Development of a readiness for hospital discharge assessment tool for stroke patients accuracy in predicting hospital readmission: a prospective cohort study

Objective: A readiness for hospital discharge assessment tool was developed to determine the risk factors predicting readmission and improve accuracy in predicting hospital readmission within thirty days among stroke patients. Method: A prospective cohort study among 348 first-time stroke patients, recruited from neurosurgery wards and stroke units of university hospitals in the central regions of Thailand, between February and September 2020 were followed from discharge until thirty days. Data were gleaned from an extensive literature review and interviews with stroke experts to generate tool component consisting of 23 items. The tool tested for content validity and reliability by content validity index (CVI) and Cronbach’s alpha coefficient (). Interrater reliability of internal consistency was assessed by the intraclass correlation coefficient (ICC). The tool was tested for accuracy to predict hospital readmission within thirty days among stroke patients by the receiver operating characteristic (ROC) curve, sensitivity, specificity, positive predictive value, and negative predictive value, with chi-square and logistic regression used to analyze. Results: The readiness for hospital discharge tool. The readiness for hospital discharge composed of: neurological signs, clinical signs and symptoms, and physical function in activity daily living. The content validity index (CVI) .85, Cronbach’s alpha coefficient of the overall scale was .94. Interrater reliability showed the optimal level (ICC = .92). The predictors explained 88.6% of the variance and correctly classified 93.4% of the results. Odds ratio (OR) and confidence interval (CI) of independent risk factors that best predicted readmission within thirty days after hospital discharge for stroke patients were age (OR = 1.024; 95%, CI: 1.006-1.042), comorbidity (OR = 5.036; 95%, CI: 1.376-18.492), dyslipidemia (OR = .600; 95%, CI: .376-.959) and time of illness onset until hospital arrival (OR = 1.742; 95%, CI: 1.082-2.804). The cut point score of readiness for hospital discharge shown by the receiver operating characteristic (ROC) curve was 13.5, sensitivity was 88.9% and specificity 83.7%, with positive predictive value 90.9% and negative predictive value 80.5%.Logit P (readmission) = +18.462 -.529 (bladder control) -2.313 (bathing) -2.325 (self-feeding) -.979 (mobility) -.481 (dysphagia) -1.673 (visual problem) -1.803 (motor power right leg) -.810 (verbal response) -.610 (level of consciousness) Conclusions: Findings identified the risk factors involved in the readmission of stroke patients in Thailand. The tool was proved valid and reliable in terms of patient readiness for hospital discharge, and accuracy in predicting hospital readmission within thirty days among discharged stroke patients. This tool can be used by healthcare providers to assess discharge conditions for stroke patients with greater accuracy.