5th Millenium development goal
Upcoming SlideShare
Loading in...5
×
 

Like this? Share it with your network

Share

5th Millenium development goal

on

  • 646 views

5th Millenium development goal

5th Millenium development goal

Statistics

Views

Total Views
646
Views on SlideShare
646
Embed Views
0

Actions

Likes
0
Downloads
2
Comments
0

0 Embeds 0

No embeds

Accessibility

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

5th Millenium development goal Document Transcript

  • 1. WHO South-East Asia Journal of Public Health 2012;1(3):279-289 279Original researchIndia’s progress towards the Millennium DevelopmentGoals 4 and 5 on infant and maternal mortalityHanimi Reddya, Manas R Pradhana, Rohini Ghosha, A G KhanbBackground: India is in a race against time to achieve the Millennium Development Goals (MDGs)4 and 5, to reduce Infant Mortality Rate (IMR) to ‘28’ and Maternal Mortality Ratio (MMR) to ‘109’,by 2015. This study estimates the percent net contribution of the states and the periods in shapingIndia’s IMR/MMR, and predicts future levels.Methods: A standardized decomposition technique was used to estimate each state’s and period’spercent share in shaping India’s decline in IMR/MMR between two time points. Linear and exponentialregression curves were fitted for IMR/MMR values of the past two decades to predict IMR/MMRlevels for 2015 for India and for the 15 most populous states.Results: Due to favourable maternal mortality reduction efforts in Bihar/Jharkhand (19%) andMadhya Pradesh/Chhattisgarh (11%), Uttar Pradesh (33%) - India is predicted to attain the MDG-5target by 2016, assuming the pace of decline observed in MMR during 1997-2009 continues to followa linear-trend, while the wait may continue until 2023-2024 if the decline follows an exponential-trend. Attaining MDG-4 may take until 2023-2024, due to low acceleration in IMR drop in Bihar/Jharkhand, Uttar Pradesh/Uttarakhand and Rajasthan. The maximum decline in MMR during 2004-2009 coincided with the launch and uptake of the National Rural Health Mission (NRHM).Conclusions: Even though India as a nation is not predicted to attain all the MDG 4 and 5 targets,at least four of its 15 most populous states are predicted to do so. In the past two decades, MMRreduction efforts were more effective than IMR reduction efforts.Key words: Achievement, infant mortality rate, maternal mortality ratio, Millennium DevelopmentGoals, net contribution, regression equation, target, India.aSouth Asia Network for Chronic Disease (SANCD), Public Health Foundation of India (PHFI), C1/52, First Floor, Safdarjung DevelopmentArea, New Delhi – 110 016bDepartment of Sociology Degree College, Aland, Gulbarga, Karnataka-585 101.Correspondence to Hanimi Reddy (email: hanimi.reddy@phfi.org)IntroductionGlobally, maternal and child mortality are indecline, although the pace of decline is notsufficient to attain Millennium DevelopmentGoals (MDGs) 4 and 5 for 128/137 developingcountries.1Due to slow progress in reducinginfant and maternal mortality and the moralurgency of reinvigorating efforts to tackle slowprogress; the United Nations (UN) launchedthe Global Strategy for Women’s and Children’sHealth in 2010.2As part of this strategy, Indiacommitted to spend US$ 3.5 billion annually,for strengthening maternal and child healthservices in 235 districts, which account fornearly 70% of infant and maternal deaths.2Book 1.indb 279 12-Dec-2012 9:02:24 AM
  • 2. WHO South-East Asia Journal of Public Health 2012;1(3):279-289280India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.In 2010, India recorded 56 000 maternal3and1.3 million infant deaths,4the highest for anycountry.5, 6, 7India’s MDG 4 target is to reduce IMR bytwo-thirds between 1990 and 2015, i.e., from80 infant deaths per 1000 live births in 1990to ‘28’ by 2015. Under MDG 4, another targetis to improve the proportion of one-year-oldchildren immunized against measles from42%8in 1992-1993 to 100% by 2015.9India’smain MDG 5 target is to reduce MMR by three-quarters between 1990 and 2015, i.e., from437 maternal