Table 4.1). Moreover, Chhattisgarh is experiencing a high economic

growth rate of 16%, which is higher than the national growth rate of 12%.43
Health
Chhattisgarh has some of India’s worst health indicators. The infant mortality rate (70.8) in

2005 was among the highest in India and substantially higher than the national average (57).

Chhattisgarh also has one of the highest levels of child malnutrition in the country; among
children under three years of age in 2005, 53% were stunted and 48% underweight (NFHS-

3). However, there have been improvements; in 1997, the prevalence of stunting (61%) and
18

underweight (53%) in children was substantially higher. Life expectancy at birth is also

below the national average. Chhattisgarh is also a malaria endemic area; in 2006, it

contributed 7% of the total malaria cases and 11% P. falciparum cases reported in the country

(NVBDCP,2007).
Though full immunization coverage in Chhattisgarh is higher than the national average, 1 in 2

children are still not fully immunized. Remarkably, immunization levels are higher than the
national average and testimony to efforts made by the state to strengthen the health workforce
at the community level. Institutional deliveries (14%) in 2005 were low and remarkably

lower than the national average (Table 4.1).

The state government has taken major initiatives to improve health services. An important
initiative of the state is the deployment of 60,000 mitcinins or community health volunteers
(CHV) in 2002. This has been one of India’s most successful CHV programs. The mitanins
undergo 13 rounds of training including in IMNCI and giving first level curative care. Further
rounds of training are also being planned. The mitanin program has influenced the design of

ASHA (Accredited Social Health Activist) scheme under the National Rural Health Mission

launched by Government of India. Another important human resource innovation by the state

has been the introduction of a carder of allopathic doctors with short duration of training, the
Rural Medical Assistant (RMA), to serve in PHCs (Box 1.4).

19

Chapter 4
Study Design and Sample
4.1 Questionnaires

Five questionnaires were used to collect information related to the study objectives. These
questionnaires were used to collect information about clinical care providers at PHCs, their

patients and the communities where the sampled PHCs were located.

In addition,

information was collected on structural attributes of the PHCs and the village where the PHC

was located. These five questionnaires were first produced in English after which they were
translated to Hindi and the translation verified. All questionnaires were subjected to several
rounds of pretesting. The questionnaires used in the study are described below.

Form 1 - Clinical Vignettes: This assesses the competence of PHC clinical care providers.

The form contains six clinical vignettes (cases) developed for conditions commonly seen at
PHCs. These six cases are described below:

Case

Primary Complaint

Case I

A 35 year old female patient having high fever and a Malaria
headache

Case II

An 8 month old male baby having diarrhea for the last Diarrhea with severe
dehydration
two days

Case III

A 2 year old girl having a cough

Case IV

A 24 year old woman, who is about 6-7 months Pregnancy induced
hypertension
pregnant, having a severe headache

Case V

A 50 year old man having cough, fever and fatigue

Tuberculosis

Case VI

A 46 year old man complains of exhaustion and
frequent urination in large volumes

Diabetes Mellitus

Diagnosis

Pneumonia

20

Each of these cases contained individual sections on patient history, examinations,

diagnostics, diagnosis, prescription, referral and advice for home care. These cases were
selected based on the disease profile of Chhattisgarh, conditions commonly seen at PHCs and

health priorities of the state. The vignettes were developed based on a review of the literature,

standard treatment guidelines and common practices suggested by treating physicians at
PHCs in Chhattisgarh, experts from All India Institute of Medical Sciences in Delhi, Raipur

Medical College, practitioners at district hospitals and private doctors in Chhattisgarh (see
Annex 4). The questionnaire was pretested with a sample of practitioners from all four types

of clinical care providers under study.

Form 2 - Patient Exit Interview: This form was used to collect information from patients
exiting the PHC. It asked about the patient’s background, information about their
consultation

i.e.

what the clinical care providers had asked and done, the patient’s

prescription, the patient’s satisfaction and perception of service quality and the
socioeconomic status of the patient.
Form 3 - Clinical Care Provider Assessment: This form collected background information
about the main clinical care providers at the PHC, the training they have received during

service, the cases they have seen in the past three months and the procedures they have
performed. Further, the questionnaire collected information on the clinical care provider’s
intention to continue service and the reasons for wanting and not wanting to do so. The final

section of the questionnaire asked about the clinical care providers job satisfaction. This was
a self-administered questionnaire.

Form 4 - Household Survey: This form was used to collect information from households in
the village where the PHC was located. Sampled households were asked about illness of

family members in the past month, whether treatment was sought, if so where it was sought
and how much was paid. This form also collected information about household member
opinions on the PHC provider and services. Information on household asset ownership was

also recorded in this form.

Form 5 - Health Facility Assessment: This form was used to extract information from the
PHC registers on monthly visits to the PHC during the past three months and the main

21

complaints of patients visiting the health center in the past week. The form was also used to
record information on the physical characteristics of the PHC building, the cleanliness of the

PHC and whether equipments and drugs were present.

4.2 Survey design and sample

Data for the study was collected in two phases between July and September 2009. In the first
phase, data on provider competence and job satisfaction was collected i.e. Form 1 and 3, by
inviting PHC clinical care providers to designated interview centers. In the second phase,

PHCs were visited to collect information from patients, households living in the village
where the PHC was located and the PHC.

Selection ofPHCs

To get a representative sample of different types of clinical care providers serving at PHCs in

Chhattisgarh, first, a listing ofPHCs and the staff present at these PHCs was compiled based
on information supplied by the State Programme Management Unit, NRHM, Department of

Health and Family Welfare, Chhattisgarh. This list contained information on the staff
currently posted and in-position at PHCs in the state. Second, this information was verified
with officials located in the PHC’s district like the Chief Medical and Health Officer

(CMHO), the Block Medical Officer (BMO). Third, PHCs that were located in areas
experiencing armed insurgency and unsafe for surveyors to visit were excluded from the list.
This final list constituted the sampling frame for PHCs.

The staffing pattern of PHCs indicated that several types of clinical care providers - regular
Medical Officer, contractual Medical Officer, AYUSH Medical Officer and RMAs - were
the main providers of clinical services at PHCs in Chhattisgarh. Regular Medical Officers

possess at least a MBBS degree and are in regular government service. Contractual Medical
Officer possess at least a MBBS degree and have been hired on a contractual basis. AYUSH

Medical Officers, in this case ayurvedic physicians, have at least a BAMS degree, have been
hired on a contractual basis as the second Medical Officer in the PHC. RMAs are allopathic
clinical care providers with a short duration of training and have been hired on a contractual
basis. Where these clinical care providers were not present, the PHCs were functioning with
22

pharmacists, staff nurses, dressers, auxiliary-nurse midwives (ANM) and other lower level
health workers. This information allowed PHCs in Chhattisgarh to be classified into the
following six groups or strata according to who was primarily responsible for providing

clinical services - regular Medical Officer, contractual Medical Officer. AYUSH Medical
Officer, RMAs, Paramedicals (pharmacists and staff nurse) and others (dressers, ANM and

other lower level staff) (Table 4.1 and Figure 4.1).

Figure 4.1

Sampling Design
?’IK ' m(.'hh(HU’'garhf

___
\\ I SH
Medical
Olficer
J

Regular
Medical Ofikei
(210i

Phase 1

RMA

V
Sampled
(37)

Sampled PHCs
(35)

Sanipk.-d FHCs

/
Patisms
(269)

Househ&td
(525)

Patients
(296)

<

...... / j \

HousehcM l I PHi< nt*
(273)
(503)

I :!• iseh >ld
|

I itiw-J-S

on

Saa.pled
(39)

/V

\
Phase 2

I’uramedkah
(5M

/\
/
Panel
(..‘-w;

Household
(525]

There are two issues with this classification. First, several PHCs had more than one type of
clinical care providers present. For example, several PHCs had both AYUSH and regular
Medical Officers on their staff. In these cases, we considered the regular Medical Officer to
be the main clinical care providers there. Secondly, it was not possible to know if the health

worker identified as the main clinical care providers at the PHC is regularly present. For

instance, a PHC might have a regular Medical Officer posted but the pharmacist provides
clinical services because the Medical Officer is typically absent. To minimize
misclassification health workers at the selected PHC were telephonically contacted to verify
who provides clinical services.

In this study, only PHCs in the regular Medical Officer, AYUSH Medical Officer, RM A, and

Paramedical group were sampled. Contractual Medical Officers were not sampled because
23

they are similarly qualified as their regular counterparts. PHCs where the only staff present
was dressers, auxiliary-nurse midwifes and lower level health workers were excluded
because these health workers do not generally provide clinical care. To get a representative
sample of clinical care providers in each group, within each of these strata, simple random

sampling without replacement was used to select 40 PHCs in each strata.

Selection ofpatients

A convenience sample of ten outpatients was selected as they exited the PHC. Only those
patients who were visiting the PHC for the first time for their current illness were eligible for

interviews. However, they could have consulted other health providers for the same condition
prior to visiting the PHC. In PHCs where more than one clinical care provider was present

(e.g. regular Medical Officer and AYUSH Medical Officer or two Medical Officers), only

patients of the main clinical care provider who was interviewed in the first phase were
selected. Patients were interviewed after taking informed consent.

Selection of households

In Chhattisgarh, villages are divided into paras or clusters of households based on caste and

religious affiliation. With the assistance of PHC staff, a listing of all paras in the village
where the PHC was located and the approximate number of households in each para was
compiled. From this list one para was randomly selected and 15 households in the para
sampled. Households were selected as follows. At the boundary of the para the surveyors

selected a random direction by spinning a pen. Every second household in this direction was
sampled till 15 households were completed. If a selected household refused to be

interviewed, the neighbouring household was included instead. In each household visited, a
competent adult family member was identified to serve as the main respondent. Households

were interviewed after taking informed consent.
Sample

Table 4.1 and Figure 4.2 show the location of PHCs in Chhattisgarh. One notable feature is
that there is substantial geographical clustering of clinical care provider groups. For instance,

24

PHCs headed by AYUSH Medical Officers, RMAs and Paramedicals are clustered in a few

districts while Medical Officer headed PHCs tend to cluster in central districts of

Chhattisgarh. This geographical clustering of clinical care providers reflects both purposive
placement and self-selection. For instance, all the AYUSH Medical Officers are located
exclusively in the tribal areas of Chhattisgarh and RMAs have been placed only in a few
districts. The geographical clustering of Medical Officers is due to their self-selection to

remain in the central regions of Chhattisgarh which are more developed than other parts of
the state.

Around 30% of PHCs are headed by regular Medical Officers, 7% contractual Medical

Officers, 24% AYUSH Medical Officers, 9% RMAs, 8% Paramedicals and 12% by others.
With more than half the PHCs in Chhattisgarh without a Medical Officer, contractual or

otherwise, this distribution highlights the difficulty in recruiting and retaining physicians at

PHC.

The sampling frame includes 456 PHCs from which PHCs for the study sample were
selected. PHCs not included in the sampling frame were those where the main clinical care

provider was not of interest to the study (e.g. contractual allopathic doctors) or where no

clinical care provider was present (e.g. dressers, auxiliary midwifes and lower level health
workers) and PHCs that were too dangerous for field surveyors to visit due to the ongoing
insurgency in the state.
The target sample covered 19% of the PHCs headed by regularized allopathic doctors, 24%

of the PHCs headed by AYUSH doctors, 64% of the PHCs headed by RMAs and 76% of the
PHCs headed by Paramedical staff.

91% of the target sample size of 160 PHCs was

achieved. Across groups, at least 88% of the target sample size was achieved. The target

sample size was not completely achieved because some clinical care providers could not be

contacted. Further, three PHCs (2 AYUSH Medical Officer and I RMA) which were sampled
in the first phase could not be surveyed in the second phase.
Patients and households were sampled in the second phase of the study. In this phase three
PHCs covered in the first phase could not be reached because of poor roads or bad security.

