This article was downloaded by: [182.73.85.194]On: 10 October 2014, At: 19:21Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Health Care for Women InternationalPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/uhcw20

Relevance of Health Knowledge inReporting Maternal Health Complicationsand Use of Maternal Health Care in IndiaShraboni Patraa, Perianayagam Arokiasamyb & Srinivas Golica Senior Research Scholar, International Institute for PopulationSciences, Mumbai, Maharashtra, Indiab Department of Development Studies, International Institute forPopulation Sciences, Mumbai, Maharashtra, Indiac Department of Development Studies, Giri Institute of DevelopmentStudies, Lucknow, IndiaAccepted author version posted online: 15 Aug 2014.Publishedonline: 08 Oct 2014.

To cite this article: Shraboni Patra, Perianayagam Arokiasamy & Srinivas Goli (2014): Relevance ofHealth Knowledge in Reporting Maternal Health Complications and Use of Maternal Health Care inIndia, Health Care for Women International, DOI: 10.1080/07399332.2014.946509

To link to this article: http://dx.doi.org/10.1080/07399332.2014.946509

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Health Care for Women International, 0:1–19, 2014Copyright © Taylor & Francis Group, LLCISSN: 0739-9332 print / 1096-4665 onlineDOI: 10.1080/07399332.2014.946509

Relevance of Health Knowledge in ReportingMaternal Health Complications and Use

of Maternal Health Care in India

SHRABONI PATRASenior Research Scholar, International Institute for Population Sciences, Mumbai,

Maharashtra, India

PERIANAYAGAM AROKIASAMYDepartment of Development Studies, International Institute for Population Sciences, Mumbai,

Maharashtra, India

SRINIVAS GOLIDepartment of Development Studies, Giri Institute of Development Studies, Lucknow, India

We measured levels of women’s health knowledge and their asso-ciation with the reporting of maternal health complications andrelated health care use. We found that women with higher levels ofhealth knowledge reported more pregnancy and postnatal compli-cations, and used more maternal health care services. Educationhas a positive impact on health, but education alone is not enoughto ensure recognizing and reporting of health complications andincreasing the demand for maternal health care services. We con-clude that the provision of health education for women will helpthem to identify maternal health complications and improve theirreporting and related health care use.

Self-reported measures of morbidity in developing countries have beenviewed with considerable scepticism by the researchers because these mea-sures of morbidity are misleading (Manesh, Sheldon, Pickett, & Carr-Hill,2008; Subramanian, Subramanian, Selvaraj, & Kawachi, 2009). Based on theirempirical assessment, Jain and colleagues (2012) showed that self-reportedmorbidity data are failing to show existing socioeconomic variations in ma-ternal morbidity by socioeconomic standing of women. According to Sen

Received 23 July 2013; accepted 16 July 2014.Address correspondence to Shraboni Patra, Senior Research Scholar, International Institute

for Population Sciences, Govandi Station Road, Deonar, Mumbai, 400088 Maharashtra, India.E-mail: shrageo@gmail.com

1

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

2 S. Patra et al.

(2002), an individual assessment of health is directly contingent on thesocial experience, and this leads to less reporting of illness among sociallydisadvantaged people as they fail to perceive the presence of health deficits.The majority of the poor and illiterate women do not report morbidity anddo not go to health facilities because they fail to recognize that they have amorbid condition. Thus, the lack of health knowledge and different healthbeliefs have been considered to be the main reasons for less reporting ofmorbidity.

Health knowledge is a relatively new concept in health promotion re-search. It is a much broader term than health literacy (Pfizer Inc., 1998). Whilethe World Health Organization (WHO, 1998, p. 10) described “health literacyis the achievement of a level of knowledge, personal skills and confidence totake action to improve personal and community health by changing personallifestyles and living conditions”; until now, however, there has been no uni-versally accepted definition of health knowledge. Nevertheless, researchersin a global context found that health knowledge is critically important forboth reporting of health problems and health care use (Nutbeam, 1998). Forinstance, Parker (2000) identified several patients with the most general andconvoluted health care problems are at greater risk for misunderstandingtheir diagnosis, medications and instructions on how to take care of theirhealth problems. Other researchers also found that by improving people’saccess to health information and awareness is critical to maximizing the ac-quiescence to use health care facilities (Nutbeam, 2008; Singh & Patra, 2013).

Health knowledge, regardless of education, can have an impact onreporting of health problems and decision making of women in seekinghealth care services (Jahan, 2000; Ratzan, 2001; United States Departmentof Health and Human Services [USDHHS], Office of Disease Prevention andHealth Promotion, 2010). Moronkola and colleagues (2006) revealed thatthe impact of belief in personal and community health practices is verystrong because belief may not always be scientific and it may make oneright or wrong in access to health care. In the same study, the authorsalso found that reproductive health knowledge is important for women tounderstand a woman’s health and well-being. Researchers in public healthalso suggested the consequences of health knowledge that include improvedself-reported health status, lower health care costs, increased understandingof health problems, shorter hospitalizations, and less frequent use of healthcare services (Altindag, Cannonier, & Mocan, 2010; Bhatt, Reid, Lewis, &Asnani, 2011; Kickbusch et al., 2001).

The concept of health knowledge, especially in terms of maternalhealth among women in India, is important because 72% of women stillbelieve that medical attention is not necessary during childbirth (Interna-tional Institute for Population Sciences [IIPS] & Macro International, 2007).Maternal health care use in terms of key indicators, such as three ormore antenatal care services (ANCs) use and institutional deliveries, is

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Relevance of Health Knowledge 3

not satisfactory. On average, one out of every five women is not re-ceiving antenatal care services for their last birth (IIPS & Macro Inter-national, 2007). With the introduction of cash incentive scheme JananiSuraksha Yojana under the National Rural Health Mission, however,the percentage of births that were delivered in health facilities increasedsteadily from 26% in 1993 to 58% in 2009 (IIPS & Macro Internationals, 2007;Registrar General of India [RGI], 2010). Nevertheless, it is substantially loweras compared with the developed countries (Goli & Jaleel, 2014; Lim et al.,2010). Even in the urban areas, despite the proximity of health care services,use of ANCs and institutional delivery, even among educated women, is farbelow the expected levels (Goli, Riddhi, & Arokiasamy, 2013). Further, suchpoor treatment-seeking attitude could be attributable to a low level of healthknowledge among Indian women (Carroll et al., 2007; Desai & Alva, 1998).

