- Email: [email protected]
THE REGIONALIZATION OF RADICAL CYSTECTOMY TO SPECIFIC MEDICAL CENTERS
0022-5347/05/1744-1385/0 THE JOURNAL OF UROLOGY® Copyright © 2005 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 174, 1385–1389, October 2005 Printed in U.S.A.
DOI: 10.1097/01.ju.0000173632.58991.a7
THE REGIONALIZATION OF RADICAL CYSTECTOMY TO SPECIFIC MEDICAL CENTERS BRENT K. HOLLENBECK,* DAVID A. TAUB, DAVID C. MILLER, RODNEY L. DUNN, JAMES E. MONTIE† AND JOHN T. WEI‡ From the Department of Urology, University of Michigan, Ann Arbor, Michigan
ABSTRACT
Purpose: Regionalization of high risk surgical procedures to larger teaching hospitals has been suggested as a means to improve the quality of care. We established a novel framework for characterizing regionalization, implemented it to determine the extent to which regionalization of radical cystectomy has occurred and delineated whether specific patient characteristics are associated with this phenomenon. Materials and Methods: We used the Nationwide Inpatient Sample to identify 22,088 patients who underwent radical cystectomy for bladder cancer from 1988 to 2000. Regionalization was assessed using 5 structural hospital measures, including teaching status, urban location, discharge volume, cystectomy volume and bed capacity. Adjusted models were developed to identify the significance of temporal trends and assess the association of demographic factors with structural qualities. Results: Compared with 1988 to 1990 subjects were more likely to undergo cystectomy at teaching hospitals (OR 1.8), high cystectomy volume hospitals (OR 1.2), high discharge volume hospitals (OR 1.7) and large bed capacity medical centers (OR 1.4) in 1998 to 2000. The concentration of cystectomy to urban medical centers during the study years was 90% to 92%. The proportion of subjects undergoing partial cystectomy decreased from 23.9% to 16.6% as regionalization occurred. Older subjects were less likely to be treated at these regionalized centers. Conclusions: Without broad legislation from health care payers radical cystectomy has increasingly regionalized to specific medical centers. Despite this regionalization disparities in its use exist among specific, vulnerable patients. Addressing this may facilitate further concentration of this procedure. KEY WORDS: bladder, bladder neoplasms, outcome assessment (health care), cystectomy, quality of care
Anticipated health care expenditures in the 21st century will comprise a significant proportion of the United States gross domestic product.1 Surgical services comprise approximately 40% of all hospital expenses2 and, thus, identifying potential causes of increased resource use provide a mechanism to curtail potentially avoidable expenses. These increasing health care expenditures have prompted payers to consider using quality of care to direct reimbursement3 and advocates of volume-outcome relationships have encouraged patients undergoing certain surgical procedures to seek treatment at high procedural volume centers.4 Surgeon and hospital volume are not strictly quality of care indicators per se. Rather, they are structural characteristics that are reflective of the setting in which health care is delivered5 and commonly used surrogate measures of quality because of expediency. Furthermore, whether measured at the hospital or surgeon level volume has been associated with lower morbidity6 and mortality7, 8 for various surgical procedures, including radical cystectomy.7, 8 As a consequence of these data, there is a push among payers to concentrate
specific surgical procedures (but not radical cystectomy) to a relatively few medical centers. The result of such regionalization will have significant implications to the patient and physician. Thus, the objectives of this study were 3-fold, namely to 1) develop a novel framework for measuring regionalization, 2) assess the degree to which regionalization of cystectomy for bladder cancer has occurred throughout the United States by applying this framework and 3) delineate patient characteristics associated with any potential trends in regionalization. METHODS
Subjects. Data from the Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS) were abstracted for 1988 through 2000. The NIS represents a 20% stratified sample of all hospital discharges in the United States. The database is maintained by the Agency for Health Care Research and Quality, and consists of uniform hospital discharge summaries from 994 nonfederal acute care hospitals throughout the United States. Data on eligible subjects were abstracted from the NIS using procedural terminology based on International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9) codes for cystectomy, including 577 for total cystectomy, 5771 for radical cystectomy, 576 for partial cystectomy and 5779 for other total cystectomy. In addition, ICD-9 codes for principle diagnoses were used to limit the sample to subjects with bladder cancer (188, 1880 to 1889, 2337, 2367 and 2394). Clinical Classification Software (principle diagnostic code 32), used in the context of the NIS
Submitted for publication January 7, 2005. Presented at annual meeting of American Urological Association, San Francisco, California, May 8 –13, 2004. * Correspondence: Department of Urology, Taubman Health Care Center, Room 3875, University of Michigan Health System, 1500 East Medical Center Dr., Ann Arbor, Michigan 48109-0330 (telephone: 734-615-0563; FAX: 734-936-9127; e-mail: bhollen@ umich.edu). † Financial interest and/or other relationship with AstraZeneca. ‡ Financial interest and/or other relationship with Sanofi, Laserscope, Calypso and Boehringer Ingelheim. 1385
1386
REGIONALIZATION OF RADICAL CYSTECTOMY
to partition subjects into relevant categories, was used to validate our sampling methodology. Secondary ICD-9 diagnostic codes were abstracted to enumerate comorbid conditions according to the Romano modification of the Charlson comorbidity index.9 Demographic information, including subject age, race, sex, income based on median income in the subject zip code, hospital admission type, treatment year, geographic region of treatment based on United States Census regions and insurance type were also abstracted. In accordance with the Code of Federal Regulations, Title 21, Section 46.101, subparagraph 4 institutional review board approval was not sought for this study. Measures of regionalization. To test the null hypothesis that cystectomy procedures did not migrate to select referral centers during the study course we developed a novel framework to measure regionalization within the NIS by identifying medical centers with specific hospital qualities where cystectomies may concentrate. These 5 qualities were 1) teaching hospital status based on 1 of certain factors, namely the presence of any American Medical Association approved residency program, membership status to the Council of Teaching Hospitals or a ratio of full-time equivalent interns and residents to beds of greater than 0.25, 2) urban hospital status, as described by location within a metropolitan statistical area based on the United States Census, 3) largest hospital status, based on the tertile of hospitals within a given geographic region containing the greatest hospital bed capacity, 4) high cystectomy volume center status, as indicated by hospitals where the most cystectomies were performed annually (9 or more yearly, that is the top 40th percentile) and 5) high discharge volume status, corresponding to hospitals with greater than 24,677 discharges among all diagnoses yearly (the top 20th percentile). Construction of a hospital cystectomy volume variable. The number of procedures performed at each hospital from 1988 to 2000 was ascertained using a unique hospital identification code. All years were considered together as 1 entity and each hospital was assigned a volume group for each year that it participated. Each individual hospital-year was considered independently and then ranked in order of increasing annual case volume. Five volume groups were defined by selecting whole number cutoffs for annual hospital volume that most closely sorted patients into groups of equal size (quintiles). In this manner the numerical thresholds for hospital volume were 1 to 3 cases yearly—very low, 4 or 5—low, 6 to 8 —medium, 9 to 14 — high and more than 14 —very high. For the purposes of regionalization assessment the top 2 quintiles were combined to identify high and very high volume centers. Statistical analyses. For the purposes of demonstrating changes in the measure of regionalization with time in multivariable modeling the treatment year was categorized into certain groups, including 1988 to 1991, 1991 to 1994, 1995 to 1997 and 1988 to 2000. Because the definition of a teaching hospital in the NIS changed beginning in 1998, longitudinal analysis of teaching hospital status was not possible using the complete data set. To allow contingent analysis of factors associated with teaching and urban hospital status (nested within the teaching status variable) we used a subset of the data that incorporates a uniform definition of teaching and urban hospitals. Only hospitals classified consistently throughout the study period were included. This decresed the number of cystectomy procedures in the analysis from 22,088 to 17,630 discharges. The analysis of this smaller data set was not weighted due to the skewed sampling required to obtain the uniform teaching/urban hospital definition. All models and bivariate analyses not involving the teaching/ urban status of a hospital were weighted to reflect the national universe of 119,491 cystectomy discharges during the study course. Unadjusted cystectomy use rates at medical centers, as described by the various measures of regionalization, were calculated by treatment year. Bivariate analyses
