Accesso libero

Extension and Validation of the Self-care Index to Predict Transfer to a Post-acute Care Institution in Internal Medicine Patients / Erweiterung und Validierung des Selbstpflegeindexes bei internistischen Patienten mit dem Ziel eine Verlegung in eine post-akute Nachsorgeinstitution vorauszusagen

INFORMAZIONI SU QUESTO ARTICOLO

Cita

A systematic screening of patients with a need for post-acute care is a helpful support for interprofessional discharge planning teams. We aimed to test self-care abilities, measured by the self-care index (SPI) as predictors of post-acute care transfer and to update the existing SPI prediction model.

We analysed data from a prospective, observational cohort study conducted at the Kantonsspital Aarau between February and October 2013. We updated the SPI model, adding age and gender using a training and validation data set. Logistic regression models were run on the outcome “transfer to a post-acute care facility” and judged based on their AUC (area under curve), AIC (Akaike information criterion), and BIC (Bayesian information criteria) values. ROC curves (receiver operating characteristic) were derived from the models; and cut-points for the linear predictors of the models were defined (thus defining the new scores). Sensitivities and specificities were calculated.

This study included 1372 adult internal medicine in-patients admitted from home, who either returned home or were transferred to a post-acute care institution. The total SPI score was a significant predictor for post-acute care referral (p < 0.001). Including age and gender in the SPI model increased the AUC to 0.85 (training) and 0.84 (validation). An improvement in the AUC by 3% (0.81 [95% CI: 0.77–0.85] to 0.84 [95% CI: 0.80–0.87]), compared to the original SPI was achieved (p = 0.004). The new score reached a sensitivity of 81% and specificity of 74% compared to a sensitivity of 64% and specificity of 84% for the original score.

The extended SPI can be used as a tool for individualised discharge organisation of internal medicine patients with higher accuracy.

eISSN:
2296-990X
Lingue:
Inglese, Tedesco
Frequenza di pubblicazione:
Volume Open
Argomenti della rivista:
Medicina, Medicina clinica, altro