Skip to main content
. 2006 Dec;41(6):2267–2289. doi: 10.1111/j.1475-6773.2006.00581.x

Table 5.

Linear Probability (Ordinary Least Squares) and IV Estimates of the Effect of “For-Profit” Status of Initial Dialysis Center on 1-Year Survival, 1996–1999

Estimation Method and Specification Number of Patient Groups Specified Using Differential Distance Instrument Group Partial F-Statistic (p-Value) Hausman F-Statistic (p-Value) Estimated Effect (Std Error, p-Value)
LPM; no covariates NA NA NA −0.0031 (0.00328; .3450)
LPM; Devereaux covariates* NA NA NA −0.01221 (.00319;<.0001)
LPM; Devereaux covariates plus NA NA NA −0.00715 (0.00366; .0511)
LPM; Devereaux covariates plus without state dummies NA NA NA −0.01315 (0.00321;<.0001)
LPM; Devereaux covariates plus without previous health care utilization NA NA NA −0.00759 (0.00368; .0395)
LPM; Devereaux covariates plus without zip code socioeconomic variables NA NA NA −0.00726 (0.00366;0.0475)
LPM; Devereaux covariates plus without distance to nearest dialysis center NA NA NA −00702 (0.00366; .0554)
2SLS 2 18535.00 (<.0001) NA 0.00072 (0.00933; .939)
2SLS 5 7909.59 (<.0001) 0.64 (0.5877) 0.00248 (0.00754; .746)
2SLS 10 3991.69 (<.0001) 0.99 (0.4379) −0.00012 (0.00719; .986)
2SLS 20 2150.53 (<.0001) 0.82 (0.6625) −0.00024 (0.007151; .973)
2SLS 40 1003.79 (<.0001) 1.23 (0.1583) 0.00049 (0.007113; .945)
*

Also specified patient age, gender, comorbidity, and race.

Also specified patient age, gender, comorbidity, race, state of residence, number of inpatient admissions and inpatient days in the year prior to dialysis, zip code-level measures of poverty, per capita income, and percent rural, distance to nearest dialysis center.

LPM, linear probability model; 2SLS, two-stage least squares.