Table 3.

Regression Estimates of Survival on Inputs for Alternative Specifications

	Input	Period of Analysis	Adj. for Diffusion	(A) All Hospitals	(B) Larger Hospitals (N > 50)
1	Log(Expend)	86–04	No	−0.015 (0.003) [Undefined]	−0.026 (0.004) [Undefined]
2	Log(DRG)	86–04	No	0.017 (0.004) [$301,000]	−0.016 (0.006) [Undefined]
3	Log(DRG)	86–04	Continuous Measure	0.022 (0.004) [$227,000]	−0.009 (0.006) [Undefined]
4	Log(DRG)	86–04	Diffusion Quintiles	0.023 (0.004) [$216,000]	−0.009 (0.006) [Undefined]
5	Log(DRG)	86–04	Hospital Fixed Effects	0.053 (0.003) [$94,000]	0.032 (0.005) [$155,000]
6	Log(DRG)	86–94	Hospital Fixed Effects	0.069 (0.004) [$72,000]	0.044 (0.013) [$113,000]
7	Log(DRG)	95–04	Hospital Fixed Effects	0.043 (0.004) [$115,000]	0.029 (0.005) [$171,000]

Notes: See notes to Table 2. Each row of the table reports the coefficient on a measure of inputs from separate regressions of risk-adjusted 1-year survival on inputs using different specifications as indicated. Column (A) reports estimates for all hospitals, while Column (B) is limited to hospitals with at least 50 AMI admits in the year. All regressions control for year dummies. The cost-effectiveness ratios are in brackets and assume 2004 average costs of $26,063 and an average length of life among incremental survivors of 5.25 years.