At six PHCs no patients were available when the surveyors visited during clinic hours. These

25

factors led to only 68% of the target sample size of patients being achieved. For households,
89% of the household target sample size was achieved.

Table 4.1

Target and Achieved Sample Size

District

Bastar
Bilaspur
Dantewada
Dhamtari
Durg
Janjgir Champa
Jashpur
Kabirdham
Ranker
Korba
Koria
Mahasamund
Raigarh
Raipur
Rajnandgaon
Surguja
TOTAL

Number
ofPHCs

53
75
37
23
71
39
32
22
28
40
27
28
50
63

42
76
706

Sampling frame
PHC (target)
PHC (actual)

456
160

Patients (target)
Patients (actual)
Households (target)
Households (actual)

Regular
Medical
Officer
19(4)
32(6)
4(1)
7(3)
16(3)
12(1)
10(1)
4
7(D
18(4)
4(1)
8(2)
17(1)
24 (9)
10(3)
18
210

Contractual
Medical
Officer
0
8
0
5
28
6

4
4
5
5
0
19
2
19

8
10
123

AYUSH
Medical
Officer
34(9)
0
33(7)
0
0
0
18(6)
0
16(3)
0
21(7)
0
0
0
0
47(8)
169

RMA

Paramedical

0

0

0

0

8(8)
0
5(4)
0
16(10)
0
0
0
0
0
0
6(6)
14(10)
4(2)
0
53

27

0
0
0
0
0

2(1)

0
17(11)
1
0
25(15)
0
18(13)
0
63

205
40
35

135

63

40
37

40
35

53
40
39

1,600
1,082

400

400

269

296

400
273

400
244

2,400
2,124

600

600
503

600

600

571

525

146

525

Note: Figures in parenthesis indicate the target sample in the district

26

Others

0
6

27
5
0
12
0
0
1
1
0
6
2
1

88

Figure 4.2

Geographical Location of Sampled PHCs
N

A

't

y*
KORIYA*"

0*5* SARGUJA
j

*/

• A
• AA
M

W w ......
KORBA
*A AA .
i'^
1V
: BiLASPUR


/

44

M■

1 •

-G

Aa !;

▲ 1>
AA /

M '7

L

.....
A

O

'■

Q ''

A

0



/■■■

MAHASAM

-Q

DURG

JAMJGIR CHA!.1PA-g|

xo
i^-ipur

O

F->iGARH

"

a

©

*

_ A
A, .
.

• •


A

£A

/

JASHPUR *

6HAMTARi ’ • <

X

'<'z-

»t

.

KANKER*

<

#



BASTER

/

0

*

*

*3

*

.__r

*

Medical Officer

bir

DANTEWADA*

o

LEGEND
Ayush Medical Officer

i
27

A

RMA



PARAMEDICS

f

There were two significant deviations from the sampling plan. One PHC which had been

classified as belonging to the Paramedical group later turned out to have a regular Medical
Officer present. This PHC was classified under the regular Medical Officer group. Another

PHC which was originally classified as regular Medical Officer was later found to have a

contractual Medical Officer. This PHC was retained in the sample.

4.3 Data collection and processing

Field woi’k

Data collection for the study was conducted in two phases between July and September 2009.

In the first phase data on provider competence and job satisfaction was collected i.e. Form 1
and 3. Five regional centers were identified - namely Raipur, Bilaspur, Korba, Sarguja and
Bastar districts - the main clinical care provider in the selected PHC was asked to visit a

designated center at a specific date and time. Selected clinical care providers were registered
as they arrived and then assigned to one of the six interview teams. Consent was taken from

all participants. The interviewed clinical care providers were offered an honorarium,
according to government norms, for participating in the study.

The interviews were conducted in the privacy of a room. At least two study investigators
were present at each of the data collection centers to monitor and supervise the process. The
investigators observed interviews at random without interruption or interference. At the end

of the interview every completed form was checked by both supervisors to correct any errors.
Completed forms were brought back safely and stored at the SHRC office in Raipur.

Interviewers were mostly graduates with degrees in physiotherapy, pharmacy and social
work. All the interviewers were fluent in Chhattisgarhi (local dialect). Altogether 21

surveyors participated in the study. Training of interviewers for the first phase was conducted
over five days and included both classroom sessions and field training. The initial sessions on
the first day were spent orienting the participants to the study objectives, general methods of

collecting survey data, basic protocols to follow and the like. The field staff then, with help of
a training manual, explained the structure of the vignettes and how each section of the
28

questionnaire would be administered. The meaning of some of the technical terms which

recurred frequently in the six cases such as auscultation, skin pinch test, oedema, stridor were
explained to the trainees.

In the second phase of data collection, PHCs contacted in the first phase were visited to carry

out exit interviews and the household survey. Interviewers from the first phase were used in
this phase and underwent four days of additional training to administer these questionnaires.

The training involved a daily routine of in-class review of the questionnaires followed by
visits to nearby PHCs for field testing. Training manuals were provided to all the surveyors

during the training.

The twenty one interviewers were divided into seven three-member survey teams. Each team
had two surveyors and one field supervisor. The job of the field supervisor was to facilitate
the exit or household interview, check the completed forms for errors and coordinate the field
work. Each team was also closely observed by the study investigators during the course of

data collection. At the end of each survey day, the study investigators and the interviewers
together went over each questionnaire to identify and rectify errors.

Data processing
Completed survey forms were stored in Raipur in a secure place. The forms were entered into

a database created using CSPro44. All forms were double entered by independent data entry
operators. No personal identifiers were entered into the database. Five different databases

corresponding to each of the five survey questionnaires were created. These datasets were

converted to the STATA 1045 format for data analysis.

29

Chapter 5

Sample Characteristics

This chapter describes the main characteristics of the sampled PHCs, the clinical care

providers at these PHCs, their patients and the households in the village where the PHC was

located. Further, characteristics of the sampled PHCs and the villages where they are located
are also described.

5.1 PHC characteristics
This section examines the characteristics of the sampled PHCs. Described below are the
characteristics of the villages where the sampled PHCs were located, the condition of these

PHCs in terms of their physical structure and condition, service availability and the mix of

cases treated.

Location

About 40% of the sampled PHCs were located in areas exclusive to tribal populations, 19%

in non-tribal areas and 41% in exclusively non-tribal areas (Table 5.1). Most (74%) of the
PHCs headed by AYUSH Medical Officers were in exclusively tribal areas, by far the highest

in any group. This is due to the purposive placement of AYUSH Medical Officers in the
tribal areas of Chhattisgarh. In the remaining groups, the distribution of PHCs between
tribal/non-tribal areas is generally the same.

All the villages where the sampled PHCs were located had a primary school. Secondary
schools were also present in most of these villages. The majority of villages also had a high

school. However, this was most frequent in the paramedic group, followed by Medical
Officers, RMA and AYUSH Medical Officers.

The mean distance from the nearest town was around 23 Kms. PHCs headed by RMAs were

closest (19 Km) to the nearest town, followed by Medical officers, AYUSH Medical Officers
and Paramedical staff groups. In terms of the other infrastructure indicators like regular

30

electricity, piped water, daily bus service, villages in the Medical Officer and Paramedical
group appear to be the most developed, followed by RMAs and AYUSH Medical Officer.

Table 5.1

Characteristics of Village Where the PHC is Located

Medical
Officer

RMA

Paramedical
Staff

All

33

26

40

Tribal population (%)

26

AYUSH
Medical
Officer
74

Non-tribal population (%)

29

0

10

37

19

Mixed population (%)

45

26

57

37

41

100

100

100

100

100

Primary school (%)

100

100

100

100

100

High school/ Sr. Secondary school
(%)
Regular electricity (8 hrs/day) (%)
Piped water (%)

77

66

67

80

72

89
51

60
20

74
36

89
54

78
40

Daily bus service (%)

86

77

80

83

81

Mobile connectivity (%)

97

83

92

91

91

21.2
(15.63)
1.6
(4.24)
35

27.1
(20.52)
2.1
(5.32)
35

19.0
(12.21)
1.7
(4.55)
39

27.4
(41.24)
13.4
(67.41)
35

23.6
(24.89)
4.5
(33.46)
144

Distance from nearest town (Km)
Distance from nearest proper road
(Km)
N

Note: Figures in parenthesis is SD

Infrastructure
Overall, the sampled PHCs have poor infrastructure; less than half of them have basic
amenities like 24 hours electricity and water supply. Less than one in ten had a generator
present. In terms of physical infrastructure, less than half had toilets for patients and only one

third had living quarters for the resident clinical care provider. Significantly, only 21% had a

separate laboratory. One reason for the poor infrastructure at these PHCs is that many of
them were sub-centers which have, on paper, been designated as PHCs but their physical

infrastructure is yet to be upgraded. In general, PHCs in the Medical Officers group had the
best infrastructure, followed by RMAs, AYUSH Medical Officers and Paramedical staff
(Table 5.2).

31

\

10

Table 5.2

PHC Infrastructure

Medical
Officer

RMA

Paramedical
Staff

All

44

50

44

24 hours electricity supply (%)

57

AYUSH
Medical
Officer
26

24 hours water supply (%)

60

29

42

35

41

Generator present (%)

20

9

3

0

8

Toilet for patients (%)

63

43

59

47

53

Living quarters for doctors (%)

46

26

39

21

33

Delivery room (%)

63

56

66

41

57

Separate examination room (%)

'll

71

67
75

56
62

64

Separate drug storage room (%)

57
69

Separate laboratory (%)

20
6

26
6

15
9

21

Separate cold chain room (%)

23
23

Cold chain for vaccine storage

83

69

100

73

80

35

35

39

35

144

69

11

(%)

N

Note: Drug stock outs in the past one year calculated as mean over three months (April, May, June)

Table 5.3

PHC Functionality

Medical
Officer
Number of drug stock outs
annually
Medicines present (%)
Functioning laboratory (%)

N

RMA

Paramedical
Staff

All

4
(3.04)
94
32

AYUSH
Medical
Officer
3.7
(2.78)
94
26

2.6
(1.00)
94
28

3.8
(2.62)
94

18

3.4
(2.38)
94
26

35

35

39

35

144

Note: Figures in parenthesis is SD

PHCs in the sample experienced stock outs of one or more drugs for less than two months a

year, on average (Table 5.3). This does not differ much across clinical care provider type
groups. However, the overwhelming majority of the PHCs had drugs present in storage when
the survey team visited and this is found to be consistently high across groups. The presence

of a working laboratory (functional microscope and reagents) presented a rather troubling
32

situation - most (74%) of the PHCs sampled did not have one. Across groups, PHCs in the
Medical Officer group had the highest proportion (32%) of functional laboratories and those

in the Paramedical group had the lowest (18%).

Overall, the majority of PHCs did not need repairs of their interior walls, floor and outside
walls (Table 5.4). Among these, the condition of the outside walls needed the most attention.

Further, the majority of PHCs were rated having clean doctor consultation rooms. However,

only a small proportion of the PHCs had clean surroundings. On the whole, in terms of the

condition of the PHC building and cleanliness, PHCs in the Medical Officers group
performed the best and those in the Paramedical staff group, performed the worst.

Table 5.4

Condition of the PHC Buildings
RMA
AYUSH
Medical
Officer
Major repairs needed

Medical
Officer

Paramedical
Staff

All

Interior wall (%)

34

47

25

50

39

Floor (%)
Outside wall (%)

20
43

29
56

31
42

27
56

27
49

Cleanliness
Doctor consultation room (%)

74

66

62

45

62

Area surrounding PHC (%)

40

43

39

27

37

N

35

35

39

35

144

Patient case mix
Table 5.5 summarizes the distribution of the main complaint recorded in the patient register
for the month of June 2009 at the sampled PHCs. Cough, fever, diarrhea and stomach pain

were the most commonly presented complaints. TB and animal/snake/scorpion bite cases
formed a small proportion of the presenting complaints. The distribution of complaints was

similar across groups suggesting that PHCs headed by all four provider types received

similar cases. Further, the proportion of both antenatal care cases and the number of
deliveries was highest for Medical Officers with an average of two deliveries across

providers.

