Methodology

This report is based on Human Rights Watch field investigations and consultations with key stakeholders in India between November 2008 and August 2009. Where available, the accounts gathered through our field investigations have been corroborated by data from government surveys, and reports or studies by nongovernmental organizations, international agencies, and public health experts in India.

Based on our preliminary consultations with 55 public health specialists, lawyers, and representatives from local nongovernmental organizations working on the right to health and women's rights across India, Human Rights Watch chose to focus on Uttar Pradesh state in north India.

Uttar Pradesh was chosen as a case study because, being the most populous state, it accounts for the highest number of maternal deaths in India.[3] It is also one of several states that had issued a 2004 governmental order seeking investigations into maternal deaths.

Human Rights Watch also examined southern Tamil Nadu's relatively stronger system of investigating maternal deaths.

The primary field investigations took place in Rae Bareilly, Unnao, Chitrakoot, Lucknow, and Barabanki districts of Uttar Pradesh in February, March, and June 2009; New Delhi in March 2009; and in Tamil Nadu in April 2009. We supplemented these field investigations with telephone interviews between June and August 2009.

Human Rights Watch researchers interviewed 191 people; 95 in individual interviews and the remainder in group interviews. These included:

In Uttar Pradesh:

• Fifty-six women and men from villages, including individuals from nine families in which maternal deaths had occurred.
• Thirty-four health staff from government health facilities, peripheral field-based health workers including auxiliary nurse-midwives (ANMs), accredited social health activists (ASHAs) or female community health aides, anganwadi workers (female workers tasked with providing early childhood care and education and nutritional supplements for pregnant women), and traditional birth attendants.
• Forty-five officials including heads of village level councils, panchayat mitras (literally, friends of the village council), chief medical officers (highest health authority at the district level), officials from the directorates of family welfare and medical and health services, and members of the district and state project management units of the National Rural Health Mission (NRHM), India's flagship rural healthcare program.
• Seven doctors from the private sector, including representatives from the Uttar Pradesh chapters of the Federation of Obstetric and Gynecological Societies of India and the Indian Medical Association. 
• Thirty journalists and representatives from nongovernmental and intergovernmental organizations including Vatsalya, Mamta, SAHAYOG, Healthwatch Forum, Jan Swasthya Abhiyaan-UP chapter, CARE-India, the John Hopkins University project on infant and maternal health, Vanangana, PATH, and UNICEF.

In NewDelhi:

• Eleven officials including officials from the Office of the Registrar General of India, the NRHM directorate, and representatives from the National Health Systems Resource Center, a technical resource center set up under the NRHM.

In Tamil Nadu:

• Four former and four present government officials overseeing maternal mortality reviews.
• Nine activists, including grassroots-level workers and a professor who participates in the maternal mortality review meetings.

Human Rights Watch had hoped to include the perspectives of doctors or health workers who were suspended, dismissed, or arrested following complaints about maternal health care in Uttar Pradesh. Unfortunately, we were able to trace only one such health worker, a hospital staff nurse.

Health workers and nongovernmental organizations providing services to villagers assisted Human Rights Watch in identifying pregnant women and families to interview; we further developed contacts and interview lists through references from interviewees.

Interviews lasted between 20 minutes and three hours and were conducted in English, Hindi, dialects of Hindi such as Awadhi or Bundelkhandi, or Tamil depending on the interviewee's preference. The primary investigator from Human Rights Watch is also fluent in spoken Hindi and Tamil. In cases where the interviewees chose to communicate in Awadhi or Bundelkhandi, the interviews were conducted with the assistance of female interpreters.

All interviews during field investigations in Uttar Pradesh, Tamil Nadu, and New Delhi were conducted after orally obtaining informed consent. Human Rights Watch has respected the choices of all interviewees to be identified, not identified, or have their experiences and views left out of the report entirely, and has assigned pseudonyms or withheld identifiable information accordingly. Pseudonyms have been assigned randomly, and do not correspond to the religion, caste, or tribe of the interviewee.