Although identification of health problems and health care use largelydepends on health beliefs and knowledge, across the globe there have beenfew efforts to link health knowledge of women with their self-reported healthstatus and health care use. To our knowledge, however, there is no studyin India that examined health knowledge of women and its association withself-reported health status and health care use. There is a need, therefore, tomeasure the extent and pattern of health knowledge among Indian women.Further, an assessment of the interaction among these three aspects: level ofhealth knowledge, reporting of maternal health problems, and use of medicalcare in an Indian context is important for planning and preparing strategies toimprove maternal health outcomes, thereby achieving the Millennium Devel-opment Goals-5 (MDG-5). Thus, we have looked at the twofold objectives forthis study. The first objective is to examine whether the reporting of maternalhealth complications is defined by the level of women’s health knowledge.The second one is to study the extent of disparity in the use of maternalhealth care services by the levels of health knowledge among women.

METHODS

Data Source

We have used information available in the India Human Development Survey(IDHS; Desai et al., 2007) for the assessment of the objectives of this study.The IHDS is the collaborative project of researchers from the University ofMaryland and National Council of Applied Economic Research (NCAER),New Delhi. For the first time, the IHDS 2005 has provided information re-garding women’s health belief and practices. In this survey, women in theage group of 15–49 years were asked to report their health beliefs, hygiene,and sanitation practices along with regular maternal health information likeprenatal, natal, and postnatal complications and health care use (Desai et al.,2010a).

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

4 S. Patra et al.

Sample Design

The IHDS was administered to a nationally representative sample of 41,554households located across all the states and union territories of India withthe exception of Andaman and Nicobar and Lakshadweep and covers theurban as well as the rural sample. One woman of reproductive age group(15–49 year) from each household was selected for an interview, to whomthe questions regarding health knowledge and maternal health complicationsand health care use were asked (Desai et al., 2010b).

Villages and urban blocks (comprising 150–200 households) formed theprimary sampling units (PSUs) from which the households were selected. Theurban and rural PSUs were selected by means of different sample design.In order to draw a random sample of urban households, all the urban areasin the states were listed in the order of their size, with a number of blocksselected from each urban area allocated based on Probability Proportionalto Sizes. Once the number of blocks for each urban area was determined,the enumeration blocks were drawn randomly. From each of the CensusEnumeration Blocks, complete household listing was conducted, and a sam-ple of 15 households was selected per block. For sampling purposes, somesmaller states were combined with nearby larger states. The rural samplecontains about half of the households that were interviewed initially by theNational Council of Applied Economic Research (NCAER, New Delhi) in1993–94, in a survey titled the Human Development Profile of India (HDPI).The other halves of the samples were drawn from both districts surveyed inHDPI as well as from the districts situated in the states and union territoriesnot covered in HDPI. The original HDPI was a random sample of 33,230households, located in 16 major states, 195 districts, and 1,765 villages. Instates where the 1993–94 survey was conducted and recontact details wereavailable, 13,593 households were randomly selected and reinterviewed in2005 (Desai et al., 2010b).

Variables

Predictor variables. The main predictor variable of this study is “healthknowledge.” We have constructed a health knowledge index (HKI) basedon questions asked to women about their health beliefs, practices, and re-productive knowledge, and required the respondents to specify each item as“true” or “false.” The list of questions asked to women in relation to healthknowledge follows: Do you know in which part of the menstrual cycle awoman is less likely to get pregnant? Is it good to drink 1–2 glasses of milkevery day during pregnancy? When you were pregnant, did you squeezeout milk from your breast to feed your child? Do men become physicallyweak even months after sterilization? Do you think that the first thin milkthat comes out after a baby is born is good for the baby? Although all these

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Relevance of Health Knowledge 5

questions do not directly reflect knowledge of maternal health, in the ab-sence of direct data they could be important proxies of reproductive healthknowledge of a woman.

We have used socioeconomic and demographic characteristics of thewomen as control variables, however, which are recoded for the purpose ofanalyses and effective comparison of the results: age group of women (15–24,25–34, and 35–49 years); place of residence (rural, urban); literacy (illiterateand literate); education level in terms of completed years of schooling, i.e.,no education, primary (1–4years), upper primary and secondary (5–10 years)and higher (above 10 years); religion (Hindu, Muslim, Christian, and Oth-ers, including Sikh, Buddhist, Jain, Tribal, Other, None); caste (“scheduledcaste” and “scheduled tribe,” “other backward caste,” and “Others,” includ-ing Brahmin and others); and wealth quintile (lowest, second, middle, fourthand highest for poorest, poorer, middle, richer, and richest, respectively).

Dependent variables. Dependent variables used in the study are alsorecoded purposively, such as antenatal check-up (no check-up, less thanthree check-ups and three or more check-ups), postnatal check-up within 2months of delivery (no check-up, check-up within 2 months, and check-upsafter 2 months of delivery, for the last 5 years), place of delivery (institu-tional and noninstitutional), pregnancy complications (include night blind-ness, blurred vision, and convulsions not from fever, excessive fatigue, ane-mia, and vaginal bleeding), and postnatal complications (include excessivevaginal bleeding and very high fever).

Statistical Analyses

We have performed the statistical analysis of the study in two stages: first,differentials in health knowledge by sociodemographic background char-acteristics of women, such as age group of women, place of residence,educational level of women, caste, religion, and wealth quintile of women,are assessed using multinomial logistic regression analyses. Further, the Mul-tiple Classification Analysis (MCA) conversion model is used to convert betacoefficients into adjusted percentages. In the second stage, the influence ofhealth knowledge on pregnancy complications; postdelivery complications;and antenatal care, delivery care, and postnatal care are examined. Bivariateand trivariate cross-tabulation and binary logistic regression models are usedto estimate the effects of health knowledge on reporting of pregnancy andpostnatal complications and ante-natal care, delivery care, and postnatal careuse. Below, we have presented mathematical forms of the models used inthis study.