33

Table 5.5

Main Complaints of Patients at PHCs
Medical
Officer

AYUSH
Medical
Officer
100

RMA

Paramedical
Staff

All

100

100

100

97

100
97

Cough (%)

100
100

Injury/accident (%)

88

94

81

82

86

Stomach pain (%)

88

100

92

82

91

Diarrhoea (%)

94

100

97

97

97

TB (%)

39

20

11

12

20

Animal/snake/scorpion bite (%)

18

14

11

15

15

Antenatal care (%)

97

66

76

62

75

Deliveries conducted mean (SD)

4.7
(8.9)
35

1.8
(2.61)

2.2
(4.24)

35

39

1.1
(2.0)
35

2.4
(5.29)
144

Fever (%)

N

99

Note: Figures in parenthesis is SD

5.2 Characteristics of sampled clinical care providers

Among the sampled clinical care providers - Medical Officers, AYUSH Medical Officers,

RM As and Paramedicals -

less than one fifth of the providers are female (Table 5.6).

However, among RMAs slightly more than a third of the providers are female. The average

age of these clinical care providers is 34 years and is lowest for RMAs (26 years) and highest
for Medical Officers (42 years). The young age of RMAs reflect their recent induction to the

public system. The majority were married. However, lower proportion of RMAs (21%) and

Paramedicals (57%) were married. Among the married, 62% were living with their spouse.
Further, while 72% of Medical Officers and 84% of Paramedicals were living with spouse,

only 47% of AYUSH doctors and 38% of RMAs were doing the same. Medical Officers and

Paramedicals had worked the longest at the PHC level and at the sampled PHCs. Expectedly,

the RMAs had worked the least since they had recently been inducted into service.

34

Table 5.6

Provider Characteristics

Male (%)
Age

Married (%)
Married and staying with
spouse (%)
Living where PHC is
located8 (%)
Months working at PHC
level
Months working at
current PHC
Regular employment (%)

N

Medical
Officer
83
42.31
(7.55)
91
72

AYUSH
Medical Officer
94
35.08
(5.15)
92
47

RMA

Paramedical

All

64
25.17
(1.76)
21
38

83
33.2
(11.28)

57
84

81
34
(9.33)
64
62

60

78

72

69

70

148.4
(88.99)
90.3
(68.85)
97

41
(21.34)
35.78
(21.5)
0

10.78
(1.74)
10.39
(2.37)
0

110.88
(130.47)
67.26
(88.01)
100

77.96
(95.46)
50.90
(63.82)
47

35

37

39

35

146

Medical
Officer
(%)

AYUSH
Medical
Officer

Note: Standard Deviations are shown in parentheses

Table 5.7

Educational Qualifications of Providers

Educational qualification

RMA

Paramedical

(%)

(%)

(%)

MBBS
Post graduate/diploma/DMV
BAMS
Diploma in Modern and Holistic Medicine
Diploma in General Nursing & Midwifery
Bachelors of Phannacy
Diploma in Pharmacy
Other
Total

83
17
0
0
0
0
0
0
100

0
0
100
0
0
0
0
0
100

0
0
0
100
0
0
0

0
100

0
0
0
0
6
6
77
11
100

N

35

37

39

35

Medical Officers had either a MBBS (83%) or a post graduate degree (17%) in allopathic

medicine (Table 5.7). All AYUSH Medical Officers in the sample had a BAMS degree and

8 This refers to whether the providers were living in the ‘village’ where the PHC was located

35

all RMAs had a Diploma in Holistic and Modem Medicine. Paramedicals had a diversity of

degrees, the most common being the diploma in pharmacy.

5.3 Patient Characteristics

More than half the sampled patients were male with an average age was 27 years (Table 5.8).

Patients visiting AYUSH doctors were younger (22 years) and those visiting RMAs were
older (30 years). Nearly a third of the patients were illiterate. The percentage of illiterate
patients (34%) visiting Paramedicals was the highest. A majority of patients owned a cycle

and cattle. Only one percent of the patients owned a car. More than half of the patients
visiting Medical Officers, RMAs and Paramedicals also owned a fan.

Table 5.8

Patient Characteristics

Male (%)
Age (years)
Not literate (%)
Own fridge (%)
Own car (%)
Own fan (%)
Own radio (%)
Own TV (%)
Own cycle (%)
Own motorcycle (%)
Own cattle (%)
Own phone (%)

N

Medical
Officer
57
27
29
7
1
55
17
46
77
17
52
36

AYUSH
Medical Officer
62
22
26
3
0
20
13

269

Paramedical
57
28
34
2
2
55
16
42
76

23

RMA
62
30
26
6
0
53
22
36
84
18
64
36

57
20

All
60
27
29
5
1
45
17
35
80
15
64
31

296

273

244

1081

18

82
13
81

14

5.4 Household Characteristics

Table 5.9 describes the characteristics of the sampled households and their members living in

the same village as the PHC selected for the study. Members in the sampled households were
young (average age 27 years), about half were male and the majority were able to read.
Household size was large with 5.6 members on average per household. These characteristics
36

were similar across the clinical care provider groups, the only exception being the AYUSH
Medical Officer group, where a lower proportion reported being able to read.

Table 5.9

Household Characteristics

Male (%)
Age (years)

Able to read (%)
Household size
Electricity for lighting
Gas for cooking
Own fridge (%)
Own car (%)
Own sewing machine
Own fan (%)
Own radio (%)
Own TV (%)
Own cycle (%)
Own motorcycle (%)
Own cattle (%)
Own phone (%)

Medical
Officer
51
27
(18.73)
67
5.8
(2.63)
92
6
7
1

19
68
17
57
80
23
56
45

AYUSH
Medical
Officer
50
27
(18.31)
57
5.4
(2.27)
72
3
3
1
8
28
15
29
77
17
78
25

RMA
50
28
(19.17)
68
5.7
(2.48)

Paramedical
52
27
(18.87)
65
5.8
(2.33)

91

91
4

5
6
2
15
55
17
49
86
20
70
41

8
2
16
61

21
51
76
25
64
40

All
51
27
(18.79)
64
5.6
(2.44)
87
5
6
1

15
54
18
47
80
21
67
38

The distribution of household assets shows that the majority of households had electricity;

very few used gas for cooking or owned a fridge, car, sewing machine, radio and
motorcycles. However, the ownership of electric fans, TV, cycles, cattle and phones was

more frequent. The household asset distribution is also similar for the Medical Officer, RMA

and Paramedical group. However, the AYUSH Medical Officer group generally has lower
levels of asset ownership, particularly for more valuable assets like TV, fridge, sewing
machine, electric fan and phones. This and the lower literacy in the AYUSH Medical Officer
group suggest that the economic condition of households in this group was lower than

households in the other clinical care provider groups. This fits well with the fact that most of
the AYUSH Medical Officers were in tribal areas.

37

Chapter 6

The Quality of Clinical Care at Primary Health Centers
Assessments of health care quality can focus on the structural, process or outcomes of care.46

Linking health outcomes to health programs can be technically demanding, particularly, in
determining how much or which component of a program is responsible for the observed

changes in outcomes. The effort expended in collecting such information can be substantial.
Further, since many health programs (e.g. management of childhood illnesses) have proven

efficacy, quality assessments have typically focused on the structural and process aspects of
care. Structural quality assessments involve audits of health facilities to record the physical
presence or condition of equipment, drugs, supplies, building, staff and staff training. While

structural quality metrics inform on the ability of health facilities to produce health services,

they provide no information on the quality of health workers and the care they produce. How
much health workers know, what they do and how well they do it are in many ways of more

immediate relevance to patient welfare. For instance, a patient faced with a health facility
without drugs or diagnostic services can usually purchase them from the market. The

profusion of pharmacies and diagnostic service providers that are a common sight outside
government run health facilities in India are an example of private initiative filling in the
structural deficiencies in public health facilities. However, a patient may have little choice of

recourse when faced with a poor quality health provider because of the patient’s inability to
fully evaluate the technical aspects of care or simply because there is no other competent
health provider available.

Measures of process quality assess the degree to which providers perform health care

processes designed to achieve positive health outcomes. These processes are important
because following them is believed to influence health outcomes positively. And the validity
of process measures is determined by the degree to which they capture processes. Several

methods to measure process quality exist. These include abstraction from patient records,
measuring patient perceptions of quality, employing standardized patients, observing patient­

provider interactions and using clinical vignettes (this study). These methods provide
information on what clinical care providers do (e.g. observations of provider-patient
interactions, patient perceptions, standardized patients) or how much they know (e.g. clinical

38

vignettes). Choosing between these methods involves a trade-off between the degree of

comparability across provider types and the range of cases for which process quality can be

evaluated (Box 6.1).

Box 6.1

Measuring Process Quality

Abstraction from patient records is a relatively straightforward way for collecting information
on process quality. However, it has limited use in developing country contexts because of
poor record keeping. Further, patient records frequently do not carry the level of detail
necessary to make judgments on the technical aspects of care. Asking patients to rate
different aspects of the care received after an encounter with a clinical care provider is
another method to collect information on the technical aspects of care. However, it is subject
to the vagaries of patient recall and provides better information on the inter-personal aspects
of care (e.g. politeness of the provider) than on the technical quality of care received. Further,
while it allows collecting information on a range of cases, differing case-mix limits
comparability across providers and settings. Direct observation of patient-provider interaction
is one of the standard techniques employed for assessing technical quality. However, it does
compromise patient privacy and in certain types of cases observers may not be welcome.
Further, because of variations in the case mix and patients, comparing providers can be a
questionable exercise. The use of standardized patients is considered as the gold standard for
measuring process quality.47 This method involves ‘fake patients’ who present themselves to
providers with specific complaints and conditions and then record the activities of the
consultation. Since the patient and the presenting case is standardized it allows for direct
comparison between providers and settings. However, the standardized patients can represent
only a limited number of conditions.

6.1 Clinical vignettes and process quality
Vignettes have been employed in a variety of fields for measuring process quality.1317 47 48
Several types of vignettes have been used in health research and all of them involve

presenting a standardized case to a health provider and then recording how he or she manages
the case. The form of clinical vignette used in this study involved two persons interviewing a

clinical care provider. One played the part of the patient and the second recorded responses.

The ‘patient’ presents a case (e.g. I am a mother who has brought her 8 month old male baby
having frequent loose stools to you) and the provider is asked to proceed as in a consultation

and sequentially ask history questions and conduct examinations and investigations.

Whenever a relevant question, examination or investigation is mentioned by the clinical care

39

provider, a standard response is given by the ‘patient’. For example, if the provider asked
liFor how many days has he had these loose stools?” the standard response is “For two

days. ” After the history, exam and diagnostic sections is completed the clinical care provider
is asked to state their diagnosis, treatment (i.e. write a prescription), advice for home care and
recommendations for referral and follow-up. The vignettes are structured in a manner that if

the provider asked the relevant history questions, examinations and investigations then there

is only one correct diagnosis that can be arrived at. Clinical care providers were asked to
respond to the vignettes as if they were in a fully functional PHC and the patient would

comply fully with their prescriptions and recommendations. Because clinical vignettes
involve the use of standardized cases they offer the important advantage of allowing cases to
be compared across providers and settings.

The clinical care provider’s responses to the items contained in the vignette - history,
examinations, diagnostics, diagnosis, prescription and home recommendations - are judged

with reference to protocols of best practices according to standard treatment guidelines or
expert opinion. How well a clinical care provider is able to cover these items reflects the
degree to which his or her responses correspond to best practices. Therefore performance on
the vignettes reflects provider knowledge of best practices i.e. his or her competence in

managing the case.