deaths per 100 000 live births to‘109’,9while it has also committed to improvethe ‘proportion of births attended by skilledhealth personnel’. With only three years left toachieve MDGs 4 and 5 targets, there is a needto understand the progress made by India andas well as its 15 most populous states.To a large extent, India shapes the globalMDGs 4 and 5 targets, because of its share ofthe global burden of child (23%) and maternalmortality (19%).1, 3Moreover, during the pasttwo decades, the 15 most populous states,which account for 95% of India’s population,have made variable progress on infant and/ormaternal mortality reduction efforts.5State-wise analysis of IMR/MMR decline provides usan opportunity for learning which strategiesdid and did not work. In this context, thespecific objectives of the study are:To estimate percent net contribution of(1)the 15 most populous states and differentperiods, in shaping India’s IMR and MMRdecline;To fit linear and exponential regression(2)curves, and understand how IMR/MMRhas declined in India and in the 15 mostpopulous states;To use the fitted regression estimates,(3)extrapolate the year by which India’sMDGs 4 and 5 targets of IMR ‘28’ and MMR‘109’ would be achieved by India and the15 most populous states.MethodsQuantifying the contribution of states’decline during a time period upon overalldecline: We have partitioned the difference inIMR/MMR of a state between t1and t2into twocomponents: Component-1 is the differencedue to variations in state-specific IMR/MMR.Component-2 is the difference due to variationsin state-specific distribution of live births. Wemade this computation for each state by usingthe formula developed by Fleiss10and refinedby Buehler et al.11(equation-1).( + )P P1 2 ( + )R R1 22 2{[ x {[( – )R R1 2 ( – )P P1 2]+ ] x (1)where,P1andP2represent the proportion of live birthsin a state at t1and t2R1and R2represent IMR/MMR of a state at t1and t2We added two components of equation-1to arrive at ‘net-excess deaths at t2ascompared to t1’ for each state.10, 11We finallycalculated what percentage of the total net-excess deaths between t1and t2in India wascontributed by each of the 15 populous-states.Using Sample Registration System (SRS) data,periodic changes in IMR were measured forthe following durations: 1990 and 1996; 1996and 2001; 2001 and 2006; 2006 and 2010;and for the whole period 1990 and 2010.Using SRS data, periodic changes in MMRwere measured for the following durations:1997–1998 and 1999–2001; 1999–2001 and2001–2003; 2001–2003 and 2004–2006;2004–2006 and 2007–2009; and for the totalperiod 1997–1998 and 2007–2009.Estimation of trends in IMR/MMR:For understanding the trends in IMR/MMRBook 1.indb 280 12-Dec-2012 9:02:24 AM
  • 3. WHO South-East Asia Journal of Public Health 2012;1(3):279-289 281India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.decline, we fitted regression curves betweenIMR/MMR values and their reference datesusing Ordinary Least Square (OLS) method,as OLS offers a greater degree of objectivity,in the absence of outliers.12It is globallyassumed; IMR/MMR declines are non-linearand approximate to exponential.13, 14If thedecline is exponential, it reflects the fact thatmortality cannot keep declining linearly belowzero. However, there is no evidence that IMR/MMR decline is best modelled as exponential.15In contrast, if decline follows a linear pattern,IMR/MMR declines at a constant rate over adefined period and can decline below zero.Hence, we fitted linear and exponentialregression-curves, separately for India andfor 15 populous states, using SRS data. Chi-squared goodness-of-fit test was used forassessing the appropriateness of the fittedregression curve.For fitting regression curves of IMR,we used a moving average figure of threeconsecutive years as the IMR value for mid-year. Regression curves of IMR were fittedby using moving average IMRs for 21 years,during 1990–2010. By using regressionestimates, we have predicted IMR figure for2015, and/or the year by which it would reach‘28’ per 1000 live births. As MMRs are periodicestimates, we