574 Assessment of the Relationship Between Mental Illness and Inpatient Length of Stay

Abstract Introduction Mental illness and burn injury requiring hospitalization are uniquely intertwined. While literature has shown that burn patients are more likely to experience acute stress disorder, post-traumatic stress disorder, and depression, the impact of pre-existing mental illness on the duration of inpatient hospitalization for burn patients is unclear. We aim to assess the impact of acute mental illness requiring treatment by a psychiatrist during admission to a burn unit on patient length of stay (LOS). Methods A retrospective analysis was conducted using burn registry data to evaluate patients age 18 to 65 admitted between January 1, 2017 and December 31, 2018. Patients admitted with burns, inhalation injury, frostbite and Steven Johnsons Syndrome were included. Patient age, total body surface area (TBSA), LOS, and psychiatry consult (PC) incidences were collected. Patients were separated into psychiatric (P) and non-psychiatric (NP) groups to compare the impact of PC on LOS. Multiple linear regression analysis was conducted in both groups to evaluate the impact of TBSA and age on LOS. Two-sample t-test assuming unequal variances was conducted to compare age, TBSA, and LOS in both groups. Results Multiple regression in the P (n=71) group indicated both age and TBSA explained 48.4% of the variance of LOS (R²= 0.48, F(2,68)= 31.96, p< 0.005). TBSA significantly predicted LOS (β= 275.9, p< 0.005), but age was insignificant (β= 0.177, p=0.4). Multiple Regression in the NP (n=617) group found that both age and TBSA explained 11% of the variance of LOS (R²= 0.11, F(2,613)=39.9, p< .005). TBSA and age significantly predicted LOS (β = 59.9, p< .005), (β = 0.09, p= 0.05), respectively. T-test assuming unequal variance found there was no significant difference in age for NP group (m= 41.73, SD= 14.02) and P group (m= 43.77, SD= 13.53; t(87)= 1.98, p = 0.11). TBSA held no significant difference between both P (m=0.05, SD= 0.08) and NP (m=0.05, SD= 0.09; t(93)= 1.98, p= 0.94). However, LOS was significantly different for P (m=26.20, SD= ±32.92) than NP (m=9.58, SD= ±17.57; t(75)= 1.99, p < 0.005) Conclusions It is an expected finding that TBSA had a significant impact on LOS, and age did not. Since the TBSA and age between N and NP are not significantly different, the impact of this correlation cannot account for the difference between LOS in the two groups. The correlation between PC and increased LOS may be used to more effectively estimate duration