Indian Journal of Community Medicine

: 2021  |  Volume : 46  |  Issue : 4  |  Page : 631--636

Development and utility of tools to identify preventable perinatal deaths: results from a community-based interventional study in two districts of Karnataka State, India

HN Harsha Kumar1, Shantaram B Baliga2, Pralhad Kushtagi3, Nutan Kamath2, Sucheta S Rao2,  
1 Department of Community Medicine, Kasturba Medical College, Mangalore, Manipal Academy of Higher Education, Manipal, India
2 Department of Pediatrics, Kasturba Medical College, Mangalore, Manipal Academy of Higher Education, Manipal, India
3 Department of Obstetrics and Gyneacology, Kasturba Medical College, Mangalore, Manipal Academy of Higher Education, Manipal, India

Correspondence Address:
Dr. H N Harsha Kumar
Department of Community Medicine, Kasturba Medical College, Mangalore, Manipal Academy of Higher Education [MAHE], Deemed to be University, Manipal, Karnataka State


Background: The presently used perinatal death certificate devised by the World Health Organization is incomplete and does not help in identifying “preventability.” Objective: To develop tools that can help identify (1) preventable perinatal death and (2) preventable and/or avoidable cause for perinatal deaths. Materials and Methods: As a prerequisite for conduct of a community-based interventional study in two different districts of Karnataka state, two information tools, Perinatal death reporting form (PeNDReF) and perinatal death audit report (PeNDAR), were designed. The process involved series of preparatory and review meetings, before and after the field work to list facilities available, categorize facilities, identify causality, assess risk factors, and assert preventability of a perinatal death. The process was repeated over a period of 6 months and the information tools, PeNDReF and PeNDAR, were finalized. Doctors and paramedical personnel of both the districts were trained to fill the tools, which were analyzed to ascertain contributing risk factors and identify preventable perinatal death. Results: The use of PeNDAR led to identification of 5.7% of perinatal deaths as “preventable” and 19% as “possibly preventable.” The use of PeNDReF helped in the identification of risk factors (maternal anemia 49.6%, age of marriage <20 years 18.7%, and maternal weight <50 kg 9.1%), avoidable/preventable factors related to quality of care, transport, and referral. Conclusions: These tools are useful for identifying “preventable” perinatal deaths and avoidable/preventable factors.

How to cite this article:
Harsha Kumar H N, Baliga SB, Kushtagi P, Kamath N, Rao SS. Development and utility of tools to identify preventable perinatal deaths: results from a community-based interventional study in two districts of Karnataka State, India.Indian J Community Med 2021;46:631-636

How to cite this URL:
Harsha Kumar H N, Baliga SB, Kushtagi P, Kamath N, Rao SS. Development and utility of tools to identify preventable perinatal deaths: results from a community-based interventional study in two districts of Karnataka State, India. Indian J Community Med [serial online] 2021 [cited 2022 May 17 ];46:631-636
Available from:

Full Text


The perinatal death certificate (PDC) devised by the World Health Organization captures information which focuses on “maternal and neonatal factors.”[1] To know if a perinatal death is preventable, it is important to capture a lot of information in addition to “maternal and neonatal factors.” Existence of avoidable/preventable factors that contribute to perinatal deaths in resource-poor settings has been reported.[2],[2],[3],[4],[5] Some avoidable factors which played a role in perinatal deaths have also been reported from India.[2] PDC used in India does not have provision to document such factors and identify “preventability” of a perinatal death. Identification of avoidable and/or preventable factors is necessary to frame strategies to reduce preventable perinatal deaths.

In India, perinatal mortality rate remains at 26 per 1000 births (for the year 2013) with stillbirths at 4 per 1000 births and is an underestimate.[6] Although all of these perinatal deaths are not preventable, it stands to reason that some may be. Even though the Government of India has made efforts to improve health-care infrastructure under the National Health Mission, poor health-care infrastructure and inefficiency in the health-care delivery in rural areas have been reported.[7] Survey of maternal and neonatal care facilities in a backward district of Karnataka state in India has revealed deficiency of infrastructure for the management of high-risk cases.[8] An analysis of newborn deaths reported from India has revealed the existence of factors such as “unsupervised care” and errors of judgment during management.[2] Hence, some perinatal deaths have contributory factors related to infrastructure and quality of care making them “preventable.” However, there is no information about the magnitude of preventable perinatal deaths as there is no tool (s) and system to capture this information.

As the PDC cannot be used to identify preventable perinatal deaths and some perinatal deaths in India are preventable, there is a need to develop some tools to identify (1) preventable perinatal deaths and (2) avoidable/preventable factors in cases of perinatal death. This paper describes the development of two tools which were developed as a part of community-based interventional study carried out over a period of 3 years in two districts of Karnataka state, India.

