This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.

Two different interventions in the care of neonatal infants were assessed. Village health workers (VHW) were trained to treat asphyxia using either a tube and mask made of silicon rubber, with a safety valve to prevent excessive pressure, or a bag and mask with a safety valve, (interventions from Phoenix, Chennai, India). Both of these interventions were compared with the use of traditional birth attendants (TBAs) who would attempt resuscitation of the infants using a mouth-to-mouth intervention.

Diagnosis and treatment.

Cost-effectiveness analysis.

The study population comprised newborn infants.

The setting was community care. The economic study was carried out in Gadchiroli, India.

The effectiveness evidence came from 1988 to 2003. The resource evidence referred to 1996 to 2003. No price year was given.

The effectiveness data were derived from a single study.

The same patients provided both the cost and the effectiveness data.

No power calculations were reported. All deliveries in 39 villages were included in the part of the study that was used in the economic analysis. There were 5,651 deliveries in the intervention group where VHWs were used and 782 in the control group where trained TBAs were used. The authors did not provide numbers for the two kinds of intervention used by the VHWs.

This was a multi-centre, comparative study with historical controls. There was no follow-up after the birth.

All births entered in the study were included in the analysis. There was no assessment of comparability at baseline. The following outcomes were used to assess the intervention:

the proportion of home deliveries in which a VHW was present;

the incidence of birth asphyxia (mild and severe);

the case-fatality rate for severe asphyxia;

the asphyxia-specific mortality rate; and

the fresh stillbirth rate.

A fresh stillbirth was an intrapartum foetal death, or a severely asphyxiated neonate who did not cry or breathe and could not be resuscitated.

The incidence of mild asphyxia was 14.2% (1995 - 1996) with treatment by a TBA and 8.4% (1996 - 1998), 5.9% (1998 - 2000) and 5.7% (2000 - 2003) with treatment by a VHW.

The incidence of severe asphyxia was 4.6% with treatment by a TBA and 2.4%, 3.7% and 4.9% with treatment by a VHW.

The case-fatality rate for severe asphyxia went from 38.5% in the mouth-to-mouth group to 28.3% in the tube plus mask group and 17.2% in the bag plus mask group.

The asphyxia-specific mortality rate went from 10.5% (1995 - 1996) in the mouth-to-mouth group to 3.5% (1996 - 1999) in the tube plus mask and 3.7% (1999 - 2003) in the bag plus mask group.

The stillbirth rate was not recorded for the control group. It was 18.4% in the tube plus mask group and 12.4% in the bag plus mask group.

The authors concluded that the trained VHWs reduced neonatal deaths. In addition, the use of a bag and mask reduced deaths more than the use of a tube and mask.

The measure of benefit used was the deaths avoided.

No discounting was carried out as the costs were incurred during less than two years. The costs of a tube plus mask and a bag plus mask were given. It was assumed that each delivery that used this equipment used one unit. The costs were not based on the actual costs incurred but on the unit cost of this equipment. No other costs were estimated. No price year was given.

No statistical analysis of the costs was carried out.

No indirect costs were estimated.

US dollars ($).

No sensitivity analysis was carried out.

Thirty-one deaths were avoided by the introduction of VHWs using either a tube and mask or a bag and mask. If the VHWs had used a bag and mask rather than a tube and mask, 30 additional stillbirths would have been avoided during the study period.

The cost of each tube and mask was $10. The cost of each bag and mask was $30. No other costs were estimated.

The cost-effectiveness of using a bag and mask compared with mouth-to-mouth resuscitation was $13 per averted death. It was not absolutely clear, but it seems that the incremental measure of effect included pooled data for the tube as well as bag and mask for the study period and the stillbirth data extrapolated for the study period.

The use of a bag and mask by village health workers (VHWs) reduces mortality in neonates. The cost of each life saved was $13.

The comparator used in the effectiveness analysis, which was relevant to the economic analysis, was implicitly justified. It had been proven to be best practice in the area given the resources available.

The source of the effectiveness data was a comparative study with historical controls. A randomised controlled trial would have been a better design. The reported data showed changes in outcomes during the intervention period, which illustrates the disadvantages of this kind of study. The authors were aware of this drawback. Since all deliveries were included in the study, the neonates must have been representative of the study population. There was no assessment of comparability at baseline. There were drawbacks in the effectiveness analysis: not all effectiveness outcomes were presented clearly for the three kinds of interventions.

The measure of health benefit in the cost-effectiveness ratio for bag plus mask compared with mouth-to-mouth resuscitation appears to have combined tube plus mask and bag plus mask data, although this was not entirely clear. Both showed similar effectiveness with regard to asphyxia deaths, but it may have been better to have extrapolated the data for the bag and mask only for the study period.

From the cost perspective adopted (i.e. the health system), only one cost component was included, that of the equipment. The cost of training the VHWs was not included. The unit cost of the equipment was given, but not the mean cost per delivery of having this equipment available. The resource use quantities were taken from a single study, while the prices were taken from the authors' setting. No other sources were used for the resource quantities. No statistical, sensitivity or any other kind of analysis of the quantities or prices was carried out. Neither the price year, nor the currency in which the prices were expressed, were given.

The authors made some comparisons of their results with those from other studies. The issue of generalisability to other settings was not addressed. The authors did not present their results in a way that made it easy to evaluate their conclusions, and their conclusion on the cost of a life saved was not derived clearly. The economic evaluation was presented more as an afterthought in the paper. The authors reported other limitations of their study. First, they only looked at deaths and not the effects on morbidity of asphyxia. Second, they did not have data on all relevant effectiveness outcomes for all the time periods studied. Finally, the measurement of asphyxia is likely to be imprecise in community-based field studies, despite attempts to make it consistent.

The authors clearly thought that home-based neonatal care with the use of bag and mask was a cost-effective strategy. They emphasised that semi-skilled VHWs would need to be recruited and trained, and that that may be a difficult undertaking.

Supported by the John D and Catherine T McArthur Foundation, the Ford Foundation, Saving Newborn Lives, Save the Children USA, and the Bill and Melinda Gates Foundation.