were performed using the chi-square test and ANOVA. Logistic regression analyses were performed to adjust for the various covariates and determine the propensity for cystectomy use at the medical centers described by our regionalization measures. An indicator variable for missing data was included for race (9% missing) to minimize observational exclusions from the models. Multicollinearity diagnostics were performed to ensure the lack of strong linear tendencies among the explanatory variables. All testing was completed using SAS, version 8.2 computerized statistical software (SAS Institute, Cary, North Carolina), was 2-tailed and was performed at the 5% significance level. RESULTS
The mean age ⫾ SD of subjects in our study was 68.6 ⫾ 10.6 years, which did not significantly vary among years (p ⫽ 0.78). Table 1 lists other patient characteristics and changes in them with time. Table 2 shows unadjusted trends in the proportion of cystectomies performed at hospitals, characterized by the 5 measures of regionalization measures. Table 3 shows the final adjusted models, illustrating the association of covariates with the propensity of subjects to be treated at these regional cystectomy centers. DISCUSSION
Less common but more complex surgical procedures have been characterized as carrying a greater risk of adverse events and increasing public scrutiny has been placed on patient safety and defining metrics for quality of surgical care.10 Consequently there is a nationwide movement afoot calling for the regionalization of certain surgical procedures that looms over various surgical disciplines.3 From a patient perspective the implications of such a directive in terms of travel time have been addressed.11 However, it is possible that the evolution of care for patients requiring certain surgical procedures has naturally progressed to specific medical centers and such a notion has not been explored. Potential causes of regionalization in this context are selective referral based on hospital/surgeon reputation, financial pressures to efficiently use time (eg a surgeon may be reimbursed more for performing multiple simple procedures in lieu of a single complex surgery) and patient recognition of centers of excellence. In such an environment further legislated regionalization would have questionable impact on health care quality. Radical cystectomy is 1 such procedure in which regionalization may eventually be mandated because of higher mortality rates at low volume centers7 and relatively high postoperative morbidity rates (28% to 37%).12⫺14 In the context of the quality of care paradigm, as proposed by Donabedian,15 we used 5 novel measures to provide a framework for describing regionalization. Our findings clearly demonstrate that relative to the earliest period cystectomy was more likely to be performed at teaching hospitals, high cystectomy volume hospitals, large hospitals in terms of bed size and high discharge hospitals in 1998 to 2000. Concentration of procedures at hospitals with these characteristics suggest that regionalization of cystectomy has already occurred to some degree but ample opportunity for further regionalization to such institutions still exists. Furthermore, these data illustrate that cystectomy has largely been isolated to urban medical centers because the proportion performed at these institutions was 0.90 to 0.92 during the study course, suggesting that urbanization has already occurred. Small temporal decreases in the proportion of those treated at urban centers likely reflect regression to the mean since the majority of patients are already treated at such centers. Similar trends have been described for surgeries such as hepatic resection,16 implying that current referral patterns for various complex procedures are funneling patients to centers of excellence. Furthermore, the almost com-
1387
REGIONALIZATION OF RADICAL CYSTECTOMY
TABLE 1. Sample by period, weighted to reflect American Hospital Association (AHA) universe of hospital discharges between 1988 and 2000 Covariate (level) Cystectomies (No.)* Age: Younger than 60 60–Younger than 70 70–Younger than 80 80 or Older Sex (female) Admission status (urgent) Insurance: Medicare Private Other Race: White Black Other Missing data Geographic region: Northeast Midwest South West Modified Charlson index9: Low comorbidity (0–1) Moderate comorbidity (2–3) High comorbidity (greater than 3) Partial cystectomy * Not weighted to reflect AHA universe.
% 1988–1991
% 1992–1994
% 1995–1997
% 1998–2000
p Value for Trend
29 (6,453)
23 (5,066)
24 (5,294)
24 (5,275)
⬍0.0001
15.1 33.7 38.4 12.8 21.3 30.0
16.7 31.4 39.4 12.5 20.8 29.9
17.7 29.6 39.4 13.3 20.9 27.5
19.6 27.6 38.8 14.0 20.5 18.2
⬍0.0001
65.3 20.2 8.5
66.5 26.8 6.7
64.8 28.4 6.8
61.4 31.4 7.2
⬍0.0001
33.2 0.8 63.9 2.2
69.7 3.0 25.1 2.2
74.1 4.0 18.1 3.8
68.9 3.7 4.9 22.5
⬍0.0001
20.7 19.2 38.1 22.0
22.7 20.9 37.9 18.5
21.6 22.5 36.2 17.7
23.2 26.4 31.7 18.7
⬍0.0001
48.9 49.8 1.4 23.9
41.3 56.1 3.6 21.0
37.1 58.4 4.6 19.0
36.6 59.4 4.1 16.6
⬍0.0001
0.0880 ⬍0.0001
⬍0.0001
TABLE 2. Unadjusted trends in cystectomies performed at specific medical centers Hospital Measure
Hospital Measure Description
Teaching hospital*
% 1988–1991
Any American Medical Association approved residency program, membership status to Council of Teaching Hospitals, or 0.25 resident/bed ratio Urban hospital* Metropolitan statistical area based on United States Census Largest capacity† Tertile of hospitals containing most beds in given geographic region High cystectomy vol† Hospitals in which 9 or more procedures were performed/yr (upper 40th percentile of all hospitals) High discharge vol† Of all diagnoses this corresponds to 24,678 or more discharges/yr (upper 20th percentile of all hospitals) * Not weighted to reflect AHA universe due to sampling changes made in 1998. † Weighted to reflect the AHA universe of 119,491 discharges following cystectomy
plete urbanization of cystectomy suggests that further regionalization to specific centers with demonstrable quality benefits will have little additional impact on patient preferences for location of care since the majority of procedures have migrated out of rural communities.17 Prior to mandating regionalization of these procedures efforts should first be directed toward establishing metrics of quality and processes of care that can accurately distinguish centers of excellence from other institutions, thereby, facilitating pilot programs aimed at concentrating case volume at these facilities. The underlying etiologies of the observed spontaneous regionalization are enigmatic. However, potential causes may be influenced by reimbursement, difficult postoperative management and evolving data regarding the effects of structural characteristics on outcomes.7, 8 Future study that elicits the cause for this regionalization will help elucidate whether this phenomenon is a consequence of self-governance (because of perceived better outcomes at medical centers with the mentioned structural characteristics), of financial considerations (eg lower reimbursement) or of patient specific factors (eg word of mouth referral). Because of significant variation in practice patterns with time,18 we examined several common factors associated with
% 1992–1994
% 1995–1997
% 1998–2000
p Value for Trend
34.4
35.0
40.6
45.3
⬍0.0001
91.6
90.6
89.9
90.5
0.0143
61.2
62.8
56.2
68.0
⬍0.0001
35.5
35.5
34.5
44.0
⬍0.0001
15.2
20.1
26.2
26.4
⬍0.0001
that occurred during the study period.
