Against Medical Advice and Improved Heart Failure Care

Background: The hospital to home transition provides a critical opportunity for ensuring progressive heart failure (HF) care; absent for those patients (pts) who signed out against medical advice (AMA). We sought to determine the characteristics of pts hospitalized with HF who leave AMA compared with pts discharged (d/c) via other modes. Method and Results: Retrospective analysis of pts admitted from April 2003 to Dec 2008 with HF. Of the 3450 pts, 91% black, mean age 68 ± 34; mean ejection fraction (EF) 35%±18; 2% signed out AMA. Six percent were discharged to alternate care facilities (UCF) and the remainder d/c to home. Differences by gender, EF and re-hospitalization rates separated the three d/c groups as shown in table 1. Age also differed such that the percent of pts older than 60 were 25, 69, 82 in the AMA, home, and UCF groups, respectively. The AMA group appeared sicker with lower EF, higher BNP level and more renal insufficiency. Twenty one percent of the pts who left AMA were re-hospitalized within 30 days. This is in contrast to the re-hospitalization rate of 14% for pts d/c to home and the 19% for pts d/c to an UCF. Although the number of pts signing out AMA is limited the overall sample size is large and the appraised period extensive, over 5 years.

	AMA, D/C Home and D/C Alternate Care Facility
	%	AMA (2%)	∗p-value	D/C Home (92%)	∗∗p-value	Alternate Facility (6%)
	Sex (female%)	25%	<0.001	57%	0.013	66%
	Age (mean)	54 ± 14	0.003	68 ± 35	0.021	74 ± 14
	Mean EF	29 ± 17	0.023	35 ± 17	0.11	37 ± 19
	Median BNP	1769	<0.001	1160	0.009	1413
	Creat ≥ 2	43%	0.002	24%	NS	28%
	DM %	33%	NS	44%	NS	37%
	HTN %	77%	0.013	89%	NS	87%
	^CAD	26%	NS	20%	0.01	29%
	Re-hospitalization % (≤ 30 days)	21%	0.13	14%	0.06	19%

∗ p-value: AMA vs. D/C Home; ∗∗p-value: Alternate Care Facility (includes nursing home, rehab, hospice) vs. Home; ^CAD: CABG, PCI, h/o MI, increased troponin.

Conclusions: Evaluation of a predominantly black inner city population hospitalized with HF reveals that pts who leave AMA are re-hospitalized, greater than one in five, within 30 days. This may be an underestimation of the extent of the problem since a pt's admission to another facility would not be captured in our registry. Pts leaving AMA were younger, more likely male, and with features consistent with worse systolic dysfunction and prognosis. Our study has potential policy implications and further study of these high risk pts who leave AMA is warranted.