Indian J Pediatr DOI 10.1007/s12098-014-1426-3
COMMENTARY
Decline in Immunization Coverage Across Well-performing Districts in India: An Urban Conundrum? Rajib Dasgupta & Purnamita Dasgupta & Ankush Agrawal
Received: 1 December 2013 / Accepted: 19 March 2014 # Dr. K C Chaudhuri Foundation 2014
Introduction Inequities in immunization persist in India in spite of the established importance of the national immunization program in preventing deaths, disability and morbidity arising from vaccine preventable diseases (VPDs) [1]. The role of programmatic complexities, political and social contexts as determinants of immunization has been explored in understanding why these inequities exist [2]. The DLHS3 reported a decline in the proportion of fully immunized children (12–23 mo) in several states which had been reporting high overall state level immunization coverage [3]. Possible explanations include the adverse impact of the polio eradication campaign, and social resistance in some states such as Tamil Nadu and Kerala due to reports of deaths and Adverse Events Following Immunization (AEFI) [4]. These explanations seem inadequate as the states reporting declines in full immunization coverage were nonendemic for polio and had only one or two pulse polio rounds annually, while the scale of AEFI is not large enough [5]. In this study the authors apply an alternative methodological approach, namely, the area effects framework, to analyze a combination of individual and socio-economic factors that can effectively explain the variations in the observed decline in district-level immunization coverage. The area effects framework, consisting of compositional, collective and contextual factors was applied to explain variations in immunization coverage observed at the district level R. Dasgupta (*) Centre of Social Medicine & Community Health, Jawaharlal Nehru University, New Delhi 110067, India e-mail:
[email protected] P. Dasgupta Institute of Economic Growth, University of Delhi, Delhi, India A. Agrawal Indian Institute of Technology of Delhi, Delhi, India
between the two rounds of the DLHS survey (rounds 2 and 3) [6]. The factors selected were (i) household factors: educational attainment and poverty status; (ii) factors that influence local physical and social environment: distance from nearest town, availability and accessibility of health infrastructure, and urbanization; (iii) community-level contextual factors: religion, caste and tribe. The empirical analysis uses a logistic regression to identify the significant explanatory factors that can explain the observed decline in coverage. Empirical district level data on explanatory factors was taken from DLHS3 (2007–8). Good governance at state level is expected to play a key role in successful implementation of the immunization program. Consequently, nine states were selected for the analysis where several districts have reported declines, despite being states deemed to have good governance in terms of their performance in several socio-economic aspects such as infrastructure availability, law and order, judicial services, and educational achievements [7]. These states are Punjab, Haryana, Andhra Pradesh, Tamil Nadu, Gujarat, Karnataka, Maharashtra, Kerala, and Himachal Pradesh. The analysis was done for all the districts (206 districts) in these nine states. Districts which report a decline in immunization coverage between DLHS2 (2002–04) and DLHS3 (2007–08) were defined as decline districts and the rest served as the control districts in each selected state. The proportion of fully immunized children as per the Government of India’s national immunization program was the indicator used for judging the coverage achieved in each district. Nearly 58 % (119) districts of a total of 206 districts in these nine states reported a decline in coverage of fully immunized children in DLHS3 as compared to DLHS2. Within a state, the highest proportion of districts reporting a decline was from Tamil Nadu at 87 %, where 26 out of the 30 districts covered in the survey reported a decline (Fig. 1). In two of the nine states, Punjab and Maharashtra, disparity in coverage across districts increased between DLHS2 and 3. The extent
Indian J Pediatr 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0
90 80 70 60 50 40 30 20 10 0
No. of districts with decline
Total districts
Districts with decline (%)
Fig. 1 Decline of full immunization across districts in selected states. The vertical axis on the left indicates the total number of districts and the number of districts where decline has been observed and the vertical axis on the right, the share of districts where decline has occurred. Sources: authors’ computations based on DLHS3 and DLHS2
of decline among districts reporting decline in immunization coverage within a state was also substantial: in Maharashtra it varied from 17 % (lowest reported decline in a district) to 92 % (highest reported decline in a district); followed by Haryana with a range of 11–79 %. The signs and significance levels of the coefficients obtained for the explanatory variables in the index results (Table 1) were as per expectations, as explained below. As the proportion of population in the poorest wealth quintile increased, the probability of experiencing a decline in coverage increased. The probability of decline expectedly reduced as the proportion of villages with a PHC increased. In this specification, when districts across the country were pooled together, the proportion of Muslims and the educational level of the head of the household were however not significant determinants of Table 1 Results of logit regression Variable
Coefficient t value