its use at medical centers described by these structural characteristics where the procedure was most concentrated. An interesting phenomenon was that the use of radical cystectomy relative to partial cystectomy, which is a less technically demanding and less morbid procedure,19 was 60% to 100% more likely at regionalized centers, namely teaching, urban, high cystectomy volume and high total discharge volume hospitals. Furthermore, because this regionalization occurred during the study course, the proportion of subjects undergoing partial cystectomy continually decreased. Given the relative rare indications for partial cystectomy (about 6% of all patients with cystectomy),20 it is not surprising that regionalization has likely brought about more appropriate care to a larger patient population. Not surprisingly subjects undergoing cystectomy had more medical comorbidities in the later study years, which may represent the narrowing of contraindications to surgery due to advances in perioperative critical care. As a consequence, teaching institutions, urban hospitals and high cystectomy volume centers had a greater propensity for using cystectomy in patients with greater infirmity. This suggests that these centers had intrinsic resources/infrastructure/processes of care to provide for this population, leading to selec-
1388
REGIONALIZATION OF RADICAL CYSTECTOMY
TABLE 3. Final adjusted models ascertaining clinical and structural factors associated with propensity to undergo radical cystectomy at hospitals characterized by 5 regionalization qualities Level
Elective admission Sex (female) Insurance: Private Other Race: Black Hispanic/other Missing data Geographic region: Midwest Northeast South Charlson index: Greater than 3 2 or 3 Subject age: Older than 60–70 Older than 70–80 Older than 80 Treatment yr: 1998–2000 1995–1997 1992–1994 Radical cystectomy
Teaching Hospital*
Reference Level
Urban Medical Center*
Highest Cystectomy Vol†
Largest Hospitals (No. beds)†
Hospitals With Most Discharges (all diagnoses)†
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
Emergent admission Male
0.8
0.8–1.0
1.4
1.2–1.6
1.3
1.3–1.4
—
—
1.2
1.1–1.2
—
—
—
—
0.8
0.8–0.8
—
—
—
—
Medicare
1.1 1.3
1.0–1.2 1.2–1.5
2.0 1.2
1.7–2.4 0.9–1.5
—
—
—
—
1.1 1.2
1.0–1.1 1.1–1.3
White
2.7 1.2 0.8
2.2–3.3 1.1–1.3 0.7–1.0
1.2 0.7 0.5
0.8–1.7 0.6–0.7 0.4–0.6
0.8 0.9 0.8
0.7–0.8 0.8–0.9 0.7–0.8
1.0 1.1 1.2
0.9–1.0 1.1–1.2 1.2–1.3
1.4 1.0 0.9
1.3–1.5 1.0–1.1 0.9–1.0
West
1.3 3.7 0.6
1.2–1.5 3.4–4.1 0.6–0.7
0.8 2.6 1.9
0.7–0.9 2.2–3.2 1.6–2.2
0.7 0.9 0.9
0.7–0.7 0.8–0.9 0.8–0.9
2.2 1.3 2.1
2.1–2.2 1.3–1.3 2.0–2.2
1.2 1.7 1.5
1.2–1.3 1.6–1.8 1.5–1.6
0 or 1
1.2 0.9
1.1–1.3 0.9–1.0
1.2 0.9
1.0–1.3 0.8–1.0
1.2 0.9
1.2–1.2 0.9–0.9
—
—
—
—
60 or Younger
0.8
0.7–0.9
1.0
0.8–1.2
0.8
0.8–0.9
—
—
0.9
0.8–0.9
0.6
0.6–0.7
0.9
0.8–1.2
0.8
0.7–0.8
0.8
0.8–0.9
0.5
0.4–0.6
0.8
0.6–1.0
0.7
0.7–0.8
0.7
0.7–0.7
1.8 1.5 1.1 2.0
1.6–2.1 1.4–1.7 1.0–1.2 1.8–2.2
0.7 0.7 0.7 1.4
0.6–0.9 0.6–0.8 0.6–0.9 1.2–1.6
1.2 0.9 0.9 1.9
1.1–1.2 0.9–0.9 0.9–1.0 1.8–1.9
1.7 1.9 1.4 1.6
1.4–1.5 1.9–2.1 1.3–1.4 1.6–1.7
1988–1991
Partial cystectomy * Based on 17,630 patients who underwent cystectomy, not weighted to reflect entire AHA universe. † Based on 22,088 patients who underwent cystectomy, weighted to reflect entire AHA universe.
