- Email: [email protected]
Unequal Use of New Technologies by Race: The Use of New Prostate Surgeries (Transurethral Needle Ablation, Transurethral Microwave Therapy and Laser) Among Elderly Medicare Beneficiaries
Outcomes/Epidemiology/Socioeconomics Unequal Use of New Technologies by Race: The Use of New Prostate Surgeries (Transurethral Needle Ablation, Transurethral Microwave Therapy and Laser) Among Elderly Medicare Beneficiaries Xinhua Yu, A. Marshall McBean* and Debra S. Caldwell From the Division of Health Services Research and Policy, University of Minnesota School of Public Health, Minneapolis, Minnesota
Purpose: We compared the availability and use of transurethral microwave therapy, transurethral needle ablation, contact or noncontact laser therapy and transurethral resection of the prostate among elderly black and white Medicare beneficiaries. Materials and Methods: We examined 100% Medicare Inpatient, Outpatient, Carrier and Denominator files of men 65 years old or older who underwent these procedures in 1999 through 2001. White-to-black race rate ratios for each procedure were computed for the entire United States, as well as for a restricted set of counties in which procedures were available to black beneficiaries. Results: A total of 170,067 TURP, 16,953 TUMT, 5,353 TUNA and 12,134 Laser procedures were performed during 3 years. Nationally there was only a 3% difference in the age adjusted TURP rates between white and black men (6.13 and 5.94 per 1,000 person-years, respectively). However, the age adjusted rates for TUMT and TUNA among white men were about twice those among black men (0.63 vs 0.31 and 0.20 vs 0.10 per 1,000 person-years, respectively). Laser rates were 17% higher among white men than among black men (0.44 vs 0.38 per 1,000 person-years). Large geographic variation existed in the new procedure rates. Negative binomial regression analysis confirmed the national findings in those counties in which the procedures were available to black men. Adjusted white-to-black rate ratios were 1.96 (95% CI 1.70 –2.25) for TUMT, 2.33 (95% CI 1.87–2.90) for TUNA and 1.36 (95% CI 1.16 –1.59) for Laser. Conclusions: After controlling for availability, elderly black Medicare beneficiaries were less likely to undergo the new BPH procedures than white beneficiaries, while the usage difference for TURP remained small. Key Words: prostatic hyperplasia; continental population groups; Medicare; surgical procedures, operative; socioeconomic factors
any studies have documented lower rates of surgical procedures among elderly black Medicare beneficiaries than among white men.1– 4 Despite the fact that the prevalence of BPH among elderly black and white men is not significantly different,5 the rate of traditional surgical treatment for BPH, transurethral resection of prostate, has always been lower among elderly black Medicare beneficiaries than among white beneficiaries. However, the racial difference in the use of TURP has decreased during last few decades such that by the end of the 1990s TURP was performed about 10% more frequently among elderly white men than among black men.6 However, little is known about the availability and use of the new BPH procedures TUMT, TUNA and contact or noncontact laser therapy by race. In this study we compared the availability and use of these 3 new procedures among elderly black and white Medicare beneficiaries. Our hypothesis was that the white-to-black
M
Submitted for publication June 8, 2005. Supported by the Centers for Medicare and Medicaid Services Contract HCFA 500-01-0043. * Correspondence: Division of Health Services Research and Policy, University of Minnesota School of Public Health, MMC 97 D369 Mayo Memorial Building, 420 Delaware St., S.E., Minneapolis, Minnesota 55455 (telephone: 612-625-6175; FAX: 612-378-4866; e-mail: [email protected]).
0022-5347/06/1755-1830/0 THE JOURNAL OF UROLOGY® Copyright © 2006 by AMERICAN UROLOGICAL ASSOCIATION
rate ratio for each of the new procedures was equal to the rate ratio for TURP. That is, the rates for TURP by race would provide information on the usual difference in BPH surgical treatment by race.
MATERIALS AND METHODS Data Sources and Case Definition/Identification We used the Medicare 100% Inpatient, Outpatient and Carrier files (use files) from the National Claims History repository, as well as the 100% Denominator files for 1999 through 2001. Identification of patients undergoing surgery or procedures was based on current procedural terminology codes for TURP (52601, 52612, 52614), TUMT (53850), TUNA (53852) and Laser (52647, 52648), and International Classification of Disease version 9 Clinical Modification procedure codes for TURP (60.29), TUMT (60.96), TUNA (60.97) and Laser (60.21). Claims were then extracted from Carrier, Inpatient or Outpatient files if any BPH procedure appeared in the claim. Information from records with identical individual health insurance claim numbers, procedure dates and procedure codes was consolidated to eliminate inappropriate double counting of cases. This process resulted in 328,096 potential cases for analysis.