We supplemented our field investigations with official data provided by the Indian central and Uttar Pradesh state governments in response to several applications filed by Human Rights Watch under the Right to Information Act, 2005.[4]

Human Rights Watch also convened an India advisory group whose purpose was to provide inputs and feedback on methodology, support in reaching out to relevant networks and groups, and review this report.[5]

Scope and Limitations

This report uses a human rights framework to examine maternal health care, setting out several specific steps we believe officials should take to better integrate accountability into maternal healthcare programs and ensure their implementation through the health system. It does not explore all available tools for accountability including external surveys assessing quality of health care, public hearings, social audits of budgets, or community-based monitoring. The NRHM, India's flagship rural healthcare program, sets out a three-pronged accountability framework of external surveys, community-based monitoring, and stringent internal monitoring. This report's focus is on the last of these three prongs, the state's internal monitoring of policies, practices, and performance. While the arguments presented in this report address the specific issue of preventable maternal mortality and morbidity, accountability as a human rights principle is central to the right to the highest obtainable standard of health more generally.

Uttar Pradesh state is one of eight Empowered Action Group (EAG) states with poor socio-economic indicators.[6]The maternal health parameters of the eight states are comparable, and activists and public health experts in India say they exhibit similar recurring health system shortcomings.[7] The concerns raised in this report in chapters III-V have been examined in the context of Uttar Pradesh, but apply to many states, particularly those in the Empowered Action Group. Financial barriers to care are a problem in many states in India including non-Empowered Action Group states.

While interviewing bereaved families, Human Rights Watch's researchers gathered information on the fulfillment of government standards for maternal health care. Detailed identification and analysis of all the socio-economic or medical causes that contributed to each of these maternal deaths is beyond the scope of this report.

Note on Estimates

All data used in this report are estimates. The indicator most often cited in this report is the maternal mortality ratio (MMR) which is an estimate of the number of maternal deaths per 100,000 live births. The Indian government has released maternal mortality estimates based on various measures including MMR. Maternal mortality data for Indian states and for the country as a whole going back to 1997 have been provided by the Indian government in two recent reports. Data for the period 1997-2003 was released in 2006 in a publicly issued special report on maternal mortality and its causes.[8] A second data set, released in mid-2009, provides preliminary information for the period 2004-2006, pending release of the full report.[9] The in-country estimates used in this report are drawn from the interim 2004-2006 data. Information about medical causes of maternal deaths is drawn from the 1997-2003 data.

International data presented in this report are drawn from the latest available international estimates of maternal mortality that date from 2005.[10]  The Maternal Mortality Working Group (MMWG) comprised of the WHO, UNICEF, UNFPA, the UN Population Division, and the World Bank, as well as several outside technical experts, developed a methodology to create comparable country, regional, and global estimates of maternal mortality. The MMWG concluded that the data provided by the Indian government's survey is uncertain, containing biases such as ill-defined cause-of-death codes. This group calculated that the Indian survey underestimates the national MMR by 50 percent. 

The national MMR figures used in this report should be interpreted with the caveat that they do not fully reflect the changes brought on by the Indian government's flagship healthcare program, the NRHM, which has only been in effect since mid-2005 when the figures were compiled for the period 2004-2006. Furthermore, these figures represent point estimates within a larger range.[11] While the point estimates taken alone suggest a discernible reduction in MMR, the overlap in its ranges makes it difficult to gauge the extent of maternal mortality reduction in many states, particularly the EAG states.[12]

Note on Terminology

"Health workers": We use this phrase to refer to three categories of field-based peripheral workers-ANMs, anganwadi workers, and ASHAs.

"Investigating maternal deaths": We use this phrase to refer to procedures that identify health system shortcomings in addressing the causes, socio-economic as well as medical, of maternal deaths.

The World Health Organization (WHO) has outlined several methods of conducting maternal death reviews including community-based (verbal autopsies) and facility-based reviews, and confidential inquiries.[13] Human Rights Watch does not have the expertise to recommend a particular method of investigating deaths. Confidential inquiries have the merit of covering deaths irrespective of the place of occurrence and of not limiting the inquiry to identifying personal, family, or community factors as is the case with verbal autopsies.

"Lower castes": The phrase "lower castes" has been used to describe two types of castes in India, scheduled castes and so-called "other backward classes." Under Indian law scheduled caste refers to Dalits or so-called "untouchables" who are traditionally considered "outcastes," beneath the lowest caste in the four-caste hierarchy. Indian law uses "other backward classes" to refer to the lowest caste within the four-caste hierarchy, the Shudras. Both groups continue to face historical discrimination and have high rates of poverty.