Table 4.

Estimates of the Effect of the Premier Hospital Quality Demonstration on Risk-Adjusted Outlier Classification

	AMI Outlier			Heart Failure Outlier			Pneumonia Outlier			CABG Outlier
Description	PHQID Effect	n	R2	PHQID Effect	n	R2	PHQID Effect	N	R2	PHQID Effect	N	R2
Hospital fixed effects	0.004 [14.2%]	14,540	0.191	0.006 [41.8%]***	16311	0.149	0.004 [24.9%]***	17,349	0.146	0.016 [11.3%]	6,379	0.215
	(0.004)	{2,479}		(0.002)	{2,780}		(0.002)	{2,957}		(0.010)	{1,136}
Hospital fixed effects in PHQID-eligible subsample	0.000 [−0.7%]	1,948	0.267	0.003 [17.3%]	2,052	0.184	0.003 [19.0%]	2,099	0.235	−0.013 [−10.2%]	1,024	0.312
	(0.005)	{325}		(0.004)	{343}		(0.003)	{351}		(0.017)	{175}
	std(AMI Outlier)			std(Heart Failure Outlier)			std(Pneumonia Outlier)			std(CABG Outlier)
DDD
Stroke	0.013 [1.3%]	14,519	0.087	0.001 [0.1%]	16,260	0.097	0.003 [0.3%]	17,289	0.063	0.048 [4.8%]	6,368	0.068
	(0.035)	{2,476}		(0.011)	{2,772}		(0.009)	{2,951}		(.037)	{1,134}
Gastrointestinal	0.022 [2.2%]	14.516	0.163	0.015 [1.5%]	16,267	0.025	0.010 [1.0%]	17,307	0.025	0.061 [6.1%]	6,364	0.161
hemorrhage	(0.038)	{2,476}		(0.008)	{2,775}		(0.006)	{2,955}		(0.039)	{1,135}

***p<.01. **p<.05. *p<.10.

Percentage change (semielasticity) in risk-adjusted outlier payment is displayed in brackets [ ].

Heteroskedasticity-consistent standard errors are displayed in parentheses ().

Unique numbers of hospital units are displayed in {} under model n.

PHQID and reference conditions are z-transformed in the DDD models.

AMI, acute myocardial infarction; CABG, coronary-artery bypass grafting; DDD, difference-in-difference-in-differences; PHQID, premier hospital quality incentive demonstration.