• Psychiatric readmission

    The first of many papers I intend to read in the area of inpatient readmission for mental health conditions is a 2007 review article by Janet Durbin et al. titled “Is Readmission a Valid Indicator of the Quality of Inpatient Psychiatric Care?” The review itself is a helpful guide to the literature, but the conclusion is necessarily noncommittal on the question posed in the title.

    Although the authors restricted the review to research published in the last decade, most of the studies were based on data from the mid-90s and predated several important changes in service delivery, including the use of newer antipsychotic medications and the expansion of community care, especially intensive community support. The evidence base was small and mainly from the United States, and measures of hospital care processes varied widely and were typically poorly defined. Readmission is a function of multiple factors, yet most of the studies used fairly simple analytic models that only controlled for demographics and diagnoses. With cohort designs, there is always a risk that uncontrolled and unmeasured variables may have accounted for readmission, especially related to post discharge experiences. Most of the study samples presented a mixed diagnostic picture and the analyses did not assess whether there were differences in risk and protective factors across diagnostic groups.


    • I have started skimming this as well. I thought it could not be questioned that inpatient readmission was a predictor of poor care quality … and then I realized that this was psych inpatient only (as opposed to readmission to a general hospital), and for patients under age 65 who had no medical comorbidities complicating the psych admission.

      I believe that for most seniors with severe mental illness, most hospital admissions don’t have an SMI diagnosis coded as the primary diagnosis (so they are coded as secondary). As most seniors have medical comorbidities, this indicates that psych conditions are complicating factors.

      Furthermore, my understanding is that whatever you think about how psych readmissions relate to care, it would be best for people with SMI to not be hospitalized. They would likely prefer it (better than having a psychotic break, be dragged to an unfamiliar place, be poked and prodded, have their freedom of movement restricted, etc). It would likely be cheaper. And the piece below reviews literature (mainly EU context, I think) showing that treatment adherence and outcomes are better in the community setting.