Factor analysis with the Principal Component Analysis Method. In gen-eral, the factors analysis is a method that reduces data dimensionality byperforming a covariance analysis between factors. We have performed the

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

6 S. Patra et al.

factor analyses using the Principal Component Analysis (PCA) method. Eachquestion used to compute health knowledge index has been assigned aweight (factor score) generated through the PCA method, and the resultingscores are standardized in relation to the normal distribution with a mean ofzero and standard deviation of one (Gwatkin, Rutstein, Johnson, Pande, &Wagstaff, 2000). Women are ranked according to score, with a higher scorerepresenting a higher level of health knowledge. Based on HKI scores, threegroups of women have been determined: no or low knowledge, mediumknowledge, and high knowledge. An alpha test with p < .05 significancelevel is applied to examine the internal consistency among the variableschosen for computing HKI.

Multinomial logistic regression and MCA conversion model. The multi-nomial logistic regression is commonly used when the independent variablesinclude both numerical and nominal measures and the outcome variable(dependent variable) encompasses three or more than three categories. Forinstance, health knowledge index, in this study, includes three categories:no or low knowledge, medium knowledge, and high knowledge. With thisdependent variable, we have written a mathematical form of multinomiallogistic regression and MCA as below:

Z1 = L og

(P1

P3

)= a1 +

∑b

1 j∗X j ,

Z2 = L og

(P2

P3

)= a2 +

∑b

2 j∗X j,

and

P1 + P2 + P3 = 1,

whereai i = 1,2: constants,

• bij i = 1,2; j = 1,2 . . . .n: multinomial regression coefficient.• P1 = Estimated probability of reporting no or low health knowledge by

women aged 15 to 49 years.• P2 = Estimated probability of reporting medium health knowledge by

women aged 15 to 49 years.• P3 = Estimated probability of reporting high health knowledge by women

aged 15 to 49 years is the reference category.

For the sake of simplicity in the interpretation of results, we have convertedthe multinomial logistic regression coefficients into adjusted percentages.

The procedure consists of following steps:

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Relevance of Health Knowledge 7

Step 1:

• By using the regression coefficient and mean values of independent vari-ables, the probability is computed as:

• Pi = exp(Zi ){1+∑exp(Zi )} , i = 1, 2, 3 and P3 = 1 – P1 + P2 where Z was the estimated

value of response for all categories of each variable.

Step 2:

• To obtain the percentage values, the probability P was multiplied by 100.

Binary logistic regression analysis. The rationale behind the use ofbinary logistic regression is almost similar to that of multinomial logisticregression, but it is typically applied when the dependent variable forms abinary variable. The advantage of logistic regression analysis is that it requiresno assumption about the distribution of the independent variables, and theregression coefficient can be interpreted in terms of odds ratios. The binarylogistic regression model is commonly estimated by maximum likelihoodfunction. For the dependent variables, the logistic model takes the followinggeneral form:

Logit p = b0 + b1x1 + b2x2 + b3x3 + . . . . . . . . . . . . . . . . . . bkxk + ek

Log p/1 − p = b0 + b1x1 + b2x2 + b3x3 + . . . . . . . . . . . . . . . . . . bkxk + ek

where b0 are intercepts and b1, b2, b3, . . . . . . . . . . . . . . . bk represents the co-efficients of each of the predictor variables in the model while ek is an errorterm. The natural logarithms of the odds of the outcomes are represented byek.

RESULTS

Women’s Health Knowledge by Background Characteristics

The adjusted percentages, estimated from multinomial logistic regressionanalysis, indicate socioeconomic differences in the level of a woman’s healthknowledge. The percentage of women with higher health knowledge isfound to be increased with an increase in the age of women, which indicatesthat knowledge is an accumulation of experiences that grow with age. Thepercentage of women with higher health knowledge (34%) is significantlyhigher in the urban area as compared with women in the rural area (33%).By education level, health knowledge is found greater among women withhigher education (35%) than women with less education or no education(32%). Thus, the level of education shows a clear positive association withthe level of health knowledge. Similarly, women belonging to the richestwealth quintile have a higher level of health knowledge (37%) than women

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

8 S. Patra et al.

TABLE 1 Adjusted Percentage of Women Aged 15–49 Years by Health Knowledge1 andBackground Characteristics, India, 2005

Adjusted percentage of women by healthknowledge

Backgroundcharacteristics

No orlow Medium High

Sample ofwomen

Age group (years)15–24 R© 39.1 27.7 33.2 21,19425–34 38.6 27.5 33.9 16,48035–49 40.1 26.4 33.4 18,575

Place of residenceRural R© 38.8 27.9 33.3 36,064Urban 40.6 25.3∗∗∗ 34.0∗ 20,185

Education levelNo education R© 40.7 27.0 32.4 20,501Primary 40.3 25.7 34.0 3,881Upper primary andsecondary

38.4 27.3∗∗ 34.3∗∗∗ 22,985

Higher education 36.6 28.8∗∗∗ 34.6∗∗∗ 8,564Wealth quintile

Poorest R© 44.9 25.0 30.1 7,256Poor 45.4 23.1 31.6 7,139Middle 40.9 26.1 33.1∗∗∗ 7,802Rich 37.1 28.7∗∗∗ 34.1∗∗∗ 8,715Richest 31.1 31.6∗∗∗ 37.3∗∗∗ 9,803

CasteScheduled castes

and scheduledtribe R©

44.6 23.4 32.0 11,772

Other backwardclasses

40.0 26.7∗∗∗ 33.3∗∗∗ 16,286

Others 33.6 31.4∗∗∗ 35.0∗∗∗ 13,496Religion

Hindu R© 39.7 26.6 33.7 33,527Muslim 41.6 28.0∗∗ 30.5∗∗∗ 4,788Christian 24.8 41.1∗∗∗ 34.1∗∗∗ 1,376Others 37.1 26.2 36.7∗∗ 16,558Total 39.3 27.2 33.5 56,249

Note: 1Health Knowledge Index is predicting women’s reproductive health status (pregnancy complica-tions and postnatal complications);Total includes women with missing information on wealth and caste who are not shown separately;reference categories are R© and no or low knowledge; significance level: ∗∗∗p < .001; ∗∗p < .01; ∗p < .05.

of the poorest wealth quintile (30%). Compared with women belonging toother religions, the percentage of women with the higher level of healthknowledge is lower among the Muslim women (Table 1).