tive referral to these centers, or they were less selective in their patient pool and, hence, more procedures were performed. Younger patient age was also commonly associated with medical centers with structural characteristics suggesting regionalization. Relative to patients 60 years or younger, older subjects are less likely to seek treatment at medical centers where cystectomy has concentrated. Collectively these data highlight that among certain vulnerable populations, namely the elderly and the infirm (multiple comorbidities), disparities in the use of cystectomy at centers of regionalization exist and determining potential factors associated with these discrepancies may elicit health care access barriers in these groups. This notion is further supported by the fact that patients with Medicare, which is presumably the most reliable payer, were less commonly treated at regionalized centers, suggesting that reimbursement may have an underlying role in health services delivery. Future study should focus on determining the principle causes of these disparities in cystectomy use, namely access to care (income, insurance and level of education) and selective treatment patterns (younger age), before further regionalization of care can address societal needs. The use of administration data is not without its inherent limitations. Ideal case mix adjustment is limited because of data collection methodology and constraints, although the application of a widely used comorbidity index in the context of administrative datasets facilitates comparison.9 Our findings that cystectomy has concentrated to medical centers with specific structural characteristics in the last 13 years do not validate this regionalization as a means of improving health care quality. Rather, they highlight trends in practice patterns, of which the causes have yet to be elicited. Finally, the relationship between potential structural characteristics (eg teaching status, annual case volume, high discharge volume, urban medical centers status and large bed size capacity) and common adverse events (eg prolonged ileus and
1.4 0.8 1.1 1.0
1.3–1.4 0.8–0.9 1.1–1.1 1.0–1.1
wound complications) following cystectomy has not been established and further study in this context will help determine the usefulness of these structural indicators. CONCLUSIONS
This study provides a novel framework for assessing the regionalization of health services delivery to specific medical centers. The usefulness of this framework is illustrated by the observed regionalization of radical cystectomy to specific institutions with certain structural traits that has occurred in the last 13 years. Furthermore, the urbanization of cystectomy is almost complete with approximately 90% of all procedures occurring in metropolitan hospitals throughout study course. Concentration of cystectomy to these institutions (teaching, large bed capacity, high cystectomy volume, high discharge volume and urban) has been associated with a concurrent decrease in the proportion of subjects undergoing partial cystectomy, which may reflect more appropriate care. Nonetheless, disparities among cystectomy use in vulnerable patient populations exist. Eliciting and eliminating the etiologies of these disparities should minimize the obstacles to the further spontaneous regionalization of cystectomy. REFERENCES
1. Chernew, M. E., Hirth, R. A. and Cutler, D. M.: Increased spending on health care: how much can the United States afford? Health Aff (Millwood), 22: 15, 2003 2. Macario, A., Vitez, T. S., Dunn, B., McDonald, T. and Brown, B.: Hospital costs and severity of illness in three types of elective surgery. Anesthesiology, 86: 92, 1997 3. Birkmeyer, J. D.: Should we regionalize major surgery? Potential benefits and policy considerations. J Am Coll Surg, 190: 341, 2000 4. Birkmeyer, J. D.: High-risk surgery—follow the crowd. JAMA, 283: 1191, 2000 5. Birkmeyer, J. D., Dimick, J. B. and Birkmeyer, N. J.: Measuring
1389
REGIONALIZATION OF RADICAL CYSTECTOMY
6. 7. 8.
9. 10.
11. 12. 13.
14.
15. 16.
the quality of surgical care: structure, process, or outcomes? J Am Coll Surg, 198: 626, 2004 Begg, C. B., Riedel, E. R., Bach, P. B., Kattan, M. W., Schrag, D., Warren, J. L. et al: Variations in morbidity after radical prostatectomy. N Engl J Med, 346: 1138, 2002 Birkmeyer, J. D., Siewers, A. E., Finlayson, E. V., Stukel, T. A., Lucas, F. L., Batista, I. et al: Hospital volume and surgical mortality in the United States. N Engl J Med, 346: 1128, 2002 Birkmeyer, J. D., Stukel, T. A., Siewers, A. E., Goodney, P. P., Wennberg, D. E. and Lucas, F. L.: Surgeon volume and operative mortality in the United States. N Engl J Med, 349: 2117, 2003 Romano, P. S., Roos, L. L. and Jollis, J. G.: Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol, 46: 1075, 1993 Goodney, P. P., Siewers, A. E., Stukel, T. A., Lucas, F. L., Wennberg, D. E. and Birkmeyer, J. D.: Is surgery getting safer? National trends in operative mortality. J Am Coll Surg, 195: 219, 2002 Birkmeyer, J. D., Siewers, A. E., Marth, N. J. and Goodman, D. C.: Regionalization of high-risk surgery and implications for patient travel times. JAMA, 290: 2703, 2003 Frazier, H. A., Robertson, J. E. and Paulson, D. F.: Complications of radical cystectomy and urinary diversion: a retrospective review of 675 cases in 2 decades. J Urol, 148: 1401, 1992 Stein, J. P., Lieskovsky, G., Cote, R., Groshen, S., Feng, A. C., Boyd, S. et al: Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol, 19: 666, 2001 Koch, M. O. and Smith, J. A., Jr.: Influence of patient age and co-morbidity on outcome of a collaborative care pathway after radical prostatectomy and cystoprostatectomy. J Urol, 155: 1681, 1996 Donabedian, A.: The definition of quality and approaches to its assessment. In: Explorations in Quality Assessment and Monitoring. Ann Arbor: Health Administration Press, vol. 1, 1980 Dimick, J. B., Wainess, R. M., Cowan, J. A., Upchurch, G. R., Jr., Knol, J. A. and Colletti, L. M.: National trends in the use and outcomes of hepatic resection. J Am Coll Surg, 199: 31, 2004
17. Finlayson, S. R., Birkmeyer, J. D., Tosteson, A. N. and Nease, R. F., Jr.: Patient preferences for location of care: implications for regionalization. Med Care, 37: 204, 1999 18. Konety, B. R. and Joslyn, S. A.: Factors influencing aggressive therapy for bladder cancer: an analysis of data from the SEER program. J Urol, 170: 1765, 2003 19. Weinstein, R. P., Grob, B. M., Pachter, E. M., Soloway, S. and Fair, W. R.: Partial cystectomy during radical surgery for nonurological malignancy. J Urol, 166: 79, 2001 20. Utz, D. C., Schmitz, S. E., Fugelso, P. D. and Farrow, G. M.: Proceedings: a clinicopathologic evaluation of partial cystectomy for carcinoma of the urinary bladder. Cancer, 32: 1075, 1973 EDITORIAL COMMENT The data in this study confirmed what most people have observed anecdotally, ie that radical cystectomy and other major surgical procedures in urology are increasingly being concentrated at tertiary and high volume hospitals. As the authors point out, the reasons are multifactorial but the relative financial disincentives for complex surgery are certainly contributory. Across the board urology is differentiating into high volume surgical practice or office based practice. This may be a good thing for patient care because numerous studies show improved outcomes when major surgical procedures are performed by high volume surgeons at high volume hospital settings. Although radical cystectomy is accepted as the most effective treatment for muscle invasive bladder cancer, multiple studies show that only around 30% to 40% of patients with documented muscle invasive urothelial cancers ever undergo cystectomy. Again, there are many reasons for this but concerns about operative morbidity and mortality are paramount. As noted by these authors, the regionalization of care may help address the problem. If greater experience generally equates with improved outcome, perhaps more patients will be considered eligible for surgery. Joseph A. Smith, Jr. Department of Urologic Surgery Vanderbilt University Nashville, Tennessee
Vol. 174, 1385–1389, October 2005 Printed in U.S.A.
