Table 3. Characteristics Associated With Experience of Potential Care Cascade Among Fee-for-Service Medicare Beneficiaries Receiving Preoperative EKG for Cataract Surgery.
| Characteristic | Potential Cascade, No. (%) | Adjusted OR (95% CI)a | |
|---|---|---|---|
| Experienced (n = 1978) | Did Not Experience (n = 10 430) | ||
| Age, y | |||
| 66-74 | 938 (47.4) | 5920 (56.8) | 1 [Reference] |
| 75-84 | 859 (43.4) | 3794 (36.4) | 1.42 (1.28-1.57)b |
| ≥85 | 181 (9.2) | 716 (6.9) | 1.54 (1.29-1.84)b |
| Female sex | 1277 (64.6) | 6868 (65.8) | 0.93 (0.84-1.03) |
| Race | |||
| White | 1643 (83.1) | 8890 (85.2) | 1 [Reference] |
| Black | 134 (6.8) | 562 (5.4) | 1.13 (0.92-1.39) |
| Hispanic | 101 (5.5) | 458 (4.4) | 1.17 (0.93-1.47) |
| Other | 92 (4.7) | 520 (5.0) | 0.91 (0.71-1.15) |
| Medicaid enrollment | 145 (7.3) | 575 (5.5) | 1.19 (0.97-1.46) |
| Setting of residence | |||
| Metropolitan | 1672 (84.5) | 8674 (83.2) | 1 [Reference] |
| Micropolitan | 157 (7.9) | 978 (9.4) | 0.91 (0.75-1.10) |
| Suburban | 84 (4.2) | 440 (3.2) | 1.09 (0.85-1.39) |
| Rural | 65 (3.3) | 338 (3.2) | 1.12 (0.85-1.48) |
| Elixhauser condition count, mean (SD) | 1.39 (1.48) | 1.04 (1.31) | 1.18 (1.14-1.22)b |
| Physician ordering preoperative EKG | |||
| Primary care physician | 1364 (69.0) | 7557 (72.5) | 1 [Reference] |
| Cardiac specialist | 374 (18.9) | 1701 (16.3) | 1.26 (1.10-1.43)b |
| Other | 240 (12.1) | 1172 (11.2) | 1.14 (0.97-1.33) |
| Practice region | |||
| Northeast | 705 (35.6) | 3379 (32.4) | 0.98 (0.82-1.18) |
| South | 634 (32.1) | 3352 (32.1) | 1.03 (0.87-1.22) |
| West | 326 (16.5) | 1877 (18.0) | 1.00 (0.82-1.22) |
| Midwest | 313 (15.8) | 1822 (17.5) | 1 [Reference] |
| Cardiologists per 10 000 residents in HRR, mean (SD)c | 7.57 (2.1) | 7.36 (2.0) | 1.05 (1.02-1.09)b |
Abbreviations: EKG, electrocardiogram; HRR, hospital referral region; OR, odds ratio.
a
Among beneficiaries who received a preoperative EKG, we compared those who did or did not experience a potential cascade using univariate analyses and multivariable logistic regression in which the primary outcome was experience of the cascade and the covariates were all characteristics included in the table and HRR random effects.
b
Statistically significant at P < .05.
c
Dartmouth Atlas, 2011.38