Properties of the Picker Patient Experience questionnaire in a randomized controlled trial of long versus short form survey instruments.

BACKGROUND: The purpose of this study was to compare the performance of the 15-item Picker Patient Experience questionnaire (PPE-15) when embedded in a short form instrument as compared with a longer form measure. METHODS: A postal questionnaire survey of patients recently discharged from two hospital trusts was carried out. Patients were randomized to receive the PPE-15 in either a four-page or a 12-page survey instrument. RESULTS: A total of 1445 questionnaires were mailed to patients in either four- or 12-page formats. A total of 949 (65.67 per cent) forms were returned. No difference in response rate was found between the two versions of the questionnaire. Item completion and psychometric properties of the PPE-15 were not found to differ significantly between the two arms of the trial. CONCLUSION: In this survey the length of questionnaire in which the PPE-15 was embedded had no impact in terms of response rate or data quality. Consequently, the results suggest that length of questionnaire, up to the 108 items included in the 12-page survey, is unlikely to adversely affect results on the PPE-15.

Factors relating to patients' reports about hospital care for coronary heart disease in England.

OBJECTIVES: All health care providers in England are required to conduct surveys of their patients' experience of health care. Data from such surveys contribute to the 'star rating' performance indicators. However, there are concerns that these subjective measures may be influenced more by characteristics of patients than by true variations in the quality of care. The purpose of this paper is threefold: to determine the relationship between demographic characteristics and an index measure of patients' reported experience; to explore the extent to which patients' experiences may be accounted for by the particular National Health Service (NHS) trust they attended; and to assess how meaningful a summary index is in terms of its ability to discriminate between providers. METHODS: Data from patients in the National Survey of National Health Service Patients treated for coronary heart disease in 194 NHS trusts. Patients were sent questionnaires after discharge, with a covering letter and a prepaid stamped addressed envelope. Up to two reminders were sent to non-responders. Multi-level linear regression models were used to estimate the extent to which patients' experiences differed between trusts and the association of demographic variables with the summary index. RESULTS: In total, 116,872 patients were sent questionnaires, but 3399 proved to be ineligible for the survey. Responses were gained from 84,310 (74.3% of eligible respondents). Age and sex were most strongly associated with reported patients' experiences. However, the actual impact of age and sex on patients' experience is small, accounting for less than 3% of the variance. The proportion of the variance that was accounted for by the hospital trust in which patients were treated was only 5%. CONCLUSIONS: Demographic characteristics do not appear to account for differences between hospital trusts in patients' experience of health care. However, there is considerable variation in patients' experience within each provider. This would suggest that summary indices of patients' experience should not be used to rank providers, although detailed information from patient surveys have a useful role in determining priorities for quality improvement within individual hospitals and for assessing changes over time.