 Materials and Methods

Study setting

The two districts chosen were (1) Dakshina Kannada, an economically, educationally better developed and with better health-care infrastructure and (2) Koppal with poorer indices on these parameters. This would help know the utility of the tools in better and poor conditions.[9],[10]

The details of all the government and private hospitals in both the districts which met the inclusion criteria for the project are described elsewhere.[8]

Development of tools

PDC considers only the maternal and neonatal factors limiting its utility to identify avoidable/preventable factors. Hence, the first step in the process was to “identify types of information” in addition to maternal and neonatal factors which could help in ascertaining preventability of perinatal deaths. Combination of review of literature and an initial preinterventional survey provided us the preliminary information that need to be considered.[8] This was followed by several rounds of meetings by the investigators. Each meeting was followed by relevant field work in both the districts. Feedback received from the field work was discussed in the next meeting. This process repeated several times over a period of 6 months helped finalize the types of information required. It was clear that some types of information helped in identifying “risk factors,” “high-risk” cases, and maternal and neonatal care practices. Based on this information, it is possible to identify “risk factors and avoidable/preventable factors related to all aspects of care provided to mother and baby. Some types of information helped in identifying contributors to perinatal death and asserting preventability. As it was clear that “one tool” cannot capture all the information, two were developed: (1) perinatal death reporting form (PeNDReF) and (2) perinatal death audit review (PeNDAR).

Types of information identified and considered in PenDRF

The focus was to collect all the types of information necessary for the identification of “risk factors, avoidable/preventable factors/issues” on all aspects of care. Hence, it included basic sociodemographic information, details of antenatal, neonatal care, transport, and referral and documentation of records.

Types of information identified and considered in PenDAR

The focus of this tool is to provide an overview of the events preceding perinatal death along with identification of contributors and help in assertion of preventability along with reasons. Hence, the details included type of perinatal loss, investigation details, autopsy findings, and contributors identified.

Field testing and modification of both the tools

These tools were field tested in Dakshina Kannada and Koppal districts. Repeated field testing led to identification of problems and several modifications were done, which are summarized in [Table 1] and [Table 2].{Table 1}{Table 2}

Perinatal death reporting form

The final modified tool captures 76 different “types of information” which were classified under the following heads: (a) socioeconomic background; (b) antenatal referral details; (c) antenatal care details including investigations and treatment; (d) intrapartum referral details; (e) newborn details at birth; (f) status of the child at the time of admission to special neonatal care unit or neonatal intensive care units; (g) outborn neonatal referral details; and (h) neonatal transport details. In addition to these, it also captures (i) type of documents examined and their status and (j) referral-related factors. The form also has a provision to note any other details relevant to, but not covered under the above heads.

Perinatal death audit review

The final modified tool consists of the following components: (a) patient–parent identification details; (b) details of perinatal loss; (c) cause determination, group classification including autopsy reports if available; (d) contributor identification; and (e) preventability assertion along with reasons.

Personnel and training

Familiarization and training sessions were held in workshop module in batches for doctors, nurses, and paramedical personnel at both the districts. Doctors were trained to fill in both the forms, PeNDAR because they are the only ones who are supposed file them and PeNDReF to enable supervision of filling by other health-care personnel. Nursing and paramedical personnel were trained to fill PeNDReF.

Outcome measures

(1) Proportion of perinatal deaths as preventable, possibly preventable, and not preventable using PeNDReF and PeNDAR; (2) identification of avoidable/preventable factors using PeNDReF; and (3) identification of contributors to perinatal death using PeNDAR.

Operational definitions

The following operational definitions were used: (1) Perinatal deaths were classified as “preventable,” possibly preventable, and not preventable. The doctor/specialist who managed the case of perinatal death is expected to assert preventability after considering the information in PeNDReF, PeNDAR, and using their clinical judgment. This had to be documented at the end of PeNDAR. (2) Avoidable/preventable factors: As identified by doctors and paramedical personnel who were involved in managing the case of perinatal death and filled in the PeNDReF. (3) Contributors: The doctor/specialist who managed the case of perinatal death is expected to identify contributors by a combination of information in PeNDAR and clinical judgment.

Data collection

Necessary clearances were taken from the government before beginning of the project. Field workers who were employed and trained under the project visited the government and private hospitals in both the districts to collect the filled forms. The collected forms were verified by the investigators and sent back for any clarifications/missing entries in the forms.

Data analysis

The collected forms were analyzed. Problems/issues and contributors that were identified are described in [Table 1] and [Table 2] and discussed. The results are expressed as proportions in [Table 3] and [Table 4].{Table 3}{Table 4}


There were 1070 perinatal deaths which were filled in the tools with 444 from 63 health-care facilities in Dakshina Kannada and 626 from 65 centers of Koppal district.


