Psychiatric illness and length of stay in general hospitals: do case finding methods matter?

Objective: Several prior studies have investigated whether patients with "non-cognitive" mental disorders (i.e., organic disorders, substance abuse, delirium, and psychotic disorders excluded) have longer Length Of Stay (LOS) than mentally healthy individuals in nonpsychiatric hospital settings. These studies yielded contrasting results. The present paper aims to examine whether methods of psychiatric case finding can explain these differences.
Methods: Using the Clinical Interview Schedule (CIS) and the General Health Questionnaire (GHQ), 462 in-patients of medical, surgical, gynecological, and rehabilitation departments were assessed for the presence of psychiatric disorders.
Results: In multiple regression analysis, all CIS-cases together did not show an association with LOS. Of the diagnostic groups assessed by CIS only major depression showed a significantly prolonged LOS. Using the GHQ sum-score as a continuous variable, LOS was significantly increased while using the GHQ as a dichotomous variable did not show such an association. After removing those suffering from multiple psychiatric diagnoses (such as major depression co-morbid with organic mental illness), none of the case definitions showed a significant association with LOS.
Conclusions: It seems that different case finding methods yield different results concerning the association of psychiatric disorders with LOS. When interpreting these results the small size of some subsamples must be taken into consideration.