Women’s Health Knowledge Versus Reporting of PregnancyComplications

Table 2 presents the percentage of women who reported pregnancy com-plications and postnatal complications by their health knowledge level and

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

TA

BLE

2Per

centa

geof

Wom

enA

ged

15to

49Y

ears

Who

Exp

erie

nce

dA

ny

Pre

gnan

cyCom

plic

atio

ns

or

Post

nat

alCom

plic

atio

ns

Acc

ord

ing

toH

ealth

Know

ledge

ofW

om

enby

Thei

rB

ackg

round

Char

acte

rist

ics,

India

,20

05

1 Pre

gnan

cyco

mplic

atio

ns

(any)

2 Post

nat

alco

mplic

atio

ns

(any)

Bac

kgro

und

char

acte

rist

ics

No

or

low

know

ledge

Med

ium

know

ledge

Hig

hkn

ow

ledge

Chi-

squar

eN

oor

low

know

ledge

Med

ium

know

ledge

Hig

hkn

ow

ledge

Chi-sq

uar

e

Age

group

(yea

rs)

3.03

0.86

15–2

444

.945

.550

.314

.715

.711

.925

–34

44.7

42.2

49.4

9.4

10.9

15.4

35–4

948

.248

.252

.210

.614

.615

.7Pla

ceofre

siden

ce0.

330.

03Rura

l45

.445

.551

.112

.114

.314

.1U

rban

47.7

45.3

49.7

11.3

12.8

14.4

Wom

en’s

educa

tion

0.43

7.42

∗

No

educa

tion

47.2

46.5

48.2

13.3

16.5

15.1

Prim

ary

educa

tion

43.9

42.8

58.7

10.5

6.6

11.1

Upper

prim

ary

and

seco

ndar

y42

.645

.953

.210

.212

.915

.5

Hig

her

educa

tion

53.9

42.8

46.6

9.8

13.5

7.2

Wea

lthquin

tile

10.6

1∗∗4.

15Poore

st44

.747

.854

.910

.913

.718

.1Poor

50.5

44.7

52.4

8.6

16.9

14.2

Mid

dle

45.0

46.8

55.9

10.1

11.7

10.1

Ric

h47

.749

.948

.916

.121

.414

.8Ric

hes

t42

.439

.343

.716

.75.

511

.9(C

onti

nu

edon

nex

tpa

ge)

9

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

TA

BLE

2Per

centa

geof

Wom

enA

ged

15to

49Y

ears

Who

Exp

erie

nce

dA

ny

Pre

gnan

cyCom

plic

atio

ns

or

Post

nat

alCom

plic

atio

ns

Acc

ord

ing

toH

ealth

Know

ledge

ofW

om

enby

Thei

rB

ackg

round

Char

acte

rist

ics,

India

,20

05(C

onti

nu

ed)

1 Pre

gnan

cyco

mplic

atio

ns

(any)

2 Post

nat

alco

mplic

atio

ns

(any)

Bac

kgro

und

char

acte

rist

ics

No

or

low

know

ledge

Med

ium

know

ledge

Hig

hkn

ow

ledge

Chi-

squar

eN

oor

low

know

ledge

Med

ium

know

ledge

Hig

hkn

ow

ledge

Chi-sq

uar

e

Cas

te1.

923.

48Sc

hed

ule

dca

stes

and

sched

ule

dtrib

es

41.0

40.3

49.9

10.0

11.7

14.4

Oth

erbac

kwar

dcl

asse

s

47.5

42.7

53.1

12.7

10.1

13.2

Oth

ers

50.2

52.2

48.2

13.0

21.5

15.7

Rel

igio

n18

.99∗∗

∗9.

02∗∗

Hin

du

45.4

42.3

51.6

12.1

14.4

13.5

Musl

im52

.469

.550

.811

.215

.321

.1Christ

ian

31.3

45.3

31.6

2.0

0.5

0.0

Oth

ers

40.2

31.7

43.2

11.5

12.9

15.8

Tota

l46

.545

.450

.411

.713

.55

14.2

5

Not

e:1 P

regn

ancy

com

plic

atio

ns

incl

ude

nig

htblin

dnes

s,blu

rred

visi

on,c

onvu

lsio

nnotfrom

feve

r,ex

cess

ive

fatig

ue,

anem

iaan

dva

ginal

ble

edin

g.Ref

eren

ceper

iod:

Last

5ye

ars

pre

cedin

gth

esu

rvey

.Tota

lin

cludes

wom

enw

ithm

issi

ng

info

rmat

ion

on

wea

lthan

dca

ste

who

are

notsh

ow

nse

par

atel

y.2 P

ost

nat

alco

mplic

atio

ns

incl

ude

vagi

nal

ble

edin

gan

dve

ryhig

hfe

ver;

refe

rence

per

iod:

last

5ye

ars

pre

cedin

gth

esu

rvey

;to

tal

incl

udes

wom

enw

ithm

issi

ng

info

rmat

ion

on

wea

lthan

dca

ste

who

are

notsh

ow

nse

par

atel

y;Si

gnifi

cance

leve

l:∗∗

∗ p<

.001

;∗∗

p<

.01;

∗ p<

.05.

10

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Relevance of Health Knowledge 11

background characteristics. By level of health knowledge, the percentagesof women who reported pregnancy complications are 46% and 50%, respec-tively, among no or low and high health knowledge women. The highestlevel of health knowledge is associated with the highest reported pregnancycomplications. Our key finding from these results is that within the same so-cioeconomic status, such as a rural place of residence, poor economic status,and no education, the reported pregnancy complications by women is in-creased significantly with an increase in the level of health knowledge. Thispattern was even sustained among women with higher education. The oddsratios of the logistic regression analysis of self-reported pregnancy compli-cations by women are presented in Table 3. In comparison with the womenwith no health knowledge (OR = 1.000, p < .05), women with high healthknowledge are significantly more likely to report pregnancy complications(OR = 1.139, p < .01).