have measured interpolatedvalue of two consecutive periods as the MMRfor the mid-period using linear interpolation.16Regression curves of MMR were fit using nineMMR data points, during 1997–2009. Usingregression estimates, MMR figure for 2015and/or the year by which it would reach ‘109’were derived.Our analysis is confined to 15 states, asSRS does not provide MMR estimates forsmaller states/union territories, and these15 states (Andhra Pradesh, Assam, Bihar/Jharkhand, Gujarat, Haryana, Karnataka,Kerala, Madhya Pradesh/Chhattisgarh,Maharashtra, Orissa, Punjab, Rajasthan, TamilNadu, Uttar Pradesh/Uttarakhand, and WestBengal) cover approximately 95% of India’spopulation; hence changes in these states willfairly change the national scenario.ResultsChanges in MDG 4: Trends in IMR of Indiaand of the 15 most populous states - duringthe past four decades are shown in Figure 1.It depicts a uniform declining trend acrossthe states, although the pace of decline wasmore rapid during 1976–1991 and again in2006–2010. Decline was the highest in UttarPradesh/Uttarakhand, followed by Gujarat andTamil Nadu. Across all the states, excludingKerala - two types of declining trends werevisible: 1) higher decline when the rates arehigh (in early 1970s); and 2) a steady rate ofdecline during 2001–2010.Table 1 provides changes in IMR, percentnet contribution of the states and the periodsto overall decline in IMR of India, and predictedlevels of IMR. In the last two decades, IMR ofIndia has declined by around 40% and numberof infant deaths from around 2.2 mi llion to1.3 million.4Percent decline was ≥50% onlyin three out of 15 states (Maharashtra, TamilNadu and West Bengal), while absolute annualdecline was above the national average inOrissa, Madhya Pradesh/Chhattisgarh, UttarPradesh/Uttarakhand and West Bengal.If total net decline in IMR of India duringdifferent periods of 1990-2010 is consideredas 100% (last row of Table-1): maximumdecline occurred during 1990-1996 (36%)followed by 2001-2006 (26%). On the otherhand, if total decline in IMR of India during1990-2010 among the 15 populous states isconsidered as 100% (8th column of Table-1):Uttar Pradesh/Uttarakhand contributedmaximum (20%) to this decline; followedby Madhya Pradesh/Chhattisgarh (14%);West Bengal (11%); Maharashtra (9%);Book 1.indb 281 12-Dec-2012 9:02:24 AM
  • 4. WHO South-East Asia Journal of Public Health 2012;1(3):279-289282India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.etc. By comparing states’ per cent share intotal net decline in IMR of India with state’spercent share in live births – one can estimatewhether a particular state has contributed‘favourably to net decline’ (state’s percentshare in net decline > state’s percent sharein live births) or ‘unfavourably to net-decline’(state’s percent share in net decline ≤ state’spercent share in live births) in IMR of Indiaduring a particular period (4th to 8th columnvs 9th column, of Table-1). During 1990-2010,Madhya Pradesh/Chhattisgarh contributedmost favourably to the net decline (with itsshare of 10% to live births it has contributed14.4% to total net decline in IMR of India).Andhra Pradesh, Karnataka, Maharashtra,Orissa, Tamil Nadu and West Bengal werethe other states which contributed favourablyto net decline. In contrast, Bihar/Jharkhandcontributed most unfavourably to net decline.However, during 2006-2010, Bihar/Jharkhand,Madhya Pradesh/Chhattisgarh, Rajasthan andUttar Pradesh/Uttarakhand transitioned fromunfavourable to favourable states; contributing57% to net decline in IMR of India when theirshare to live births was 52% (Table 1).If the declining trend in IMR observedduring 1990-2010 continues linearly, India’sIMR would be 42 per 1000 live births (95%CI: 38-45) by 2015 and MDG 4 target level of‘28’ would be achieved in 2023–2024. If thedecline follows an exponential trend, India’sIMR would be 45 per 1000 live births (95%CI: 41–49) by 2015, and MDG 4 target wouldbe achieved in 2033–2034. Unless specialefforts are made to reduce IMR in Assam,Bihar/Jharkhand, Haryana, Rajasthan andUttar Pradesh/Uttarakhand – it may takeat least up to 2023–2024 for India to reachthe MDG 4 target, and much longer for theaforementioned states.Book 1.indb 282 12-Dec-2012 9:02:25 AM