1830
Vol. 175, 1830-1835, May 2006 Printed in U.S.A. DOI:10.1016/S0022-5347(05)00997-3
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES Beneficiaries included in the study had to be 65 years old or older at surgery, continuously enrolled in Medicare Part A and Part B throughout the study period and reside in 1 of the 50 states or the District of Columbia. Similar to most studies using Medicare administrative data, those enrolled in managed care during the year of surgery as well as those with end stage renal disease or prostate cancer during the study were excluded from analysis. A total of 123,589 patients were excluded, leaving 204,507 for analysis. Information on the type, dates and location of the BPH surgery was obtained from the use files, and the beneficiary age, race and county of residence were obtained from the Denominator files. County level socioeconomic information, namely race specific median household income and the percent of the population with less than a high school education, was obtained from United States Census 2000 Summary File 3. Analytical Methods Since the new BPH procedures were only performed in a limited number of geographic areas, separate analyses were performed for the total United States elderly male population as well as for those geographic areas (counties) in which the new technologies were available (procedure available counties). Availability was functionally defined, that is the county of residence of patients who underwent the procedure, not the county in which the procedure was performed. Furthermore, to reduce confounding due to population structure differences, to be included the counties had to have more than 10 black male beneficiaries residing in them during 3 study years (10 person-years of residence). Race specific procedure rates are reported. There were 3 age groups used in computing age specific rates and age adjusted rates, namely 65 to 74, 75 to 84 and 85 years old or older. Direct standardization was used for age adjustment with the 3-year total population as the standard. Negative binomial regression was used to compute the white-to-black procedure rate ratios adjusted for age group and county socioeconomic information. This model also took into account the heterogeneity within data such as overdispersion (observed variance greater than expected). Parameters were estimated using the generalized estimate equation to adjust for the geographic heterogeneity of procedure rates. Robust standard errors in Proc GENMOD of SAS® version 8.2 were used for constructing tests and confidence intervals. RESULTS From 1999 through 2001, 170,067 TURP, 16,953 TUMT, 5,353 TUNA and 12,134 Laser procedures were performed among black and white Medicare beneficiaries (table 1). The national age adjusted TURP rate among white men (6.13 per 1,000 person-years) was 3% higher than among black men (5.94 per 1,000 person-years). The TURP rates were highest in the 75 to 84-year-old age group for black and white men. The age adjusted rates for TUMT, TUNA and Laser among white men were 0.63, 0.20 and 0.44 per 1,000 person-years, respectively, about twice as high for TUMT and TUNA (0.31 and 0.10 per 100,000 person-years), and 17% higher for Laser (0.38 per 100.000 person-years) than among black men. When all 3 new procedures were combined the white-
1831
TABLE 1. Age specific and age adjusted rates of prostate procedures/1,000 among elderly Medicare beneficiaries, 1999 to 2001 Age Group
No. White Men (rate)
No. Black Men (rate)
TURP: 65–74 78,108 (5.66) 6,384 (5.52) 75–84 66,336 (6.91) 4,166 (6.51) 85⫹ 14,038 (5.75) 1,035 (6.07) Total No. ⫹ age adjusted rate 158,482 (6.13) 11,585 (5.94) TUMT: 65–74 9,397 (0.68) 383 (0.33) 75–84 5,926 (0.62) 193 (0.30) 37 (0.22) 85⫹ 1,017 (0.42) Total No. ⫹ age adjusted rate 16,340 (0.63) 613 (0.31) TUNA: 65–74 3,130 (0.23) 119 (0.10) 75–84 1,785 (0.19) 67 (0.10) 12 (0.07) 85⫹ 240 (0.10) Total No. ⫹ age adjusted rate 5,155 (0.20) 198 (0.10) Laser: 65–74 6,293 (0.46) 474 (0.41) 75–84 4,261 (0.44) 231 (0.36) 45 (0.26) 85⫹ 830 (0.34) Total No. ⫹ age adjusted rate 11,384 (0.44) 750 (0.38) TUMT, TUNA or Laser: 65–74 18,820 (1.36) 976 (0.84) 75–84 11,972 (1.25) 491 (0.77) 94 (0.55) 85⫹ 2,087 (0.86) Total No. ⫹ age adjusted rate 32,879 (1.27) 1,561 (0.79)
Rate Ratio (white/black) 1.03 1.06 0.95 1.03 2.06 2.04 1.92 2.04 2.20 1.77 1.40 1.99 1.11 1.23 1.29 1.17 1.62 1.62 1.55 1.61
to-black ratio of the age adjusted rates was 1.61. Among white and black men the new procedures were more likely performed on the youngest of the elderly. TUMT and TUNA were not performed among residents of all counties (figs. 1 to 3). There were counties in which the TUMT rate was more than twice the national average and counties in which the TUNA rate was more than 5 times the national average. Of the counties 16% had TUMT rates higher than 1.5 per 1,000 person-years. These counties included 11% of the white elderly male population and 6% of the black population. Similarly, for TUNA the counties with procedure rates greater than 1.5 per 1,000 person-years had 3% of the white population and 1% of the black population, and for Laser the counties with procedure rates greater than 1.5 per 1,000 had 6% of the white population and 5% of the black population. As expected the analyses using only the procedure available counties with more than 10 black male residents increased the procedure rates for the new BPH procedures (table 2). However, the white-to-black procedure rate ratios were similar to those using all United States male beneficiaries as the study cohort (table 1), and had little effect on TURP rates among each race group. Table 3 presents the white-to-black rate ratios and 95% CI for the procedure available counties adjusted for the age, geographic variation and county level socioeconomic information. The adjusted rate ratios ranged from 1.36 (95% CI 1.16 –1.59) for Laser to 2.33 (95% CI 1.87–2.90) for TUNA. The rate ratio was 1.65 (95% CI 1.50 –1.81) for TUMT, TUNA and Laser combined. For comparison the adjusted rate ratio for TURP was 1.08 (95% CI 1.04 –1.11).