Table 2 shows the percentage of women reporting postnatal complica-tions by their health knowledge level and background characteristics. Thereported postnatal complication is increased with an increase in the level ofhealth knowledge, and such a pattern is consistent for all the socioeconomicgroups. The result of logistic regression analysis, after controlling for othersocioeconomic background characteristics, has also shown that women withhigher health knowledge are significantly more likely to report greater post-natal complications (OR = 1.378, p < .01) than women with no or low healthknowledge (OR = 1.000, p < .05; Table 3).

Women’s Health Knowledge Versus Maternal Health Care Use

The bivariate cross-tabulation analyses (Table 4) show that the percentage ofwomen who reported receiving at least three antenatal check-ups increasedwith an increase in health knowledge level. Whereas ANCs use is only 52%among women with no or low health knowledge, it increased to 72% forwomen with high health knowledge. Further, the trivariate cross-tabulationanalyses indicate that with the same socioeconomic and demographic back-ground, the percentage of women who attended at least three antenatalcheck-ups varied greatly by their health knowledge levels. An importantfinding from the trivariate analyses is that, compared with lower socioeco-nomic groups, women in higher socioeconomic groups have shown greatervariation in antenatal care services use by their health knowledge level. Wehave found that the women who received the highest level (90%) of threeand more antenatal check-ups belong to the richest wealth quintile withhigher health knowledge. In contrast, the women who received the lowest(42%) of three and more antenatal check-ups belonged to the poorest wealthquintile with no or low health knowledge (Table 4). The odds ratios, esti-mated from the logistic regression analysis (Table 5), also support the factthat women with medium (OR = 1.581, p < .001) and high (OR = 1.434,p < .001) health knowledge are significantly more likely to receive at least

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

12 S. Patra et al.

TABLE 3 Results of Logistic Regression Analysis of Pregnancy Complications and PostnatalComplications Experienced by Women Aged 15–49 Years by Their Health Knowledge andBackground Characteristics, India, 2005

Maternal health problems

Backgroundcharacteristics

Pregnancycomplications Expβ (CIs

of Exp β)Postnatal complications

Expβ (CIs of Exp β)

Health knowledgeNo or low R© 1 1Medium 1.057 (0.930,1.202) 1.049 (0.749,1.468)High 1.139∗∗ (1.013,1.280) 1.378∗∗ (1.032,1.841)

Age group (years)15–24 age R© 1 125–34 age 0.912 (0.805,1.034) 0.805 (0.581,1.115)35–49 age 1.007 (0.890,1.140) 0.937 (0.691,1.270)

Place of residenceRural R© 1 1Urban 1.052 (0.942,1.176) 1.144 (0.865,1.512)

Level of educationNo education R© 1 1Primary education 0.978 (0.799,1.198) 0.567∗∗ (0.322,1.001)Upper primary

and secondaryeducation

0.957 (0.850,1.076) 0.756∗ (0.563,1.016)

Higher education 0.956 (0.807,1.133) 0.678∗ (0.435,1.055)Wealth quintile

Poorest R© 1 1Poor 1.005 (0.851,1.187) 0.830 (0.564,1.221)Middle 0.947 (0.803,1.118) 0.808 (0.548,1.191)Rich 0.865∗ (0.732,1.021) 0.910 (0.617,1.341)Richest 0.781∗∗∗ (0.662,0.920) 0.622∗∗ (0.404,0.956)

CasteScheduled caste

and scheduledtribe R©

1 1

Other backwardclass

1.096 (0.966,1.243) 1.101 (0.801,1.512)

Others 1.129∗ (0.980,1.300) 1.449∗∗ (1.007, 2.086)Religion

Hindu R© 1 1Muslim 1.303∗∗∗ (1.110,1.529) 1.252 (0.881,1.780)Christian 0.875 (0.634,1.206) 0.000 (0.0)Other 1.318∗∗ (1.009,1.723) 1.276 (0.649,2.506)

Note: R©Reference category of different characteristics; Significance level: ∗∗∗p < .001; ∗∗p < .01; ∗p <

.05.

three or more antenatal check-ups as compared with women with no or lowhealth knowledge.

In Table 4, we observe the similar patterns of results in case of insti-tutional delivery coverage. From the analyses, we found that, among thewomen with the same socioeconomic background, institutional delivery

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

TA

BLE

4Per

centa

geof

Wom

enA

ged

15–4

9Y

ears

Who

Rec

eive

dA

nte

nat

alChec

k-ups,

Inst

itutio

nal

Del

iver

y,an

dPost

-Nat

alM

edic

alCar

eby

Thei

rH

ealth

Know

ledge

and

Oth

erB

ackg

round

Char

acte

rist

ics,

India

,20

05

Atle

astth

ree

Inst

itutio

nal

Post

nat

alch

eck-

up

ante

nat

alch

eck-

ups

del

iver

y(w

ithin

2m

onth

sofbirth

s)

Bac

kgro

und

No/

low

Med

ium

Hig

hChi-

No/

low

Med

ium

Hig

hChi-

No/

low

Med

ium

Hig

hChi-

.ch

arac

terist

ics

know

ledge

know

ledge

know

ledge

squar

ekn

ow

ledge

know

ledge

know

ledge

squar

ekn

ow

ledge

know

ledge

know

ledge

squar

e

Age

group

(yea

rs)

1.53

3.79

2.26

15–2

455

.174

.666

.138

.957

.164

.611

.718

.116

.425

–34

52.7

67.8

64.4

41.0

53.9

64.1

12.1

16.2

16.7

35–4

947

.667

.370

.935

.658

.456

.610

.815

.016

.5Pla

ceofre

siden

ce0.

292.

025.

45∗

Rura

l51

.366

.169

.438

.456

.560

.311

.716

.416

.4U

rban

53.4

70.8

73.8

38.0

57.0

67.3

11.0

16.9

17.3

Wom

en’s

educa

tion

1.03

4.25

16.3

0∗∗

No

educa

tion

50.7

70.5

67.8

40.3

55.6

61.8

12.0

15.4

16.2

Prim

ary

educa

tion

57.4

69.4

60.4

46.0

54.4

55.2

14.4

16.9

13.1

Upper

prim

ary

and

seco

ndar

yed

uca

tion

53.7

70.6

67.8

35.0

57.0

62.1

10.5

17.8

17.6

Hig

her

educa

tion

47.0

70.1

67.7

36.4

60.7

65.6

9.9

16.9

17.1

Wea

lthquin

tile

381∗

∗∗70

5.15

∗∗∗

188.