The use of PeNDReF led to identification of risk factors for perinatal death which are presented in [Table 3]. Anemia (29.1% and 65.1%), low age of marriage (<20 years) (13.5% and 22.6%), weight <50 kgs (16.4% and 3.6%), fetal distress (34.5% and 3.1%), and prematurity (67% and 44.3%) were prominent.


It was possible to identify that 5.7% of all perinatal deaths are “preventable” and 19% are “possibly preventable” [Table 4]. Although the identification of “preventability” is low for all perinatal deaths, it was much better for “antenatal (≈3%), followed by neonatal (1.9%) and intranatal fetal deaths (0.8%) [Table 4]. Identification of “preventable” and “possibly preventable” perinatal deaths was better in Dakshina Kannada district as compared with Koppal district. Information about contributors identified during perinatal period in Dakshina Kannada are as follows (n = 153, many cases had more than one contributor): maternal 63 (41.2%), neonatal 59 (38.6%), fetoplacental 42 (27.4%), health system related 12 (7.8%), referral related 4 (2.6%), and socioeconomic background related 3 (≈2%). Because of poor documentation, there was no information about contributors in Koppal district.


One published study from India reports problems with quality of care in cases of perinatal death though there is no information about any tools that were used for the purpose.[2] Efforts to find out avoidable/preventable factors are not new. Most of the researchers have used the model of “mortality audit committee” which meets at regular intervals to analyze and find out the avoidable factors.[11],[12],[13],[14],[15],[16],[17],[18] However, these researchers did not develop and report any tools for that purpose.[11],[12],[13],[14],[15],[16],[17],[18] One study from Tanzania has reported the use of structured protocols to audit without providing details of the tools used.[19] To our knowledge, this is the only study reporting the development and utility of tools with no others for comparison. These tools were developed and tried in two different types of districts, i.e., Dakshina Kannada which is well developed and Koppal which lags behind in development.

The use of PeNDReF resulted in the identification of “risk factors” the pattern of which is similar in both the districts though the magnitude differs [Table 3]. Similar patterns of avoidable/preventable factors have been reported by several studies which adopted the “mortality audit committee” model.[11],[12],[13],[14],[15],[16],[17],[18] Although the details of structured protocols in Tanzania are lacking, they have reported the existence of similar avoidable/preventable factors.[19] Hence, the pattern of “risk factors” and avoidable/preventable factors obtained with PeNDReF matches with the information obtained from the “mortality audit committee” model. This establishes the utility of PeNDReF as a useful tool for the identification of “risk factors and avoidable/preventable factors. As information pertaining to many aspects of care are captured, it helps in localizing the problems and provides inputs for planning “preventive strategies.”

The use of PeNDAR resulted in the identification of “preventable” and “possibly preventable” perinatal deaths. This is unique and not comparable to other studies based on “mortality audit committee” model.[11],[12],[13],[14],[15],[16],[17],[18] Identification of preventable and possibly preventable deaths in Dakshina Kannada district (8.3% and 37.6%) was better than Koppal (3.8% and 6.1%). Poor documentation and reporting of perinatal deaths in Koppal district is known.[20] For the same reason, majority (87.1%) of perinatal deaths in Koppal district remained unclassified [Table 4] and contributor identification was not possible. Very low proportion (6% and 19%) of deaths were identified as “preventable and possibly preventable.” About 56% of deaths were “unclassified.” Doctors who filled up this tool were providing information instead of deciding about the preventability of perinatal deaths, resulting in low proportion of identified “preventable perinatal deaths.” Information provided by PeNDReF and PeNDAR are complimentary and help identify problems and frame strategies to reduce preventable perinatal deaths.

Most of the studies which follow the “mortality audit committee” model are limited to a single hospital/health facility.[11],[12],[13],[14],[15],[16],[17],[18] This community-based approach included government and private health-care facilities in two different types of districts (i.e., well developed and backward), making the tools useful for different types of conditions in India.

There are some limitations. As the information about all aspects of care could not be obtained/available from one source [Table 1], it takes time and effort to capture complete information. If the documentation is poor/nonexistent, the time and effort required to capture information from multiple sources would increase. Improving the documentation in hospital case records is essential to identify avoidable/preventable factors. This could be reduced to some extent by (1) training the doctors to improve the documentation of case records and (2) taking information from multiple sources though the time and effort increase. Low proportion of identified preventable deaths could addressed by (1) helping the doctors to understand their role and (2) constituting an expert committee which could use these tools to identify preventable deaths and contributors (as a part of the intervention carried out during this project, an expert committee was constituted at both the districts which audited perinatal deaths using these tools. The proportion of identified preventable deaths increased and unclassified deaths reduced. These details are beyond the scope of this paper).