DOI: 10.1097/01.ju.0000173632.58991.a7
THE REGIONALIZATION OF RADICAL CYSTECTOMY TO SPECIFIC MEDICAL CENTERS BRENT K. HOLLENBECK,* DAVID A. TAUB, DAVID C. MILLER, RODNEY L. DUNN, JAMES E. MONTIE† AND JOHN T. WEI‡ From the Department of Urology, University of Michigan, Ann Arbor, Michigan
ABSTRACT
Purpose: Regionalization of high risk surgical procedures to larger teaching hospitals has been suggested as a means to improve the quality of care. We established a novel framework for characterizing regionalization, implemented it to determine the extent to which regionalization of radical cystectomy has occurred and delineated whether specific patient characteristics are associated with this phenomenon. Materials and Methods: We used the Nationwide Inpatient Sample to identify 22,088 patients who underwent radical cystectomy for bladder cancer from 1988 to 2000. Regionalization was assessed using 5 structural hospital measures, including teaching status, urban location, discharge volume, cystectomy volume and bed capacity. Adjusted models were developed to identify the significance of temporal trends and assess the association of demographic factors with structural qualities. Results: Compared with 1988 to 1990 subjects were more likely to undergo cystectomy at teaching hospitals (OR 1.8), high cystectomy volume hospitals (OR 1.2), high discharge volume hospitals (OR 1.7) and large bed capacity medical centers (OR 1.4) in 1998 to 2000. The concentration of cystectomy to urban medical centers during the study years was 90% to 92%. The proportion of subjects undergoing partial cystectomy decreased from 23.9% to 16.6% as regionalization occurred. Older subjects were less likely to be treated at these regionalized centers. Conclusions: Without broad legislation from health care payers radical cystectomy has increasingly regionalized to specific medical centers. Despite this regionalization disparities in its use exist among specific, vulnerable patients. Addressing this may facilitate further concentration of this procedure. KEY WORDS: bladder, bladder neoplasms, outcome assessment (health care), cystectomy, quality of care
Anticipated health care expenditures in the 21st century will comprise a significant proportion of the United States gross domestic product.1 Surgical services comprise approximately 40% of all hospital expenses2 and, thus, identifying potential causes of increased resource use provide a mechanism to curtail potentially avoidable expenses. These increasing health care expenditures have prompted payers to consider using quality of care to direct reimbursement3 and advocates of volume-outcome relationships have encouraged patients undergoing certain surgical procedures to seek treatment at high procedural volume centers.4 Surgeon and hospital volume are not strictly quality of care indicators per se. Rather, they are structural characteristics that are reflective of the setting in which health care is delivered5 and commonly used surrogate measures of quality because of expediency. Furthermore, whether measured at the hospital or surgeon level volume has been associated with lower morbidity6 and mortality7, 8 for various surgical procedures, including radical cystectomy.7, 8 As a consequence of these data, there is a push among payers to concentrate
specific surgical procedures (but not radical cystectomy) to a relatively few medical centers. The result of such regionalization will have significant implications to the patient and physician. Thus, the objectives of this study were 3-fold, namely to 1) develop a novel framework for measuring regionalization, 2) assess the degree to which regionalization of cystectomy for bladder cancer has occurred throughout the United States by applying this framework and 3) delineate patient characteristics associated with any potential trends in regionalization. METHODS
Subjects. Data from the Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS) were abstracted for 1988 through 2000. The NIS represents a 20% stratified sample of all hospital discharges in the United States. The database is maintained by the Agency for Health Care Research and Quality, and consists of uniform hospital discharge summaries from 994 nonfederal acute care hospitals throughout the United States. Data on eligible subjects were abstracted from the NIS using procedural terminology based on International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9) codes for cystectomy, including 577 for total cystectomy, 5771 for radical cystectomy, 576 for partial cystectomy and 5779 for other total cystectomy. In addition, ICD-9 codes for principle diagnoses were used to limit the sample to subjects with bladder cancer (188, 1880 to 1889, 2337, 2367 and 2394). Clinical Classification Software (principle diagnostic code 32), used in the context of the NIS
Submitted for publication January 7, 2005. Presented at annual meeting of American Urological Association, San Francisco, California, May 8 –13, 2004. * Correspondence: Department of Urology, Taubman Health Care Center, Room 3875, University of Michigan Health System, 1500 East Medical Center Dr., Ann Arbor, Michigan 48109-0330 (telephone: 734-615-0563; FAX: 734-936-9127; e-mail: bhollen@ umich.edu). † Financial interest and/or other relationship with AstraZeneca. ‡ Financial interest and/or other relationship with Sanofi, Laserscope, Calypso and Boehringer Ingelheim. 1385
1386
REGIONALIZATION OF RADICAL CYSTECTOMY
to partition subjects into relevant categories, was used to validate our sampling methodology. Secondary ICD-9 diagnostic codes were abstracted to enumerate comorbid conditions according to the Romano modification of the Charlson comorbidity index.9 Demographic information, including subject age, race, sex, income based on median income in the subject zip code, hospital admission type, treatment year, geographic region of treatment based on United States Census regions and insurance type were also abstracted. In accordance with the Code of Federal Regulations, Title 21, Section 46.101, subparagraph 4 institutional review board approval was not sought for this study. Measures of regionalization. To test the null hypothesis that cystectomy procedures did not migrate to select referral centers during the study course we developed a novel framework to measure regionalization within the NIS by identifying medical centers with specific hospital qualities where cystectomies may concentrate. These 5 qualities were 1) teaching hospital status based on 1 of certain factors, namely the presence of any American Medical Association approved residency program, membership status to the Council of Teaching Hospitals or a ratio of full-time equivalent interns and residents to beds of greater than 0.25, 2) urban hospital status, as described by location within a metropolitan statistical area based on the United States Census, 3) largest hospital status, based on the tertile of hospitals within a given geographic region containing the greatest hospital bed capacity, 4) high cystectomy volume center status, as indicated by hospitals where the most cystectomies were performed annually (9 or more yearly, that is the top 40th percentile) and 5) high discharge volume status, corresponding to hospitals with greater than 24,677 discharges among all diagnoses yearly (the top 20th percentile). Construction of a hospital cystectomy volume variable. The number of procedures performed at each hospital from 1988 to 2000 was ascertained using a unique hospital identification code. All years were considered together as 1 entity and each hospital was assigned a volume group for each year that it participated. Each individual hospital-year was considered independently and then ranked in order of increasing annual case volume. Five volume groups were defined by selecting whole number cutoffs for annual hospital volume that most closely sorted patients into groups of equal size (quintiles). In this manner the numerical thresholds for hospital volume were 1 to 3 cases yearly—very low, 4 or 5—low, 6 to 8 —medium, 9 to 14 — high and more than 14 —very high. For the purposes of regionalization assessment the top 2 quintiles were combined to identify high and very high volume centers. Statistical analyses. For the purposes of demonstrating changes in the measure of regionalization with time in multivariable modeling the treatment year was categorized into certain groups, including 1988 to 1991, 1991 to 1994, 1995 to 1997 and 1988 to 2000. Because the definition of a teaching hospital in the NIS changed beginning in 1998, longitudinal analysis of teaching hospital status was not possible using the complete data set. To allow contingent analysis of factors associated with teaching and urban hospital status (nested within the teaching status variable) we used a subset of the data that incorporates a uniform definition of teaching and urban hospitals. Only hospitals classified consistently throughout the study period were included. This decresed the number of cystectomy procedures in the analysis from 22,088 to 17,630 discharges. The analysis of this smaller data set was not weighted due to the skewed sampling required to obtain the uniform teaching/urban hospital definition. All models and bivariate analyses not involving the teaching/ urban status of a hospital were weighted to reflect the national universe of 119,491 cystectomy discharges during the study course. Unadjusted cystectomy use rates at medical centers, as described by the various measures of regionalization, were calculated by treatment year. Bivariate analyses