22∗∗

∗Poore

st42

.443

.648

.225

.927

.444

.18.

914

.69.

3Poor

44.2

60.0

43.6

26.5

36.8

42.0

10.3

14.7

11.3

Mid

dle

52.1

64.0

67.5

38.7

52.5

53.8

10.5

15.5

18.2

Ric

h53

.576

.277

.242

.968

.770

.013

.717

.317

.3Ric

hes

t71

.785

.489

.665

.179

.185

.213

.717

.921

.4Cas

te Sched

ule

dca

stes

and

sched

ule

dtrib

es

48.5

55.0

51.6

187.

57∗∗

∗31

.943

.747

.235

6.17

∗∗∗

10.6

16.8

14.4

82.8

6∗∗∗

Oth

erbac

kwar

dcl

asse

s48

.368

.769

.835

.456

.255

.412

.515

.716

.2

Oth

ers

63.5

83.1

76.9

53.8

68.9

79.9

11.0

17.2

18.5

Rel

igio

n25

.73∗

∗∗53

.06∗

∗∗39

.96∗

∗∗H

indu

52.7

67.5

66.7

39.1

56.2

59.2

11.5

14.9

15.8

Musl

im49

.280

.970

.538

.354

.066

.713

.024

.124

.9Christ

ian

45.7

90.6

86.5

17.2

84.8

92.4

5.4

21.2

20.8

Oth

ers

44.4

72.1

55.5

28.9

52.4

66.3

9.5

25.6

10.8

Tota

l52

.468

.571

.638

.256

.863

.811

.416

.716

.9

Not

e:Ref

eren

cePer

iod:

Last

5ye

ars

pre

cedin

gth

esu

rvey

;to

tal

incl

udes

wom

enw

ithm

issi

ng

info

rmat

ion

on

wea

lthan

dca

ste

who

are

not

show

nse

par

atel

y;Si

gnifi

cance

leve

l:∗∗

∗ p<

.001

;∗∗

p<

.01;

∗ p<

.05.

13

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

14 S. Patra et al.

TABLE 5 Results of Logistic Regression Analysis of Received Antenatal Check-Up, Institu-tional Delivery, and Postnatal Check-Up by Women Aged 15–49 Years, by Health Knowledgeand Background Characteristics, India, 2005

Backgroundcharacteristics

Antenatal check-upExpβ (CIs of Expβ)

Institutional deliveryExpβ (CIs of Expβ)

Postnatal check-upExpβ (CIs of Expβ)

Health knowledgeNo or low R© 1 1 1Medium 1.581∗∗∗(1.371,1.823) 1.732∗∗∗(1.510,1.987) 1.91∗∗∗(1.628,2.241)High 1.434∗∗∗(1.260,1.633) 1.430∗∗∗(1.262,1.620) 1.893∗∗∗(1.633,2.195)

Age group (years)15–24 R© 1 1 125–34 0.951(0.826,1.094) 0.927(0.810,1.061) 1.130(0.964,1.324)35–49 0.929(0.810,1.067) 0.907(0.795,1.036) 0.946(0.810,1.104)

Place of residenceRural R© 1 1 1Urban 0.965(0.852,1.094) 0.894∗∗(0.793,1.008) 1.078(0.936,1.242)

Level of educationNo education R© 1 1 1Primary

education1.116(0.893,1.395) 1.136(0.913,1.414) 1.068(0.830,1.376)

Upper primaryand secondaryeducation

1.056(0.925,1.206) 1.003(0.883,1.139) 0.928(0.799,1.077)

Highereducation

0.930(0.769,1.125) 1.110(0.923,1.334) 1.047(0.843,1.299)

Wealth quintilePoorest R© 1 1 1Poor 1.153(0.967,1.375) 1.061(0.885,1.272) 1.016(0.821,1.257)Middle 1.613∗∗∗(1.353,1.924) 1.742∗∗∗(1.463,2.076) 1.480∗∗∗(1.199,1.828)Rich 2.192∗∗∗(1.831,2.625) 2.437∗∗∗(2.044,2.906) 1.962∗∗∗(1.588,2.424)Richest 3.541∗∗∗(2.943,4.261) 4.851∗∗∗(4.052,5.809) 2.544∗∗∗(2.065,3.135)

CasteScheduled caste

and scheduledtribe R©

1 1 1

Other backwardclass

1.404∗∗∗(1.224,1.611) 1.481∗∗∗(1.294,1.695) 1.089(0.928,1.277)

Others 1.923∗∗∗(1.641,2.254) 2.469∗∗∗(2.121,2.874) 1.520∗∗∗(1.269,1.820)Religion

Hindu R© 1 1 1Muslim 0.912(0.762,1.091) 0.791∗∗∗(0.668,.938) 1.067(0.827,1.306)Christian 1.780∗∗∗(1.182,2.681) 2.317∗∗∗(1.562,3.435) 1.985∗∗∗(1.275,3.093)Other 1.310∗(0.965,1.777) 1.199(0.893,1.612) 1.576∗∗(1.104,2.251)

Note: R©-Reference category of different characteristics; Significance level: ∗∗∗p < .001; ∗∗p < .01;∗p < .05.

coverage increases with an increase in the level of health knowledge (38%,57%, and 64% for women with no or low, medium, and high health knowl-edge, respectively). The combined effect of health knowledge and socioe-conomic characteristics further increases the use of institutional deliveries.We have found institutional delivery coverage is high among women withthe highest level of health knowledge and who belong to the richest wealth

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Relevance of Health Knowledge 15

quintile (85%), are Christian (92%), and have a high level of education (66%).In contrast, institutional delivery coverage is the lowest among women withno or low health knowledge living in rural areas (38%) and who are inthe poorest wealth quintile (26%) and scheduled caste and scheduled tribe(32%). The results of logistic regression analysis further strengthen the evi-dence that women’s health knowledge acts as a key predictor of institutionaldelivery coverage in India (Table 5). The odds ratios of logistic regression re-veal that after controlling for other background characteristics, women withmedium (OR = 1.732, p < .001) and the highest level of health knowledge(OR = 1.430, p < .001) are significantly more likely to receive institutionaldelivery coverage than those women with no or low health knowledge (OR= 1.000, p < .05).

