RESUMO
BACKGROUND: Length of stay (LOS) for inpatient psychiatric services is an important factor with serious drawbacks when it is extended more than needed. Impacts on economy, social functioning, and stigma can hamper improvement and affect the patients' experiences on future mental healthcare. Predictions of which patients have a higher chance for prolonged LOS have been extensively researched. Previous systematic reviews found consistent predictors of both longer and shorter LOS. However, they do not provide an estimate from the pooled effect sizes. Furthermore, to our knowledge, there is no meta-analysis on the influence of these factors. The primary objective of this study will be to provide point estimates on the effect sizes of all studied predictors of the LOS of psychiatric inpatients. METHODS: We will conduct a systematic search in PubMed, MEDLINE, EMBASE, and PsycINFO for observational studies evaluating the effect size of independent factors on the length of stay of psychiatric inpatients. Prospective and retrospective cohorts that assess the influence of predictors through the reporting of standardized regression coefficients will be included. We will provide a qualitative synthesis of the findings from each study and perform a meta-analysis from pooled regression coefficients that were adjusted for other variables or confounders in order to obtain a point estimate and confidence interval for all factors extracted from the included studies. DISCUSSION: The results from this study may provide more accurate predictions for mental health institutions, psychiatrists, mental health service providers, patients, and families on the prognosis regarding the length of stay for needed inpatient care. This information may be used to anticipate individuals with a higher chance for prolonged hospitalization to plan the necessary interventions for these specific situations. Considering both the benefits and disadvantages of longer and shorter stays, the pooled estimates for independent factors may be used by mental healthcare providers and patients for informed decision-making. The results from this study will also update results presented in previous studies and identify the strengths and limitations from the current available evidence. SYSTEMATIC REVIEW REGISTRATION: PROSPERO ID CRD42020172840.