Methods: In accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) we used the following databases: CINAHL, PsycINFO, Web of Science, and PubMed (from January 1, 2012 to December 27, 2017) to identify relevant articles on the relationship between SMI diagnosis and medical/surgical readmissions. We used the National Institute of Health’s Quality Appraisal Tool for Observational Cohort and Cross-Sectional Studies guidelines to appraise studies and assess risk of bias. Data were narratively synthesized and a pooled random effects unadjusted odds ratio was estimated using meta-analysis. Heterogeneity was investigated using subgroup analysis and meta-regression.
Results: Our search yielded 424 articles after removing duplicates. Fourteen met inclusion criteria. All studies were retrospective observational cohort studies. A wide range of medical/surgical cohorts were investigated. Study methodology varied with little agreement on the definition of SMI, data sources, medical/surgical diagnoses or demographic/clinical variables. Twelve studies found significant relationships between SMI and readmissions. Two studies did not support the significant relationship. The meta-analysis showed that people with SMI have greater odds of readmission than people without SMI (pooled OR 1.38, CI 1.23-1.56, I2 = 98.6%). There was heterogeneity in patient cohorts, study methodology, and definition of SMI. No significant possibility of publication bias was detected (Classic fail-safe N = 3480).
Discussion: Our results suggest that patients with SMI have higher rates of medical/surgical readmissions than patients without SMI. Given the prevalence of SMI in patients hospitalized for medical/surgical problems and the heterogeneity of evidence, further research on the relationship between SMI and readmissions is critically needed.
See more of: Invited Posters