Ethical Considerations in Health Services Research in Developing Countries1

Background

Neonatal mortality accounts for nearly half of the global child mortality in developing countries; 63 per cent of newborns in developing countries are born at home, where it can be more challenging to respond to complications in pregnancy, labour, delivery and the neonatal period (0-28 days of life).

It has been documented that, globally, 38 per cent of childhood deaths are in the neonatal period, and 450 newborns die each hour.Almost all (99 per cent) neonatal deaths globally are in low-middle income countries, yet most epidemiologic neonatal research has been conducted in rich countries.

In developing countries, the majority of neonates are born and cared for in rural homes. Nonetheless, the majority of data on neonatal morbidity and mortality are from hospital data. Most health statistics related to neonates in rural settings have been collected unsystematically, usually whenever the health worker in the area arrives. Not only are there limited data regarding morbidity and mortality data from rural areas, there are even fewer data available regarding unmet healthcare needs. In preparation for designing a field trial to assess a neonatal care delivery programme in a rural area, researchers conducted a prevalence study to determine the extent of the problem of neonatal morbidity and mortality in this region and to determine the extent of unmet need.

The prevalence study

The prevalence study was conducted in a rural district of a poor country.This district is extremely undeveloped; farming of rice and forestry are the main source of household incomes. Government health services include a male and female paramedic worker for every 3000 people and a primary care centre for every 20 000 people. No specialised neonatal care is available in the area. Private rural medical practitioners, herbalists, and magic healers are the main sources of care. Traditional birth assistants often help women with deliveries. In addition, a non-governmental organisation (NGO) has been training village health workers (VHWs) to work in this district for ten years. The VHWs are women who live in the village, have five to ten years' education and are provided with three weeks of intensive training related to infant health. Specifically, VHWs are trained to take history, observe the process of labour and examine newborns for a series of 'warning signs' of neonatal morbidity or distress. They provide basic education to the mothers and they refer mothers to other sources of care if they see a baby who seems very sick.

The prevalence study takes place in 39 villages already served by VHWs.In these villages, VHWs are asked to record the findings of their observations.The VHW in each village is to observe all pregnant women periodically, observe the birth, and visit the neonates approximately every three days for the first 28 days of life (or until death, or until the mother and baby left the village, whichever occurs first).

The purpose of the prevalence study is to estimate the number and types of existing neonatal morbidities in this district and also the number of neonatal mortalities over the period of one year. In addition, investigators want to document the number and proportion of neonates with various different indications for health care who do not end up receiving health care, as well as the number and proportion of infants with various, different indications for health care who do actually receive such care. The rationale for this research is that, in order to plan for appropriate research and care in the region, it is critical to have estimates of the burden of neonatal morbidities as well as the unmet need for care. The majority of studies up to this time have focused on either single conditions or on neonatal care received in hospitals. A community meeting was held in each village during which the VHW described the study; community consent was obtained in each village through this meeting.

The neonates received care according to their usual practice, including care from family members and traditional birth assistants, care from government nurses and private doctors, if invited by the family. If a doctor recommended that a family take a baby to the hospital, the NGO working on the study offered to provide the ambulance service for transporting the sick baby, but the final decision was left to the family. Any care received or recommended, and the sources of care received, were recorded by the VHW.

One year was devoted to the prevalence study. During this year, VHWs observed neonatal health 'with minimum interventions'.Of the 763 neonates observed during the year, 54.4 per cent had indications for health care, according to the VHWs, including 48 per cent suffering from high-risk health conditions.However, only 2.6 per cent of the neonates with indications for health care were seen and treated by a doctor, and only three (0.4 per cent) were hospitalised.

Out of the 763 neonates, 40 died during the year. Of the 40 neonates who died, 38 had been characterised as having high-risk health conditions.That is, approximately 10 per cent of those characterised as having high-risk health conditions died.Of the three sick neonates who were hospitalised by the families, none died.

Important data on prevalence and incidence of neonatal morbidities and mortalities were learned, which proved helpful for designing an intervention study that followed.

Questions
1. Discuss the ethics of this prevalence study.

2. Investigators conducted this study in order to obtain a rigorous understanding of the background conditions, in order to plan high quality, responsive, and relevant interventions.