1832
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES
FIG. 1. County specific rates of TUMT procedures among elderly Medicare beneficiaries, 1999 to 2001
FIG. 2. County specific rates of TUNA procedures among elderly Medicare beneficiaries, 1999 to 2001
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES
1833
FIG. 3. County specific rates of laser procedures among elderly Medicare beneficiaries, 1999 to 2001
DISCUSSION The findings of greatest concern in this study are the 2-fold and higher rates of TUMT and TUNA among white elderly Medicare beneficiaries compared with black beneficiaries. These findings were surprising given the relative parity in TURP rates which we found which has been improving over time.6,7 These differences persisted even after adjusting for socioeconomic status and geographic availability. Healthy People 2010 and the more recent Institute of Medicine report Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care unequivocally confirm the existence of racial disparities in health and health care in the United States.8,9 Access to Health Care in America and the Agency for Healthcare Research and Quality’s National Healthcare Disparities Report, 2004 identified multiple reasons/barriers for differences in health care use.10,11 These include the availability of the service, financial barriers such as health insurance coverage and reimbursement, patient perception including culture, language and acceptance of treatment, and provider beliefs and behaviors. We attempted to control for the availability of the new procedures by studying only counties in which residents had undergone the new procedure and in which there were black beneficiaries. These areas contained a greater proportion of the black male population (94%) than the white male population (82%) (data not shown but available from the authors). This may be because a large percentage of the facilities and providers of the new procedures were located in metropolitan areas (figs. 1 to 3). However, limiting the analysis to those geographic areas did not appreciably re-
TABLE 2. Geography and age adjusted prostate surgery procedure rates/1,000 person- years among elderly Medicare beneficiaries in procedure available counties with more than 10 black male Medicare beneficiaries, 1999 to 2001
TURP: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate TUMT: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate TUNA: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate Laser: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate TUMT, TUNA or Laser: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate
No. White Men (rate)
No. Black Men (rate)
Rate Ratio (white/black)
62,360 (5.62) 53,951 (6.86) 11,201 (5.69)
5,898 (5.41) 3,851 (6.42) 941(6.00)
1.04 1.07 0.95
127,512 (6.09)
10,690 (5.84)
1.04
7,635 (0.76) 4,820 (0.67) 785 (0.44)
379 (0.38) 190 (0.34) 36 (0.25)
2.02 1.95 1.71
13,240 (0.70)
605 (0.35)
1.98
2,711 (0.39) 1,562 (0.30) 205 (0.16)
117 (0.18) 66 (0.18) 12 (0.14)
2.19 1.65 1.15
4,478 (0.33)
195 (0.18)
1.90
5,659 (0.59) 3,789 (0.55) 730 (0.43)
472 (0.50) 230 (0.44) 44 (0.33)
1.19 1.24 1.27
10,178 (0.56)
746 (0.46)
1.21
16,005 (1.44) 10,171 (1.29) 1,720 (0.87)
968 (0.89) 486 (0.81) 92 (0.59)
1.62 1.60 1.49
27,896 (1.33)
1,546 (0.83)
1.60
Geographic adjusted rate ratios were also adjusted for age.
1834
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES
TABLE 3. Age, county socioeconomic status and geography adjusted white/black rate ratios of prostate procedures
TURP TUMT TUNA Laser TUMT, TUNA or Laser
Adjusted Rate Ratio
Lower 95% CI
Upper 95% CI
1.08 1.96 2.33 1.36 1.65
1.04 1.70 1.87 1.16 1.50
1.11 2.25 2.90 1.59 1.81
All value p ⬍0.0001.
duce the black-to-white rate ratios. One possible reason it had little impact is that the practices of the providers of the new procedures serve different patient populations than those who continue to perform only TURP. Defining the practice populations of individual urology practices is beyond the scope of this report. Using Medicare data one can determine the racial composition of the population actually treated by a practice, but defining the actual population at risk as being treated by the practice is not possible, particularly since urology is largely a referral specialty. It is unlikely that financial barriers were the cause of the differences by race. In our study all beneficiaries were covered by Medicare Part A and Part B. In addition, the Medicare program has set the physician payment rate of new procedures lower than that of TURP, reducing the potential out-of-pocket expenses paid to the physician by the beneficiary. The majority of the new procedures were performed in outpatient settings or physician offices (94% of TUMT, 93% of TUNA and 72% of Laser) compared to TURP, which is overwhelmingly performed in an inpatient setting (91%). Thus, in the majority of instances the total cost of new procedures to the patient would not include the first day hospital co-payment ($792 in 2001, for example) which must be paid by hospitalized patients. We calculated the total patient financial responsibility to the urologist and to the institution (hospital, outpatient, ambulatory surgical center, etc) for the different procedures during the study period. The median amounts for inpatient TURP ($954) and outpatient TURP ($973) were higher than the amounts for noninpatient hospital TUMT ($506) and TUNA ($509), and similar to noncontact Laser ($929) and contact Laser ($965). These costs would likely encourage the use of 2 of the new procedures among those of lower socioeconomic status, yet these procedures were the ones for which we found the greatest racial differences. While we controlled for socioeconomic status in our analysis, other personal factors and perceptions may also have a key role in decision making on the part of patients and the use of the new procedures. Cultural differences as well as attitudes and beliefs regarding the new procedures may differ between black and white men, and are cited as possible reasons for differences in health care use.12,13 However, they are not always sufficient to explain the differences.13 Kressin et al have studied racial differences in patient preference regarding cardiac catheterization.14,15 In their initial study they found little difference in measures of health beliefs and attitudes between black and white patients at the Department of Veterans Affairs. Subsequently they reported that health beliefs as well as sociodemographic and clinical variables did not explain differences in the use of cardiac catheterization between black
and white patients. Unfortunately we know of no information in the literature regarding racial differences in attitudes or beliefs regarding the new BPH procedures, and the Medicare administrative data provide little insight into these areas. Thus, we could not investigate whether they had a role in the use of the new prostate procedures. Because the procedures are new, physician perception of the appropriateness of the procedure for each individual patient may be more important than in well established procedures. This communication between patients and physicians is likely to affect the decision to undergo TUMT, TUNA or Laser. However, it is difficult to understand why urologists would recommend the new procedures to one racial group compared to the other unless there is a difference in disease severity or another clinical factor at surgery. As summarized by Wei et al there is no variation in the prevalence of obstructive urinary symptoms by race/ethnicity, but overall lower urinary tract symptoms appear to occur with greater severity in black men, which might lead to a recommendation or a patient preference for TURP.16 We believe that the large, national sample and the geographic analysis are strengths in this study. The latter helped us control for the restricted geographic availability of the new procedures. However, there are some limitations. It is difficult to examine the severity of BPH based on claims data. The lack of information about supplemental insurance coverage and payments prohibited us from examining the actual out-of-pocket cost of the procedures for patients.