  • 5. WHO South-East Asia Journal of Public Health 2012;1(3):279-289 283India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.Table1:State-wisechangesinIMR,contributionofstatesandperiodstoIndiasIMRdeclineandyearofachievingIndiasMDG-4targetChangesinIMR1between1989-91(1990)and2009-10(2010)%NetcontributionofthestatestoperiodicdeclineinIMRofIndiabetween:%statessharetototallivebirthsduring1990–2010IfIMRdeclinefollowsalineartrendIfIMRdeclinefollowsanexponentialtrendState%DeclineAnnualabsolutedecline1989–91(1990)and1995–97(1996)1995–97(1996)and2000–2002(2001)2000–02(2001)and2005–07(2006)2005–07(2006)and2009–10(2010)1989–91(1990)and2009–10(2010)ExpectedIMRin2015(95%CI)YearofachievingMDG2-4targetIMRlevelof28ExpectedIMRin2015(95%CI)YearofachievingMDG2-4targetIMRlevelof28AndhraPradesh36.41.48.86.47.46.37.56.445.3(42.6-48.1)2031-3247.1(41.5-52.7)2044-45Assam28.11.21.31.02.12.61.82.856.7(54.7-58.7)2043-4458.0(53.1-62.9)2065-66Bihar/Jharkhand36.11.40.811.8-1.216.55.713.247.2(44.8-49.7)2032-3349.0(43.8-54.1)2045-46Gujarat39.21.56.5-0.83.74.74.35.042.7(39.6-45.8)2028-2944.4(38.9-50.0)2039-40Haryana32.21.20.31.52.12.01.42.247.3(44.7-49.8)2032-3349.9(42.8-55.1)2046-47Karnataka47.81.611.0-1.05.44.76.34.732.4(28.9-35.9)2018-1936.2(29.4-43.0)2023-24Kerala30.60.31.01.1-0.20.30.52.110.5(9.0-12.0)1988-8910.8(7.6-14.0)1988-89MadhyaPradesh/Chhattisgarh43.92.517.319.110.711.314.410.052.5(51.0-54.1)2025-2658.0(52.7-63.3)2041-42Maharashtra50.01.58.412.110.85.58.88.122.4(20.8-23.9)2010-1126.5(21.4-31.6)2010-11Orissa47.32.98.713.17.24.17.83.548.6(45.5-51.7)2022-2356.0(49.9-62.1)2037-38Punjab39.31.21.91.72.81.82.12.034.8(32.3-37.2)2021-2236.5(31.2-41.8)2027-28Rajasthan34.01.5-4.35.09.27.43.46.754.0(50.0-58.1)2032-3356.6(47.6-65.6)2050-51TamilNadu57.61.84.55.68.25.15.84.521.2(18.2-24.2)2009-1026.4(19.0-33.8)2009-10UttarPradesh/Uttarakhand40.82.120.211.021.222.219.622.153.6(51.1-56.1)2029-3057.2(53.0-61.4)2046-47WestBengal54.51.913.712.510.75.410.76.722.1(20.5-23.7)2012-1327.7(22.3-33.1)2014-15India42.11.8100.0100.0100.0100.0100.0100.041.5(38.4-44.7)2023-2445.0(41.0-49.0)2033-34%Net-contributionofdifferentperiodstoIndiasIMRdropduring1990and2010----1202.8(35.5%)485.5(14.3%)871.8(25.7%)826.5(24.4%)3386.6(100.0%)----------1:IMR:Infantmortalityrateper1000livebirths:three-yearmovingaveragesforalltheyears,exceptfor2010whichisatwo-yearaverage(2009,2010)2:MDG:MillenniumDevelopmentGoalBook 1.indb 283 12-Dec-2012 9:02:25 AM
  • 6. WHO South-East Asia Journal of Public Health 2012;1(3):279-289284India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.Measles immunization rates: India isdoing well on the other MDG 4 indicator, asthe percentage of 12–23 month old childrenimmunized against measles improved from42% in 1992–19938to 74% by 2009.17Goodness of fit of linear and exponentialregression curves on IMRs for India during1990-2010 were appropriate, with respectivechi-square values of 0.79 (p<1.000) and 1.36(p<1.00).Changes in MDG 5: Figure 2 depictschanges in MMRs of India and of 15 populousstates, during 1997–2009. During this period,there was a precipitous decline in MMR of Indiafrom 398 to 212 per 100 000 live births, eventhough variations in the base (1997–1998)MMR levels of states were mainly responsiblefor recent (2007–2009) variations.Table 2 provides changes in MMR, percentnet contribution of 15 populous states andperiods to overall decline in MMR of India,and predicted levels of MMR. During 1997–2009, MMR of India declined by 47%, withan annual absolute decline of 15.5 points.Maternal deaths decreased from around100 000 to 60 000, assuming MMRs providedby SRS are correct. If total decline in MMRof India during different periods of 1997–2009 is considered as 100%, the maximumdecline occurred between 2004-2006 and2007–2009 (32%), followed by 2001-2003and 2004–2006 (27%). Uttar Pradesh/Uttarakhand contributed most favourably tonet-decline in MMR of India (with its shareof 18% to the live births, contributed 33%to MMR decline). Bihar/Jharkhand, MadhyaPradesh/Chhattisgarh, Rajasthan and Assamalso contributed favourably to net decline.It is encouraging that between 2004-2006and 2007–2009, Bihar/Jharkhand, MadhyaPradesh/Chhattisgarh, Rajasthan and UttarPradesh/Uttarakhand – together contributedBook 1.indb 284 12-Dec-2012 9:02:25 AM
