Characterizing long-term outcomes following AMA discharges after assault-related penetrating trauma.

Abstract

Patients discharged against medical advice do not receive adequate treatment and have a greater risk of readmission. This study assessed the rate of discharges against medical advice following assault-related penetrating trauma, with secondary aims to evaluate long term pre/post-injury hospitalizations and mortality. Adult assault-related penetrating injuries admitted to a Level 1 Trauma Center were identified in the prospectively maintained database. Chart review was conducted for hospitalizations ± 5 years from index injury and statewide mortality data was used to identify deaths outside of hospital care. Out of a total of 1,744 assault-related penetrated injuries, 3.2% (52/1630) of survivors discharged against medical advice. Reasons for discharge against medical advice included: unknown (38%), home/child/family/pets (25%), unhappy with care/restrictions (23%), and work/money/other (13%). Post-discharge mortality did not differ between routine (6.5%) and against medical advice discharge (3.9%). Against medical advice and routine discharge had similar rates of any hospitalization (38.5 v 28.2%) and trauma hospitalization in prior 5-years (35 v 36%). However, significantly more against medical advice discharges had prior hospitalizations involving drug or alcohol abuse (65 v 38%), but not mental health diagnosis (55 v 55%). Significantly more against medical advice discharges have post-injury hospitalizations compared to routine discharges (48 vs 26.5%); however, include similar rates of repeat traumatic injury (36 v 32%). Those with against medical advice discharges were significantly more likely to have prior hospitalizations involving drug or alcohol abuse and significantly higher rates of post-injury hospitalizations. However, we did not see an increase in repeat traumatic injury or post-discharge mortality in those with against medical advice discharges when compared to those with routine discharges.