Do you think the study could have been conducted any other way?Why or why not?What would the implications have been for the scientific findings?What would the implications have been for the cost?What approach would you recommend?

3. Do your beliefs about the ethics of this study vary, depending on whether it was conducted by local health researchers, or by US or European researchers, in conjunction with local researchers?Why or why not?

4. What do you think was meant by the phrase 'community consent was obtained in each village by the VHW'. Do you think this approach is appropriate?If not, how would you modify it?

The intervention trial

The prevalence study indicated that this population of neonates suffered a significant disease burden and a substantial level of unmet healthcare need.One finding in the prevalence study was that almost 20 per cent of neonates suffered from sepsis (a systemic infection, characterised by high fever and significant illness). Investigators chose to focus their clinical trial on the prevention and treatment of sepsis and other neonatal morbidities in home-cared, rural neonates in the same population. Previous to this trial, there have been various studies confirming the effectiveness of home-based health workers identifying and treating health problems successfully, and reducing morbidity and mortality. Further, the US Agency for International Development (USAID), World Health Organization (WHO), and the United Nations have all supported the training of local healthcare workers to provide additional services and care in impoverished areas. Village health workers are known to provide a low cost and efficient means of improving health outcomes, at least for certain indications.Indeed, the highly successful WHO strategy of Integrated Management of Childhood Illnesses (IMCI) relies heavily on identification of ill children in the home through local health workers. No study of such an approach had been conducted previously to reduce mortality from neonatal sepsis and other neonatal morbidities.

The intervention trial was designed as a controlled field trial with the 39 villages from the above prevalence study receiving intervention and 47 nearby villages, also in the 'catchment area' of the NGO, serving as controls. The VHWs in the intervention villages received training in home-based management of neonatal conditions including birth asphyxia, premature birth, low birth weight, hypothermia and breast-feeding problems, and they were trained to treat neonatal sepsis with antibiotics. In the control arm, VHWs were trained to collect data, going door to door during pregnancy, birth and the first 28 days of the neonatal period.

The study was conducted for three years with the hypothesis that the intervention villages would see a 25 per cent reduction in neonatal mortality by the third year.

The intervention study found a statistically significant 50 per cent reduction in neonatal and infant mortality in the intervention villages (p < 0.001). Treatment for sepsis by the VHW reduced mortality by almost 14 per cent (p < 0.01).

Questions
A few years after the completion of the above study, another researcher wants to replicate this study in another poor country that also has alarmingly high rates of neonatal morbidity and mortality. Given the published results from the study, above, and given that the concept of local health workers has been so well validated, is it justified to repeat the above study?

On the other hand, this is the first study testing this concept to reduced neonatal sepsis and neonatal mortality; what are the risks of not repeating the study? Why would a country want to replicate a study with its own people?

Reference list

Bang AT, Bang RA, Baitule SB, Reddy MH, Deshmukh MD. Effect of home-based neonatal care and management of sepsis on neonatal mortality: field trial in rural India.Lancet. 1999 Dec 4;354(9194):1955-61.

Bang AT, Bang RA, Baitule S, Deshmukh M, Reddy MH. Burden of morbidities and the unmet need for health care in rural neonates—a prospective observational study in Gadchiroli, India. Indian Pediatr. 2001 Sep;38(9):952-65.

Bryce J, Victora CG, Habicht JP, Vaughan JP, Black RE. The multi-country evaluation of the integrated management of childhood illness strategy: lessons for the evaluation of public health interventions. Am J Public Health. 2004 Mar;94(3):406-15.

Tulloch J. Integrated approach to child health in developing countries. Lancet. 1999 Sep;354 Suppl 2:SII16-20. Review.

Lawn JE, Cousens S, Zupan J; Lancet Neonatal Survival Steering Team. 4 million neonatal deaths: when? Where? Why? Lancet. 2005 Mar 2;365(9462):891-900.

Armstrong Schellenberg J, Bryce J, de Savigny D, Lambrechts T, Mbuya C, Mgalula L, Wilczynska K; Tanzania IMCI Multi-Country Evaluation Health Facility Survey Study Group. The effect of Integrated Management of Childhood Illness on observed quality of care of under-fives in rural Tanzania. Health Policy Plan. 2004 Jan;19(1):1-10.