  • 7. WHO South-East Asia Journal of Public Health 2012;1(3):279-289 285India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.Table2:State-wisechangesinMMR,contributionofstatesandperiodstoIndiasMMRdeclineandyearofachievingIndiasMDG-5targetChangesinMMR1between1997-2001and2006-09%NetcontributionofthestatestoperiodicdeclineinMMRofIndiabetween:%statessharetototallivebirthsduring1997and2009IfMMRdeclinefollowsalineartrendIfMMRdeclinefollowsanexponentialtrendState%DeclineAnnualabsolutedecline1997–98and1999–011999–01and2001–032001–03and2004–062004–06and2007–091997–98and2007–09ExpectedMMRin2015(95%CI)YearofachievingMDG2-5targetMMRlevelof109ExpectedMMRin2015(95%CI)YearofachievingMDG2-5targetMMRlevelof109AndhraPradesh32.05.3–2.87.97.22.53.47.8101(57–144)2014115(83–147)2017–18Assam31.314.814.2–10.10.85.63.72.7365(244–486)2046374(295–453)2082–83Bihar/Jharkhand50.822.537.93.415.015.218.611.0119(81–156)2016184(172–196)2024–25Gujarat29.55.20.75.01.21.21.75.1107(82–131)2015116(106–126)2017–18Haryana–12.5–1.4–2.20.5–1.01.4–0.32.1158(117–199)2050159(141–177)2073–74Karnataka27.35.6–2.56.62.43.32.25.4151(114–188)2022162(135–188)2029–30Kerala46.05.8–0.14.41.10.61.23.338(15–60)200358(41–75)2003–04MadhyaPradesh/Chhattisgarh39.014.36.714.07.814.210.68.4199(171–228)2022231(205–257)2035–36Maharashtra37.35.22.05.24.04.33.89.878(56–100)200990(73–107)2008–09Orissa25.47.3–5.010.74.53.42.93.7216(119–313)2026234(166–302)2039–40Punjab38.69.06.70.10.20.61.92.5120(52–187)2017132(89–174)2021–22Rajasthan37.415.8–0.212.09.69.57.65.8232(181–283)2023270(228–312)2038–39TamilNadu26.02.8–5.27.43.51.31.36.377(34–119)200785(55–114)2007–08UttarPradesh/Uttarakhand40.820.631.526.534.436.533.018.0253(219–287)2023300(268–333)2040–41WestBengal52.113.218.26.59.20.38.18.235(0–78)201084(67–102)2012–13India46.715.5100.0100.0100.0100.0100.0100.0117(109–125)2016159(153–166)2023–24%Net-contributionofdifferentperiodstoIndiasMMRdropduring1997and2009––3654.0(24.2%)2545.7(16.8%)4141.9(27.4%)4770.8(31.6%)15112.4(100.0%)–––––1:MMR:Maternalmortalityratioper100,000livebirths:two/threeyearmovingaverage2:MDG:MillenniumdevelopmentgoalBook 1.indb 285 12-Dec-2012 9:02:26 AM
  • 8. WHO South-East Asia Journal of Public Health 2012;1(3):279-289286India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.75% to the net decline in India’s MMR, whentheir collective share in live births was only43%.If the declining trend in MMR observedduring 1997-2009 continues linearly, Indiawill be very close to achieving the MDG 5target level of ‘109’ by 2016, as the MMRis predicted to be 117 (95% CI: 109–125)in 2015. However, if the decline follows anexponential trend, India’s MMR would be 159(95% CI: 153–166) in 2015, and MDG-5target would only be reached in 2023–2024.Kerala, Tamil Nadu and Maharashtra hadalready reached MMR of 109 by 2009, and ifthe same pace of decline continues linearlyor exponentially, Andhra Pradesh, Gujaratand West Bengal have a good chance oftouching 109 MMR, by 2017-2018. India isdoing well on ‘percent deliveries attended byskilled health personnel’, with an improvementfrom 33% in 1992-19938to 76% by 2009.17Goodness-of-fit of linear regression curve onMMRs of India during 1997–2009 suggestsmoderate deviation from the observed values[chi-squared=0.2 (p<0.032)], while theexponential fit suggests negligible deviation[chi-squared=0.36 (p<0.055)].DiscussionIf the pace of linear decline in MMR during1997-2009 continues, India will be veryclose to attaining MDG-5 target level of ‘109’per 100 000 live births by 2016. However,India’s IMR would be hovering around 42in 2015 if the decline follows a linear trend,and the MDG 4 target level of ‘28’ would onlybe achieved in 2023-2024. If the declines inIMR/MMR follow