Postnatal check-ups within 2 months of delivery are essential for thehealth of both mothers and children. We have also found that the percent-ages of women who received postnatal check-ups vary considerably withtheir level of health knowledge. Among women with no or low knowledge,the rate of postnatal check-ups within 2 months of delivery is much lower(11%) than among women with the highest level of health knowledge (17%).Postnatal care is the lowest among women with no or low health knowledgeand among women who belong to the poorest wealth quintile (9%). On theother hand, postnatal care is higher among women with the highest level ofhealth knowledge and who belong to the richest wealth quintile (21%).

Similar results from our logistic regression analyses (Table 5) also sup-port the findings from trivariate cross tabulation. The women who havemedium (OR = 1.910, p < .001) and the highest level of health knowledge(OR = 1.893, p < .001) are significantly more likely to receive postnatal careservices than those women who have no or low health knowledge (OR =1.000, p < .01). Overall, both trivariate and logistic regression analyses sug-gest positive associations between women’s health knowledge levels and thepostnatal care they receive.

DISCUSSION

In this study, we made the first attempt to quantify the level of health knowl-edge among Indian women in age group 15–49 years. The concept of “healthknowledge” has been empirically examined, and the difference betweenhealth literacy and health knowledge is discussed. We have constructed a“health knowledge index” based on women’s health beliefs, practices, ex-periences, and reproductive knowledge. For the first time, we also assessedthe association between health knowledge levels of women and reportedmaternal health complications and use of maternal health care. With the nu-anced assessment of health knowledge and its effect on reported pregnancycomplication and maternal health care use, we have provided a number

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

16 S. Patra et al.

of intriguing findings. We have found that women’s health knowledge isvaried by their education level, economic status, and place of residence.Similarly, our findings also suggest a positive association between women’shealth knowledge and maternal health care use. Attending at least three ormore antenatal check-ups and having an institutional delivery and postnatalcheck-ups within 2 months of delivery are found to be increased with theincrease in health knowledge level of women.

Nevertheless, even within the same socioeconomic background, womenreported pregnancy and postnatal complications, and the uses of maternalhealth care are varied by their level of health knowledge. The highly edu-cated women along with those with high health knowledge reported moreabout their pregnancy complications and postnatal complications, and theysought more medical assistance such as antenatal care, institutional deliverycare, and postnatal care. We attribute this pattern of results to a compoundeffect of education and health knowledge on reporting of maternal healthcomplications and use of maternal health care. An important and interestingpattern of results of our study is a greater disparity in the use of maternalhealth care services among the women with higher socioeconomic statusthan women with lower socioeconomic status by their health knowledgelevel. Based on our findings, therefore, we promote that health knowledgeis a critical predictor of reporting of illness, health care seeking behavior,and health outcomes.

Overall, the findings of our study advance that although there is a pos-itive impact of education on health, still women need proper reproductivehealth knowledge to understand their own health status and health-relatedproblems. Otherwise, misconception may lead to adverse health outcomes.Our findings are in tune with those of earlier studies that said, educationmakes people knowledgeable, but even within educated people, improperhealth conception exists (Chakraborty, Islam, Chowdhury, Bari, & Akhter,2003; Glewwe, 1999; Kendel, 1991; Lubbock & Stephenson, 2008; Phoxay,Okumura, Nakamura, & Wakia, 2001; Shewry, Smith, & Tunstall-Pedoe,1990).

CONCLUSION AND IMPLICATIONS

We conclude that improvement in literacy alone will not help women tounderstand their own health problem and seek necessary health care for it.Health knowledge of women is an important aspect for understanding theirhealth problems and seeking proper health care. Specifically, focusing onhealth education programs for women can help improve their reproductivehealth knowledge, thereby increasing reporting of pregnancy and decreasingdelivery and postdelivery health complications. We have the same opinionof other researchers that MDG-5 can be achieved through the proper use

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Relevance of Health Knowledge 17

of maternal health care services, that is, antenatal care, institutional deliverycare, and postnatal care. We further state, however, that the achievement ofMDG-5 would become easier when a woman understands her own healthproblem correctly and seeks proper health care services. Increasing women’shealth knowledge can speed up the process to reach the goal of MDG-5.Providing health education to women through Auxiliary Nurse Midwives,Accredited Social Health Activists, and Anganwadi Workers, therefore, willmake women more knowledgeable about their health problems and improvehealth care seeking behavior. Additionally, the Ministry of Health and FamilyWelfare of the Government of India should set priorities for greater fundingallocation for health promotion research as well as health services in thepublic sector, which remain underfunded compared with many countries inthe world. In fact, public health funding with less than 2% of gross domesticproduct in India is one of the lowest in the world. In a country like India,however, the social equity in public health cannot be ignored, especially withrespect to the distribution of health facilities, service delivery, and medicalassistance under health promotion programs. In this process, establishingaffordable health care among poor and socially disadvantaged communitiesand their exposure to mass media is essential.

REFERENCES

Altindag, D. T., Cannonier, C., & Mocan, N. H. (2010). The impact of education onhealth knowledge (National Bureau of Economic Research Working Paper No.16422, JEL No. I1, I12, I18, I21). Cambridge, MA: National Bureau of EconomicResearch.

Bhatt, K., Reid, M. E., Lewis, N. A., & Asnani, M. R. (2011). Knowledge and healthbeliefs of Jamaican adolescents with sickle cell disease. Journal of PaediatricBlood and Cancer, 57, 1044–1048.

Carroll, J., Epstein, R., Fiscella, K., Volpe, E., Diaz, K., & Omar, S. (2007). Knowl-edge and beliefs about health promotion and preventive health care amongSomali women in the United States. Health Care for Women International, 28,360–380.

