Table 7.

Analysis of the Association between Change In-Hospital Quality Assessment and the Proportion of “Complications” Date Stamped as Being Present on Admission

	AAA		CABG		CEA		PTCA		THR
	OR	p-Value	OR	p-Value	OR	p-Value	OR	p-Value	OR	p-Value
Change to higher quality	1.64	.39	1.20	.46	1.88	.15	1.51	.056	0.77	.48
No change in outlier status	Reference		Reference		Reference		Reference		Reference
Change to lower quality	1.66	.18	1.64	.44	0.86	.18	0.72	.37	0.52	.35

An odds ratio>1 indicates that “complications” were more likely to be coded as being present on admission compared to the reference population. An odds ratio<1 indicates that “complications” were less likely to be coded as being present on admission compared to the reference population. Assuming that misclassifying “complications” as being present on admission accounted for hospitals being reclassified as higher quality outliers using risk adjustment models incorporating date stamp information, then those hospitals would be expected to have an odds ratio>1

CABG, coronary artery bypass graft surgery; PTCA, coronary angioplasty; CEA, carotidendarterectomy; AAA, abdominal aortic aneurysm surgery; THR, total hip replacement.