WHO South-East Asia Journal of Public Health
  • 321
  • Home
  • Print this page
  • Email this page
Home About us Editorial board Search Ahead of print Current issue Archives Submit article Instructions Subscribe Contacts Login 

 Table of Contents  
Year : 2012  |  Volume : 1  |  Issue : 4  |  Page : 446-456

Institutionalizing district level infant death review: an experience from southern India

1 Health Specialist, United Nations Children’s Fund, Field Office, Hyderabad, India
2 Senior Consultant, Child Health, United Nations Children’s Fund, Bangalore, India
3 Mission Director, National Rural Health Mission, Karnataka, India
4 Professor, Department of Community Medicine, Sri Manakuala Vinayagar Medical College and Hospital, Pondicherry, India
5 Professor, Department of Community Medicine, Mahatma Gandhi Institute of Medical Sciences, Sewagram, India

Date of Web Publication25-May-2017

Correspondence Address:
Sanjeev Upadhyaya
Health Specialist, United Nations Children’s Fund, Field Office, Hyderabad
Login to access the Email id

DOI: 10.4103/2224-3151.207047

PMID: 28615610

Rights and Permissions

Background: An Infant Death Review (IDR) programme was developed and implemented in two districts of Karnataka.
Objective: We explored the processes that led to the development of the IDR programme with a view to improving the existing pilot programme and to ensuring its sustainability.
Methods: A sequential mixed-methods design was followed in which quantitative data collection (secondary data) was followed by qualitative data collection (in-depth interviews). Quantitative data were entered using EpiInfo (version 3.5.1) software and qualitative data were analysed manually.
Results: Apart from ascertaining the cause of infant deaths, the IDR Committee discusses social, economic, behavioural and health system issues that potentially contribute to the deaths. As a result of the IDR programme, key actors perceived an improvement in infant death reporting at district level, the development of a rapport with the local community, and elaboration of a feedback system for corrective actions. This has led to improved health care during pregnancy.
Conclusions: We found that involvement of the different stakeholders in planning and implementing the IDR programme offered a platform for collective learning and action. Impediments to the success of the programme need to be addressed by corrective actions at all levels for its future sustainability.

Keywords: Verbal autopsy, infant death, rural, district, India.

How to cite this article:
Upadhyaya S, Shetty S, Kumar SS, Dongre A, Deshmukh P. Institutionalizing district level infant death review: an experience from southern India. WHO South-East Asia J Public Health 2012;1:446-56

How to cite this URL:
Upadhyaya S, Shetty S, Kumar SS, Dongre A, Deshmukh P. Institutionalizing district level infant death review: an experience from southern India. WHO South-East Asia J Public Health [serial online] 2012 [cited 2022 Jan 22];1:446-56. Available from: http://www.who-seajph.org/text.asp?2012/1/4/446/207047

  Introduction Top

The infant mortality rate (IMR) is universally regarded as an important indicator of the health and economic status of communities, and the effectiveness of maternal and child health services. According to the Sample Registration System (SRS) of India, IMR in Karnataka, a southern state of India, was 38 per 1000 live births in 2010.[1] India’s National Population Policy (2000) envisions reducing IMR to less than 30 per 1000 live births,[2] echoing Millennium Development Goal 4 to reduce under-five mortality by two thirds between 1990 and 2015.[3] Cause-of-death data are critical to formulating good public health programmes; developing national, regional, and global policies; and implementing and evaluating public health action.[4],[5] In developing countries, most deaths are neither attended by physicians nor certified medically, so two thirds to three quarters of the world’s population remain outside any systematic mortality surveillance.[6],[7],[8],[9],[10] Studies in Asia and Africa in the 1950s and 1960s used systematic interviews with physicians to assess causes of death,[11] a technique that the Narangwal project in India later labelled verbal autopsy (VA); the method subsequently evolved, particularly during the 1970s when the World Health Organization (WHO) discussed lay reporting of health information by people with no medical background.[11],[12],[13],[14] Today, VA remains the best available approach to assess the cause of death in communities in which most deaths occur at home. Separate questionnaires for neonatal, child, maternal, and other adult deaths have been suggested.[15] The need for an infant death audit system was therefore crucial to map the rise or decline in infant deaths, to identify common causes of death, as well as for prioritization of actions, fund allocation and monitoring and evaluation.