exponential trends, reachingMDGs 4&5 targets gets further postponed byIndia and most states. India’s MMR declineduring 1997-2009 may mainly be attributed tofavourable contributions from Uttar Pradesh/Uttarakhand, Bihar/Jharkhand, MadhyaPradesh/Chhattisgarh, and Rajasthan. Due tounfavourable contributions in IMR reductionefforts by Bihar/Jharkhand, Uttar Pradesh/Uttarakhand, Rajasthan and Assam, India’srun for MDG-4 target is delayed.How comparable are our IMR/MMRpredictions with others? Were the state-specific changes in IMR/MMR during aparticular period reflected in national levelchanges in a standardized way or not? Whichone of the two regression curves (linear/exponential) Indian IMRs/MMRs fits better?Due to paucity of reliable MMR estimatesfor India/states prior to 1997,18predictionswere based on IMR. A recent Lancet series,5projected India’s IMR and MMR in 2015 as 43and 153 respectively, while another report19projected India’s IMR in 2015 as 46–49.These predictions matched with our IMRpredictions. Our IMR extrapolations for 2015are in synchronization with predictions of theCentral Statistical Organization, Governmentof India.14For the states of Assam, Orissa,and Rajasthan, there is good comparabilitybetween our MMR predictions and the AnnualHealth Survey (AHS, 2010-2011).20For MMR,our exponential regression-based predictionsmatch closely with other findings.3,7Lozano et al.,1in their paper on trackingthe progress of MDGs 4 and 5 in 163 countriespredicted India’s IMR and MMR in 2011respectively as 49 (95% CI: 41–56) and187 (95% CI: 142–238), while our linearand exponential regression-based estimatesrespectively were 48 (95% CI: 47–49) and 50(95% CI: 46-54); and 178 (95% CI: 173–184)and 196 (95% CI: 195–197). As compared toreferred predictions1,3,4,5,7,14,18,19ours are morerobust due to the use of more recent IMR/MMR data, moving-average IMR estimates,and predicted IMRs/MMRs through the use oflinear and exponential regression curves.A trend analysis of IMRs in India during1970–2000 by Saikia et al.,21concluded thatthe decline was much steeper during the 1970sBook 1.indb 286 12-Dec-2012 9:02:26 AM
  • 9. WHO South-East Asia Journal of Public Health 2012;1(3):279-289 287India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.and 1980s, and that IMR had stagnated during1996–2000, 2000–2004 and 2002–2006. Wealso noticed that the decline in IMR was thelowest during 1996–2001. Our decompositionof the decline in India’s IMR among the statesalso matched this finding.21. Analysis of India’sIMRs during 1981–19976found that the declinetended to stagnate for brief periods and wasoften followed by a subsequent rapid decline.We are of the view that India’s IMR declinestill followed a similar phenomenon, as wenoted that a plateau during 1996–2001 wasfollowed by a rapid decline during 2001–2006(Table 1). Our findings agree with previousstudies for Kerala, Maharashtra, Tamil Nadu,and West Bengal as the only states likely toachieve MDG 4 and 5 targets.5,14Contribution of state-specific changesin rate/ratio onto national-level change:If populations are similar with respect tofactors associated with the event under study,there is no problem in comparing eventsacross states. If populations are not similarlyconstituted, direct comparison of the overallevents may be misleading.10In the presentanalysis, rather than measuring changes inIMR/MMR of India between two time points asmere percent/absolute change, we estimatedit as: net-effect of state-wise distribution oflive births and state-specific mortality rates/ratios - a technique used extensively22,23forunderstanding birth weight-specific or regionaldifferences in mortality.Linear versus exponential regression?While estimating the progress made bydifferent countries including India towardsMDGs 4 and 5, linear regression curveswere used.1,5However, for understandingthe decline in IMRs of India and Nepal in thepast three decades, exponential regressioncurves were found to be more appropriate.13IMR decline in 18 Western nations in the 20thcentury was, for the most part, neither