CONCLUSIONS To our knowledge this is the first study to report the unequal use of new surgical BPH procedures by race. These differences should be monitored, particularly if studies such as the Minimally Invasive Surgical Therapies Treatment Consortium for Benign Prostatic Hyperplasia trial demonstrate that the new procedures have advantages compared to other treatments.17 From 1999 through 2001 white elderly Medicare beneficiaries were 61% more likely to undergo TUMT, TUNA or Laser for BPH than black beneficiaries. For comparison TURP was performed among white men at a 3% greater rate than among black men. There were large geographic variations in procedure rates. After restricting analysis to those procedure counties with more than 10 black male beneficiaries, and adjusting for geographic variation and area socioeconomic status, racial differences in the use of the new procedure persisted.
Abbreviations and Acronyms BPH Laser TUMT TUNA TURP
⫽ ⫽ ⫽ ⫽ ⫽
benign prostatic hyperplasia contact or noncontact laser therapy transurethral microwave therapy transurethral needle ablation transurethral resection of the prostate
REFERENCES 1. McBean, A. M. and Gornick, M.: Differences by race in the rates of procedures performed in hospitals for Medicare beneficiaries. Health Care Financ Rev, 15: 77, 1994
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES 2. Gornick, M. E., Eggers, P. W., Reilly, T. W., Mentnech, R. M., Fitterman, L. K., Kucken, L. E. et al: Effects of race and income on mortality and use of services among Medicare beneficiaries. N Engl J Med, 335: 791, 1996 3. Ayanian, J. Z.: Race, class, and the quality of medical care. JAMA, 271: 1207, 1994 4. Epstein, A. M. and Ayanian, J. Z.: Racial disparities in medical care. N Engl J Med, 344: 1471, 2001 5. Platz, E. A., Smit, E., Curhan, G. C., Nyberg, L. M. and Giovannucci, E.: Prevalence of and racial/ethnic variation in lower urinary tract symptoms and noncancer prostate surgery in U.S. men. Urology, 59: 877, 2002 6. Wasson, J. H., Bubolz, T. A., Lu-Yao, G. L., Walker-Corkery, E., Hammond, C. S., Barry, M. J. et al: Transurethral resection of the prostate among Medicare beneficiaries: 1984 to 1997. J Urol, 164: 1212, 2000 7. Lu-Yao, G. L., Barry, M. J., Chang, C. H., Wasson, J. H. and Wennberg, J. E.: Transurethral resection of the prostate among Medicare beneficiaries in the United States: time trends and outcomes. Prostate Patient Outcomes Research Team (PORT). Urology, 44: 692, 1994 8. Healthy People 2010, United States Department of Health and Human Services, 2000. http://www.healthypeople.gov/ document. Accessed June 4, 2005 9. Smedley, B. D., Stith, A. Y. and Nelson, A. R.: Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, D. C.: The National Academies Press, 2003 10. Millman, M.: Access to Health Care in America. Washington, D. C.: The National Academies Press, 1993 11. National Healthcare Disparities Report, 2004. Washington, D. C.: Agency for Healthcare Research and Quality, 2005 12. Whittle, J., Conigliaro, J., Good, C. B. and Joswiak, M.: Do patient preferences contribute to racial differences in cardiovascular procedure use? J Gen Intern Med, 12: 267, 1997 13. Ayanian, J. Z., Cleary, P. D., Weissman, J. S. and Epstein, A. M.: The effect of patients’ preferences on racial differences in access to renal transplantation. N Engl J Med, 341: 1661, 1999 14. Kressin, N. R., Clark, J. A., Whittle, J., East, M., Peterson, E. D., Chang, B. H. et al: Racial differences in health-related beliefs, attitudes, and experiences of VA cardiac patients: scale development and application. Med Care, suppl., 40: 172, 2002 15. Kressin, N. R., Chang, B. H., Whittle, J., Peterson, E. D., Clark, J. A., Rosen, A. K. et al: Racial differences in cardiac catheterization as a function of patients’ beliefs. Am J Pub Health, 94: 2091, 2004
1835
16. Wei, J. T., Calhoun, E. A. and Jacobsen, S. J.: Benign prostatic hyperplasia. In: Urologic Diseases in America. Edited by M. S. Litwin and C. S. Saigal. United States Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Washington, D. C.: United States Government Publishing Office, NIH Publication No. 04 –5512, chapt. 2, pp. 43– 67, 2004 17. National Institute of Diabetes and Digestive and Kidney Diseases. Minimally Invasive Surgical Therapies (MIST) Treatment Consortium for Benign Prostatic Hyperplasia (BPH). Available at http://www.niddk.nih.gov/patient/mist/mist.htm. Accessed September 30, 2005
EDITORIAL COMMENT Yu et al demonstrate disparities in access to minimally invasive therapies for BPH from 1999 to 2001. Specifically they demonstrate that black men were less likely to undergo TUMT, TUNA or laser therapy than white men. These findings beg the greater question of why these racial differences in access to care exist. The authors suggest a number of different explanations, including differences in patient populations of “early-adopter” urologists, cultural differences as well as differences in attitudes and beliefs among black and white men, or differences in clinical presentation that might make these procedures more appropriate in some patients than others. While all of these explanations are interesting, it is my belief that the racial differences in procedure rates found in this study are due almost exclusively to differences in the patient populations of the urologists who first adopted