Quality

AMA discharge linked to increased readmissions, discontinuity of care


 

Background: AMA discharges are common (1%-2% of all U.S. discharges) and disproportionately affect vulnerable patient populations, specifically those of lower socioeconomic status and the uninsured. Previous studies have been insufficiently powered to assess the effects of AMA discharge on 30-day readmission rates at a national level.

Dr. Chase J. Webber is a hospitalist at Vanderbilt University Medical Center, Nashville, Tenn.

Dr. Chase J. Webber


Study design: Retrospective cohort.

Setting: Community and teaching hospitals in 22 states.

Synopsis: With use of the 2014 Nationwide Readmissions Database of 23,110,641 index hospitalizations of patients 18 years or older, this study found that AMA discharge occurred with 1.3% of admissions. AMA discharge was associated with greater than twice the odds of 30-day readmission, compared with routine discharge. Of patients discharged AMA, 20.2% had an unplanned readmission within 30 days, compared with 10.1% of patients discharged routinely (OR, 2.25; 95% CI, 2.20-2.30; P less than .001).

Patients who were discharged AMA had almost 20 times the odds of undergoing repeat AMA discharge at readmission (OR, 18.41; 95% CI, 17.46-19.41; P less than .001) and twice the odds of presenting to a different hospital (OR, 2.35; 95% CI, 2.22-2.49; P less than .001). The study did not capture readmissions in a different state than that of the index hospital and was limited to the 22 states participating in the 2014 Readmissions Database.

Bottom line: Discharge AMA is associated with significantly higher odds of 30-day readmission, subsequent AMA discharge and presentation to another hospital, compared with routine discharge.

Citation: Kumar N. Burden of 30-day readmissions associated with discharge against medical advice among inpatients in the United States. Am J Med. 2019 Jun;132(6):708-17.

Dr. Webber is a hospitalist at Vanderbilt University Medical Center, Nashville, Tenn.

Next Article:

   Comments ()