The United Nations Children’s Fund (UNICEF), Hyderabad, India in partnership with the Department of Health and Family Welfare Services, Karnataka, has been implementing a pilot Infant Death Review (IDR) programme to strengthen the district health system review of infant deaths. So far, the research on verbal autopsy has focused on technical aspects such as the development of standardized tools, validated algorithms and computerized algorithms.[14],[16] However, for VA methods to extend their reach successfully from research to routine application, feasibility demonstration at local level and implementation research have been suggested to resolve emerging challenges as the programme develops.[17] Hence, we explored the processes leading to the development of the IDR programme within the health system, and the perspectives of key programme actors at various levels. We hope that our findings may be useful to refine the pilot programme, and for health-care systems that wish to implement a similar programme in a developing country context.

  Material and methods Top


The present study on the Infant Death Review programme was carried out in Raichur and Dakshina Kannada districts of Karnataka. These two districts were chosen in consultation with the Director of the National Rural Health Mission, Karnataka, based on the performance of its health system. Raichur district, located 413 km north of the state capital Bangalore with a population of 1 924 773, has poor health indicators and a female literacy rate of 49.6%. On the other hand, Dakshina Kannada, 347 km west of Bangalore, has a population of 2 083 625 with better health indicators and 84% female literacy. The crude birth rate of Karnataka was 19.2 per 1000 population with 34.1% home deliveries.[1],[18] This review was done during February and March 2012.


A sequential mixed-methods design was followed where quantitative data collection (secondary data) was followed by qualitative data collection (in-depth interviews).

To begin with, verbal autopsy forms and the minutes of meetings and reports related to the programme were reviewed. Reports from 566 infant deaths from both districts from January to June 2011 were obtained from the district health system. Following quantitative analysis of the verbal autopsy records, 38 in-depth interviews (IDI) were conducted at district and sub-district level with health-care providers who were actively involved in programme activities and willing to participate freely [Table 1]. Semi-structured guidelines were used to conduct the interviews, the purpose of which was to explore the respondents’ experiences related to the IDR programme. Interviews were conducted at a date, time and place convenient for the respondents. With their permission, interviews were audio recorded and transcribed verbatim. The interviewer (second author) was sensitized to qualitative data collection.
Table 1: Category of respondents and number of in-depth interviews conducted

Click here to view

Data analysis

Quantitative data was entered using Epi-Info (version 3.4.3) software. The proportional mortality was calculated for neonatal and post-neonatal periods. The inter-observer reliability between the Medical Officer and the district Infant Death Review Committee for causes of death was assessed by kappa statistics.[16] This measures the agreement between two or more observers taking into account that they may agree or disagree purely by chance. For qualitative data analysis, structural coding was done using codes such as process, questionnaire design, its administration, mortality classification, ascertainment of cause of death and forces ‘for’ and ‘against’ the IDR programme. Transcripts were reviewed and a list of codes worked out. Manual content analysis was done and text data were presented under each coding category. Data collection and analysis was supervised by a public health expert who was trained in handling qualitative data. Findings of the analysis were reviewed by other authors to ensure their completeness and trustworthiness. Statements ‘for’ and ‘against’ the programme were calculated for different levels of health-care providers.

Informed consent was obtained from the respondents of the IDI. Permission for this review was obtained from the Director of the National Rural Health Mission, Karnataka.

  Results Top

Analysis of verbal autopsy records

Based on the prevalent infant mortality rates applied by the state to the rural district of Raichur and the urban district of Dakshina Kannada, we expected 1208 infant deaths to be reported during January to June 2011. However, only 566 infant deaths were reported – 447 (79%) from Raichur and 119 (21%) from Dakshina Kannada – for which we obtained verbal autopsy information.

Of the 556 infant deaths, 316 (55.8%) were male and 250 (44.2%) were female; 451 (79.7%) belonged to parents who had below-poverty-line cards provided by the state government. Scheduled castes\scheduled tribes (SC\ST) accounted for 239 (42.2%) of the deceased infants, more than half of whom occurred in Raichur. Noteworthy is the breakdown between the two districts: 42% infant deaths in Raichur took place at home, the next highest proportions being in private health facilities (28.8%), government health facilities (19.9%) and in transit (9.6%). In Dakshina Kannada, 36.2% infant deaths took place in government health facilities followed by private health care facilities (32.8%), at home (18.9%) and in transit (12.1%) [Table 2].
Table 2: Sociodemographic characteristics of deceased infants in Raichur and Dakshina Kannada districts, January to June 2011

