The contribution of case mix, skill mix and care processes to the outcomes of community hospitals

Abstract

Abstract Background The rapid increase in the proportion of older people underscores the need for new organizational models to face the unmet needs of frail patients with multiple conditions. Community hospitals (CHs) could be a solution to tackle these needs and foster integration between acute and primary care. The aim of this study was to investigate which patients' characteristics and which care processes affect clinical outcomes, in order to identify who could benefit the most from CH care and the best skill mix to deliver in this setting of care. Methods This study included all patients aged ≥65 and discharged in 2017 from the 16 CHs of Emilia-Romagna, northern Italy. Data sources were the regional CH informative system and hospital discharge records. CH skill mix and processes of care were collected with a survey; 3 non-respondent CHs were excluded. The study outcome was in-hospital variation of the Barthel index (BI) (≥10 vs. <10). We performed a 2-level random-intercept logistic regression analysis, and used the variance partition coefficient (VPC) to quantify the proportion of BI improvement that lay at CH level. Results Of the 13 CHs, 7 admitted ≥150 patients, 8 had a general practitioner medical support model, and 6 had >12 nurses' working hours/week/bed. Overall, 53% of the patients had a BI improvement ≥10 (4% to 71% across CHs). The patient case mix (i.e. baseline BI, female, older age, transfer from acute care) explained a portion of variability across CHs, as shown by the VPC that decreased from 0.32 to 0.26. Skill mix and processes of care were not associated with BI change, and the VPC resulting from controlling for these variables was virtually unchanged (0.28). Conclusions Patients' characteristics explained part of between-CH variation in BI improvement. Professional skill mix and processes of care, albeit fundamental to achieve appropriate care and respond to the unmet needs of the frail elderly, did not account for differences in CH-specific outcomes. Key messages A combination of quantitative and qualitative methods might better explain the outcome variability across intermediate care services. Multidisciplinary CH teams and services can be helpful to address the unmet needs of older people, but further studies are necessary.