Patient satisfaction with computer-assisted structured initial assessment facilitating patient streaming to emergency departments and primary care practices: results from a cross-sectional observational study accompanying the DEMAND intervention in Germany

BACKGROUND: Patient numbers in emergency departments are on the rise. The DEMAND intervention aims to improve the efficacy of emergency services by computer-assisted structured initial assessment assigning patients to emergency departments or primary care practices. The aims of our study were to evaluate patient satisfaction with this intervention and to analyse if reduced patient satisfaction is predicted by sociodemographic data, health status or health literacy. METHODS: We conducted a cross-sectional patient survey in emergency departments and co-located primary care practices. Each intervention site was planned to participate for two observation periods, each with a duration of one full week. Study participants were recruited by the local staff. The patients filled out a written questionnaire during their waiting time. Patient satisfaction was assessed by agreement to four statements on a four point Likert scale. Predictors of patient satisfaction were identified by multilevel, multivariable logistic regression models adjusted for random effects at the intervention site level. RESULTS: The sample included 677 patients from 10 intervention sites. The patients had a mean age of 38.9 years and 59.0% were women. Between 67.5% and 55.0% were fully satisfied with aspects of the intervention. The most criticised aspect was that the staff showed too little interest in the patients’ personal situation. Full satisfaction (“clearly yes” to all items) was reported by 44.2%. Reduced patient satisfaction (at least one item rated as “rather yes”, “rather no”, “clearly no”) was predicted by lower age (odds ratio 0.79 for ten years difference, 95% confidence interval 0.67/0.95, p = 0.009), presenting with infections (3.08,1.18/8.05,p = 0.022) or injuries (3.46,1.01/11.82,p = 0.048), a higher natural logarithm of the symptom duration (1.23,1.07/1.30,p = 0.003) and a lower health literacy (0.71 for four points difference, 0.53/0.94,p = 0.019). CONCLUSIONS: The patients were for the most part satisfied with the intervention. Assessment procedures should be evaluated a) regarding if all relevant patient-related aspects are included; and whether patient information can be improved b) for patients with strong opinions about cause, consequences and treatment options for their health problem; and c) for patients who have problems in the handling of information relevant to health and healthcare. TRIAL REGISTRATION: German Clinical Trials Register (https://www.drks.de/drks_web/setLocale_EN.do) no. DRKS00017014. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12875-022-01825-5.