Abstract
This analysis of data from US Renal Data System examined patient- and center-level variables for patients who did and did not undergo cardiac stress testing to determine which were associated with stress testing 18 months prior to renal transplantation.
Approximately 40% of patients with end stage renal disease (ESRD) have ischemic heart disease.1 Given this high prevalence, many renal transplant centers assess patients for ischemic heart disease in the pretransplant evaluation. Once accepted as a transplant candidate and placed on the waiting list, there is little guidance regarding subsequent stress testing in asymptomatic individuals. We examined the predictors of and variation in stress testing in the 18 months prior to renal transplantation.
Methods
The Washington University Human Research Protection Office exempted this study from institutional review board oversight. From the United States Renal Data System,2 we identified patients with ESRD who underwent a first renal transplant between July 1, 2006, and November 30, 2013, were 40 years or older, and had primary Medicare insurance for at least 18 months prior to the transplant. Patient-level and facility-level characteristics were compared between patients who did and did not undergo stress testing using the t test and Pearson χ2 test for continuous and categorical variables, respectively. Nonnormal and ordinal variables were summarized by the median (first quartile, third quartile) and compared using the Mann-Whitney U test.
Variables with P < .10 on univariate analysis were entered into a multivariable logistic regression model that included both patient-level and center-level variables. A hierarchical model approach was used with the center as a second-level clustering variable. The median odds ratio (MOR) was determined to describe the likelihood of undergoing a stress test if a patient hypothetically moved from a center with a lower probability of having a stress test to one with a higher probability.3 All analyses were conducted in SAS version 9.3 (SAS Institute Inc), and P < .05 (2-sided) was considered significant. Data were analyzed from June 7, 2017, to February 1, 2019.
Results
In 26 694 patients from 217 facilities, the percentage who underwent stress testing in the 18 months prior to renal transplantation by center ranged from 11.1% to 96.2% (median, 60.4%; interquartile range, 31.2%) (Figure). Individuals who underwent stress testing were older with higher percentages of white patients, female patients, recipients of kidneys from living donors, congestive heart failure, valvular disease, peripheral vascular disease, diabetes, hypertension, prior myocardial infarction, and coronary artery disease (Table).
Figure. Percentage of 24 694 Patients Undergoing Stress Testing at 217 Facilities.
Table. Baseline Characteristics.
| Characteristic | No. (%)a | ||
|---|---|---|---|
| Overall (N = 24 694) | No Stress Test (n = 9967) | Stress Test (n = 14 727) | |
| No. of transplants performed at facility during the index transplant year, median (IQR) | 107.0 (1.0-385.0) | 101.0 (3.0-385.0) | 114.0 (1.0-385.0) |
| Age at transplant, mean (SD), y | 57.35 (9.7) | 56.00 (9.7) | 58.26 (9.6) |
| Female | 9434 (38.2) | 3886 (39.0) | 5548 (37.7) |
| Race/ethnicity | |||
| Native American | 372 (1.5) | 184 (1.8) | 188 (1.3) |
| Asian | 1459 (5.9) | 537 (5.4) | 922 (6.3) |
| Black | 8900 (36.0) | 3822 (38.3) | 5078 (34.5) |
| White | 13 774 (55.8) | 5335 (53.5) | 8439 (57.3) |
| Unknown/other | 189 (0.8) | 89 (0.9) | 100 (0.7) |
| BMI, mean (SD) | 28.14 (5.4) | 28.10 (5.4) | 28.17 (5.4) |
| Donor type | |||
| Cadaveric/unknown | 21 972 (88.9) | 9016 (90.4) | 12 956 (88.0) |
| Living | 2722 (11.0) | 951 (9.5) | 1771 (12.0) |
| Diabetes | 13 111 (53.1) | 4687 (47.0) | 8424 (57.2) |
| Hypertension | 21 519 (87.1) | 8026 (80.5) | 13 493 (91.6) |
| Tobacco use | 4529 (18.3) | 1593 (16.0) | 2936 (19.9) |
| Prior myocardial infarction | 1311 (5.3) | 253 (2.5) | 1058 (7.2) |
| Coronary artery disease | 7905 (32.0) | 1650 (16.6) | 6255 (42.5) |
| Congestive heart failure | 5173 (20.9) | 1509 (15.1) | 3664 (24.9) |
| Valvular disease | 2996 (12.1) | 617 (6.2) | 2379 (16.2) |
| Peripheral vascular disease | 5416 (21.9) | 1615 (16.2) | 3801 (25.8) |
| Cardiomyopathy | 1669 (6.8) | 421 (4.2) | 1248 (8.5) |
| Cerebrovascular disease | 2304 (9.3) | 657 (6.6) | 1647 (11.2) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); IQR, interquartile range.