The form of clinical vignette used in this study attempts to bring a degree of realism to the
assessment with one of the interviewers interacting with the provider as a patient might do in

an actual consultation. However, the clinical care provider knows that the patient is

imaginary and has to rely on questions and answers for all aspects of the consultation
including physical examinations and diagnostics. Performance on the vignettes is different
from what providers do in practice. What they do in practice is influenced by how much they
know (i.e. competence), the structural constraints and incentive environment within which

they operate, and their expectations of patient compliance. For instance, providers who work
in settings where x-ray machines don’t work or certain drugs and key staff are not available

will suitably alter their case management. Provider behavior can also be influenced by
incentives that make them prescribe unnecessary drugs or investigations. In such situations

what provider’s do in practice can deviate from what they know they should do (i.e.
competence) and from standard procedures.

40

A few studies have compared performance on vignettes with other measures of process

quality. One study from Tanzania found correlation between performance on vignettes and

provider practice (Leonard and Masatu 2005).48 Further, vignette performance was better
correlated with the best practices of physicians than what they typically (average) did in

practice, leading the authors to conclude that vignettes are useful for measuring ability but

not practice. Another study set in the United States examined physician outpatient
performance on vignettes with standardized patients and chart abstraction.47 Performance on

standardized patients was consistently higher than on the vignettes suggesting that practice

was better than competence. The mean performance scores of vignettes were closer to that of
standardized patients (the gold standard) than to chart abstraction leading the authors to
conclude that it is a valid measure of practice quality. A study set in Delhi, India found that
the difference between performance on vignettes and practice varied by provider type and the
sector (public or private).13 Private sector medical doctors did more of what they knew

compared to public sector doctors. Further, non-qualified doctors did most of what they
knew. The authors attribute these differences to the incentives clinical care providers face.

These assessments, though few, suggest that performance on vignettes is associated with

provider practice, and the distance between them depends on provider effort.
Developing the vignettes

The clinical vignette used in this study involved presenting six standardized cases to the main
clinical care provider at PHCs. The six cases were malaria in an adult woman, diarrhea with
severe dehydration in an infant boy, pneumonia in a young girl, pregnancy induced
hypertension, adult male with TB and an adult male with type II diabetes. Several factors

influenced the selection of cases. First, these cases were selected based on the disease profile
of Chhattisgarh and conditions typically seen at PHCs. Results from a study on illness
patterns conducted by the Chhattisgarh State Health Resource Centre (SHRC) suggested that
malaria among adults, ARI and diarrhea among children to be the most common illnesses

suffered by these age-groups respectively. In addition, Chhattisgarh also has a high incidence
of tuberculosis. The high maternal mortality in the state and India made it important to

include the case of pregnancy induced hypertension and the growing burden of noncommunicable diseases nationally made the diabetes case necessary. Second, the cases were
selected so that a clinical care provider in a fully functional PHC can diagnose and treat these
cases, except for pregnancy induced hypertension, where the correct practice is to refer to a

41

higher level facility. Third, the cases were selected to gauge the ability of clinical care

providers at PHCs to treat a range of patients (adults, children, and women) and conditions
(communicable diseases, chronic disease and reproductive health).

The specific questions, examinations, investigations and home care recommendations for

each case was developed in a multi-step process. In the first step, standard treatment
guidelines prescribed both by the state and national health agencies were reviewed for

formulating an initial set of items to be asked for patient history, examinations and

investigations, prevention advice, danger signs for worsening conditions and duration for
follow-up. In the second stage, after an initial draft of the vignettes was prepared on the basis
of standard treatment guidelines, each case was shown to an expert in the field working at the

All India Institute of Medical Sciences in New Delhi. Concurrently, five Medical Officers
working at PHCs in Chhattisgarh were also consulted on the types of questions,
examinations, investigations and home recommendations they give in each case. Based on
these inputs, the questionnaire was modified and adapted to fit a clinical setting. This

adaptation involved omitting unnecessary questions or adding questions and responses that
commonly occurred in clinical practice. In the third stage, the questionnaire was tested on

Medical Officers, AYUSH physicians, RMAs and pharmacists in PHCs .located in

Chhattisgarh. This was an important step because it ensured that the instrument was flexible
for different types of clinical care providers. As a general rule, if the ‘patient’ was asked

about symptoms not specified in the vignette the standard reply was “No” and if questions
were asked about the pattern, duration and type of illness not already specified in the
vignette, the standard reply would be “/ cannot remember. ”

An important consideration in designing the vignettes was their applicability to providers
from different traditions, principally AYUSH doctors. The AYUSH doctors serving at PHCs

in Chhattisgarh are all trained in ayurveda and had completed a five and a half years course in
Bachelors of Ayurvedic Medicine and Surgery (BAMS) (see Box 1.3). However, their
medical coursework includes training in allopathic medicine. In addition, many had also

received in-service training in allopathic medicine. Finally, all the AYUSH doctors in the
sample reported practicing allopathic medicine, either solely (32%) or in combination with
ayurvedic medicine (68%). This allowed for the vignettes tool to be applicable to AYUSH

doctors. In addition, during the pre-testing phase, specific attention was given to how the

42

vignettes performed when applied to AYUSH doctors and this experience suggested that the

existing set of questions and their standard responses were adequate.

6.2 Analytical methods
In

each

case,

history

questions,

examinations

and

diagnostic

tests

and

home

recommendations of critical importance were scored. Items of clinical importance were

identified by three independent experts in the areas of internal medicine, gynecology and
pediatrics from Raipur Medical College, a district hospital and among private doctors in

Chhattisgarh. History, examination, diagnostic and home care items for each case in the
vignette were rated by the experts as ‘must be asked’ or ‘should be asked’ or ‘not important’.
Items that at least one expert felt ‘must be asked’ and the other ‘should be asked’ were

scored.

Each vignette consisted of five broad sections - history, examinations, diagnostics, diagnosis,

prescription and home recommendations. Items within the history, examinations and

diagnostic section received a score of one if they were asked or mentioned by the provider
and experts had identified it as being of clinical relevance. Correct responses to items in the

diagnosis, prescription and home recommendation sections, as identified by experts, received
a score of one. Items were scored in a manner which gave each of these sections a weight of

one, which meant that individual items in a section would have the same weight but items in

different sections would have differing weights. Section scores were summed across cases
and scaled so that the maximum attainable score was one hundred. Scores were standardized

and expressed as standard deviations from the overall sample mean.

The prescriptions written by the providers were first classified into three groups Recommended (prescription strictly follows the standard treatment guidelines); Minor errors

(prescription deviates from standard treatment guidelines); Major error (significantly reduces
the effectiveness of the prescription and/or increases the risk of harm and/or does not cure the

disease). Prescriptions in the middle group were given to a panel of three experts in the areas
of internal medicine, gynecology and pediatrics who classified this set of prescriptions as

being ‘effective’ and Tess effective’ or ‘ineffective’ based on their judgment of the
prescriptions. These five groups were re-classified into two groups by combining
prescriptions that matched standard treatment guidelines with those marked as ‘effective’ or

43

‘less effective’ and the remaining prescriptions were classified as being ‘ineffective’. No
prescriptions were evaluated for the TB case because, national guidelines require that

suspected TB patients be referred to a Designated Microscopy Center for sputum testing and
if positive then they are enrolled for DOTS treatment at the nearest DOTS center and given a

kit containing all the medicines they need to take. A second issue was referrals. For the
preeclampsia and particularly the diabetes case, several health providers, especially among

AYUSH and paramedics, indicated that they would directly refer the case instead of
prescribing medicines. For cases of preeclampsia, the recommended course of action for a

PHC clinical care provider is to refer the case; therefore these prescriptions were not
evaluated. For diabetes, a patient going to a fully staffed and properly equipped PHC should

expect to receive treatment; consequently referrals were categorized as ineffective for the
prescription analysis.

6.3 Results

Figure 6.1 presents box plots of the clinical care provider competence scores (range 0 to 100)
for each of the six cases and overall. Clinical care providers performed their best in the
malaria (i.e. highest median competence9) and TB case and their worst in the pneumonia and
preeclampsia case. The competence scores also had a large range indicating that provider

ability varies substantially within cases. In every case, Medical Officers or RMAs had the

highest median scores followed by AYUSH Medical Officers and Paramedicals. Median
scores of Medical Officers and RMAs were similar in all cases. Paramedics had the lowest

median scores in all cases. Further, AYUSH Medical Officers did particularly poorly in the
preeclampsia case where their median scores are substantially lower than Medical Officers

and RMAs and marginally better than Paramedicals. The combined scores from all six cases
indicate that Medical Officers and RMAs have the highest overall median competence.

However, with their median score being around 50, even the best performers had substantial

room for improvement. AYUSH Medical Officers had substantially lower overall median
scores compared to these two providers and Paramedics had the lowest median score.

9 In the box plot, the median is represented by the horizontal line which divides the box.

44

Figure 6.1

Provider Competence Score by Case
o
o

] Medical Officer |

o
co
CD
O
O
O)
CD
O

IT 1

Q

Q
CL

T

8

• J_
r
T

T

| Paramedical

n-r

o
co

c

AYUSHRMA
Medical Officer |

r

ir a

-

L

1

JuL

E
O

O

e

I.J

e
e
-re

..

o
CM

J

o

Malaria

Diarrhea

TB

Pneumonia

Preeclampsia Diabetes

All

All providers tended to do poorly in history taking and physical examinations - no clinical

care provider was able to correctly respond to more than 44% of the relevant history or

examination items (Table 6.1). However, clinical care providers correctly identified the
majority of relevant diagnostic tests and case diagnosis. One reason for this disjoint between

performance on history/examinations and investigations/diagnosis is that health providers

typically arrive at a good idea about the underlying condition after asking only a few history
questions and doing a few examinations. Further, the high proportion of correct

investigations and diagnosis suggests the greater reliance of providers on diagnostic tests for

arriving at diagnosis. Clinical care provider performance on prescription practices was
moderate with 61% writing non-ineffective prescriptions. Similarly, no more than 37% of the

essential home care advice items were identified by the best performing group.

45

Table 6.1

Proportion of Correct Responses and Overall Competence

History
Examinations
Investigation
Diagnosis
Prescriptions
(effective)10
Home care

Competence score
(range 0 to 100)
Competence score
(standardized and
adjusted)
Number of providers

Medical
Officer
44% (2170)
42% (1365)
81% (105)
86% (210)

AYUSH
Medical
Officer
32%*(2294)
29%*(1443)
75% (111)
66%* (222)

RMA
43% (2418)
40% (1521)
80% (117)
86% (234)

Paramedical
25%*(2170)
21%*(1365)
61%*(105)
54%*(210)

61% (140)
37% (1155)

51% (148)
31%* (1221)

61% (156)
37% (1287)

33%*(140)
26%*(1155)

56.5 (8.42)

43.7 (8.36)

55.8 (9.10)

33.8 (9.43)

0.51 (0.47)
35

0.46* (0.20)
37

0.44 (0.22)

-1.24* (0.19)
35

39

Note: Figures are % (number of relevant items) or mean (SD); ** Significantly different from Medical Officer at
alpha=0.05

In every area of competence, Medical Officers and RMAs have the highest proportion of

correct responses, followed by AYUSH Medical Officers (Table 6.1). In every component,

Paramedicals do the worst. Overall competence scores indicate that Medical Officers and
RMAs have similar competence scores, followed by AYUSH Medical Officers and

Paramedicals. However, the best performers are not ‘best’ with average competence scores
peaking at around 60, leaving considerable room for improvement. ANOVA results indicated

that the different clinical care provider groups significantly differ from each other in
standardized average competence. Multiple comparisons of group differences (Bonferonni

corrected) indicated that there was no statistically significant difference between the mean
competence scores of Medical Officers and RMAs. AYUSH Medical Officers and

Paramedics had average competence scores significantly different from Medical Officers and
RMAs. Average competence of AYUSH Medical Officers was significantly different from

Paramedics.

10 Effective prescriptions include those classified as 'effective’ and ‘less effective’. The proportions reported
here slightly differ from those reported in Chapter 7 ‘Prescription Analysis’ because of the differences in the
cases included.