were performed using the chi-square test and ANOVA. Logistic regression analyses were performed to adjust for the various covariates and determine the propensity for cystectomy use at the medical centers described by our regionalization measures. An indicator variable for missing data was included for race (9% missing) to minimize observational exclusions from the models. Multicollinearity diagnostics were performed to ensure the lack of strong linear tendencies among the explanatory variables. All testing was completed using SAS, version 8.2 computerized statistical software (SAS Institute, Cary, North Carolina), was 2-tailed and was performed at the 5% significance level. RESULTS
The mean age ⫾ SD of subjects in our study was 68.6 ⫾ 10.6 years, which did not significantly vary among years (p ⫽ 0.78). Table 1 lists other patient characteristics and changes in them with time. Table 2 shows unadjusted trends in the proportion of cystectomies performed at hospitals, characterized by the 5 measures of regionalization measures. Table 3 shows the final adjusted models, illustrating the association of covariates with the propensity of subjects to be treated at these regional cystectomy centers. DISCUSSION
Less common but more complex surgical procedures have been characterized as carrying a greater risk of adverse events and increasing public scrutiny has been placed on patient safety and defining metrics for quality of surgical care.10 Consequently there is a nationwide movement afoot calling for the regionalization of certain surgical procedures that looms over various surgical disciplines.3 From a patient perspective the implications of such a directive in terms of travel time have been addressed.11 However, it is possible that the evolution of care for patients requiring certain surgical procedures has naturally progressed to specific medical centers and such a notion has not been explored. Potential causes of regionalization in this context are selective referral based on hospital/surgeon reputation, financial pressures to efficiently use time (eg a surgeon may be reimbursed more for performing multiple simple procedures in lieu of a single complex surgery) and patient recognition of centers of excellence. In such an environment further legislated regionalization would have questionable impact on health care quality. Radical cystectomy is 1 such procedure in which regionalization may eventually be mandated because of higher mortality rates at low volume centers7 and relatively high postoperative morbidity rates (28% to 37%).12⫺14 In the context of the quality of care paradigm, as proposed by Donabedian,15 we used 5 novel measures to provide a framework for describing regionalization. Our findings clearly demonstrate that relative to the earliest period cystectomy was more likely to be performed at teaching hospitals, high cystectomy volume hospitals, large hospitals in terms of bed size and high discharge hospitals in 1998 to 2000. Concentration of procedures at hospitals with these characteristics suggest that regionalization of cystectomy has already occurred to some degree but ample opportunity for further regionalization to such institutions still exists. Furthermore, these data illustrate that cystectomy has largely been isolated to urban medical centers because the proportion performed at these institutions was 0.90 to 0.92 during the study course, suggesting that urbanization has already occurred. Small temporal decreases in the proportion of those treated at urban centers likely reflect regression to the mean since the majority of patients are already treated at such centers. Similar trends have been described for surgeries such as hepatic resection,16 implying that current referral patterns for various complex procedures are funneling patients to centers of excellence. Furthermore, the almost com-
1387
REGIONALIZATION OF RADICAL CYSTECTOMY
TABLE 1. Sample by period, weighted to reflect American Hospital Association (AHA) universe of hospital discharges between 1988 and 2000 Covariate (level) Cystectomies (No.)* Age: Younger than 60 60–Younger than 70 70–Younger than 80 80 or Older Sex (female) Admission status (urgent) Insurance: Medicare Private Other Race: White Black Other Missing data Geographic region: Northeast Midwest South West Modified Charlson index9: Low comorbidity (0–1) Moderate comorbidity (2–3) High comorbidity (greater than 3) Partial cystectomy * Not weighted to reflect AHA universe.
% 1988–1991
% 1992–1994
% 1995–1997
% 1998–2000
p Value for Trend
29 (6,453)
23 (5,066)
24 (5,294)
24 (5,275)
⬍0.0001
15.1 33.7 38.4 12.8 21.3 30.0
16.7 31.4 39.4 12.5 20.8 29.9
17.7 29.6 39.4 13.3 20.9 27.5
19.6 27.6 38.8 14.0 20.5 18.2
⬍0.0001
65.3 20.2 8.5
66.5 26.8 6.7
64.8 28.4 6.8
61.4 31.4 7.2
⬍0.0001
33.2 0.8 63.9 2.2
69.7 3.0 25.1 2.2
74.1 4.0 18.1 3.8
68.9 3.7 4.9 22.5
⬍0.0001
20.7 19.2 38.1 22.0
22.7 20.9 37.9 18.5
21.6 22.5 36.2 17.7
23.2 26.4 31.7 18.7
⬍0.0001
48.9 49.8 1.4 23.9
41.3 56.1 3.6 21.0
37.1 58.4 4.6 19.0
36.6 59.4 4.1 16.6
⬍0.0001
0.0880 ⬍0.0001
⬍0.0001
TABLE 2. Unadjusted trends in cystectomies performed at specific medical centers Hospital Measure
Hospital Measure Description
Teaching hospital*
% 1988–1991
Any American Medical Association approved residency program, membership status to Council of Teaching Hospitals, or 0.25 resident/bed ratio Urban hospital* Metropolitan statistical area based on United States Census Largest capacity† Tertile of hospitals containing most beds in given geographic region High cystectomy vol† Hospitals in which 9 or more procedures were performed/yr (upper 40th percentile of all hospitals) High discharge vol† Of all diagnoses this corresponds to 24,678 or more discharges/yr (upper 20th percentile of all hospitals) * Not weighted to reflect AHA universe due to sampling changes made in 1998. † Weighted to reflect the AHA universe of 119,491 discharges following cystectomy
plete urbanization of cystectomy suggests that further regionalization to specific centers with demonstrable quality benefits will have little additional impact on patient preferences for location of care since the majority of procedures have migrated out of rural communities.17 Prior to mandating regionalization of these procedures efforts should first be directed toward establishing metrics of quality and processes of care that can accurately distinguish centers of excellence from other institutions, thereby, facilitating pilot programs aimed at concentrating case volume at these facilities. The underlying etiologies of the observed spontaneous regionalization are enigmatic. However, potential causes may be influenced by reimbursement, difficult postoperative management and evolving data regarding the effects of structural characteristics on outcomes.7, 8 Future study that elicits the cause for this regionalization will help elucidate whether this phenomenon is a consequence of self-governance (because of perceived better outcomes at medical centers with the mentioned structural characteristics), of financial considerations (eg lower reimbursement) or of patient specific factors (eg word of mouth referral). Because of significant variation in practice patterns with time,18 we examined several common factors associated with
% 1992–1994
% 1995–1997
% 1998–2000
p Value for Trend
34.4
35.0
40.6
45.3
⬍0.0001
91.6
90.6
89.9
90.5
0.0143
61.2
62.8
56.2
68.0
⬍0.0001
35.5
35.5
34.5
44.0
⬍0.0001
15.2
20.1
26.2
26.4
⬍0.0001
that occurred during the study period.
