A5083 - 30 Day Readmission Following Weight Loss Surgery: Can Psychological Factors Predict Non-Specific Indications for Readmission?

Abstract

Although bariatric surgery is considered a safe and efficacious treatment option for severe obesity, 5% or more of patients are readmitted due to complications in the first 30 days. Some of these complications relate directly to surgical risks whereas others relate to more vague complaints or patient non-adherence. Such non-specific indications for surgery may reflect risks outside of the surgical procedure. A large literature has demonstrated that bariatric surgery patients are a psychiatrically vulnerable population, though no studies to date have investigated whether pre-surgical psychosocial factors are related to readmission. Patients who had been initially evaluated by our program and readmitted within 30 days post-bariatric surgery during a 4 year period (2012-2015) were identified (n=102). These patients were matched (2:1) to 204 non-readmitted patients on pre-operative BMI (48.1 vs. 48.9 kg/m2), age (48.1 vs. 48.9 years), gender (73% vs. 74% female) and race (67% vs. 65% White). Psychiatric diagnoses, treatment history, binge eating pathology (as measured by the Binge Eating Scale; BES) and psychological testing (Minnesota Multiphasic Personality Inventory-2-Restructured Form; MMPI-2-RF) at intake were compared between the 2 groups using Chi-Square and t-test analyses. Further, amongst those readmitted, the indication for hospitalization was investigated via electronic medical record. Those with specific complications (e.g., leak, infection, DVT; n=68) were further delineated from those with vague complaints (e.g., pain, nausea) or complications that could relate to non-adherence (e.g., dehydration, misuse of medications; n=34). The 3 groups (no readmission, specific indication, non-specific indication) were compared on demographics, psychiatric variables and psychological testing via Chi-Square and analysis of variance (ANOVA).Individuals who were readmitted had less education (13.4 vs. 14.22 years; t=2.15; em>p<.02) than non-readmitted patients. Readmitted patients were less likely to be in outpatient psychiatric care (48%) compared to non-readmitted patients (64%; X2=6.76; p<.02) and were on fewer psychiatric medications compared to the readmitted patients (M=.60 vs. 1.39; t=22.74; p<.001). Although matched on age and gender initially, those with non-specific readmissions were significantly younger (43.4 years vs. 49.25 in specific readmission vs. 48.9 in non-readmitted; F=2.66; p=.05) and more likely to be female (88% vs. 69% vs. 74%; X2=6.08; p=.05). Significant differences were found on the Uncommon Virtues scale of the MMPI-2-RF which reflects a tendency to under-report psychiatric symptoms using a response style in which the patient denies minor imperfections that most individuals are willing to admit (F=13.43; p<.001). Those who were readmitted with non-specific complaints had significantly higher under-reporting scores (M=64.14) compared to those with specific indications (M=59.74) which were greater than those not readmitted (M=55.11). Other MMPI-2-RF group differences and the effect of gender were no longer significant once the effect of under-reporting was removed. No significant differences were found on psychiatric diagnoses, the BES or number of weekly binge episodes although there was a trend for greater history of inpatient psychiatric hospitalizations in those were readmitted (10.3%) compared to those who were not (5.4%; X2=2.45; p=.09). Readmitted patients were more likely to psychologically present themselves in an overly positive manner, particularly among those who had non-specific indications for readmission. The tendency to under-report on psychological testing has been linked with under-reporting during the clinical interview and may also influence the veracity presented with nutrition, medicine and surgery (e.g., denying comorbidities, reporting adherence although not following treatment plan). This under-reporting style may impact the bariatric team's ability to identify risk factors that could be ameliorated prior to surgery. Similarly, patients who are readmitted were less likely to be receiving mental health care and psychopharmacology although they may be more likely to have had a significant past history of psychiatric hospitalization. Such ongoing treatment may increase monitoring and/or adherence. Finally, younger patients were more likely to be readmitted for non-specific reasons. Future research should examine better strategies for identifying patients at risk for readmission and for reducing an under-reporting style during the evaluation period.