WP 8

Health economic analysis policy in Europe and South Asia

WP leader: Pablo Rebollo / Anita Patel 

The aim of WP8 was to assess the feasibility of data collection to inform a full-scale cost-effectiveness study, and to provide preliminary estimates of the cost-effectiveness of a complex intervention (dietary supplementation plus lifestyle advice –WP3) compared to lifestyle advice alone for pregnant women with vitamin deficiencies in Bangladesh.

The objectives of the economic evaluation were:

1.     To test the feasibility of collecting health economics data;

2.     To estimate the cost of the intervention per participant;

3.     To estimate the use of health care resources (and associated costs) by study participants in the intervention and control groups;

4.     To provide preliminary estimates of cost-effectiveness of dietary supplementation plus lifestyle advice compared to lifestyle advice alone.

5.     To identify factors that potentially predict health care costs in the trial sample.

Outcomes of the health economics analysis included:

  1. Estimate of the cost of intervention;
  2. Estimate of the cost of healthcare services per participant, including and excluding the cost of the intervention;
  3. Estimate of the total cost per participant (cost of intervention plus cost of healthcare services use plus cost of travel);
  4. Identification of potential predictors of costs;
  5. Estimate of the incremental cost-effectiveness ratio, linking total cost and the effectiveness outcome (subject to further analysis).

The pilot study provided estimates of intervention costs and health care costs in the intervention and control groups. Since this pilot study was not powered to detect significant differences between the intervention and control groups, our results were not compared statistically. We identified several issues associated with data collection that it would be important to account for when designing future evaluations of the intervention in other settings. For example, lack of detailed records concerning the inputs and costs associated with delivering tailored nutritional advice as part of the intervention – collection of these records in future would require further specific data collection to enable reliable cost estimates to be made; lack of specific data collection to inform the distinction between research-related health care tests/investigations and those that would be part of routine care – again this requires the use of specific data collection tools or more details in a trial protocol to set out clearly what additional tests/investigations would take place and  in what context; problems with coding of missing data – this would need to be addressed through appropriate database set-up and training of relevant research staff.  Further analyses will be undertaken on receipt of additional data from our collaborators. 

Acceptability and feasibility of the trial

Acceptability: Both the local investigators and the participants opted for oral compared to injectable dosing of micronutrients. The local investigator preference was also for intermittent supervised oral dosing of vitamin D to ensure 100% concordance with medication.  It was also not thought best practice by the obstetric units and local investigators to randomise pregnant women who had normal levels of both vitamin D and B12 and therefore only those with vitamin deficiencies were randomised to oral replacement. Similarly, the dieticians felt it important that the pregnant women allocated to not taking oral supplementation (control group 1) should have personalised advice on supplementing their diet with foods containing the relevant vitamin.

Feasibility: It is clear that the original protocol cannot be used for a larger scale trial in the current setting (Dhaka), and seems unlikely that it can be exported to another setting.  However, the modified protocol, could be considered for a future trial assuming the primary endpoint of replenishment of the individual vitamin has been achieved (results are known but cannot be divulged until publication).  If the primary end-point is not achieved for vitamin D then daily dosing of vitamin D will need to be considered with a change in trial design.