its use at medical centers described by these structural characteristics where the procedure was most concentrated. An interesting phenomenon was that the use of radical cystectomy relative to partial cystectomy, which is a less technically demanding and less morbid procedure,19 was 60% to 100% more likely at regionalized centers, namely teaching, urban, high cystectomy volume and high total discharge volume hospitals. Furthermore, because this regionalization occurred during the study course, the proportion of subjects undergoing partial cystectomy continually decreased. Given the relative rare indications for partial cystectomy (about 6% of all patients with cystectomy),20 it is not surprising that regionalization has likely brought about more appropriate care to a larger patient population. Not surprisingly subjects undergoing cystectomy had more medical comorbidities in the later study years, which may represent the narrowing of contraindications to surgery due to advances in perioperative critical care. As a consequence, teaching institutions, urban hospitals and high cystectomy volume centers had a greater propensity for using cystectomy in patients with greater infirmity. This suggests that these centers had intrinsic resources/infrastructure/processes of care to provide for this population, leading to selec-
1388
REGIONALIZATION OF RADICAL CYSTECTOMY
TABLE 3. Final adjusted models ascertaining clinical and structural factors associated with propensity to undergo radical cystectomy at hospitals characterized by 5 regionalization qualities Level
Elective admission Sex (female) Insurance: Private Other Race: Black Hispanic/other Missing data Geographic region: Midwest Northeast South Charlson index: Greater than 3 2 or 3 Subject age: Older than 60–70 Older than 70–80 Older than 80 Treatment yr: 1998–2000 1995–1997 1992–1994 Radical cystectomy
Teaching Hospital*
Reference Level
Urban Medical Center*
Highest Cystectomy Vol†
Largest Hospitals (No. beds)†
Hospitals With Most Discharges (all diagnoses)†
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
OR
95% CI
Emergent admission Male
0.8
0.8–1.0
1.4
1.2–1.6
1.3
1.3–1.4
—
—
1.2
1.1–1.2
—
—
—
—
0.8
0.8–0.8
—
—
—
—
Medicare
1.1 1.3
1.0–1.2 1.2–1.5
2.0 1.2
1.7–2.4 0.9–1.5
—
—
—
—
1.1 1.2
1.0–1.1 1.1–1.3
White
2.7 1.2 0.8
2.2–3.3 1.1–1.3 0.7–1.0
1.2 0.7 0.5
0.8–1.7 0.6–0.7 0.4–0.6
0.8 0.9 0.8
0.7–0.8 0.8–0.9 0.7–0.8
1.0 1.1 1.2
0.9–1.0 1.1–1.2 1.2–1.3
1.4 1.0 0.9
1.3–1.5 1.0–1.1 0.9–1.0
West
1.3 3.7 0.6
1.2–1.5 3.4–4.1 0.6–0.7
0.8 2.6 1.9
0.7–0.9 2.2–3.2 1.6–2.2
0.7 0.9 0.9
0.7–0.7 0.8–0.9 0.8–0.9
2.2 1.3 2.1
2.1–2.2 1.3–1.3 2.0–2.2
1.2 1.7 1.5
1.2–1.3 1.6–1.8 1.5–1.6
0 or 1
1.2 0.9
1.1–1.3 0.9–1.0
1.2 0.9
1.0–1.3 0.8–1.0
1.2 0.9
1.2–1.2 0.9–0.9
—
—
—
—
60 or Younger
0.8
0.7–0.9
1.0
0.8–1.2
0.8
0.8–0.9
—
—
0.9
0.8–0.9
0.6
0.6–0.7
0.9
0.8–1.2
0.8
0.7–0.8
0.8
0.8–0.9
0.5
0.4–0.6
0.8
0.6–1.0
0.7
0.7–0.8
0.7
0.7–0.7
1.8 1.5 1.1 2.0
1.6–2.1 1.4–1.7 1.0–1.2 1.8–2.2
0.7 0.7 0.7 1.4
0.6–0.9 0.6–0.8 0.6–0.9 1.2–1.6
1.2 0.9 0.9 1.9
1.1–1.2 0.9–0.9 0.9–1.0 1.8–1.9
1.7 1.9 1.4 1.6
1.4–1.5 1.9–2.1 1.3–1.4 1.6–1.7
1988–1991
Partial cystectomy * Based on 17,630 patients who underwent cystectomy, not weighted to reflect entire AHA universe. † Based on 22,088 patients who underwent cystectomy, weighted to reflect entire AHA universe.