46

Observed differences in average competence between clinical care provider types could be

due to reasons other than their training and knowledge. The characteristics of the sampled

providers, the PHCs they work in and the location of the PHCs can also be responsible for the
observed differences when these characteristics systematically differ across provider groups.
In such circumstances estimates of provider competence will not be unbiased. Provider
experience is one such factor; those with more experience might manage cases differently
than those with lesser experience. In the study sample, RMAs had the least experience at

PHCs because they were recently inducted. PHCs which frequently experience drug
stockouts may induce providers to prescribe differently than those located in well stocked

PHCs. PHCs located in less remote areas might attract better qualified providers or be better
equipped or have better availability of drugs and supplies - all of which can influence

observed provider competence. In the study sample, AYUSH Medical Officers are located in
the tribal areas of Chhattisgarh while the Medical Officers, RMAs and Paramedicals are

located in non-tribal areas.

To control for factors which influence provider competence, standardized competence scores

were regressed on individual, PHC and location characteristics (Annex 1). The adjusted

estimates are presented in Table 6.1. The results indicate that there are no statistically
significant differences in the adjusted average standardized competence scores of Medical

Officers and RMAs. Further, the adjusted scores of AYUSH Medical Officers and
Paramedicals are significantly different from Medical Officers. Medical Officers performed
the best with adjusted standardized scores 0.51 standard deviations above the sample mean,
followed by RMAs (0.44). Adjusted competence scores for AYUSH Medical Officers was
0.46 and Paramedicals 1.24 standard deviations below the sample mean.

6.4 Discussion
This study measured the competence of different types of clinical care providers serving at
PHCs in Chhattisgarh to manage some common and important conditions relating to child,

adult and maternal health. Several aspects of provider competence was assessed - history

taking, examinations, diagnostics, diagnosis, treatment and home care. How well providers
responded to the clinically relevant items in these sections reflected their ability and

knowledge to manage these cases.

47

Findings from this study suggest that Medical Officers and RMAs are equally competent to

manage conditions commonly seen in primary care settings. AYUSH Medical Officers are
less competent than Medical Officers and RMAs. Paramedicals are the least competent. This

was observed for infectious, chronic and maternal health conditions and for a range of patient
types - infants, children and adult men and women. Further, these results hold even after
controlling for various individual, facility and location characteristics. This relative

performance is consistently found in all aspects of the outpatient consultation - history taking,

examinations, investigation, diagnosis, prescription and home recommendations. An

important implication of this is that, in terms of clinical competence for primary health care,
clinical care providers with short duration of training appear to be a viable alternative to
physicians.

Most states in India have been posting AYUSH doctors to fill vacancies of Medical Officers
at PHCs and to mainstream Indian systems of medicine. This study casts doubt on the

appropriateness of this practice when it results in the AYUSH doctor becoming the primary

clinical care provider in the PHC. The overall competence of AYUSH doctors and their
performance on different aspects of the consultation (expect for prescriptions) is below that
of Medical Officers and RMAs. Nearly half the treatment prescribed by AYUSH Medical

Officers was ineffective compared to 39% for Medical Officers. Without additional clinical
training, particularly in primary health care, AYUSH doctors do not seem to be the best

alternative to physicians at PHCs. However, they certainly are a better alternative to having

paramedical staff provide clinical care. Moreover, their competence with additional
allopathic training can be quite different and it is important to explore this potential.

The average competence of paramedics is both expected and disturbing. They do not receive
any formal training in clinical care nor are they meant to perform such activities. Yet,

because there is no physician, these paramedics continue to treat patients in numerous PHCs

across Chhattisgarh and in the rest of the country. Their activities as clinical providers are
clearly a danger to their patients and undermine trust in the public health system. However, in
other countries paramedical staff like nurses have been found to be as effective as fully

qualified physicians in providing primary health services. Paramedics in the study sample
are, however, dominated by pharmacists. Nevertheless, the potential of paramedical staff with

appropriate training to serve in PHCs is an important area of future research.

48

In several other countries, available evidence indicates that clinical care providers with
shorter duration training perform as well as those with longer training. A study on the
performance of different types of health workers trained in implementing the IMCI protocol

in Bangladesh, Brazil, Uganda and Tanzania found that there was generally no difference in
the quality of child care between long and short duration trained health workers/4 Another
study from Tanzania reported that for managing complicated deliveries and major obstetrical

operations there were no differences in outcomes, risk indicators, or quality between medical

officers (medical school graduates) and assistant medical officers.36 In as many as 47 Sub­

saharan countries non-physician health workers have been trained to diagnose and treatment
and have helped ameliorate the shortage of qualified health workers in rural areas and at

lower training and better rural retention than medical doctors.1

The findings of this study are subject to several limitations. Provider competence was
assessed only for six cases and as such the validity of the findings is limited to these cases.

While these cases represent those that are commonly seen at PHCs, the present findings need
not necessarily hold for other serious conditions such as accidents, snake bites, complicated
deliveries and more complicated conditions. However, these six cases cover a range of

patient types and illness conditions and the consistency seen in the results across these cases
makes it likely that the relative competence levels observed are robust. A second issue is that

providers were assessed only on their clinical competence. Health providers at PHCs are

responsible for a range of non-clinical duties such as preventive care and health education. It
is quite possible that these providers perform quite differently in these other activities. A third
issue has to do with the difference between competence and practice. The vignettes measured

competence but this could be, as several studies have indicated, quite different from what

these same health providers do in practice. Though competence and practice are associated, it
is not necessary that the relative competence levels observed here also hold in practice.

Clinical care providers at PHCs were better at managing certain cases. In general, everyone
did better in managing malaria and TB and tended to do poorly in diarrhea, pneumonia,

preeclampsia (AYUSH and Paramedicals) and diabetes. Both malaria and TB are the focus of

established national disease control programs and clinical care provider competence in
managing these two cases reflects the success these vertical programs have had at the

49

peripheral levels of the health system. The cases on which clinical care providers did not
perform well have a large disease burden. Both diarrhea and pneumonia are well known child

killers in India and account for a substantial portion of morbidity and deaths in children.49

The relatively poor competence of PHC clinical care providers in managing these diseases is
remarkable because it indicates how poorly equipped frontline health workers are in dealing

with this common and important problem. Another area of concern is the poor ability of
AYUSH Medical Officers and Paramedicals to manage the preeclampsia case. With these

two types of clinical care providers commonly being the sole clinical provider in PHCs,
efforts to improve maternal health and deal with pregnancy related complications need to pay

attention to appropriately skilling these providers.

50

Chapter 7
Prescription Analysis

Two sets of prescriptions were analyzed in this study. The first set is from the clinical
vignettes used to assess clinical care provider competence in managing standardized cases.

The second set is from the prescriptions of the patients of these clinical care providers. The
analysis presented here examines the quality of these prescriptions, their effectiveness in
treating the diagnosed condition and the rational use of drugs.

7.1 Prescription analysis for clinical vignettes

Clinical vignettes were used to assess the technical competence of clinical care providers to
manage the following six cases - malaria in an adult female, diarrhoea in an infant,
pneumonia in a child, preeclampsia in a pregnant woman, TB and diabetes in an adult male.

Each clinical vignette was divided into six sections - history questions, examinations,

investigations, diagnosis, treatment (prescription or referral) and home care (including follow
up). For the treatment section, clinical care providers were asked to write a prescription based

on their diagnosis of the case with the understanding that all the medications prescribed are

available and the patient will fully comply with the prescription.

7.1.1 Methods

The sampled clinical care providers were asked to write prescriptions for malaria, diarrhea,

pneumonia, preeclampsia and diabetes. TB prescriptions were not analyzed because the
standard procedure is for suspected TB patients to be referred to a TB unit for testing and

DOTS treatment. Preeclampsia cases are also to be referred, however, we include these
prescriptions in the analysis. Prescriptions where clinical care providers indicated that they

would refer the case without any treatment were not included in the analysis. The following
steps were followed in evaluating the prescriptions.

51

J

\00

)0

Stcmdard treatment guidelines

A comprehensive guideline was prepared defining standard treatments for each of the cases.
These guidelines were prepared after an exhaustive review of the relevant standard treatment
guidelines being used in India (see Box 7.1). Based on these standard treatment guidelines

prescriptions were classified as: Recommended (prescription strictly follows the standard
treatment guidelines); Minor errors (prescription deviates from standard treatment
guidelines); Major errors (significantly reduces the effectiveness of the prescription and/or

increases the risk of harm and/or does not cure the disease).

Panel of experts

Prescriptions in the ‘Minor errors’ group, which were the majority, deviated from the

standard treatment guidelines but could either be beneficial or not to the patient. To further
examine this, three panels of experts was formed. The panels represented the specializations
of internal medicine (for the malaria and diabetes cases), gynecology (for the preeclampsia

case) and pediatrics (for diarrhea and pneumonia cases). Each panel comprised of three
experts, drawn from Raipur Medical College, a district hospital and one private practitioner.

These experts were asked to classify the ‘Minor error’ prescriptions into the following three
categories: Effective (prescriptions are effective in treating the condition of the patient); Less

effective (prescriptions have little effect on the condition of the patient); and Ineffective
(significantly reduces the effectiveness of the prescription and/or increase the risk of harm
and/or do not cure the disease). Experts were blinded to the identity and qualification of the

clinical care provider who wrote the prescription. Prescriptions were rated independently. If
all three experts did not agree on the classification, the particular prescription was taken back
to the panel till a consensus was reached.

Finally, prescriptions were classified into three groups by combining the results from the

standard treatment guidelines exercise and the ratings of the expert panels. Prescriptions were
classified as ‘Effective’ if they were rated as ‘Effective’ by experts or were in the
‘Recommended’ group (i.e. followed standard treatment guidelines). Prescriptions termed
‘Less effective’ were classified as ‘Less effective’ by the expert panel. Prescriptions

52

categorized as ‘Ineffective’ were classified as ‘ineffective’ by the expert panel or were judged

to be in the ‘Major error’ group according to standard treatment guidelines.

Box 7.1

1.
2.

3.

4.

5.
6.

Standard Treatment Guidelines

Department of Health & Family Welfare, Govt, of Chhattisgarh. Standard
treatment guidelines for Medical officers, 2003
Armed Forces Medical College, Panel In Collaboration with Ministry of Health &
Family Welfare, Government of India &WH0 Country Office, India. Standard
treatment Guidelines Medical Management & Costing of Select Conditions, 2007
National Institute of Malaria Research and National Vector Borne Disease
Control Programme, Government of India. Guidelines for Diagnosis and
treatment of Malaria in India, 2009
Ministry of Health & Family Welfare, Govt, of India & World Health
Organization Child & Adolescent Health and Development (CAH). Integrated
Management of Neonatal and Childhood Illness: Physician Chart Booklet, 2003
Tripathi KD. Essentials of Medical Pharmacology, 6lh Edition, Jaypee
publications, 2009
Maternal Health Division, Department of Family Welfare, Ministry of Health &
Family Welfare, Government of India. Guidelines for Pregnancy Care and
Management of Common Obstetric Complications by Medical Officers, 2005

7.1.2 Results

Case 1: Malaria in a 35 year old woman

Out of a total of 146 prescriptions, 143 were analyzed. Two prescriptions, one by a Medical

Officer and another in the Paramedical group were not included in the analysis due to lack of
consensus among the experts in how they should be classified. One case in the Paramedical

group was referred.

The largest proportion of ‘Effective’ prescriptions were written by RMAs (64%) followed by

AYUSH Medical Officers (57%), Medical Officers (44%) and Paramedicals (27%) (Figure
7.1). Almost half of the prescriptions written by paramedics were scored as ‘Ineffective.’ The

reason why Medical Officers had a lower percentage of ‘effective’ malaria prescriptions was
because clinical care providers in other groups frequently wrote ‘blister packs’ in their
prescriptions. Because blister packs come with standardized dosage, frequency and duration.
53

they were automatically classified as following standard treatment guidelines. For example,

out of 37 AYUSII Medical Officers. 10 prescribed blister packs for malaria.

