… yet another research notebook entry.
I’m going to sprint through some older literature on this topic. But before I do, let me make one comment. Many studies in this area relate rehospitalization to some measure of the number of frequency of past hospitalizations. As expected, the correlation is positive and statistically significant. After all, frequent past rehospitalizations is very similar to the outcome of interest: rehospitalization. But of what use is knowing the value of the correlation between the two? I suppose if one wanted to know if a given patient is likely to be rehospitalized then this information is of some use. But is that the most common question one might ask in this area?
Not me. What I’d ask is, how can we help this patient not be rehospitalized. I’d want to know about how modifiable treatment factors relate to rehospitalization. Past history of hospitalization is not a modifiable treatment factor. It is, in a sense, a measure of severity of illness, though not a perfect one. In general, sicker patients will have had more hospitalizations.
But prior hospitalization rate is also a measure of quality of care (whether inpatient or post-discharge). Patients subject to lower quality care have had more hospitalizations. You know what else is related to quality of care? The outcome variable: whether or not the patient is going to be rehospitalized. In general, if an explanatory variable is very similar to the outcome variable, they are both going to be correlated with the same unobservable factors (like quality, but also unobservable patient characteristics), biasing the estimates. Pro tip: Don’t control for the outcome with the outcome or something very similar to it.
“Length of Stay, Referral to Aftercare, and Rehospitalization Among Psychiatric Inpatients,” by Thompson, Neighbors, Munday, and Trierweiler
This retrospective study explored the interrelationship among aftercare, length of hospital stay, and rehospitalization within six months of discharge in a sample of psychiatric inpatients. Methods: Data were analyzed for 1,481 patients who had received inpatient care at a state psychiatric hospital from November 1991 to July 1994. Logistic regression models were estimated to predict the likelihood of referral to aftercare and of readmission to a hospital within six months of the index discharge. Variables controlled for were patients’ characteristics; psychiatric status at the time of discharge, including length of stay; and the availability of informal support. Results: Sixteen percent of the patients received a referral to aftercare, and about 13 percent of the patients were readmitted within six months of discharge. White patients were twice as likely as African Americans to receive a referral to aftercare. Length of hospitalization and having a diagnosis of schizophrenia were also predictors of referral to aftercare. Referral to aftercare was not shown to mediate the relationship between length of stay and rehospitalization. However, having a schizoaffective disorder, a poor discharge prognosis, and a high number of previous admissions were associated with an increased risk of readmission. No other demographic characteristics were related to readmission within six months of discharge, but referral to aftercare significantly increased the risk of readmission. Conclusions: The study suggested the possibility of racial disparities in referral to aftercare and a complex relationship between referral and rehospitalization. Both these findings warrant further investigation that gives particular attention to individual-level indicators of need and system-level barriers to and facilitators of psychiatric care.
“Associated factors of rehospitalization among schizophrenic patients,” by Suzuki et al.
The purpose of the present study was to identify the associated factors of rehospitalization in [Japenese] schizophrenic patients [circa 1998]. A case-control study was conducted. The cases consisted of rehospitalized patients ( n = 67) and controls selected from the outpatients who were matched by age, gender, and the period after the last discharge ( n = 62). In the multiple logistic regression analysis, no clinic visits in the second month prior to entry, the number of clinic visits in the previous month, and junior high school graduation as education level were significantly ( P < 0.01) associated with rehospitalization after controlling their present function as assessed by the Global Assessment of Functioning. Close monitoring of clinic visits and outreach service appear to be important in preventing rehospitalization of schizophrenic patients. These identified modifiable factors suggest further needs for development and implementation of integrated mental health services in the community.