Chakraborty, N., Islam, M. A., Chowdhury, R. I., Bari, W., & Akhter, H. H. (2003).Determinants of the use of maternal health services in rural Bangladesh. HealthPromotion International, 18, 327–337. doi: 10.1093/heapro/dag414

Desai, S., & Alva, S. (1998). Maternal education and child health: Is there a strongcausal relationship? Demography, 35, 71–81.

Desai, S. B., Dubey, A., & Joshi, B. L., Sen, M., Shariff, A., & Vanneman, R. (2010a).Human development in India: Challenges for a society in transition. Oxford,UK: Oxford University Press.

Desai, S., Dubey, A., Joshi, B. L., Sen, M., Shariff, A., & Vanneman, R. (2007). IndiaHuman Development Survey (IHDS) [Computer file], ICPSR22626-v1. Ann Arbor,MI: Inter-university Consortium for Political and Social Research.

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

18 S. Patra et al.

Desai, S. B., Dubey, A., Joshi, B. L., Sen, M., Shariff, A., & Vanneman, R. (2010b).India human development survey: Design and data quality. IHDS Techni-cal Paper No. 1. New Delhi, India: Oxford University Press. Retrieved fromhttp://ihds.umd.edu/IHDS papers/technical%20paper%201.pdf

Glewwe, P. (1999). Why does mother’s schooling raise child health in developingcountries? Evidence from Morocco. The Journal of Human Resources, 34(1),124–159.

Goli, S., & Jaleel, A. C. P. (2014). What is the cause of the decline in maternal mortalityin India? Evidence from time series and cross-sectional analyses. Journal ofBiosocial Science, 46, 351–365.

Goli, S., Riddhi, D., & Arokiasamy, P. (2013). Pathways of economic inequalities inmaternal and child health in urban India: A decomposition analysis. PLOS OneMedical Journal. doi:10.1371/journal.pone.0058573

Gwatkin, D. R., Rutstein, S., Johnson, K., Pande, R. P. & Wagstaff, A. (2000). Socio-economic differences in health, nutrition and population. Washington, DC: HNPPoverty Thematic Group of the World Bank.

International Institute for Population Sciences (IIPS), and Macro International. (2007).National Family Health Survey (NFHS-3, 2005-06): India, 2007. Mumbai, India:IIPS.

Jahan, A. J. (2000). Promoting health literacy: A case study in the preventionof diarrhoeal disease from Bangladesh. Health Promotion International, 15,285–291.

Jain, K., Goli, S., & Arokiasamy, P. (2012). Are self reported morbidity deceptivein measuring socio-economic inequalities? Indian Journal of Medical Research,136, 750–757.

Kendel, D. S. (1991). Health behaviour, health knowledge and schooling. Journal ofPolitical Economy, 99, 287–305.

Kickbusch, I. S. (2001). Health literacy: Addressing the health and education divide.Health Promotion International, 16, 289–297.

Lim, S. S., Lalit, D., Hoisington, J. A., James, S. L., Hogan, M. C., & Gakidou, E.(2010). India’s Janani Suraksha Yojana, a conditional cash transfer programmeto increase births in health facilities: An impact evaluation. Lancet, 375(9730),2009–2023.

Lubbock, L. A., & Stephenson, R. B. (2008). Use of maternal health care servicesin the Department of Matagalpa, Nicaragua. Pan American Journal of PublicHealth, 24(2), 75–84.

Manesh, A. O., Sheldon, T. A., Pickett, K. E., & Carr-Hill, R. (2008). Accuracy ofchild morbidity data in demographic and health surveys. International Journalof Epidemiology, 37, 194–200.

Moronkola, O. A., Ojediran, M. M., & Amosu, A. (2006). Reproductive health knowl-edge, beliefs and determinants of contraceptives use among women attendingfamily planning clinics in Ibadan, Nigeria. African Health Sciences, 3, 155–159.

Nutbeam, D. (1998). Health promotion glossary (WHO/ HPR/ HEP/ 98.1). Geneva,Switzerland: WHO. Retrieved from http://www.who.int/hpr/NPH/docs/hp_glossary_en.pdf

Nutbeam, D. (2008). The evolving concept of health literacy. Social Science &Medicine, 67, 2072–2078.

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014

Relevance of Health Knowledge 19

Parker, R. (2000). Health literacy: A challenge for American patients and their healthcare providers. Health Promotion International, 15, 277–283.

Pfizer Inc. (1998, October). Promoting health literacy: A call to action. ConferenceProceedings, New York, NY.

Phoxay, C., Okumura, J., Nakamura, Y., & Wakia, S. (2001). Influence of women’sknowledge on maternal health care use in southern Laos. Asia Pacific Journalof Public Health, 13(1), 13–19.

Ratzan, S. (2001). Health literacy: Communication for the public good. Health Pro-motion International, 16, 207–214.

Registrar General of India (RGI). (2010). Sample registration system (SRS). New Delhi,India: Office of the Registrar General.

Sen, A. (2002). Health: Perception versus observation. British Medical Journal, 324,860–861.

Shewry, M. C., Smith, W.C.S., & Tunstall-Pedoe, H. (1990). Health knowledge andbehaviour change: A comparison of Edinburgh and north Glasgow. HealthEducation Journal, 49, 185–190.

Singh, R. K., & Patra, S. (2013). Differentials in the use of antenatal care servicesin EAG States of India. International Research Journal of Social Science, 2(11),28–32.

Subramanian, S. V., Subramanyam, M. A., Selvaraj, S., & Kawachi, I. (2009). Are selfreports of health and morbidity in developing countries misleading? Evidencefrom India. Social Science and Medicine, 68, 260–265.

United States Department of Health and Human Services (USDHHS), Office of Dis-ease Prevention and Health Promotion. (2010). Healthy people 2010. Retrievedfrom http://www.healthypeople.gov/2010/

World Health Organization (WHO). (1998). Health promotion glossary. Geneva,Switzerland: Author. Retrieved from http://www.who.int/healthpromotion/about/HPR%20Glossary%201998.pdf

Dow

nloa

ded

by [

182.

73.8

5.19

4] a

t 19:

21 1

0 O

ctob

er 2

014