this technology. The reader should remember that the study period was relatively early in our experience with minimally invasive surgical techniques for BPH. At that time many or all of the minimally invasive procedures performed in a given county likely would have been undertaken by a single or a small number of providers. These providers likely had unique clinical practices which attracted select patient populations, the majority of which apparently were white. While the racial disparities noted in the current report are disheartening, they underscore the need for further research. Specifically, now that the technology has diffused out more completely, do these racial differences in access to care still exist? David F. Penson Departments of Urology and Preventive Medicine Keck School of Medicine University of Southern California Los Angeles, California
Purpose: We compared the availability and use of transurethral microwave therapy, transurethral needle ablation, contact or noncontact laser therapy and transurethral resection of the prostate among elderly black and white Medicare beneficiaries. Materials and Methods: We examined 100% Medicare Inpatient, Outpatient, Carrier and Denominator files of men 65 years old or older who underwent these procedures in 1999 through 2001. White-to-black race rate ratios for each procedure were computed for the entire United States, as well as for a restricted set of counties in which procedures were available to black beneficiaries. Results: A total of 170,067 TURP, 16,953 TUMT, 5,353 TUNA and 12,134 Laser procedures were performed during 3 years. Nationally there was only a 3% difference in the age adjusted TURP rates between white and black men (6.13 and 5.94 per 1,000 person-years, respectively). However, the age adjusted rates for TUMT and TUNA among white men were about twice those among black men (0.63 vs 0.31 and 0.20 vs 0.10 per 1,000 person-years, respectively). Laser rates were 17% higher among white men than among black men (0.44 vs 0.38 per 1,000 person-years). Large geographic variation existed in the new procedure rates. Negative binomial regression analysis confirmed the national findings in those counties in which the procedures were available to black men. Adjusted white-to-black rate ratios were 1.96 (95% CI 1.70 –2.25) for TUMT, 2.33 (95% CI 1.87–2.90) for TUNA and 1.36 (95% CI 1.16 –1.59) for Laser. Conclusions: After controlling for availability, elderly black Medicare beneficiaries were less likely to undergo the new BPH procedures than white beneficiaries, while the usage difference for TURP remained small. Key Words: prostatic hyperplasia; continental population groups; Medicare; surgical procedures, operative; socioeconomic factors
any studies have documented lower rates of surgical procedures among elderly black Medicare beneficiaries than among white men.1– 4 Despite the fact that the prevalence of BPH among elderly black and white men is not significantly different,5 the rate of traditional surgical treatment for BPH, transurethral resection of prostate, has always been lower among elderly black Medicare beneficiaries than among white beneficiaries. However, the racial difference in the use of TURP has decreased during last few decades such that by the end of the 1990s TURP was performed about 10% more frequently among elderly white men than among black men.6 However, little is known about the availability and use of the new BPH procedures TUMT, TUNA and contact or noncontact laser therapy by race. In this study we compared the availability and use of these 3 new procedures among elderly black and white Medicare beneficiaries. Our hypothesis was that the white-to-black
M
Submitted for publication June 8, 2005. Supported by the Centers for Medicare and Medicaid Services Contract HCFA 500-01-0043. * Correspondence: Division of Health Services Research and Policy, University of Minnesota School of Public Health, MMC 97 D369 Mayo Memorial Building, 420 Delaware St., S.E., Minneapolis, Minnesota 55455 (telephone: 612-625-6175; FAX: 612-378-4866; e-mail: [email protected]).
0022-5347/06/1755-1830/0 THE JOURNAL OF UROLOGY® Copyright © 2006 by AMERICAN UROLOGICAL ASSOCIATION
rate ratio for each of the new procedures was equal to the rate ratio for TURP. That is, the rates for TURP by race would provide information on the usual difference in BPH surgical treatment by race.
MATERIALS AND METHODS Data Sources and Case Definition/Identification We used the Medicare 100% Inpatient, Outpatient and Carrier files (use files) from the National Claims History repository, as well as the 100% Denominator files for 1999 through 2001. Identification of patients undergoing surgery or procedures was based on current procedural terminology codes for TURP (52601, 52612, 52614), TUMT (53850), TUNA (53852) and Laser (52647, 52648), and International Classification of Disease version 9 Clinical Modification procedure codes for TURP (60.29), TUMT (60.96), TUNA (60.97) and Laser (60.21). Claims were then extracted from Carrier, Inpatient or Outpatient files if any BPH procedure appeared in the claim. Information from records with identical individual health insurance claim numbers, procedure dates and procedure codes was consolidated to eliminate inappropriate double counting of cases. This process resulted in 328,096 potential cases for analysis.