linearnor exponential.15For India’s IMR declineduring 1990-2010, both linear and exponentialregressions fitted well, while for MMR declines,exponential regression fits better than linear.As India’s IMR/MMR decline during the pasttwo decades is an outcome of heterogeneousprogress made by different states, sometimeslinear and sometimes exponential, and alsotaking into account high IMR/MMR levels in amajority of the states - it is difficult to concludewhich of the two regression curves is moreappropriate for the Indian scenario.Strengths/limitations: Our analysisincluded the latest IMRs/MMRs from SRS,the most reliable source for national andstate specific estimates. For fitting regressioncurves, we used moving averages or periodicestimates, instead of point estimates. Oneof the limitations of this study is that ouranalysis/interpretation relied completely onthe quality and completeness of SRS data. Anevaluation of SRS data showed omission ratesof 1.8% for births and 2.5% for deaths.24IMRsof SRS are considered to be robust,6and theymatched closely with all the three NationalFamily Health Survey (NFHS)25estimates, forIndia. However, IMRs of SRS in 2010 werelower than AHS estimates by 1–10 absolutepoints, in eight states.20India’s MMRs of SRSdeviated substantially with NFHS-226and UN18estimates. As MMR estimates of UN are usuallyindirect estimates,1,3,18we are of the opinionthat SRS estimates are still robust for India.Is maximum decline in MMR between2004-2006 and 2007-2009 due to NRHMor an artefact? Periodic analysis of MMRdecline in India during 1997–2009 indicatedmaximum drop between 2004–2006 and2007–2009. Is this finding an artefact orinfluence of NRHM? Lim et al.,27indicated thatIndia’s conditional cash transfer scheme ‘JananiSuraksha Yojana (JSY)’ of NRHM contributedto an increase in institutional deliveries,17andwas associated with reduction of about fourperinatal and two neonatal deaths per 1000Book 1.indb 287 12-Dec-2012 9:02:26 AM
  • 10. WHO South-East Asia Journal of Public Health 2012;1(3):279-289288India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.live births, while having no significant effect onmaternal mortality.27Since its launch in 2005,NRHM is credited with deploying more than750 000. Accredited Social Health Activists(ASHAs), as change agents between thewomen and the health system, resulting inimproved prospects for maternal and newborncare5as compared to the past.Conclusions: Even though the pace ofdecline in IMR accelerated during 2001–2010after a period of stagnation (1996–2001), Indiais still predicted to fall short of achieving itsMDG 4 target level of ‘28’ per 1000 live birthsby 2015, in 11 out of the 15 most populousstates, and in India as a nation. Assam, Bihar/Jharkhand, Rajasthan and Uttar Pradesh/Uttarakhand, need to put special efforts foraccelerating decline in IMR. Considering thepace of MMR decline during 1997–2009, sixout of the 15 most populous states have afairly good chance of attaining India’s MDG5 target level of ‘109’, albeit two–three yearsbehind schedule (2017–2018).AcknowledgementsWe sincerely acknowledge the active guidanceand support of Prof. Shah Ebrahim andDr Fiona Taylor. We thank Dr Ulla Sovio,Dr Jai and Dr Preet Dhillon for reviewing themanuscript and Prof PM Kulkarni and Prof RMPandey for providing valuable inputs.ReferencesLozano R, Wang H, Foreman KJ, Rajaratnam JK,1.Naghavi M, Marcus JR, et al. Progress towardsMillennium Development Goals 4 and 5 on maternaland child mortality: an updated systematic analysis.The Lancet. 2011; 378(9797): 1139-65.United Nations. Global strategy for women’s and2.children’s health. New York: UN, 2010. Availableat http://www.un.org/sg/hf/global_strategy_commitments.pdf. Accessed on 20.01.12WHO, UNICEF, UNFPA, and the World Bank estimates3.(2012). Trends in Maternal Mortality: 1990 to 2010.Geneva, World Health Organization.United Nations Children’s Fund.4. Levels & trendsin child mortality: estimates developed by the UNInter-agency Group for child mortality estimation.New York: UNICEF, 2011. http://www.childinfo.org/files/Child_Mortality_Report_2011.pdf - accessed 1August 2012.Paul VK, Sachdev HS, Mavalankar D, Ramachandran5.P, Sankar MJ, Bhandari N, et al. Reproductive health,and child health and nutrition in India: meeting thechallenge. The Lancet. 