Optimizing patient-reported sociodemographic measures in health care - a national example of the development and evaluation of a practical standardized instrument for collecting patient data in studies.

BACKGROUND: The development of patient-oriented health care is continuous subject to old and new challenges. To better identify patient needs, studies are necessary and standardized data collection is essential to describe, analyze and compare data. Associated demographic variables remain a topic of discussion in health services research. Common approaches either differ in relevant sociodemographic aspects, or their standardization is too detailed to be feasibly used in studies with a clinical or health services research focus. Our aim was to develop and evaluate a comprehensive and practical instrument for collecting sociodemographic patient data in studies. METHODS: A variant (interdisciplinary expert panel) of the nominal group technique was used as a consensus method, in addition, a focus group was held and a 2-phase pretest according to Prüfer and Rexroth (method: "Think Aloud") was used for evaluation. In the first phase we conducted 6 interviews with potential patients. The main phase was planned to enroll at least 200 patients to test the instrument with a quantitative sample as recommended for testing patient-reported outcome measures. It took place in eight departments of the University Medical Center Hamburg-Eppendorf. Exploratory analyses were conducted using SPSS. The interdisciplinary expert panel revised the instrument. RESULTS: In the main phase, 290 (51%) of 574 approached patients participated [median (minimum; maximum) age: 40 (19; 86) years; 55% female]. Of these respondents 66%/ 31%/ 3% considered the length of the questionnaire to be appropriate/ too long/ too short; 96% found it understandable. Seven questions on living and income situation (commented at least three times) showed a need for clarification and modification. Non-responders optionally provided reasons for decline of which 7 clusters could be formed (predominantly: time). Final standardized variables: biological sex; year/country of birth; marital status/biological children; living/household type/persons/age; highest school/ qualification/current profession; household income/specialties/persons; country of birth mother/father; nationalities/residence status/native languages/German skills. CONCLUSIONS: Patient involvement is helpful in the development of patient survey instruments. Our instrument is based on the patients' perspective and understanding of these measurements. The aim is to further strengthen patient-centered health care. The revised instrument will be used in studies. The expert group will continue to identify and implement socio-demographic developments.