1830
Vol. 175, 1830-1835, May 2006 Printed in U.S.A. DOI:10.1016/S0022-5347(05)00997-3
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES Beneficiaries included in the study had to be 65 years old or older at surgery, continuously enrolled in Medicare Part A and Part B throughout the study period and reside in 1 of the 50 states or the District of Columbia. Similar to most studies using Medicare administrative data, those enrolled in managed care during the year of surgery as well as those with end stage renal disease or prostate cancer during the study were excluded from analysis. A total of 123,589 patients were excluded, leaving 204,507 for analysis. Information on the type, dates and location of the BPH surgery was obtained from the use files, and the beneficiary age, race and county of residence were obtained from the Denominator files. County level socioeconomic information, namely race specific median household income and the percent of the population with less than a high school education, was obtained from United States Census 2000 Summary File 3. Analytical Methods Since the new BPH procedures were only performed in a limited number of geographic areas, separate analyses were performed for the total United States elderly male population as well as for those geographic areas (counties) in which the new technologies were available (procedure available counties). Availability was functionally defined, that is the county of residence of patients who underwent the procedure, not the county in which the procedure was performed. Furthermore, to reduce confounding due to population structure differences, to be included the counties had to have more than 10 black male beneficiaries residing in them during 3 study years (10 person-years of residence). Race specific procedure rates are reported. There were 3 age groups used in computing age specific rates and age adjusted rates, namely 65 to 74, 75 to 84 and 85 years old or older. Direct standardization was used for age adjustment with the 3-year total population as the standard. Negative binomial regression was used to compute the white-to-black procedure rate ratios adjusted for age group and county socioeconomic information. This model also took into account the heterogeneity within data such as overdispersion (observed variance greater than expected). Parameters were estimated using the generalized estimate equation to adjust for the geographic heterogeneity of procedure rates. Robust standard errors in Proc GENMOD of SAS® version 8.2 were used for constructing tests and confidence intervals. RESULTS From 1999 through 2001, 170,067 TURP, 16,953 TUMT, 5,353 TUNA and 12,134 Laser procedures were performed among black and white Medicare beneficiaries (table 1). The national age adjusted TURP rate among white men (6.13 per 1,000 person-years) was 3% higher than among black men (5.94 per 1,000 person-years). The TURP rates were highest in the 75 to 84-year-old age group for black and white men. The age adjusted rates for TUMT, TUNA and Laser among white men were 0.63, 0.20 and 0.44 per 1,000 person-years, respectively, about twice as high for TUMT and TUNA (0.31 and 0.10 per 100,000 person-years), and 17% higher for Laser (0.38 per 100.000 person-years) than among black men. When all 3 new procedures were combined the white-
1831
TABLE 1. Age specific and age adjusted rates of prostate procedures/1,000 among elderly Medicare beneficiaries, 1999 to 2001 Age Group
No. White Men (rate)
No. Black Men (rate)
TURP: 65–74 78,108 (5.66) 6,384 (5.52) 75–84 66,336 (6.91) 4,166 (6.51) 85⫹ 14,038 (5.75) 1,035 (6.07) Total No. ⫹ age adjusted rate 158,482 (6.13) 11,585 (5.94) TUMT: 65–74 9,397 (0.68) 383 (0.33) 75–84 5,926 (0.62) 193 (0.30) 37 (0.22) 85⫹ 1,017 (0.42) Total No. ⫹ age adjusted rate 16,340 (0.63) 613 (0.31) TUNA: 65–74 3,130 (0.23) 119 (0.10) 75–84 1,785 (0.19) 67 (0.10) 12 (0.07) 85⫹ 240 (0.10) Total No. ⫹ age adjusted rate 5,155 (0.20) 198 (0.10) Laser: 65–74 6,293 (0.46) 474 (0.41) 75–84 4,261 (0.44) 231 (0.36) 45 (0.26) 85⫹ 830 (0.34) Total No. ⫹ age adjusted rate 11,384 (0.44) 750 (0.38) TUMT, TUNA or Laser: 65–74 18,820 (1.36) 976 (0.84) 75–84 11,972 (1.25) 491 (0.77) 94 (0.55) 85⫹ 2,087 (0.86) Total No. ⫹ age adjusted rate 32,879 (1.27) 1,561 (0.79)
Rate Ratio (white/black) 1.03 1.06 0.95 1.03 2.06 2.04 1.92 2.04 2.20 1.77 1.40 1.99 1.11 1.23 1.29 1.17 1.62 1.62 1.55 1.61
to-black ratio of the age adjusted rates was 1.61. Among white and black men the new procedures were more likely performed on the youngest of the elderly. TUMT and TUNA were not performed among residents of all counties (figs. 1 to 3). There were counties in which the TUMT rate was more than twice the national average and counties in which the TUNA rate was more than 5 times the national average. Of the counties 16% had TUMT rates higher than 1.5 per 1,000 person-years. These counties included 11% of the white elderly male population and 6% of the black population. Similarly, for TUNA the counties with procedure rates greater than 1.5 per 1,000 person-years had 3% of the white population and 1% of the black population, and for Laser the counties with procedure rates greater than 1.5 per 1,000 had 6% of the white population and 5% of the black population. As expected the analyses using only the procedure available counties with more than 10 black male residents increased the procedure rates for the new BPH procedures (table 2). However, the white-to-black procedure rate ratios were similar to those using all United States male beneficiaries as the study cohort (table 1), and had little effect on TURP rates among each race group. Table 3 presents the white-to-black rate ratios and 95% CI for the procedure available counties adjusted for the age, geographic variation and county level socioeconomic information. The adjusted rate ratios ranged from 1.36 (95% CI 1.16 –1.59) for Laser to 2.33 (95% CI 1.87–2.90) for TUNA. The rate ratio was 1.65 (95% CI 1.50 –1.81) for TUMT, TUNA and Laser combined. For comparison the adjusted rate ratio for TURP was 1.08 (95% CI 1.04 –1.11).