Click here to view

Regarding the category of infant deaths in each district from January to June 2011, Dakshina Kannada reported 83 (69.7%) neonatal and 36 (30.3%) post-neonatal deaths, while Raichur reported 294 (65.8%) neonatal and 153 (34.2%) post-neonatal deaths ([Table 2] and [Table 3]).
Table 3: Level of agreement on cause of death ascertained by Medical Officer and by Infant Death Review Committee, Dakshina Kannada District, Karnataka

Click here to view

As ascertained by the district IDR Committee, the most common causes of infant deaths in Dakshina Kannada were pneumonia (19.3%), birth asphyxia\injury (13.4%), congenital malformations (12.6%), low birth weight (9.2%) and prematurity (8.5%). Overall, the inter-code reliability between cause of death as ascertained by medical officers and the District IDR sub-committee was good (kappa agreement 0.81; 95% confidence interval 0.73–0.88) [Table 2]. In Raichur, the most common causes of death as ascertained by medical officers were birth asphyxia\injury (26.2%), pneumonia (17.4%), and congenital malformations (11%) [Table 4].
Table 4: Causes of death ascertained by Medical Officer in Raichur, Karnataka

Click here to view

Processes leading to the development of the IDR programme

(i) Preparatory process: In March 2010, UNICEF Hyderabad organized a national consultative workshop on Infant Death Review in Bangalore in collaboration with the Government of Karnataka. Experts at the workshop discussed the VA questionnaire, mortality classification and developed guidelines for IDR implementation at district level, and in July 2010 the programme was initiated. This involved government health-care providers; programme managers from RCH and the department of Women and Child Development; paediatrics and community medicine specialists from medical colleges; and representatives from private hospitals involved in the planning process of the IDR programme. The workshop participants finalized the verbal autopsy tool and guidelines for the IDR programme.

(ii) Questionnaire design: a comprehensive verbal autopsy questionnaire for neonatal and post-neonatal deaths was developed in consultation with experts as a checklist with filters based on existing tools and literature. It was a simple colour-coded questionnaire (pink for newborn deaths and yellow for post-neonatal deaths) in the local language Kannada. The questionnaire was pre-tested on a sample of 12 deaths to incorporate local words and terms. The VA tool consisted of 17 pages including 3 pages for writing descriptive notes such as case summaries. It covered comprehensive information on socio-demographic characteristics, antenatal, natal and postnatal care and health-care seeking behaviour to capture the context in which the death occurred.

(iii) Capacity-building at various levels: Training started with a consultation of officials from the State Directorate and a few districts, followed by rapid training courses at district and sub district levels. At district level, an initial launch meeting was conducted to sensitize all stakeholders, including medical colleges, professional bodies, and other public service departments (e.g. Women and Child Development) in both districts. A separate sensitization programme on modalities of reporting infant deaths at the community and facility level was conducted for the Indian Academy of Paediatrics chapter in Dakshina Kannada and for the Indian Medical Association in Raichur. A one-day training programme was organized for all medical officers separately in each district on administration of verbal autopsy and ascertainment of cause of death.

Skill-building was carried out by reviewing all verbal autopsy questions in an interview in the local language, with illustrations on how to ascertain the cause of death based on diagnostic criteria. Following this, a few trained, skilled and motivated medical officers were identified by the district RCH officer to conduct further training for auxiliary nurse midwives, lady health visitors and other paramedical staff at taluk level. Medical officers of primary health centres sensitized anganwadi workers (AWW) and accredited social health activists (ASHA) under their service area. In order to create awareness of the IDR programme, Gram-panchayat (local self-government) members, leaders of women’s self-help groups, and representatives from local nongovernmental organizations were sensitized in specific training sessions.

(iv) Information on infant deaths: Village workers such as AWW and ASHA were asked to report details of infant deaths to the Medical Officer of the Primary Health Centre (PHC) on a short form called the First Information Report (FIR).In addition, they should send a Short Message Service (SMS) to a unique mobile number at district headquarters. In one Raichur district block, the district health system with support from UNICEF provided self-addressed and printed postal envelopes for reporting FIR.

(v) VA administration and cause of death ascertainment: The Medical Officer was asked to pay a home visit and interview the close carer of the deceased infant within 15 days of its reporting to ascertain the cause of death. The completed VA forms were then forwarded to the district Infant Death Review subcommittee (IDRSC). This committee consisted of the District Health Officer, district RCH officer, paediatricians from regional medical colleges, public health experts, district programme managers from the department of Women and Child Development, and medical officers from peripheral health centres.