Background: Although present findings about frequent users of psychiatric inpatient services vary from study to study, some potentially important predictors of frequent use were extracted. The purpose of this study was to examine the potentially contributory factors of frequent use of psychiatric inpatient services by [German] schizophrenia patients [between 1998 and 2000] and to test the influence single factors have in an overall model. Methods: A total of 307 schizophrenia patients were interviewed five times with intervals of 6 months. Data were collected about service receipt and health care costs, strength of primary diagnosis and comorbidities, as well as about patients’ needs for care and satisfaction with care. Patients with three or more psychiatric admissions within a 30- month period were defined as frequent users. Results: According to this criterion, 12% of the study population were frequent users. Compared with ordinary users, these patients accounted for significantly higher costs in hospital- and community-based care. Important predictors for frequent use of psychiatric inpatient services were the number of previous hospitalizations and current scores of psychopathology. In addition, a longitudinal analysis showed the importance of social factors for the use of psychiatric inpatient care. Therefore, a number of the frequent users’multiple admissions could also be caused by social problems. Conclusions: The mental health system should, thus, provide well-directed community- based resources,which give frequent users support to solve their social problems.
The study tests whether psychiatric services utilization may be predicted from administrative databases without clinical variables equally as well as from databases with clinical variables. Persons with a psychiatric hospitalization at an urban medical center were followed for 1 year postdischarge (N = 1384.) Dependent variables included statewide rehospitalization and the number of hours of outpatient services received. Three linear and logistic regression models were developed and cross-validated: a basic model with limited administrative independent variables, an intermediate model with diagnostic and limited clinical indicators, and a full model containing additional clinical predictors. For rehospitalization, the clinical cross-validated model accounted for twice the variance accounted by the basic model (adjusted R 2 = .13 and .06, respectively). For outpatient hours, the basic cross-validated model performed as well as the clinical model (adjusted R 2 = .36 and .34, respectively.) Clinical indicators such as assessment of functioning and co-occurring substance use disorder should be considered for inclusion in predicting rehospitalization.
“Prediction of readmission of psychiatric inpatients,” by Feigon and Hays
Using demographic and episode-based variables this study attempted to predict which patients would require frequent psychiatric hospitalizations. Records of 943 patients were randomly selected from 14,649 admissions and examined for a 5-yr. period following initial admission. Sex, ethnicity, and age at first admission were not significantly related to readmission. Among the demographic variables, marriage was inversely related to readmission. Involuntary commitment and a longer length of stay at the original admission were associated with a higher rate of readmission. Despite the results being stadstically significant, the small covariances of these effects indicate little clinical utility in the prediction of readniission for an individual patient
This study examined clinical correlates of rapid readmission to a psychiatric inpatient service (less than 3 months after discharge) compared to delayed readmissions (3–12 months) in first-admission patients diagnosed with schizophrenia, bipolar disorder with psychosis, and major depression with psychosis. After reviewing the clinical records and research summaries of all patients who were readmitted within 1 year of discharge, we compared the two readmission groups with respect to demographic and clinical characteristics and subsequent clinical course. Rapid readmission was significantly associated with instability of clinical condition at first discharge (especially mood symptoms) and, to a lesser degree, with failure to prescribe specific medication for affective psychosis patients. Regardless of duration of community tenure, readmission was strongly associated with medication nonprescription or discontinuation. The results suggest that managed care protocols aimed at preventing rapid readmission may require specific symptom assessment and pharmacotherapeutic intervention during the initial hospitalization. Readmission can be used as a quality indicator of both clinical processes (hospital and outpatient care) so long as duration of community tenure prior to readmission is taken into account.
“Factors Associated With Readmission to a Psychiatric Facility,” by Bernardo and Forchuk
This study examined patient-related factors that were associated with readmission to a tertiary care psychiatric hospital in Canada. The charts of a random sample of 200 patients were reviewed from an index discharge date in 1991 through subsequent rehospitalizations over the next three years. Eighty-eight patients (44 percent) were readmitted at least once. The only variable that significantly differentiated patients who were readmitted from those who were not was a history of admission. System variables or factors that are not patient related, such as staff attitudes and perceptions, may contribute to readmission and thus may warrant further exploration.
All emphasis above added by me.