Risk factors for long-stay in an Italian acute psychiatric ward: a 7-year retrospective analysis

Background: In West during the last decades, the phenomenon of "bed blockers" has been more frequently investigated, probably because of increasing economic constraints in the management of public health. According to most authors, the lack of rehabilitation facilities, organizational problems within the hospital, the long wait for medical consultations and diagnostic procedures would be the main causes of "delayed discharge". Early studies were carried out in long-term care, rehabilitation and post-acute geriatric wards. In Psychiatry, the few studies on this topic highlighted a wide range of causes, including both patient conditions and organizational health system problems. In Italy, the problem of psychiatric delayed discharges has become more pressing after the 180 Law, which established the closure of all psychiatric hospitals and implemented psychiatric wards inside General Hospitals to admit only 15 acute patients for a very short period. Purposes: To highlight the phenomenon of long-stay in an acute psychiatric ward and to relate it to demographic, clinical and organizational variables. Methods: The survey was conducted in the 15-bed public psychiatric ward of Modena (Italy). All admissions were retrospectively collected from the database of the Department from 1 January 2005 to 31 December 2011 (3981 hospitalizations with an average stay of 12.49 days). Demographic data, clinical variables, inpatient care problems, discharge programs were statistically related to the duration of admissions (survival analysis: log-rank test, Kaplan-Meier curves). The 3981 hospitalizations were divided into two groups according to the 90\ub0 percentile of duration: < 27 days (n=3575) and 65 27 days (n=406) and the variables of the two groups were compared (multiple logistic regression). Secondary analysis was conducted on the subgroup of the longest hospitalizations further divided into two groups according to the 90\ub0 (from 27 days to < 36 days) and 95\ub0 percentile ( 6536 days), in order to find out variables related (survival analysis: log-rank test; multiple logistic regression test). Results: The longest hospitalizations ( 65\uf02027 days) represent 11% of all admissions during the observation period. When all variables are compared to the duration of hospitalizations, most of them are statistically significantly related to the length of hospitalizations, but, when statistical analysis was focused on the comparison between the two groups of the longest hospitalizations, a smaller number of variables (\u201cgender\u201d, \u201cage\u201d, \u201crehabilitative programs\u201d, \u201cextra-psychiatric clinical activities\u201d, \u201cpharmacotherapy\u201d and \u201caggressiveness of patient\u201d) were identified by survival analysis as statistically significant correlates of long-stay (log-rank test), whereas only \u201cfemale gender\u201d and aggressiveness pf patient\u201d were the variables statistically significantly related to the length of hospitalizations evidenced by multivariate logistic regression analysis. Conclusions: Our results suggest that a wide range of factors may be responsible for the delayed discharges in psychiatry as most previous studies have already shown. However, only few factors were related to the longest duration of hospitalization and, among these, aggressiveness was the only one statistically significant correlate to long-stay in all statistic tests. This data confirms the clinical observation that aggressive behaviour can be sufficient by itself to explain the difficulty of discharging