A Cluster-Randomized Trial to Test Sharing Histories as a Training Method for Community Health Workers in Peru

Women naturally communicate using life narratives. Through systematic recall and sharing memories of their own childbearing and child rearing experiences, community health workers (CHWs) become engaged and empowered to change their own and other mothers’ health behaviors. Training CHW with sharing histories can improve capabilities as change agents for better child health.

Attendance rates at monthly workshops were 82% or higher for five of the six module topics, as shown in Table A. These were the workshops provided for community health workers (CHW) by a trained health care provider in the primary health care facility. However, these rates do not fully reflect the learning by CHW. For CHW who missed a workshop, community supervisors met with the CHW in their community to reteach the module content. b. Pretest and posttest scores for each module topic We applied knowledge tests to each CHW immediately before and after each training module. Tests were applied verbally to CHWs in a private setting with responses noted by the evaluator who was usually the health care provider trainer. CHW scores showed highly significant improvements from pretest to posttest for all module topics.
The result of community follow-up training can be seen in the posttest scores for diarrhea and infant growth & nutrition which are higher than those of other module topics (Table B), even though their workshop attendance was less than for the other topics. This was especially true for the experimental group as compared to the control group.

II. COST INFORMATION ON THE COMMUNITY HEALTH INTERVENTION FOR SCALING-UP BY THE GOVERNMENT
The principal investigator did a detailed cost estimate for the effective introduction of this full integrated community health model into a government primary health care (PHC) system. Costs are relatively low based on the assumption that Ministry of Health (MOH) staff will be involved as program supervisors and CHW trainers at no additional salary cost. The model assumes involvement of local municipalities in financial support of stipends for community supervisors (CS) as well as non-financial incentives for CSs and CHWs.
Therefore, cash costs to be assumed by the MOH to implement the model include: (1) training costs for trainers, CSs, and CHWs; (2) start-up cost of reproducing training manuals and flipcharts for eight modular topics given as job aids to trainers, CHWs, CSs and each PHC facility for counseling mothers; (3) on-going costs of reproducing monitoring and supervision forms; and (4) the partial cost of external human resource involvement which is limited to one expert training consultant to train MOH trainers for a small proportional of her time-effort in one district based on population size of the district.

Results
For a district of 25,000 inhabitants and seven PHC facilities, cost per child under age five would be USD $16.57 per child for the first year to start up. The cost is divided into $11.37 from the MOH and $5.20 for the local government contribution for community supervisor stipends. Annual maintenance costs per child under age 5 would be reduced to USD $14.55 divided between $9.35 from the MOH and $5.20 for the local government contribution. This cost includes one full-time MOH staff person per district in charge of managing the community health program, paid by the MOH; in addition, a stipend will be paid to 10 CSs by the municipal government.
Local in-kind resources not included in the costs are the percent effort of a health service network (Red de Salud in Spanish) supervisor, 10 health staff who serve as CHW/CS trainers, 167 CHWs, and 45 PHC staff who receive orientation-training to support the community health model.
For a district of 10,000 inhabitants with 3 PHC facilities and with fewer children <5 years, the estimate is marginally more at $17.34 per child per year total. The alternative cost is evidently much lower and close to zero because there is no community health program to cost out.
One could consider $16.57 (start-up) and $14.55 (maintenance) per child per year, or slightly over $1.00 per month, as expensive in one country but not in another. Costs would also vary by country according to the local cost of inputs.
Current practice is to have no community health program, or one that is in place but is not effective. In our case, the cost was the same for the model in both study groups. The experimental training clusters had better results than the other due to better training methods, but these did not increase the cost.