P < .001 for all comparisons except female (P = .04) and BMI (P = .30).
Independent predictors of stress testing included white race (odds ratio [OR], 1.13; 95% CI, 1.05-1.21), diabetes (OR, 1.14; 95% CI, 1.05-1.24), hypertension (OR, 1.70; 95% CI, 1.55-1.86), coronary artery disease (OR, 2.66; 95% CI, 2.47-2.88), valvular disease (OR, 1.85; 95% CI, 1.66-2.07), peripheral vascular disease (OR, 1.11; 95% CI, 1.03-1.20), and cardiomyopathy (OR, 1.25; 95% CI 1.09-1.43). The MOR was 2.28 (95% CI, 2.04-2.48).
Discussion
Our study demonstrates that substantial variability exists among centers in use of stress testing in the 18 months prior to renal transplant. With an MOR of 2.28, transplant center was second only to coronary artery disease in determining the use of stress testing,3 suggesting that the decision to perform a stress test was influenced more strongly by hospital culture than by all other individual patient characteristics, including those associated with coronary artery disease. Although it should be noted that while we accounted for a wide range of patient-level variables in performing our multivariable analysis, we cannot exclude the possibility that unmeasured confounders may have influenced the selection of patients for stress testing at each facility.
Variation in care is often attributed to differences in disease prevalence or access to care.4 However, neither of those factors is operative in this population of patients, all of whom had ESRD and were on the waiting list for renal transplantation. Rather, the variability in practice among centers suggests that there is no consensus regarding which patients, if any, should undergo stress testing. The lack of consensus may reflect the cognitive dissonance created by the strong association of ischemia on stress testing with adverse outcomes, the absence of data regarding the benefit of stress testing in patients with ESRD awaiting renal transplantation, and the presence of high-quality data that revascularization does not improve outcomes in other high-risk surgical populations.5 Defining the role of surveillance stress testing in patients awaiting renal transplantation requires randomized clinical trials.
References
- 1.Parfrey PS, Foley RN. The clinical epidemiology of cardiac disease in chronic renal failure. J Am Soc Nephrol. 1999;10(7):1606-1615. [DOI] [PubMed] [Google Scholar]
- 2.US Renal Data System Coordinating Center 2013 Researcher’s Guide to the USRDS Database. https://www.usrds.org/2013/rg/A_intro_sec_1_13.pdf. Accessed April 2, 2019.
- 3.Larsen K, Merlo J. Appropriate assessment of neighborhood effects on individual health: integrating random and fixed effects in multilevel logistic regression. Am J Epidemiol. 2005;161(1):81-88. doi: 10.1093/aje/kwi017 [DOI] [PubMed] [Google Scholar]
- 4.Wennberg JE. Unwarranted variations in healthcare delivery: implications for academic medical centres. BMJ. 2002;325(7370):961-964. doi: 10.1136/bmj.325.7370.961 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.McFalls EO, Ward HB, Moritz TE, et al. Coronary-artery revascularization before elective major vascular surgery. N Engl J Med. 2004;351(27):2795-2804. doi: 10.1056/NEJMoa041905 [DOI] [PubMed] [Google Scholar]