tive referral to these centers, or they were less selective in their patient pool and, hence, more procedures were performed. Younger patient age was also commonly associated with medical centers with structural characteristics suggesting regionalization. Relative to patients 60 years or younger, older subjects are less likely to seek treatment at medical centers where cystectomy has concentrated. Collectively these data highlight that among certain vulnerable populations, namely the elderly and the infirm (multiple comorbidities), disparities in the use of cystectomy at centers of regionalization exist and determining potential factors associated with these discrepancies may elicit health care access barriers in these groups. This notion is further supported by the fact that patients with Medicare, which is presumably the most reliable payer, were less commonly treated at regionalized centers, suggesting that reimbursement may have an underlying role in health services delivery. Future study should focus on determining the principle causes of these disparities in cystectomy use, namely access to care (income, insurance and level of education) and selective treatment patterns (younger age), before further regionalization of care can address societal needs. The use of administration data is not without its inherent limitations. Ideal case mix adjustment is limited because of data collection methodology and constraints, although the application of a widely used comorbidity index in the context of administrative datasets facilitates comparison.9 Our findings that cystectomy has concentrated to medical centers with specific structural characteristics in the last 13 years do not validate this regionalization as a means of improving health care quality. Rather, they highlight trends in practice patterns, of which the causes have yet to be elicited. Finally, the relationship between potential structural characteristics (eg teaching status, annual case volume, high discharge volume, urban medical centers status and large bed size capacity) and common adverse events (eg prolonged ileus and
1.4 0.8 1.1 1.0
1.3–1.4 0.8–0.9 1.1–1.1 1.0–1.1
wound complications) following cystectomy has not been established and further study in this context will help determine the usefulness of these structural indicators. CONCLUSIONS
This study provides a novel framework for assessing the regionalization of health services delivery to specific medical centers. The usefulness of this framework is illustrated by the observed regionalization of radical cystectomy to specific institutions with certain structural traits that has occurred in the last 13 years. Furthermore, the urbanization of cystectomy is almost complete with approximately 90% of all procedures occurring in metropolitan hospitals throughout study course. Concentration of cystectomy to these institutions (teaching, large bed capacity, high cystectomy volume, high discharge volume and urban) has been associated with a concurrent decrease in the proportion of subjects undergoing partial cystectomy, which may reflect more appropriate care. Nonetheless, disparities among cystectomy use in vulnerable patient populations exist. Eliciting and eliminating the etiologies of these disparities should minimize the obstacles to the further spontaneous regionalization of cystectomy. REFERENCES
1. Chernew, M. E., Hirth, R. A. and Cutler, D. M.: Increased spending on health care: how much can the United States afford? Health Aff (Millwood), 22: 15, 2003 2. Macario, A., Vitez, T. S., Dunn, B., McDonald, T. and Brown, B.: Hospital costs and severity of illness in three types of elective surgery. Anesthesiology, 86: 92, 1997 3. Birkmeyer, J. D.: Should we regionalize major surgery? Potential benefits and policy considerations. J Am Coll Surg, 190: 341, 2000 4. Birkmeyer, J. D.: High-risk surgery—follow the crowd. JAMA, 283: 1191, 2000 5. Birkmeyer, J. D., Dimick, J. B. and Birkmeyer, N. J.: Measuring
1389
REGIONALIZATION OF RADICAL CYSTECTOMY
6. 7. 8.
9. 10.
11. 12. 13.
14.
15. 16.
the quality of surgical care: structure, process, or outcomes? J Am Coll Surg, 198: 626, 2004 Begg, C. B., Riedel, E. R., Bach, P. B., Kattan, M. W., Schrag, D., Warren, J. L. et al: Variations in morbidity after radical prostatectomy. N Engl J Med, 346: 1138, 2002 Birkmeyer, J. D., Siewers, A. E., Finlayson, E. V., Stukel, T. A., Lucas, F. L., Batista, I. et al: Hospital volume and surgical mortality in the United States. N Engl J Med, 346: 1128, 2002 Birkmeyer, J. D., Stukel, T. A., Siewers, A. E., Goodney, P. P., Wennberg, D. E. and Lucas, F. L.: Surgeon volume and operative mortality in the United States. N Engl J Med, 349: 2117, 2003 Romano, P. S., Roos, L. L. and Jollis, J. G.: Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives. J Clin Epidemiol, 46: 1075, 1993 Goodney, P. P., Siewers, A. E., Stukel, T. A., Lucas, F. L., Wennberg, D. E. and Birkmeyer, J. D.: Is surgery getting safer? National trends in operative mortality. J Am Coll Surg, 195: 219, 2002 Birkmeyer, J. D., Siewers, A. E., Marth, N. J. and Goodman, D. C.: Regionalization of high-risk surgery and implications for patient travel times. JAMA, 290: 2703, 2003 Frazier, H. A., Robertson, J. E. and Paulson, D. F.: Complications of radical cystectomy and urinary diversion: a retrospective review of 675 cases in 2 decades. J Urol, 148: 1401, 1992 Stein, J. P., Lieskovsky, G., Cote, R., Groshen, S., Feng, A. C., Boyd, S. et al: Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol, 19: 666, 2001 Koch, M. O. and Smith, J. A., Jr.: Influence of patient age and co-morbidity on outcome of a collaborative care pathway after radical prostatectomy and cystoprostatectomy. J Urol, 155: 1681, 1996 Donabedian, A.: The definition of quality and approaches to its assessment. In: Explorations in Quality Assessment and Monitoring. Ann Arbor: Health Administration Press, vol. 1, 1980 Dimick, J. B., Wainess, R. M., Cowan, J. A., Upchurch, G. R., Jr., Knol, J. A. and Colletti, L. M.: National trends in the use and outcomes of hepatic resection. J Am Coll Surg, 199: 31, 2004
17. Finlayson, S. R., Birkmeyer, J. D., Tosteson, A. N. and Nease, R. F., Jr.: Patient preferences for location of care: implications for regionalization. Med Care, 37: 204, 1999 18. Konety, B. R. and Joslyn, S. A.: Factors influencing aggressive therapy for bladder cancer: an analysis of data from the SEER program. J Urol, 170: 1765, 2003 19. Weinstein, R. P., Grob, B. M., Pachter, E. M., Soloway, S. and Fair, W. R.: Partial cystectomy during radical surgery for nonurological malignancy. J Urol, 166: 79, 2001 20. Utz, D. C., Schmitz, S. E., Fugelso, P. D. and Farrow, G. M.: Proceedings: a clinicopathologic evaluation of partial cystectomy for carcinoma of the urinary bladder. Cancer, 32: 1075, 1973 EDITORIAL COMMENT The data in this study confirmed what most people have observed anecdotally, ie that radical cystectomy and other major surgical procedures in urology are increasingly being concentrated at tertiary and high volume hospitals. As the authors point out, the reasons are multifactorial but the relative financial disincentives for complex surgery are certainly contributory. Across the board urology is differentiating into high volume surgical practice or office based practice. This may be a good thing for patient care because numerous studies show improved outcomes when major surgical procedures are performed by high volume surgeons at high volume hospital settings. Although radical cystectomy is accepted as the most effective treatment for muscle invasive bladder cancer, multiple studies show that only around 30% to 40% of patients with documented muscle invasive urothelial cancers ever undergo cystectomy. Again, there are many reasons for this but concerns about operative morbidity and mortality are paramount. As noted by these authors, the regionalization of care may help address the problem. If greater experience generally equates with improved outcome, perhaps more patients will be considered eligible for surgery. Joseph A. Smith, Jr. Department of Urologic Surgery Vanderbilt University Nashville, Tennessee