Overall scores as an alternative to global ratings in patient experience surveys; a comparison of four methods
1 Netherlands Institute for Health Services Research (NIVEL), P.O. Box 1568, 3500 BN Utrecht, The Netherlands
2 Institute for Quality in Health Care, Diemen, The Netherlands / TRANZO (Tilburg University), Tilburg, The Netherlands
BMC Health Services Research 2013, 13:479 doi:10.1186/1472-6963-13-479Published: 19 November 2013
Global ratings of healthcare by patients are a popular way of summarizing patients’ experiences. Summary scores can be used for comparing healthcare provider performance and provider rankings. As an alternative, overall scores from actual patient experiences can be constructed as summary scores. This paper addresses the statistical and practical characteristics of overall scores as an alternative to a global rating in summarizing patient survey results.
Data from a 2010 patient experience survey for approximately 12,000 nursing home residents (7.5% of all Dutch nursing home residents at the time) from 464 nursing homes in the Netherlands (25% of the Dutch nursing homes) was used. Data was collected through specifically designed standardized interview surveys. The respondents’ scores for 15 established quality indicators (or composites) for nursing home care were used to calculate overall scores for each nursing home, using four different strategies. The characteristics of the overall scores were compared against each other and with the respondents’ global rating.
The individual indicators showed stronger associations with each of the four overall strategies than with the global ratings. Furthermore, the dispersion of the overall scores across nursing homes was greater. Differences between overall scores appeared limited.
Overall scores proved more valid than global ratings as a summary of the indicator scores, and also showed more pronounced differences between nursing homes. Because of the limited statistical differences between the strategies, and for practical reasons, a straightforward averaging of quality indicator scores may be preferred as an overall score.