Analysis

excluding the AYUSII Medical Officers' prescriptions with blister packs showed that RM As

still performed the best with 64% ‘Effective’ prescriptions followed by Medical Officer
(42%), AYUSH Medical Officer (41%) and Paramedicals (25%).

Rational drug use: Medical Officers prescribed the most number of recommended drugs per
prescription (1.76) followed by RMAs (1.51). AYUSH Medical Officers (1.19) and
Paramedicals (1.12) (Table 7.1). Prescriptions by Medical Officers had the lowest number of
antibiotics and injections.

Figure 7.1
100

Classification of Malaria Prescriptions


90 -

80 j

I

CO

CD
Q-

70 j
60 'j
50 i
40 -

27

44
57

27

10
0

□ Effective

24

□ Less effective

13

30 {
20 i

64

21.
32


15

30

46

□ Ineffective

i

Medical Officer• AYUSH Medical
Officer

RMA

Paramedicals

Case 2: Diarrhea with severe dehydration in an eight month old baby

145 out of 146 prescriptions were analyzed. One case was referred by Paramedicals. In
general, clinical care providers did not do well in writing effective prescriptions for diarrhea
(Figure 7.2). RMAs (10%) performed the best in terms of writing ‘effective’ prescriptions,
followed closely by Medical Officers (9%) and AYUSH Medical Officers (3%). No one in

the Paramedical group wrote an ‘Effective’ prescription for diarrhea. ‘Ineffective’
prescriptions were most frequently seen in the Paramedicals (74%) groups; followed by

AYUSH Medical Officers (68%), Medical Officers (48%) and RMAs (36%).

54

Rational drug use: Prescriptions by RMAs had the highest number of recommended drugs

(1.5) followed closely by Medical Officers (1.3) and AYUSH Medical Officers (1.2) (Table
7.1).

Number of antibiotics per prescription for treating diarrhea was highest for

Paramedicals and AYUSH Medical Officers (1.5).

Figure 7.2

Classification of Diarrhea Prescriptions

100 1
90

OJ

ro
a
aj
Q-

80
70
60
50
40
30 1
20 4

10
26

30

54

□ Effective
67

__

74

□ Less effective
□ Ineffective

48
36

10
0
Medical Officer AYUSH Medical
Officer

RMA

Paramedicals

Case 3: Pneumonia in a two year old child

All 146 cases of pneumonia were analyzed. Pneumonia prescriptions displayed a similar

pattern as diarrhea in terms of an overall poor performance in writing ‘effective prescriptions’

(Figure 7.3). Medical Officers (14%) performed considerably better than other providers in
writing ‘effective’ prescription followed by AYUSH Medical Officers (5%), RMAs (3%).

Paramedicals did not write a single ‘effective’ prescription. The majority of Paramedicals

wrote ‘ineffective’ prescriptions, followed by RMAs (46%), Medical Officers (41%) and

AYUSH Medical Officers (37%).

Rational drug use: Prescriptions by AYUSH Medical Officers had the highest number of

recommended drugs (1.29) followed closely by RMAs (1.2) and Medical Officers (1.11)
while those written by Paramedicals had the lowest (0.97) (Table 7.1). AYUSH Medical

Officers also had the highest number of antibiotics per prescription, while RMAs had the

highest average number of steroids and injections.
55

Classification of Pneumonia Prescriptions

Figure 7.3

100

40

70
<D
DO

ra

c
0J

o

a.

HH

14

90
80

51

54

49

60
50
40 j

□ Effective

□ Less effective

30

□ Ineffective

60

41

20

46

37

10
0

Medical Officer

AYUSH Medical
Officer

RMA

Paramedicals

Case 4: Preeclampsia in a six month pregnant woman
Out of a total of 146 prescriptions, 133 were analyzed as 13 cases were referred. RMAs

(27%) performed the best in terms of writing ’Effective' prescriptions, followed by Medical
Officers (24%), AYUSH Medical Officers (10%) (Figure 7.4).

Not a single 'Effective'

prescription was present in the Paramedicals group. The majority of the prescriptions written
by Paramedicals (94%) were ‘Ineffective’, followed by AYUSH Medical Officers (61%),
Medical Officers (59%) and RMAs (41%).

Classification of Preeclampsia Prescriptions

Figure 7.4
100 1
90
80
70 J
C
o

o
Q-

10

-6

24

27
29

17

60
50
40 j
30 1

32

94

□ Less effective
59

61

□ Ineffective

20

10
0

□ Effective

41

1
Medical Officer AYUSH Medical
Officer

RMA

56

Paramedicals

Rational drug use: Prescriptions by RMAs (0.65) and Medical Officers (0.47) had the highest

number of recommended drugs per prescription, while none of the prescriptions written by
Paramedicals had any recommended drug (Table 7.1). A significant number of prescriptions

had hazardous/contraindicated drugs with this being most frequent among Medical Officers
(24%). AYUSH Medical Officers prescribed steroids and injections most often (0.03 steroids

and 0.06 injections per prescription). Prescriptions written by RMAs had the highest number
of antibiotics (0.14).

Case 5: Diabetes-Type II in a 46 year old male
Out of a total of 146 prescriptions, only 97 were analyzed as the remaining 49 were referred

(Figure 7.5). Medical Officers had the highest proportion of 'Effective’ prescriptions (52%)
followed by RMAs (46%) and AYUSH Medical Officers (23%). Paramedicals (80%) had the

highest percentage of'Ineffective' prescriptions.

Rational drug use: Table 7.1 indicates that Paramedicals prescribed the highest number of

antibiotics per prescription (0.65) followed by AYUSH Medical Officers (0.27), RM As
(0.21) and Medical Officers (0.11). On the whole. Medical Officers wrote the most rational

prescriptions with the highest number of recommended drugs per prescription, least number
of antibiotics and no steroids and injections for treating diabetes cases.

Figure 7.5

Classification of Diabetes Prescriptions

100 ;

90 j
80 j
o 70

23
52

15

46

f 60

45
5 50
40
26
30
20
32
22
10 -I
0 ..
Medical OfficAYUSH Medical Officer

□ Effective

1

80

□ Ineffective

43

■i
RMA

57

□ Less effective

Paramedicals

All cases
Figure 7.6 summarizes the performance of the sampled clinical care providers in the five
cases. Overall, the percentage of ‘Effective' prescriptions was fairly low across all clinical
care providers. RMAs (30%) performed the best, followed closely by Medical Officers

(28%). AYUSFI Medical Officers (19%) and Paramedicals (7%). Paramedicals (70%) had the

most number of ‘Ineffective’ prescriptions, followed by AYUSH Medical Officers (47%),

RMAs (35%) and Medical Officers (40%).

Rational drug use: The total number of drugs per prescription ranged between 3 and 4 (Table
7.1). RMAs (1.15) and Medical Officers (1.12) most frequently prescribed recommended

drugs. Among clinical care provider groups. Medical Officers prescribed the least antibiotics
(0.65) and steroids (0.03) and had the second lowest frequency of prescribing injections

(0.48).

Figure 7.6
100
90
80 i
70 •
DD
60 j
ra
a
50 1
OJ
o.aj 40
30 :
20 10 ?
0

Overall Classifications of Prescriptions

28

___z---

19
--------

30

24



34

32

35



40

_____
47

□ Effective
69

□ Ineffective

35



Medical Officer

RMA

AYUSH Medical
Officer

58

0 Less effective

Paramedicals

Table 7.1 Rational Drug Use
Medical
Officer
Drugs
Recommended drugs
Antibiotics
Injections
Sample size (n)
Drugs
Recommended drugs
Antibiotics
Steroids
Injections
Sample size (n)
Drugs
Recommended drugs
Antibiotics
Steroids
Injections
Sample size (n)
Drugs
Recommended drugs
Antibiotics
Steroids
Injections
Hazardous/Contraindicated
drugs (%)
Sample size (n)

AYUSH
Medical Officer
Malaria
4.60
5.35
1.76
1.19
0.35
0.95
0.29
1.03
34
37
Diarrhea
4.54
4.16
1.31
1.27
1.37
1.51
0.06
0.03
1.19
1.37
35
37
Pneumonia
3.54
3.71
1.29
1.11
1.46
1.20
0.13
0.11
0.57
0.60
35
37
Preeclampsia
3.00
3.40
0.47
0.3
0.13
0.06
0.03
0
0.06
0.03
24.24
12.90
31

RMA

Paramedical

4.77
1.51
0.64
0.72
39

3.84
1.12
0.81
0.36
33

4.82
1.51
1.33
0.03
1.51
39

3.88
1.09
1.5
0
0.79
34

3.89
1.33
0.18
0.77
39

2.8
0.97
1.11
0.08
0.40
35

3.03
0.65
0.14
0
0.05
8.11

2.19
0
0
0
0.03
6.25

34

37

31

2.14
0.73
0.27
0
0.14
27

2.75
0.80
0.21
0.36
0.11
28

2.60
0.50
0.65
0.15
0.10
20

3.79
1.00
0.94
0.04
0.65

3.92
1.15
0.76
0.04
0.67

3.11
0.77
0.84
0.04
0.36

1.20

Diabetes

Drugs
Recommended drugs
Antibiotics
Steroids
Injections
Sample size (n)

1.90
0.90

Drugs
Recommended drugs
Antibiotics
Steroids
Injections_________

3.72
1.12
0.65
0.03
0.48

0.11
0
0
22

All cases

Note: All values are calculated as averages per prescription

59

7.2 Prescription analysis for provider practice

In the second phase of the study, clinical care providers sampled in the first phase were

visited at the PHCs where they work. Patients of these clinical care providers were
interviewed when they exited the PHC and information contained in their prescriptions was

recorded. At each sampled PHC, a convenience sample of ten new patients was selected for

exit interviews. Against the target of 1,460 patients, 1,082 patients were interviewed.

7.2.1 Methods

From the prescriptions of patients exiting the PHC, surveyors recorded if it mentioned the

patient’s age, sex, chief complaint, signs and symptoms, and findings of any physical

examination. Surveyors recorded details of the investigations, referral (and next
appointment), diagnosis and the name, dose, frequency and duration of the prescribed drugs
mentioned in the prescription. This information was used to estimate various indicators of

rational drug use.

7.2.2 Findings

Out of the total 1,082 patients interviewed, 954 patients received a prescription. There is little
difference between the clinical care provider groups in the proportion of prescriptions
providing information on the patient’s background-name, age, sex. Patient complaints were
most frequently mentioned in prescriptions written by RMAs (81%), followed by AYUSH

Medical Officer (73%), Medical Officers (70%) and Paramedicals (55%) (Table 7.2). Very
few prescriptions noted signs and symptoms; RMAs (13%) and Medical Officers (14%) had
about the same proportion, followed by AYUSH Medical Officers (2%). None of the
prescriptions written by Paramedicals mentioned signs/symptoms. Further, prescriptions that

mentioned the patient’s diagnosis were also fairly low across all clinical care provider
groups; Medical Officers (47%) did the best, followed by AYUSH Medical Officers (40%),

RMAs (35%) and Paramedicals (23%).

60

Table 7.2

Prescription Practices

AYUSH
Medical
Officer
296

RMA

Paramedical

275

242

99% (264)
71

100 (296)

73

89 (244)
81

62 (150)
55

Prescriptions with signs/symptoms (%)

14

2

13

0

Prescriptions with diagnosis (%)

47

40

36

23

Medical
Officer
Total number of patients interviewed

Patients with prescription N(%)
Prescriptions with complaint (%)

268

Rational drug use: The average number of antibiotics and injections per prescription was

similar across the different clinical care provider groups (Figure 7.6). AYUSH Medical

Officers had the highest average for both antibiotics and injections, followed by Medical

Officers, RMAs and Paramedicals.