In Dakshina Kannada, during monthly IDR Committee meetings, all the submitted VA forms and clinical case records of the deceased infants available from parents were discussed. In addition to ascertaining the cause of death, IDRSC discussed socioeconomic factors and events related to pregnancy, delivery, postnatal period, health-care seeking, gaps in the health system and health-care services. In Raichur, IDRSC reviewed only selected cases of infant deaths from community and hospital settings. The cases were selected on the basis of completeness of records, geographical location and cause of death. IDRSC provided feedback to peripheral health institutions regarding common causes of infant deaths and implementation of strategies tailored to local needs. The proceedings of IDRSC were submitted to the district IDR Committee to refine needs-based strategies to prevent infant deaths.

(vi) Positive and negative effects of the Infant Death Review system and options for improvement

Positive elements of the IDR system for the district health authorities and medical college faculties were their new partnership and the involvement of regional medical colleges in planning, implementation and training. This created a rapport with the local community and improved infant death reporting. Forces ‘against’ the programme were an initial lack of motivation at district and peripheral level to the new initiative, incomplete information in verbal autopsy reports, late reporting, lengthy form, and overburdened staff. Options suggested for future sustainability were to make the VA reporting form more concise, to train district level staff and medical officers and to improve health-care facilities.

Health workers of peripheral health institutions reported that the identification of gaps in the health-care system and financial incentives for AWW and ASHA to report deaths had contributed to infant death reporting. They also saw an opportunity to educate family members on the prevention of infant deaths. On the negative side, these health workers felt overburdened with their existing workload, and suffered from poor cooperation from caregivers and a lack of transport facilities. Options to redress this situation included financial incentives for medical officers to complete verbal autopsies, improved logistics, and forming sub district committees where infant deaths are high.

Anganwadi and ASHA workers noted that, as a result of the IDR programme, they had more information on the causes of infant deaths, and had feedback. This in turn had improved antenatal care. However, they have to face poor cooperation from caregivers, including occasional anger. These village workers considered that the arrival of a medical officer in the village to conduct verbal autopsy confers a sense of importance to issues related to infant deaths [Table 5].
Table 5: Forces ‘for’ and ‘against’ the IDR programme at various levels

Click here to view

  Discussion Top

Decision-makers and programme managers need simple data that can be linked to improving the provision of care. These potential users of verbal autopsy methods have different perspectives from researchers, tending to prefer simple VA instruments for their feasibility and programme relevance over technical performance. Our study explored the different perspectives of district health authorities, staff of the Primary Health Centre and village-level staff as key actors in the IDR programme. This contextual knowledge is useful for health-care managers and planners for ensuring programme sustainability. In particular, we noted that health system managers sought concrete and programme-relevant data, implying an interest in the socioeconomic determinants of mortality.

The Survey of Causes of Death by the Registrar General of India has been merged with the Sample Registration System, resulting in a modified verbal autopsy method with physician review. The current SRS sample size does not allow sub district level estimation of IMR, which varies across districts. Thus, a reduction in the average IMR in a State does not provide a complete picture of mortality decline, necessitating identification of high mortality-prone areas and planning innovative strategies for IMR reduction in these areas.[1] Hence the need for sub-district ‘infant death audit systems’ to gain accurate and reliable estimates for planning at the district and sub-district level.

Our study proposes a feasible approach to strengthen the existing health-care system at district level. We found that the involvement of stakeholders – the Department of Health, Department of Women and Child Development, local medical colleges, district hospital authority, peripheral health institutions, local representatives from WHO and UNICEF – in planning and implementing the IDR programme provided a platform for collective learning and action. The causes of deaths derived from verbal autopsies may not be as reliable as hospital-based certification, but information on risk factors and the health care sought prior to death made health-care providers more aware of the causes of infant deaths. The potential impediments identified to the success of the IDR programme need to be addressed by corrective actions at the relevant level by policy-makers and programmers for its sustainability.

  Conclusion Top

Involving the different stakeholders such as government health departments and private health-care providers in the IDR programme creates synergy and is essential for its sustainability. In addition to ascertaining the cause of infant deaths, the IDR Committee discusses social, economic, behavioural and health system issues that may have contributed to the death. As a result of the IDR programme, respondents perceived an improvement in infant death reporting at the district level, a closer rapport with the local community, and the welcome development of a feedback system for corrective actions. This has led to improved quality of care during pregnancy as perceived by health service providers.