1832
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES
FIG. 1. County specific rates of TUMT procedures among elderly Medicare beneficiaries, 1999 to 2001
FIG. 2. County specific rates of TUNA procedures among elderly Medicare beneficiaries, 1999 to 2001
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES
1833
FIG. 3. County specific rates of laser procedures among elderly Medicare beneficiaries, 1999 to 2001
DISCUSSION The findings of greatest concern in this study are the 2-fold and higher rates of TUMT and TUNA among white elderly Medicare beneficiaries compared with black beneficiaries. These findings were surprising given the relative parity in TURP rates which we found which has been improving over time.6,7 These differences persisted even after adjusting for socioeconomic status and geographic availability. Healthy People 2010 and the more recent Institute of Medicine report Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care unequivocally confirm the existence of racial disparities in health and health care in the United States.8,9 Access to Health Care in America and the Agency for Healthcare Research and Quality’s National Healthcare Disparities Report, 2004 identified multiple reasons/barriers for differences in health care use.10,11 These include the availability of the service, financial barriers such as health insurance coverage and reimbursement, patient perception including culture, language and acceptance of treatment, and provider beliefs and behaviors. We attempted to control for the availability of the new procedures by studying only counties in which residents had undergone the new procedure and in which there were black beneficiaries. These areas contained a greater proportion of the black male population (94%) than the white male population (82%) (data not shown but available from the authors). This may be because a large percentage of the facilities and providers of the new procedures were located in metropolitan areas (figs. 1 to 3). However, limiting the analysis to those geographic areas did not appreciably re-
TABLE 2. Geography and age adjusted prostate surgery procedure rates/1,000 person- years among elderly Medicare beneficiaries in procedure available counties with more than 10 black male Medicare beneficiaries, 1999 to 2001
TURP: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate TUMT: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate TUNA: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate Laser: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate TUMT, TUNA or Laser: 65–74 75–84 85⫹ Total No. ⫹ age adjusted rate
No. White Men (rate)
No. Black Men (rate)
Rate Ratio (white/black)
62,360 (5.62) 53,951 (6.86) 11,201 (5.69)
5,898 (5.41) 3,851 (6.42) 941(6.00)
1.04 1.07 0.95
127,512 (6.09)
10,690 (5.84)
1.04
7,635 (0.76) 4,820 (0.67) 785 (0.44)
379 (0.38) 190 (0.34) 36 (0.25)
2.02 1.95 1.71
13,240 (0.70)
605 (0.35)
1.98
2,711 (0.39) 1,562 (0.30) 205 (0.16)
117 (0.18) 66 (0.18) 12 (0.14)
2.19 1.65 1.15
4,478 (0.33)
195 (0.18)
1.90
5,659 (0.59) 3,789 (0.55) 730 (0.43)
472 (0.50) 230 (0.44) 44 (0.33)
1.19 1.24 1.27
10,178 (0.56)
746 (0.46)
1.21
16,005 (1.44) 10,171 (1.29) 1,720 (0.87)
968 (0.89) 486 (0.81) 92 (0.59)
1.62 1.60 1.49
27,896 (1.33)
1,546 (0.83)
1.60
Geographic adjusted rate ratios were also adjusted for age.
1834
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES
TABLE 3. Age, county socioeconomic status and geography adjusted white/black rate ratios of prostate procedures
TURP TUMT TUNA Laser TUMT, TUNA or Laser
Adjusted Rate Ratio
Lower 95% CI
Upper 95% CI
1.08 1.96 2.33 1.36 1.65
1.04 1.70 1.87 1.16 1.50
1.11 2.25 2.90 1.59 1.81
All value p ⬍0.0001.
duce the black-to-white rate ratios. One possible reason it had little impact is that the practices of the providers of the new procedures serve different patient populations than those who continue to perform only TURP. Defining the practice populations of individual urology practices is beyond the scope of this report. Using Medicare data one can determine the racial composition of the population actually treated by a practice, but defining the actual population at risk as being treated by the practice is not possible, particularly since urology is largely a referral specialty. It is unlikely that financial barriers were the cause of the differences by race. In our study all beneficiaries were covered by Medicare Part A and Part B. In addition, the Medicare program has set the physician payment rate of new procedures lower than that of TURP, reducing the potential out-of-pocket expenses paid to the physician by the beneficiary. The majority of the new procedures were performed in outpatient settings or physician offices (94% of TUMT, 93% of TUNA and 72% of Laser) compared to TURP, which is overwhelmingly performed in an inpatient setting (91%). Thus, in the majority of instances the total cost of new procedures to the patient would not include the first day hospital co-payment ($792 in 2001, for example) which must be paid by hospitalized patients. We calculated the total patient financial responsibility to the urologist and to the institution (hospital, outpatient, ambulatory surgical center, etc) for the different procedures during the study period. The median amounts for inpatient TURP ($954) and outpatient TURP ($973) were higher than the amounts for noninpatient hospital TUMT ($506) and TUNA ($509), and similar to noncontact Laser ($929) and contact Laser ($965). These costs would likely encourage the use of 2 of the new procedures among those of lower socioeconomic status, yet these procedures were the ones for which we found the greatest racial differences. While we controlled for socioeconomic status in our analysis, other personal factors and perceptions may also have a key role in decision making on the part of patients and the use of the new procedures. Cultural differences as well as attitudes and beliefs regarding the new procedures may differ between black and white men, and are cited as possible reasons for differences in health care use.12,13 However, they are not always sufficient to explain the differences.13 Kressin et al have studied racial differences in patient preference regarding cardiac catheterization.14,15 In their initial study they found little difference in measures of health beliefs and attitudes between black and white patients at the Department of Veterans Affairs. Subsequently they reported that health beliefs as well as sociodemographic and clinical variables did not explain differences in the use of cardiac catheterization between black
and white patients. Unfortunately we know of no information in the literature regarding racial differences in attitudes or beliefs regarding the new BPH procedures, and the Medicare administrative data provide little insight into these areas. Thus, we could not investigate whether they had a role in the use of the new prostate procedures. Because the procedures are new, physician perception of the appropriateness of the procedure for each individual patient may be more important than in well established procedures. This communication between patients and physicians is likely to affect the decision to undergo TUMT, TUNA or Laser. However, it is difficult to understand why urologists would recommend the new procedures to one racial group compared to the other unless there is a difference in disease severity or another clinical factor at surgery. As summarized by Wei et al there is no variation in the prevalence of obstructive urinary symptoms by race/ethnicity, but overall lower urinary tract symptoms appear to occur with greater severity in black men, which might lead to a recommendation or a patient preference for TURP.16 We believe that the large, national sample and the geographic analysis are strengths in this study. The latter helped us control for the restricted geographic availability of the new procedures. However, there are some limitations. It is difficult to examine the severity of BPH based on claims data. The lack of information about supplemental insurance coverage and payments prohibited us from examining the actual out-of-pocket cost of the procedures for patients.