Figure 7.6

Rational Drug Use

2.0 |
1.8
s !.6 1
o
K 1.4 4
p 1-2 I
ex
1.0 j
QJ
ex
0.8
OJ
0.6
<U
< 0.4
0.2
0.0

1.7

II ■
. - . ’

1.4

0.6

s

J

Medical Officer

1.4

III

Antibiotics

Injections

■i____ ,_ ■■MBg ,_ IM____ ,
RMA

AYUSH Medical
Officer

Paramedicals

7.3 Discussion

Prescriptions written as part of the standardized cases in the clinical vignettes allow

comparing prescription practices of different types of clinical care providers. Findings

suggest that Medical Officers and RMAs are equally effective in treating conditions
61

commonly seen in primary care settings. AYUSH Medical Officers are less so and
Paramedicals have the least ability to prescribe effectively. This was observed for infectious,
chronic and maternal health conditions and for a range of patient types - infants, children and
adult men and women. This implies that in terms of ability to treat conditions seen at primary

health care settings, physicians with short duration of training appear to be a viable
alternative to fully qualified physicians.

The findings also indicate that the majority of prescriptions were not effective in treating
some of the most common and important diseases in Chhattisgarh. Strikingly, in every
clinical care provider group there was a non-ignorable proportion of prescriptions that harm

the patient. Notably, almost half the prescriptions written by AYUSH Medical Officers and
the majority written by Paramedicals were ineffective. In terms of rational drug use, Medical
Officers wrote the most rational prescriptions followed by AYUSH Medical Officers, RM As

and Paramedicals.

Clinical care providers did better at prescribing effective treatments for the malaria and

diabetes cases. However, they did poorly for treating cases of diarrhea, pneumonia and
preeclampsia. Malaria is the focus of an established national disease control program and
clinical care provider ability to prescribe effectively reflects the success this vertical program

has had on frontline health workers. The cases where clinical care providers did not perform
well have a large disease burden; both diarrhea and pneumonia are well known child killers

in India and account for a substantial portion of morbidity and deaths in children. The poor
ability of PHC clinical care providers in treating these diseases is remarkable because it

indicates how poorly equipped frontline health workers are in dealing with this common and
important problem. Another area of concern is the poor ability of providers to prescribe

effective treatments for the preeclampsia case. Efforts to improve maternal health and deal

with pregnancy related complications needs to pay attention to appropriately skilling these
providers.

Analysis of the prescriptions collected from exiting patients indicates that, overall, the
sampled clinical care providers did well in mentioning information on the patient’s

background and complaint in their prescriptions. However, few prescriptions mentioned

signs/ symptoms and less than half mentioned diagnosis. In terms of rationality of drug use,

62

there was little difference between clinical care provider groups in the average number of

antibiotics and injections per prescription. AYUSH Medical Officers were most liberal in
prescribing antibiotics and injections and Paramedicals the most conservative. The greater

predisposition of AYUSH Medical Officers to prescribe antibiotics was also observed in the

prescriptions from the clinical vignettes.

The analysis presented here has several potential limitations. For one, the vignettes measure
the clinical care provider’s prescribing knowledge, but this can be quite different from what

they do in practice because of what they face in terms of drug availability and patient
compliance. Secondly, the panel of experts used for classifying the prescriptions, were mostly

serving at tertiary care centers in urban areas of Chhattisgarh. Their experience, the type of

patients they get, the constraints (or lack of) they face in terms of availability of drugs are
different from what clinical care providers at PHCs face. Consequently, what PHC clinical
care providers prescribe, given their specific contexts and constraints, might deviate from

what experts would expect. Another potential limitation was that the expert panel comprised

of allopathic doctors who lack knowledge of ayurvedic medications. However, AYUSH
physicians working at PHCs, mostly prescribe and practice allopathic medicine. This was

confirmed by the presence of only two prescriptions written by AYUSH Medical Officers
that mentioned ayurvedic medications (one for preeclampsia and one for diabetes) and none
of the prescriptions were exclusively ayurvedic.

The results presented here are revealing about the prescribing abilities of clinical care
providers at PHCs. The similar performance of Medical Officers and RMAs increases the

legitimacy of the latter to provide clinical services at PHCs. AYUSH Medical Officers don’t
perform as well as Medical Officers or RMAs, nevertheless, their performance can

potentially be much better with additional clinical training. The poor performance of
Paramedicals highlights the harmful treatment people can receive when there is no qualified

physician at a PHC. Indeed, clinical services delivered by Paramedicals, who are not properly

trained to perform this function, only serves to undermine the public’s trust in the primary
care system.

63

Chapter 9
Service Utilization, Equity and Community Perceptions

In the second phase of this study, households were sampled from the village where the PHC
was located. From these households information on their care seeking behavior, health

expenditures and perceptions of the local PHC were collected. In addition, the asset

ownership of these households was also recorded. This chapter describes findings from the
household survey. From the 146 PHCs sampled in the first phase of this study a total of

2,124 households were contacted and information on 11,929 individuals collected in the

second phase. The method used to select households is described in Chapter 4.
9.1 Care seeking behavior

Around one third of members in the sampled households had either fallen ill or suffered an

injury in the past 30 days (Table 9.1). The main complaints were fever, cough, cold and

diarrhea. The distribution of these complaints was also similar across the four clinical care
provider groups. Almost all who fell sick sought care outside their homes. Households were
asked where members sought treatment for each instance of illness in the past month for
which care was sought out side home. Visit information was collected for upto five visits

made for that particular episode of illness.11 Results indicate that treatment is mostly sought

from private providers, except in communities where the local PHC was headed by AYUSH
Medical Officers (Figure 9.1). PHCs headed by AYUSH Medical Officers received the

largest share (60%) of total visits, followed by RMAs (35%), Medical Officers (29%) and

paramedics (20%). Where there is no qualified clinical care provider, as in the Paramedical

case, the local PHC is hardly used for treatment. Indeed, the share of visits to private
providers is largest for this group. The pattern of visits to the local PHC suggest that, all else
being the same, there is at least as much public trust in AYUSH Medical Officers and RMAs
as there is in Medical Officers.

11 For example, an ill person could have visited a private doctor first, then the local PHC and then a private
hospital for treatment.

68

Burden of Illness and Care Seeking

Table 9.1

Medical
Officer

AYUSH
Medical
Officer

RMA

Paramedical

All

111 during past 30 days (%)

31

34

32

35

33

Care outside home (%)

92

93

92

93

92

N (individuals)

3,003

2,696

3,211

3,019

11,929

Fever

36

43

40

37

39

Cold

18

9

17

18

16

Cough

8

8

7

6

7

Diarrhea

6

7

7

6

7

Stomach pain

2

4

3

4

3

Injury

3

2

3

1

2

Other

26

26

24

27

26

100

100

100

100

100

936

921

1,046

1,067

3,970

Main complaint/illness
(%):

N (111 individuals)

Figure 9.1

Health Care Providers Visited for Treatment

Medical Officer

AYUSH Officer

RMA

Paramedical
20

60

40

80

100

Percent

Note PHC is local PHC and community health worker. Private is private clinician and hospital
Other is pharmacist/compounder, tranditional healer and other

69

Why people bypass the local PHC cannot be attributed only to the type of provider since
other factors could be responsible e.g. the availability of drugs and equipment, the regularity

with which staff is present, the characteristics of the area and the presence of alternative
providers. Sampled households were asked why they did not use the local PHC for treatment
(Table 9.2). Overall, the most common (i.e. the top five and excluding illness not severe,

other) reasons for not using the local PHC had to do with disliking the service at the PHC,
non-availability of the doctor/health provider, lack of drugs, and the presence of better
private providers in area. Further, the distribution of reasons for non-use are similar between

PHCs headed by Medical Officers and RMAs suggesting that there was nothing particular to
the clinical care provider, PHC infrastructure or location characteristics which differentiated
these two groups.

Table 9.2

Reasons for Not Using the Local PHC for Treatment

Illness/condition not severe enough

10

AYUSH
Medical
Officer
19

Illne-ss/condition serious

4

PHC too far/ no transport
Treatment too expensive
Doctor/ health provider not available

Reasons for not using PHC services (%)

Medical
Officer

RMA Paramedic Overall

10

8

11

13

6

4

6

2

4

5

4

1
8

0
7

5
0
10

1
28

1
15

Doctor/health provider not competent /
able to cure illness
Don’t like service

10

6

8

8

8

18

14

14

16

16

No drugs available

13

9

12

13

12

Doctor/staff does not talk politely
Doctor/staff ask for bribe

2
0

1
1

0
1

1
0

1
0

Other health providers in the area give
better service
Other
N (Non-user household members)

14

4

19

7

11

35
568

32
308

30
581

29
748

31
2205

70

4
0

The distribution of reasons for non-use of PHCs headed by AYUSH Medical Officers and
paramedics follow a different pattern. For both these clinical care provider groups, “other

health providers in the area give better service” was not a frequently cited reason for non-use.
f

This partly explains the popularity of PHCs headed by AYUSH Medical Officers, seen in
Figure 9.1, who mostly serve in the tribal areas of Chhattisgarh where alternatives to public

sector providers are few. In contrast, the lack of private alternatives does not result in the
increased popularity of the PHCs run by paramedics - their PHCs received the least visits
f

4

because no doctor/health worker was available. The remoteness of these PHCs, their average

distance from a tarred road was 13 Km compared to less than 2 Km for the other three groups
(Table 5.1), appears to be a deterrent to both public and private physicians from serving

i

there.
I
I

9.2 Equity of health service use
I

The community survey collected information on household asset information. Matching rural
/
/
t
I

assets from a representative survey of the state - the National Family Health Survey (2005-

06) for Chhattisgarh was used to create a rural asset index. Applying principal components
analysis to the selected assets and extracting the first component created this index. This rural

asset index was used to divide the population of Chhattisgarh into quintiles. Asset weights

i

from this exercise were applied to the assets from the community survey and the sampled

I

individuals placed in one of five population quintiles which were merged into three groups.

Figure 9.2 shows the economic distribution of individuals who used the local PHC for
I

treatment when sick.

/

Utilization of services at the local PHC was pro-poor. For every clinical care provider group,

the proportion of people who sought care at the local PHC declined with better economic
status. Further, there is a clear socioeconomic gradient; except for AYUSH Medical Officers,
use of the local PHC decreased use with better economic status. The ratio of PHC utilization

of the lowest and highest economic groups suggest that AYUSH Medical Officers (1.3) were

the most equitable, followed by Medical Officers (1.4), RMAs (1.6) and Paramedicals (2.6).

71

*

*
I

Mission, their posting at public health facilities, indeed the ones in the study sample are de

*

facto Medical Officers, makes them an important part of the public sector health workforce.
The opportunity for them to practice at PHCs not only enables them to build local reputations

I

but they can also professionally develop and grow within the public sector system by moving

*

within the hierarchy of PHCs, CHCs and district hospitals.

*

*

Table 10.4

Job Satisfaction Dimensions and Scores by Provider Type

Family life
Job performance

Professional
development
Overall job satisfaction

Cronbach’s
alpha
0.74
0.71

0.68
0.77

*
*

Medical
Officer
2.6
(0.40)
2.2
(0.34)

AYUSH
Medical
Officer
2.9*
(0.45)
2.9
(0.44)

RMA
2.5
(0.35)
2.2
(0.28)

Paramedical
2.7
(0.65)
2.5
(0.51)

2.5
(0.36)
2.8
(0.51)

2.7*
(0.52)
3.0*
(0.58)

1.8*
(0.32)
2.4*
(0.4)

2.4
(0.46)
2.8
(0.51)

35

37

39

35

N (clinical care
providers)

*

Note: Figures in parenthesis are SD. Significantly different from Medical Officer at alpha

0.05*

The consistently low job satisfaction level of RM As, overall and across all job satisfaction

dimensions, is due to several issues. Like AYUSH Medical Officers, they are hired on annual

contracts, but their future prospects are quite different. They are not allowed to use the title of

"doctor’ which has been a long standing contention of theirs. Their future as a cadre is
uncertain because it is unclear if the Chhattisgarh government will continue with them due to

strong lobbies, particularly from the medical community. Further, being a new cadre, they
are not fully acceptable to their co-workers or their superiors (who are all physicians).