  Acknowledgements Top

The authors thank Dr Arun Aggarwal, Professor of Community Medicine, Postgraduate Institute of Medical Education and Research, Chandigarh, for his critical review and comments. We are grateful to the Directorate of Health and Family Welfare Services, Karnataka, for the opportunity to pilot this study in two districts in the state. We also thank the District Health and Family Welfare Officers and District Reproductive and Child Health Officers in Raichur and Dakshina Kannada for allowing this study to be conducted in their districts. Finally, we express our appreciation to all staff and expert members of the IDR Committee who volunteered information, as well as to all individuals whose contributions made the work possible.

  References Top

Sample registration system of India, Sample Registration Bulletin. 2011 Dec; 46(1) - http://pib. nic.in/archieve/others/2012/feb/d2012020102.pdf - accessed 16 April 2013.  Back to cited text no. 1
Gvernment of India, National Commission on Population. National population policy 2000 - URL:http://populationcommission.nic.in/npp_obj. htm - accessed 16 April 2013.  Back to cited text no. 2
Tracking the Millennium development goals in Asia and the Pacific. 2009 - www.mdgasiapacific.org/…/ Regional_MDGs_report_2_chapter1.pdf - accessed 16 April 2013.  Back to cited text no. 3
Losos JZ. Routine and sentinel surveillance methods. East Mediterr Health J. 1996; 2(1): 45-60  Back to cited text no. 4
Byass P. Who needs cause-of-death data. PLoS Med. 2007; 4(11): e333.  Back to cited text no. 5
Beaglehole R, Bonita R. Challenges for public health in the global context—prevention and surveillance. Scand J Public Health. 2001; 29(2): 81-83.  Back to cited text no. 6
Byass P. Person, place and time—but who, where, and when? Scand J Public Health. 2001; 29(2): 84-86.  Back to cited text no. 7
Cleland J. Demographic data collection in less developed countries 1946–1996. Popul Stud (Camb). 1996; 50(3): 433-450.  Back to cited text no. 8
Begg S, Rao C, Lopez AD. Design options for sample-based mortality surveillance. Int J Epidemiol. 2005; 34(5): 1080-1087.  Back to cited text no. 9
Hill K. Making deaths count. Bull World Health Organ. 2006; 84(3): 162.  Back to cited text no. 10
Garenne M, Fauveau V. Potential and limits of verbal autopsies. Bull World Health Organ. 2006; 84(3): 164.  Back to cited text no. 11
World Health Organisation. Lay reporting of health information. Geneva: WHO, 1978.  Back to cited text no. 12
Garenne M, Fontaine O. Assessing probable causes of deaths using a standardized questionnaire: a study in rural Senegal. In: Vallin J, D’Souza S, Palloni A, editors. Measurement and analysis of mortality. Proceedings of the International Union for the Scientific Study of Population seminar, Sienna, Italy, July 7–10, 1986. Oxford, Clarendon Press, 1990. pp. 123-142.  Back to cited text no. 13
Soleman N, Chandramohan D, Shibuya K. Verbal autopsy: Current practices and challenges. Bull World Health Organ. 2006; 84 (3): 239-245.  Back to cited text no. 14
AbouZahr C. Verbal autopsy: Who needs it? Population Health Metrics. 2011; 9: 19 - http://www.pophealthmetrics.com/content/9/1/19 - 16 April 2013.  Back to cited text no. 15
Bang AT, Bang RA and SEARCH team. Diagnosis of causes of childhood deaths in developing countries by verbal autopsy: suggested criteria. Bulletin of World Health organization. 1992; 70(4): 499-507.  Back to cited text no. 16
Government of India, Ministry of Health and Family Welfare. District level household and facility survey. Fact sheet Karnataka. Mumbai: International Institute for Population Sciences - http://www.rchiips.org/pdf/ rch3/state/Karnataka.pdf - 16 April 2013.  Back to cited text no. 17
Fleiss JL. Statistical methods for rates and proportions. New York: John Wiley, 1981. pp. 38-46.  Back to cited text no. 18


  [Table 1], [Table 2], [Table 3], [Table 4], [Table 5]


Similar in PUBMED
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Material and methods
Article Tables

 Article Access Statistics
    PDF Downloaded143    
    Comments [Add]    

Recommend this journal