Proportion of households in the poorest wealth −0.05* quintile in the district Proportion of Muslims −0.02 Proportion of SC and ST 0.02** Proportion of households where respondent is −0.02 educated till at least primary level Proportion of villages located at a distance of 0.02** more than 10 km from the nearest town Proportion of villages with a primary health centre 0.02* Urbanization level (2001) −0.02* Constant 0.13 Number of observations (districts) 206 Chi-square 23.89
−1.97 −0.83 1.7 −1.32 1.94 2.06 −2.16 0.1
The dependent variable for a district is defined as 0 if there is decline in the immunization coverage between DLHS2 and DLHS3 and 1 otherwise. The p-value associated with the Chi-square test is 0.001. * and ** indicate a statistically significant coefficient at 10 and 5 % levels of significance respectively
coverage outcomes. As the proportion of SC/ST population increased, the probability of decline in coverage reduced. This finding can be attributed to special provisions made for reaching these sub-populations (e.g., coverage through outreach services and additional ANMs under Tribal Sub-Plan). Of more interest was the finding that as the proportion of villages in the district which were located at a distance of more than 10 km from the nearest town increased, the probability of decline in immunization coverage decreased; confirming the disadvantages of proximity to towns. Districts with higher levels of urbanization thus had a higher probability of showing a decline in immunization coverage. Decline in immunization is a cause for alarm since it is contrary to the country’s goal of achieving 100 % coverage for six VPDs. It also poses a challenge in achieving the Global Vaccine Action Plan’s (GVAP) goal of 90 % coverage at the national level and 80 % coverage for every district by 2015 [8]. This analysis is an important contribution towards recognizing the need for focusing on specific areas and groups of population who may otherwise get left out with too much focus on only state level average figures. Fast paced urbanization emerges as a ‘risk factor’ where the emergence of peri-urban areas and small towns [9] with relatively poor health access offers a rational explanation to the observed phenomenon of decline in immunization coverage. Often such urban areas are an extension of existing towns and cities, inhabited by migratory urban poor, and characterized by vulnerability due to concentration of population and social diversity which co-exists with rural traits of isolation and invisibility to policy makers [10]. The GVAP recognizes that coverage can also be very low in settlements of the urban poor, especially in cities with transitory migrant populations and that new strategies for reaching the urban poor and urban migrants will also be necessary. In the specific context of health service provision in urban India, peri-urban areas are not covered since they fall outside the purview of municipalities while small towns (with less than 50,000 population) are not covered within the National Urban Health Mission (NUHM). Yet, Census data shows that over 70 % of towns in these nine states are small towns with less than 50,000 population (except in Andhra Pradesh) and are consequently left out of NUHM coverage. The paper highlights that urbanization brings specific and hitherto unrecognized challenges to the sustainability of an ongoing program; state level aggregates on immunization coverage can mask alarming trends in lack of coverage among specific districts, even among those that are conventionally understood to be the best performers. These findings have direct implications for urban health planning and programs such as NUHM, as well as for meeting measles elimination and rubella control targets.
Indian J Pediatr Contribution Dr. Narendra K Arora, Executive Director, The INCLEN Trust International & CHNRI, India will act as guarantor for this paper. Conflict of Interest None. Role of Funding Source None.
References 1. Dutta AK, Ray S. Immunization in practice—clearing the cobwebs. Indian J Pediatr. 2013;80:404–10. 2. Mathew JL. Inequity in childhood immunization in India: a systematic review. Indian Pediatr. 2012;49:203–23. 3. International Institute of Population Sciences (IIPS). District level household and facility survey (DLHS-3). 2007–08, India. Mumbai. 4. UNICEF. Effective vaccine management assessment in Tamil Nadu http://www.unicef.org/india/TN-Draft2.pdf (2012). Accessed 25 Jul 2013.
5. Dasgupta R, Arora NK. An international perspective on vaccine safety. In: Chatterjee A, editor. Vaccinophobia and vaccine controversies of the 21st century. New York: Springer; 2013. p. 419–38. 6. MacIntyre S, Ellaway A, Cummins S. Place effects on health: how can we conceptualise, operationalise and measure them? Soc Sci Med. 2002;55:125–39. 7. Mundle S, Chakraborty P, Chowdhury S, Sikdar S. The quality of governance: how have Indian States performed? Working paper No. 2012–104. Delhi: National Institute of Public Finance and Policy; 2012. p. 1–104. 8. GVAP. Annex 1, recommended indicators. global vaccine action plan 2011–2020, WHO. http://www.who.int/immunization/global_vaccine_ action_plan/GVAP_Annex1.pdf?ua=1 (2013). Accessed 23 Jan 2014. 9. Iaquinta DL, Drescher AW. Defining the peri-urban: rural– urban linkages and institutional connections. Land reform, economic and social development department, food and agriculture organization of the United Nations (FAO). http://www. fao.org/docrep/003/x8050t/x8050t02.htm (2000). Accessed 15 October 2013. 10. Pelling M, Mustafa D. Vulnerability, disasters and poverty in desakota systems. Political and development working paper series, number 31. London: King’s College London; 2010. p. 1–26.