CONCLUSIONS To our knowledge this is the first study to report the unequal use of new surgical BPH procedures by race. These differences should be monitored, particularly if studies such as the Minimally Invasive Surgical Therapies Treatment Consortium for Benign Prostatic Hyperplasia trial demonstrate that the new procedures have advantages compared to other treatments.17 From 1999 through 2001 white elderly Medicare beneficiaries were 61% more likely to undergo TUMT, TUNA or Laser for BPH than black beneficiaries. For comparison TURP was performed among white men at a 3% greater rate than among black men. There were large geographic variations in procedure rates. After restricting analysis to those procedure counties with more than 10 black male beneficiaries, and adjusting for geographic variation and area socioeconomic status, racial differences in the use of the new procedure persisted.
Abbreviations and Acronyms BPH Laser TUMT TUNA TURP
⫽ ⫽ ⫽ ⫽ ⫽
benign prostatic hyperplasia contact or noncontact laser therapy transurethral microwave therapy transurethral needle ablation transurethral resection of the prostate
REFERENCES 1. McBean, A. M. and Gornick, M.: Differences by race in the rates of procedures performed in hospitals for Medicare beneficiaries. Health Care Financ Rev, 15: 77, 1994
UNEQUAL USE OF NEW BENIGN PROSTATIC HYPERPLASIA SURGERIES 2. Gornick, M. E., Eggers, P. W., Reilly, T. W., Mentnech, R. M., Fitterman, L. K., Kucken, L. E. et al: Effects of race and income on mortality and use of services among Medicare beneficiaries. N Engl J Med, 335: 791, 1996 3. Ayanian, J. Z.: Race, class, and the quality of medical care. JAMA, 271: 1207, 1994 4. Epstein, A. M. and Ayanian, J. Z.: Racial disparities in medical care. N Engl J Med, 344: 1471, 2001 5. Platz, E. A., Smit, E., Curhan, G. C., Nyberg, L. M. and Giovannucci, E.: Prevalence of and racial/ethnic variation in lower urinary tract symptoms and noncancer prostate surgery in U.S. men. Urology, 59: 877, 2002 6. Wasson, J. H., Bubolz, T. A., Lu-Yao, G. L., Walker-Corkery, E., Hammond, C. S., Barry, M. J. et al: Transurethral resection of the prostate among Medicare beneficiaries: 1984 to 1997. J Urol, 164: 1212, 2000 7. Lu-Yao, G. L., Barry, M. J., Chang, C. H., Wasson, J. H. and Wennberg, J. E.: Transurethral resection of the prostate among Medicare beneficiaries in the United States: time trends and outcomes. Prostate Patient Outcomes Research Team (PORT). Urology, 44: 692, 1994 8. Healthy People 2010, United States Department of Health and Human Services, 2000. http://www.healthypeople.gov/ document. Accessed June 4, 2005 9. Smedley, B. D., Stith, A. Y. and Nelson, A. R.: Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care. Washington, D. C.: The National Academies Press, 2003 10. Millman, M.: Access to Health Care in America. Washington, D. C.: The National Academies Press, 1993 11. National Healthcare Disparities Report, 2004. Washington, D. C.: Agency for Healthcare Research and Quality, 2005 12. Whittle, J., Conigliaro, J., Good, C. B. and Joswiak, M.: Do patient preferences contribute to racial differences in cardiovascular procedure use? J Gen Intern Med, 12: 267, 1997 13. Ayanian, J. Z., Cleary, P. D., Weissman, J. S. and Epstein, A. M.: The effect of patients’ preferences on racial differences in access to renal transplantation. N Engl J Med, 341: 1661, 1999 14. Kressin, N. R., Clark, J. A., Whittle, J., East, M., Peterson, E. D., Chang, B. H. et al: Racial differences in health-related beliefs, attitudes, and experiences of VA cardiac patients: scale development and application. Med Care, suppl., 40: 172, 2002 15. Kressin, N. R., Chang, B. H., Whittle, J., Peterson, E. D., Clark, J. A., Rosen, A. K. et al: Racial differences in cardiac catheterization as a function of patients’ beliefs. Am J Pub Health, 94: 2091, 2004
1835
16. Wei, J. T., Calhoun, E. A. and Jacobsen, S. J.: Benign prostatic hyperplasia. In: Urologic Diseases in America. Edited by M. S. Litwin and C. S. Saigal. United States Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Washington, D. C.: United States Government Publishing Office, NIH Publication No. 04 –5512, chapt. 2, pp. 43– 67, 2004 17. National Institute of Diabetes and Digestive and Kidney Diseases. Minimally Invasive Surgical Therapies (MIST) Treatment Consortium for Benign Prostatic Hyperplasia (BPH). Available at http://www.niddk.nih.gov/patient/mist/mist.htm. Accessed September 30, 2005
EDITORIAL COMMENT Yu et al demonstrate disparities in access to minimally invasive therapies for BPH from 1999 to 2001. Specifically they demonstrate that black men were less likely to undergo TUMT, TUNA or laser therapy than white men. These findings beg the greater question of why these racial differences in access to care exist. The authors suggest a number of different explanations, including differences in patient populations of “early-adopter” urologists, cultural differences as well as differences in attitudes and beliefs among black and white men, or differences in clinical presentation that might make these procedures more appropriate in some patients than others. While all of these explanations are interesting, it is my belief that the racial differences in procedure rates found in this study are due almost exclusively to differences in the patient populations of the urologists who first adopted this technology. The reader should remember that the study period was relatively early in our experience with minimally invasive surgical techniques for BPH. At that time many or all of the minimally invasive procedures performed in a given county likely would have been undertaken by a single or a small number of providers. These providers likely had unique clinical practices which attracted select patient populations, the majority of which apparently were white. While the racial disparities noted in the current report are disheartening, they underscore the need for further research. Specifically, now that the technology has diffused out more completely, do these racial differences in access to care still exist? David F. Penson Departments of Urology and Preventive Medicine Keck School of Medicine University of Southern California Los Angeles, California