2011 Jan 22; 377(9762):332-49.Claeson M, Bos ER, Mawji T, Pathmanathan I. Reducing6.child mortality in India in the new millennium. BullWorld Health Organ. 2000; 78(10): 1192-9.Chatterjee A, Paily VP. Achieving Millennium7.Development Goals 4 and 5 in India. BJOG. 2011;118(Suppl 2): 47–59.International Institute for Population Sciences (IIPS).8.National family health survey (MCH and FamilyPlanning): India 1992-93. Bombay: IIPS, 1995.Government of India (GOI). National health profile9.of India 2010. New Delhi, 2010.Fleiss JL, Levin B, Paik MC.10. Statistical methods forrates and proportions. 3rd edn. New York: John Wiley& Sons, 2003.Buehler JW, Kleinman JC, Hogue CJ, Strauss LT, Smith11.JC. Birth weight-specific infant mortality, UnitedStates, 1960 and 1980. Public Health Rep. 1987Mar-Apr; 102(2): 151-61.United Nations Children’s Fund. Background note12.on methodology for under-five mortality estimation.New York: UNICEF.Thapa S. Declining trends of infant, child and under-13.five mortality in Nepal. J Trop Pediatr. 2008 Aug;54(4): 265-8.Central Statistical Organization, Ministry of Statistics14.and Programme Implementation, Government of India.2009. Millennium Development Goals – India countryreport. 2009. Mid-Term Statistical Appraisal. http://www.searo.who.int/EN/Section1243/Section1921/Section1924.htm - accessed 5 September 2012.Bishai D, Opuni M. Are infant mortality rate declines15.exponential? The general pattern of 20th centuryBook 1.indb 288 12-Dec-2012 9:02:26 AM
  • 11. WHO South-East Asia Journal of Public Health 2012;1(3):279-289 289India’s progress towards the Millennium Development Goals 4 and 5 on infantand maternal mortalityHanimi Reddy et al.infant mortality rate decline. Popul Health Metr.2009; 7: 13.Linear interpolation equation calculator.16. http://www.ajdesigner.com/phpinterpolation/linear_interpolation_equation.php - accessed 1 August2012.United Nations Children’s Fund. 2009 coverage17.evaluation survey: national fact sheet. New Delhi:UNICEF India Country Office, 2009. http://www.unicef.org/india/National_Fact_Sheet_CES_2009.pdf - accessed 1 August 2012.United Nations Development Programme. Fact sheet:18.goal 5: improve maternal health. New York: UNDP,2008.Ram F, Mohanty SK, Usha R.19. Progress and prospectsof Millennium Development Goals in India. Mumbai:International Institute for Population Sciences,2009.Government of India. Annual health survey bulletins20.2010-11. New Delhi, 2011. http://censusindia.gov.in/vital_statistics/AHSBulletins/ahs.html - accessed7 Aug 2012.Saikia N, Singh A, Ram F. Has child mortality in India21.really increased in the last two decades? Economicand Political Weekly. 2010; XLV (51): 62-70.State of Georgia. Georgia epidemiology report:22.Quantifying the effect of low birth weight onracial differences in infant mortality rate, Georgia:Department of Public Health, 1989.Allen DM, Buehler JW, Hogue CJ, Strauss LT, Smith23.JC. Regional differences in birth weight-specific infantmortality, United States, 1980. Public Health Rep.1987 Mar-Apr; 102(2): 138-45.Registrar General of India. Report on intensive24.enquiry conducted in sub-sample of SRS units.Occasional paper No. 1. New Delhi: Office of theRegistrar General of India, Ministry of Home Affairs,1988.International Institute for Population Sciences25.(IIPS) and Macro International. National familyhealth survey (NFHS-3), 2005–06: India: Volume I.Mumbai: IIPS, 2007.International Institute for Population Sciences (IIPS)26.and ORC Macro. National family health survey(NFHS-2), 1998–99: India. Mumbai: IIPS, 2000.Lim SS, Dandona L, Hoisington JA, James SL, Hogan27.MC, Gakidou E. India’s Janani Suraksha Yojana, aconditional cash transfer programme to increasebirths in health facilities: an impact evaluation. TheLancet. 2010 Jun 5; 375(9730): 2009-23.Book 1.indb 289 12-Dec-2012 9:02:26 AM