PeNDReF and PeNDAR are data collecting and reporting tools that help identify preventable perinatal deaths, risk factors, and avoidable/preventable factors for perinatal death. They provide a comprehensive overview of different aspects of perinatal care and help plan preventive strategies.


We thank, in both the districts, district health officers, field workers, and all the hospitals for cooperation, data collection, and permission to study the documentation and reporting systems, respectively. We express our gratitude to the Government of Karnataka for giving us the permission to carry out this study.

Financial support and sponsorship

Department of Health Research, Ministry of Health Family Welfare, Government of India, Reference Number: DHR/GIA/35/2014.

Conflicts of interest

There are no conflicts of interest.


1World Healh Organization. The WHO Application of ICD-10 to Deaths during the Perinatal Period: ICD-PM; 2016. Available from: [Last accessed on 2019 Mar 9].
2Bhatt RV. Professional responsibility in maternity care: Role of medical audit. Int J Gynaecol Obstet 1989;30:47-50.
3Pattinson R, Kerber K, Waiswa P, Day LT, Mussell F, Asiruddin SK, et al. Perinatal mortality audit: Counting, accountability, and overcoming challenges in scaling up in low- and middle-income countries. Int J Gynaecol Obstet 2009;107:S113-2.
4Dumont A, Fournier P, Abrahamowicz M, Traoré M, Haddad S, Fraser WD, et al. Quality of care, risk management, and technology in obstetrics to reduce hospital-based maternal mortality in Senegal and Mali (QUARITE): A cluster-randomised trial. Lancet 2013;382:146-57.
5Kidanto HL, Wangwe P, Kilewo CD, Nystrom L, Lindmark G. Improved quality of management of eclampsia patients through criteria based audit at Muhimbili National Hospital, Dar es Salaam, Tanzania. Bridging the quality gap. BMC Pregnancy Childbirth 2012;12:134.
6Sankar MJ, Neogi SB, Sharma J, Chauhan M, Srivastava R, Prabhakar PK, et al. State of newborn health in India. J Perinatol 2016;36:S3-8.
7Upadhyay RP, Chinnakali P, Odukoya O, Yadav K, Sinha S, Rizwan SA, et al. High neonatal mortality rates in rural India: What options to explore? ISRN Pediatr 2012;2012:968921.
8Kumar HN, Balig BS, Kushtagi P, Kamath N, Rao S. Exploratory study on maternal and child health care facilities in two districts of Karnataka state A health systems research. Int J Med Public Heal 2018;8:152-7.
9Registrar General of India. Census of India 2011: District Census Handbook Dakshina Kannada. Karnataka: Directorate of Census Operations; 2014. Available from: [Last accessed on 2018 Aug 02].
10Registrar General of India. Census of India 2011: District Census Handbook Koppal. Karnataka: Directorate of Census Operations; 2014. Available from: [Last accessed on 2018 Aug 02].
11El Amin S, Langhoff-Roos J, Bødker B, Bakr AA, Ashmeig AL, Ibrahim SA, et al. Introducing qualitative perinatal audit in a tertiary hospital in Sudan. Health Policy Plan 2002;17:296-303.
12Mbaruku G, van Roosmalen J, Kimondo I, Bilango F, Bergström S. Perinatal audit using the 3-delays model in western Tanzania. Int J Gynaecol Obstet 2009;106:85-8.
13Chigbu CO, Okezie OA, Odugu BU. Intrapartum stillbirth in a Nigerian tertiary hospital setting. Int J Gynaecol Obstet 2009;104:18-21.
14Van Roosmalen J. Perinatal mortality in rural Tanzania. Br J ObsteGynaecol 1989;96:827-34.
15De Muylder X. Perinatal mortality audit in a Zimbabwean district. Paediatr Perinat Epidemiol 1989;3:284-93.
16Frost O. Maternal and perinatal deaths in an Addis Ababa Hospital, 1980. Ethiop Med J 1984;22:143-6.
17Johnstone FD, Ochiel SO. Perinatal mortality at Kenyatta National Hospital, Nairobi. East Afr Med J 1980;57:119-23.
18Hinderaker SG, Olsen BE, Bergsjø PB, Gasheka P, Lie RT, Havnen J, et al. Avoidable stillbirths and neonatal deaths in rural Tanzania. BJOG 2003;110:616-23.
19Kidanto HL, Mogren I, van Roosmalen J, Thomas AN, Massawe SN, Nystrom L, et al. Introduction of a qualitative perinatal audit at Muhimbili National Hospital, Dar es Salaam, Tanzania. BMC Pregnancy Childbirth 2009;9:45.
20Harsha Kumar HN, Balig BS, Kushtagi P, Kamath N, Rao S. Documentation and reporting of perinatal deaths in two districts of Karnataka stat India: A situational analysis. Indian Pediatrics 2020;57:1006-9.