Quick diagnosis units: predictors of time to diagnosis and costs

Financial crisis has forced health systems to seek alternatives to hospitalization-based healthcare. Quick diagnosis units (QDUs) are cost-effective compared to hospitalization, but the determinants of QDU costs have not been studied. We aimed at assessing the predictors of costs of a district hospital QDU (Hospital Plató, Barcelona) between 2009 and 2016. This study was a retrospective longitudinal single center study of 404 consecutive outpatients referred to the QDU of Hospital Plató. The referral reason was dichotomized into suggestive of malignancy vs other. The final diagnosis was dichotomized into organic vs nonorganic and malignancy vs nonmalignancy. All individual resource costs were obtained from the finance department to conduct a micro-costing analysis of the study period. Mean age was 62 ± 20 years (women = 56%), and median time-to-diagnosis, 12 days. Total and partial costs were greater in cases with final diagnosis of organic vs nonorganic disorder, as it was in those with symptoms suggestive or a final diagnosis of cancer vs noncancer. Of all subcosts, imaging showed the stronger correlation with total cost. Time-to-diagnosis and imaging costs were significant predictors of total cost above the median in binary logistic regression, with imaging costs also being a significant predictor in multiple linear regression (with total cost as quantitative outcome). Predictors of QDU costs are partly nonmodifiable (i.e., cancer suspicion, actually one of the goals of QDUs). Yet, improved primary-care-to-hospital referral circuits reducing time to diagnosis as well as optimized imaging protocols might further increase the QDU cost-effectiveness process. Prospective studies (ideally with direct comparison to conventional hospitalization costs) are needed to explore this possibility.