Importantly, they can only serve at the PHC level, irrespective of how long or how well they
serve. They have limited career prospects within the public health system.

80

*

*
*

s
*

• x

*

References
1. Mullan F. Non physician clinicains in 47 sub-African countries. Lancet 2007(370):2158
63.

2. Priya R. Public Health Services in India: A Historical Perspective. In: Gangolli LV,
Duggal R, Shukla A, eds. Review of Healthcare In India. Mumbai: Centre for Enquiry into
Health and Allied Themes, 2005: 41-74.

3. Gautham M, Shyamprasad KM. Needed: ‘basic’ doctors of modern medicine. The Hindu
November 5, 2009.
4. Rao K, Bhatnagar A, Berman P. India's health workforce: size, composition and
distribution. In: La Forgia J, Rao K, eds. India Health Beat. New Delhi: World Bank, New
Delhi and Public Health Foundation of India, 2009.
5. Doctors for Rural India. Economic & Political Weekly 2010;vol xlv no. 8.

6. Dhar A. MCI claims consensus on rural health cadre. The Hindu 2010 February 6.
7. Dhar A. Rural health care degree plan as schedule, says Azad. The Hindu 2010 May 12.
8. Dhar A. 'Rural health care course in advanced stage of finalization'. The Hindu 2010 May
29.
9. Raha S, Sundararaman T, Gupta G, Jain K, Antony KR, Rao K. Chhattisgarh Experience
with 3-Year Course for Rural Health Care Practitioners - A Case Study.

10. Raha S. It takes a village cadre. Indian Express 2010 February 10.
11. Berman P. Rethinking Health Care Systems: Private Health Care Provision in India.
World Development 1998;26(8): 1463-1479.
12. Banerjee A, Deaton A, Duflo E. Wealth, Health, and Health Services in Rural Rajasthan.
The American Economic Review 2004;94(2):326-330.
13. Das J, Hammer J. Money for nothing: The dire straits of medical practice in Delhi, India.
Journal ofDevelopment Economics 2007;83:1-36.
14. Government of India. Revised Indian National Classification of Occupations (NCO)2004: Directorate General of Employment & Training, Ministry of Labour, Government of
India.

15. Sharma B, Shekhar B. Pilot Inventories of Unqualified Private Health Care Providers in
two Districts in Andhra Pradesh: Harvard School of Public Health and University of
Hyderabad, 2003.
16. WHO. Working together for health. World Health Report 2006. Geneva: World Health
Organization, 2006.
17. Das J. Location, location, location: residence, wealth, and the quality of medical care in
Delhi, India. Health Affairs 2007;26(3):338-51.

81

18. Government of India. Bulletin on Rural Health Statistics 2008: Ministry of Health and
Family Welfare, Government of India, 2009.
19. Hammer J, Aiyar Y, Samji S. Understanding Government Failure in Public Health
Services. Economic and Political Weekly 2007;42(40):4049-4058.

20. Raha S, Berman P, Bhatnagar A. Career Preferences of Medical and Nursing Students in
Uttar Pradesh. In: La Forgia J, Rao K, eds. India Health Beat. New Delhi: World Bank, New
Delhi and Public Health Foundation of India, 2009.

21. Raha S, Berman P, Bhatnagar A. Some priority challenges of the nursing sector in India.
India Health Beat 2009;!(5).
22. Academy for Nursing Studies. Situational analysis of pulbic health nursing personnel in
India. Hyderabad: Academy of Nursing Studies, 2005.
23. Raha S, Berman P, Saran I, Rao K, Bhatnagar A. HRH: A Political Economy and
Institutional Analysis of the Indian context: World Bank, New Delhi and Public Health
Foundation of India, New Delhi, 2009.
24. Government of India. Compendium of recommendations of various committees on health
development 1943-1975. NewDelhi: Central Bureau of Health Intelligence, Directorate
General of Health Services, Ministry of Health and Family Welfare, Government of India,
1985.
25. Mullan F. Doctors For The World: Indian Physician Emigration. Health Affairs
2006;25(2):380-393.
26. Abhijit B, Deaton A, Duflo E. Wealth, Health, and Health Services in Rural Rajasthan.
The American Economic Review 2004;94(2):326-330.
27. Richards T. Medical education in India—in poor health. British Medical Journal (Clin Res
Ed) 1985;290(6475): 1132-5.

28. Supe A, Burdick W. Challenges and issues in medical education in India. Academic
Medicine 2006;81 (12): 1076-8.
29. Government of India. Task force on medical education for the National Rural Health
Mission. New Delhi: Ministry of Health and Family Welfare, Government of India, 2006.
30. Government of India. Recommendations - Medical Education: National Knowledge
Commission, Government of India, 2007.
31. Sood R. Medical education in India. Medical Teacher 2008;30(6):585-91.

32. Bansal R. Need for strengthening of internship (rotatory housemanship) training in India.
Education for Health (Abingdon) 2004 17(3):332-8.
33. Berman P, Gwatkin DR, Burger SE. Community-based health workers: Head start or
false start towards health for all? Social Science and Medicine 1987;25(5):443-459.

82

34. Huicho L. How much does quality of child care vary between health workers with diff
ering durations of training? An observational multicountry study. Lancet 2008(372):910-16.
35. Tamas Fulop MIR. Reviewing health manpower development-a method of improving
national health systems. 1987.
36. McCord C. The Quality of Emergency Obstetrical Surgery by Assisstant Medical
Officers in Tanzanian District Hospitals. African Workforce 2009.
37. Laurant M RD, Hermens R, Braspenning J, Grol R, Sibbald B. Substitution of doctors by
nurses in primary care. Cochrane Database Syst Rev. 2005 2(CD001271).
38. Sundararaman T. Community Health Worker Programs and The Public Health System.
In: Gangolli LV, Duggal R, Shukla A, eds. Review of Healthcare In India. Mumbai: Centre
for Enquiry into Health and Allied Themes, 2005: 107-116.
39. Haines A, Sanders D, Lehmann U, et al. Achieving child survival goals: potential
contribution of community health workers. The Lancet 2007;369(9579):2058-9.
40. Sundararaman T. Community health-workers: scaling up programmes. The Lancet
2007;369(9579):2058-9.
41. Central Council of Indian Medicine, 2008.

42. Chhattisgarh Human Development Report, 2005.
43. Government of India. Economic Survey 2009-2010: Ministry of Finance, 2010.
44. CSPRo [program]. 4.0’version. International Programs Center Population Division U.S.
Census Bureau Washington DC 2009.
45. 2001 S. Statistical Software: Release 10.0. College Station, TX: Stata Corporation.
46. Donabedian A. Evaluating physician competence. Bulletin of the World Health
Organization 2000;78(6):857-860.
47. John W. Peabody JL, Peter Glassman, et al Comparison of Vignettes, Standardized
Patients, and Chart Abstraction: A Prospective Validation Study of 3 Methods for Measuring
Quality. JAMA. 2000;283( 13): 1715-1722.
48. Leonard KL, Masatu MC. Variations In The Quality Of Care Accessible To Rural
Communities In Tanzania: Health Affairs, 2007.
49. Vishwanathan H, Rohde J. Diarrhea in Rural India: A Nationwide Study of Mothers and
Practitioners. New Delhi: Vision Books, 1990.

50. Rao K, Peters D, Bandeen-Roche K. Towards patient-centered health services in India—a
scale to measure patient perceptions of quality. International Journal of Quality in Health
Care 2006; 18(6):414-21.

83

Annexure

Correlates of Clinician Competence

Annex 1

Individual characteristics include age (which also controls for their experience working at
PHCs) and sex. PHC characteristics include a PHC infrastructure index composed of the sum
of 13 items indicating the presence of such things like the availability of electricity, water,
number of rooms, specific rooms for drug storage, cold chain, and consultations. A second
PHC level variable is the number of drug stock outs in the past year at the PHC. Village level
variables include a village development index composed of the sum of six items indicating
the presence of a high school, regular electricity supply, piped water, regular bus service and
cell phone connectivity. Other village level variables include the population of the area being
non-tribal and the distance of the village from the nearest town. The magnitude of the
regression coefficients and their statistical significance remains the same across models.
None of the individual and PHC level variables had any significant effect on provider
competence scores. Four different models were fit starting with the simplest case in which
only dummy variables for provider type are included. The reference category is the Medical
Officer.

Table A2

Regression Analysis of Correlates of Provider Competence

AYUSH Doctor

RMA
Paramedical

(1)

(2)

(3)

(4)

-0.986**
[0.17]
-0.054
[0.16]
-1.773**
[0-17]

-0.996**
[0.18]

-0.966**
[0.19]
-0.024
[0.21]
-1.746**

-0.973**
[0.20]
-0.074
[0.22]
-1.748**

[0.19]

[0.19]

0.061
[0-15]
0.000
[0.01]

0.154
[0.16]

Male

Age (years)

-0.051

[0.21]
-1.777**
[0.18]
0.057
[0.15]
-0.000
[0.01]

[0.12]

0.665
[0.37]

0.552
[0.41]

[0.01]
0.005
[0.02]
0.003
[0.02]
0.056
[0.05]
-0.004
[0.00]
-0.163
[0.17]
0.510
[0-47]

144
0.530

144
0.531

144
0.532

144
0.554

PHC Infrastructure Index

0.013

Drug stock out frequency in past year

[0.02]
0.007
[0.02]

Village Development Index
Distance of nearest town (Km)

Non Tribal Area
Constant

Observations
R-squared

-0.003

0.692**

Standard errors in brackets; ** p<0.01

I

Annex 2

Correlates of Patient Perceived Quality

Medical
Advice

Physician
Behaviour

Staff
Behaviour

Facility
Infrastructure

Overall
Perceived
Quality

-0.049
(-0.53)
0.144
(1.63)
-0.175
(-1.92)

-0.080
(-0.88)
0.044
(0.50)
-0.370**
(-4.10)

0.096
(1.02)
0.186*
(2.09)
0.108
(1.17)

0.113
(1.28)
0.205*
(2.43)
-0.042
(-0.48)

0.008
(0.09)
0.181*
(2.09)
-0.203*
(-2.26)

0.150**
(-2.72)
0.004*
(2.44)
0.238**
(-3.39)

-0.092
(-1.70)
0.001
(0.93)
-0.160*
(-2.30)

-0.044
(-0.80)
0.000
(0.31)
-0.096
(-1-35)

-0.099
(-1.88)
0.001
(0.55)
0.060
(0.89)

-0.134*
(-2.48)
0.002
(1-52)
-0.160*
(-2.32)

0.034**
(3-14)
-0.006
(-0.45)
0.000
(0.01)
-0.004**
(-3.23)
0.128
(1.50)
0.074
(0.44)
1074

0.029**
(2.69)
-0.008
(-0.63)
-0.012
(-0.45)
-0.002
(-1.96)
0.294**
(3.36)
-0.123
(-0.73)
1074

0.106**
(10.21)
0.038**
(3.16)
0.007
(0.29)
0.001
(0.43)
0.084
(1.01)
-0.674**
(-4.21)
1074

0.059**
(5.56)
-0.001
(-0.08)
0.009

Observations

0.015
(1.39)
-0.022
(-1.74)
0.027
(1.02)
-0.001
(-0.52)
0.147
(1.70)
-0.104
(-0.62)
1074

R-squared

0.041

0.062

0.028

0.117

0.075

AYUSH Doctor
RMA

Paramedical

Male

Age (years)

Illiterate

PHC Infrastructure Index

Drug stock out frequency in the past year
Village Development Index

Distance of nearest town (Km)
Non tribal area

Constant

Note: T-statistics in parentheses; **p<0.01,*p<0.05

II

(0.33)
-0.002*
(-1.98)
0.211*

(2.47)
-0.240
(-1.46)
1074

Media
12268.pdf

Position: 1202 (7 views)