The factors associated with length of stay on acute care psychiatry inpatient units in St. John's, Newfoundland and Labrador

Background: In St. John's, Newfoundland there are three tertiary care institutions that provide acute inpatient psychiatry services. To date, the factors associated with increased length of stay, in these institutions, has not been systematically studied. -- Some of the psychiatrists who admit patients to these facilities have speculated that the length of stay and the distribution of the factors associated with length of stay varies across the inpatient psychiatric facilities in the city. The accuracy of their opinions has not been examined previously. -- Objectives: Length of stay on acute care hospital units is influenced by many factors such as patient demographics, diagnosis, treatment regime and attitudes of the attending physician. This study has three objectives: 1) to identify the factors associated with length of stay on the acute admission wards in St. John's, Newfoundland, 2) to assess the accuracy of the psychiatrists' opinions that length of stay differs significantly among the three psychiatric institutions in St. John's and 3) to assess the accuracy of the psychiatrist's opinions with respect to the factors that impact on length of stay in St. John's, Newfoundland. -- Method: A chart review was conducted to measure length of stay by institution and to determine which factors were associated with length of stay for inpatient psychiatry units in St. John's. The data represents patients discharged from hospital between April and June 1997 inclusive. The patients were above 19 years of age, were discharged from an acute care inpatient unit during the study period, and had no other admissions during the thirty days prior to the admission or within 30 days of the discharge. Fourteen variables were examined for a potential association with length of stay. Regression analysis was used to identify significant associations between patient, illness and treatment related factors and length of stay. -- Data from the chart review was also used to assess the accuracy of the psychiatrists' opinion that length of stay varies significantly among the acute care admissions of the city. The length of stay was calculated for each admission site and then compared for significant differences. -- Psychiatrists' perceptions of the factors associated with length of stay were gathered by an opinion questionnaire. The questionnaire was administered to all psychiatrists with admitting privileges for the acute psychiatry inpatient sites in St. John's. Descriptive analysis was used to quantify the psychiatrists' opinions. These opinions were then compared to the results of the chart review for factors associated with length of stay. -- Results: Results of the chart review revealed that length of stay did not vary significantly among the three institutions, or between the two types of admission sites. Of the fourteen factors examined for an association with length of stay, patient age, primary diagnosis, presence of a discharge plan, and duration of the discharge plan were significant. The involvement of pastoral care workers in the patient care almost reached the level of significance. -- Results of the survey revealed that the majority of psychiatrists were of the opinion that patient age, gender and race had little impact on length of stay. The age and gender of the treating psychiatrist were also believed to have little impact on length of stay. Level of education, marital status, employment status, income level, and place of residence were believed to have moderate impact on length of stay. The level of patient stress, primary diagnosis, the presence of co-morbid illness, treatment issues, the expertise of the attending psychiatrist and the practice composition of the attending psychiatrist were believed to impact substantially on length of stay. This was not entirely consistent with the results of the chart review. -- Discussion: Four of the fourteen variables were significantly associated with length of stay on the inpatient psychiatry units in St. John's, Newfoundland. Several of these findings are consistent with earlier studies. Of the ten remaining factors that were not significantly related to length of stay, seven had been previously studied and shown to be associated with length of stay in other clinical settings. This suggests that the factors associated with length of stay in St. John's, Newfoundland may be unique to that region and that length of stay literature is not universally applicable. -- The psychiatrists' perception that lengths of stay varied significantly among sites was inaccurate. As well, the psychiatrists' perceptions of factors that are associated with length of stay were not consistent with the findings of the chart review. This discrepancy between opinion and reality warrants further examination if physicians are to be expected to actively reduce length of stay in the future