Elder, J. P., T. Louis, et al. (1992). "The use of diarrhoeal management counselling cards for community health volunteer training in Indonesia: The HealthCom Project." Journal of Tropical Medicine and Hygiene 95(5): 301-308.

Background: The Indonesian Ministry of Health relies on a network of over a million kader (community health volunteers) to bring primary health care the village level.

Methods: In West Java, the Department of Health's Control of Diarrhoeal Disease (CDD) Program recently carried out an extensive research and development effort to produce effective job aids for the kader in CDD and a training programme to teach their use. A set of counselling cards was produced to provide kader with a tool to diagnose and treat diarrhoea and teach the proper use of ORS. Researchers conducted a controlled evaluation in which they measured the cards' effectiveness through observations of kader performance and interviews with mothers they had counselled. In the intervention group, 15 kader underwent two days training in the use of the cards when diagnosing and advising treatment for cases of diarrhoea in their villages. The 16 control kader received comparable CDD training without the cards. Each group provided lists of local mothers they pledged to counsel during the coming weeks. Follow-up interviews were held with these mothers to test their level of knowledge on CDD and to observe their ability to mix ORS properly.

Results: Significant performance differences between the intervention kader and mothers, and the control kader and mothers, were demonstrated. The intervention kader were consistently more accurate in their diagnoses and recommendations for treatment with a mean of 83% accuracy vs 68% for the control kader. Mothers counselled by the intervention kader also prepared ORS significantly better than the mothers counselled by the control kader, with 97 vs 74% accuracy. Ministry of Health staff in West Java, Indonesia, and staff from 2 US organizations compared data on 15 community health volunteers (kader), who had undergone 2 days of training on the use of cards to advise mothers of children with diarrhea, with data on 16 kader who did not use the counseling cards. 81.7% of the intervention kader correctly followed the 9 diagnostic steps compared to 30.5% of controls (p .001). 78.3% of intervention kader correctly performed the 13 counseling steps for mild diarrhea whereas only 77% of the controls did (p .05). For severe diarrhea, 89.7% of intervention kader correctly followed the diagnostic steps, while just 25.6% of the controls did (p .005). They also did considerably better than controls for chronic diarrhea or dysentery (73.9% vs. 20.9%; p .005). Conversely, they did not perform the steps for moderate diarrhea better than controls. All intervention kader properly communicated the effects of diarrhea to mothers, compared to 84.1% of controls (p = .005). They were also more likely to show the mothers educational materials (84.8% of intervention mothers vs. 34.1% of control mothers; p .001), how to mix the oral rehydration salts (ORS) (89.1% vs. 72.7% p .05), and how to administer the ORS solution to their child (93.5% vs. 77.3%; p .05). Yet, they did not perform better than controls in advising mothers about administering fluids, soft foods, ORS, and going to the hospital. They were more likely to advise them about breast feeding, however (80.4% vs. 52.3%; p .005). After controlling for education and length of service, intervention kader still did significantly better than controls in giving diarrhea treatment directions (p .05). Intervention mothers were more likely to use boiled water (97.8% vs. 81.8%; p = .014), shake the ORS packet (100% vs. 77.3%; p .001), and stir the ORS solution with a spoon (100% vs. 81.8%; p = .002) when mixing the ORS than the controls. They were not more likely to use the correct glass than the controls, however. After controlling for education, age, and family size, intervention mothers still did a much better job at mixing ORS than controls (p .001).