Journal Article
Improving the quality of hospital care for children by supportive supervision

a cluster randomized trial, Kyrgyzstan

Objective

To determine whether periodic supportive supervision after a training course improved the quality of paediatric hospital care in Kyrgyzstan, where inappropriate care was common but in-hospital postnatal mortality was low.

Methods

In a cluster, randomized, parallel-group trial, 10 public hospitals were allocated to a 4-day World Health Organization (WHO) course on hospital care for children followed by periodic supportive supervision by paediatricians for 1 year, while 10 hospitals had no intervention. We assessed prospectively 10 key indicators of inappropriate paediatric case management, as indicated by WHO guidelines. The primary indicator was the combination of the three indicators: unnecessary hospitalization, increased iatrogenic risk and unnecessary painful procedures. An independent team evaluated the overall quality of care.

Findings

We prospectively reviewed the medical records of 4626 hospitalized children aged 2 to 60 months. In the intervention hospitals, the mean proportion of the primary indicator decreased from 46.9% (95% confidence interval, CI: 24.2 to 68.9) at baseline to 6.8% (95% CI: 1.1 to 12.1) at 1 year, but was unchanged in the control group (45.5%, 95% CI: 25.2 to 67.9, to 64.7%, 95% CI: 43.3 to 86.1). At 1 year, the risk ratio for the primary indicator in the intervention versus the control group was 0.09 (95% CI: 0.06 to 0.13). The proportions of the other nine indicators also decreased in the intervention group (P < 0.0001 for all). Overall quality of care improved significantly in intervention hospitals.

Conclusion

Periodic supportive supervision for 1 year after a training course improved both adherence to WHO guidelines on hospital care for children and the overall quality of paediatric care.

Journal Article This peer-reviewed research is available free of charge. UNU-WIDER believes that research is a global public good and supports Open Access.