Skip to main content
. Author manuscript; available in PMC: 2013 Jul 1.
Published in final edited form as: Med Care. 2012 Jul;50(7):569–577. doi: 10.1097/MLR.0b013e31824e319f

Table 3.

Pre- and Post-Reform Use of High Referral Rate Procedure by Area Income & Race/Ethnicity Cohorts

Cohort Pre-reform Procedure Rate Post-reform Procedure Rate Overall Post-Reform Change (%) in Procedure Rate Excess Overall Change among Minorities/Lower Area Income Cohorts Net Change (%): Post-reform Change (%) in procedure rate among non-elderly due to health reform
# procedures
/10,000		 95%
confidence interval		 # procedures
/10,000		 95%
confidence interval		 Difference in % Change between Non-White/White or Lower Area Income/Highest Income p-value
(Difference=0)		 % Net
Change		 95%
confidence interval
By Area Income Cohorts
Non-elderly (age 40–64 years)
 Low 156 [154,159] 169 [167, 172] 8% 4% 0.04 13% [9%, 17%]
 Medium 151 [149, 154] 166 [164, 169] 10% 6% <0.001 15% [6%, 24%]
 High 154 [153, 156] 161 [159, 162] 4% Reference 2% [−3%, 8%]
Elderly (age >=70 years)
 Low 302 [295, 308] 284 [278, 291] −6% −7% <0.001
 Medium 354 [347, 361] 340 [333, 346] −4% −5% 0.002
 High 365 [360, 370] 369 [365, 374] 1% Reference
By Race/Ethnicity Cohorts
Non-elderly (age 40–64 years)
 Hispanics 118 [113, 123] 146 [141, 151] 23% 16% <0.001 22% [5%, 38%]
 Blacks 149 [143, 155] 180 [174, 186] 21% 14% <0.001 5% [−20%, 30%]
 Whites 157 [155, 158] 167 [166, 168] 7% Reference 7% [5%, 10%]
Elderly (age >=70 years)
 Hispanics 199 [181, 216] 202 [186, 218] 2% 3% 0.926
 Blacks 234 [218, 251] 268 [251, 285] 14% 15% 0.003
 Whites 355 [352, 359] 353 [349, 356] −1% Reference

Notes:

  1. Pre-reform=[Oct 2004, Jun 2006] and post-reform=[Jan 2008, Sept 2009]
  2. Procedure rates, adjusted for sex and age differences, are estimated by direct standardization method.
  3. Overall change (%) in procedure rate is the simple % change in the pre- and post-reform procedure rates. Difference in overall change by race/ethnicity and area income are obtained by comparing overall change.
  4. Net change (%) in procedure rate among non-elderly due to health reform are calculated from a separate (county fixed effects) Poisson regression with a difference-in-difference specification using combined non-elderly and elderly cohort data. Confidence intervals are based on bootstrapped standard errors (N=1,000 iterations) to adjust for clustering of observations within county.
  5. Area Income cohorts are defined as: low=lowest quartile zip codes, medium=second quartile and high=top two quartiles