- Email: [email protected]
“Never Events”: Centers for Medicare and Medicaid Services Complications After Radical Cystectomy
Health Outcomes Research “Never Events”: Centers for Medicare and Medicaid Services Complications After Radical Cystectomy Gregory A. Joice, Christopher M. Deibert, Max Kates, Benjamin A. Spencer, and James M. McKiernan OBJECTIVE MATERIALS AND METHODS
RESULTS
CONCLUSION
To describe “never event” (NE) complications after radical cystectomy. The Centers for Medicare and Medicaid Services has denied reimbursement for 10 hospitalacquired conditions that were deemed “reasonably preventable.” These NEs do not take into consideration pre-existing risk factors that make patients more susceptible. Radical cystectomy is a complex surgery that is often necessary in a population with extensive comorbidities. In this setting, the application of an unmodified system to deny reimbursement of adverse outcomes could have a significant effect on healthcare delivery. We hypothesized that measurable patient and hospital characteristics could predict the occurrence of NE complications after radical cystectomy. Using the weighted Nationwide Inpatient Sample database, we identified 61,142 patients with bladder cancer who underwent radical cystectomy from 2002 to 2009. The NE rates were calculated, and their effect on in-hospital mortality, length of stay, and total hospital costs were determined by multivariate regression analysis. The rate for any NE was 2.42%. Vascular-catheter infections (1.25%) were the most common. Black race and comorbidities increased the likelihood of a NE. The presence of any NE increased the average length of stay (by 15 days), total costs (by $37,000), and in-hospital mortality (8.0% vs 2.2%). Centers for Medicare and Medicaid Services NEs are more likely in older patients with comorbidities. NEs strongly predict for negative patient outcomes during hospitalization (length of stay, costs, in-hospital mortality). High-risk populations could benefit from risk adjustment before implementing a significant alteration in hospital or physician reimbursement policies. UROLOGY 81: 527e532, 2013. 2013 Elsevier Inc.
B
ladder cancer is primarily a disease of the elderly, with >70% of patients diagnosed after age 65.1 In addition, nearly 70% of patients with genitourinary malignancies have 1 comorbidity, more than patients with other cancers, placing these patients at high risk of complications.2 Radical cystectomy with lymph node dissection remains the standard surgical treatment of muscle-invasive bladder cancer.3 After radical cystectomy, 5-year recurrence-free survival ranges from 66% to 73% and overall survival from 62% to 66%.4,5 Financial Disclosure: The authors declare that they have no relevant financial interests. Funding Support: This work was supported by a grant from the Doris Duke Charitable Foundation to Columbia University in support of Clinical Research Fellow M. Kates. From the Department of Urology, Columbia University College of Physicians and Surgeons, New York, NY; Department of Epidemiology, Columbia University Mailman School of Public Health, New York, NY; and New York Presbyterian Hospital, New York, NY Reprint requests: James M. McKiernan, M.D., Department of Urology, Columbia University College of Physicians and Surgeons, 161 Fort Washington Avenue, New York, NY 10032. E-mail: [email protected] Submitted: May 17, 2012, accepted (with revisions): September 20, 2012
ª 2013 Elsevier Inc. All Rights Reserved
However, radical cystectomy is a long, complex, and morbid procedure. The 90-day complication rates after radical cystectomy range from 28% to 64%, with inhospital mortality rates of 2%-3%.5-10 Advanced age and comorbidities are associated with an increased rate of complications after radical cystectomy.8 These investigators, and others, have also shown that the occurrence of 1 complications significantly increased the chances of in-hospital mortality (5.36% vs 1.47%), increased the length of stay (LOS) (13 vs 9 days), and increased the total costs ($53,342 vs $38,339).9 However, no study has specifically considered the predictive ability of these characteristics for “preventable” complications in the radical cystectomy population. As a part of the Deficit Reduction Act of 2005, the Secretary of Health and Human Services was instructed by Congress to identify hospital-acquired conditions that were (1) high cost or high volume, or both; (2) resulted in assignment to a higher paying diagnosis-related group when present as a secondary diagnosis; and (3) were reasonably preventable through the application of 0090-4295/12/$36.00 http://dx.doi.org/10.1016/j.urology.2012.09.050
527
evidence-based medicine.11 Conditions were considered reasonably preventable if guidelines were available from professional organizations or academic institutions, with evidence quantifying the extent to which applying these guidelines would result in the prevention of the conditions.12 With this charge, the Centers for Medicare and Medicaid Services (CMS) identified 10 hospital-acquired conditions, also known as “never events” (NEs) (Table 1). As of October 1, 2009, hospitals were no longer reimbursed for the additional care delivered if a patient developed any of these 10 NEs. The policy was intended to motivate hospitals to accelerate programs in patient safety but was met with much resistance.13 Major American healthcare system leaders have suggested that not all NEs are preventable.14 Furthermore, the CMS did not include any exclusions or adjustments for patient demographics or procedures that place patients at greater risk of developing an NE. Pronovost et al15 suggested that this policy might create disincentives for hospitals to treat sicker patients. With the widespread focus on quality of care and the introduction of financial consequences to poor quality, we examined the factors that might predict the development of NEs after radical cystectomy for bladder cancer. If such factors can readily be identified, they could be used to stratify hospitals and procedures according to their risk of patients developing NEs. This risk stratification could be incorporated into current and future CMS reimbursement policies. We hypothesized that patients with more comorbid conditions would experience more NEs, leading to increased in-hospital mortality, LOS, and hospital costs. We also anticipated that higher volume centers would experience fewer NEs, leading to improved outcomes.
MATERIAL AND METHODS Data Source and Patient Selection The Nationwide Inpatient Sample (NIS) is the largest all-payer inpatient database in the United States, consisting of a 20% stratified sample of community and academic hospitals. The NIS contains data elements for each discharge that include the primary and secondary diagnoses and procedures, patient demographics, and hospital characteristics. The recent update of the NIS through 2009 included approximately 1000 hospitals from 44 states.16 The NIS also contains a weighting method based on the hospital stratification to allow for creation of national estimates.17 We queried the 2002-2009 NIS for any adult aged >40 years with a diagnosis of bladder cancer (Clinical Classifications Software diagnosis code 32) who had undergone radical cystectomy (“International Classification of Disease, 9th revision, Clinical Modification” [ICD-9-CM] procedure codes 57.71 and 57.79). We identified CMS NEs by searching each of the diagnosis variables for the relevant ICD-9-CM diagnosis codes according to the most recent documentation provided by the CMS (Table 1).18 Deep vein thrombosis, pulmonary embolism, and surgical site infection are not considered a NE after radical cystectomy, only after certain orthopedic, cardiac, and gastric bypass procedures. Because many of the codes changed between 2002 and 2009, we used the historical codes provided by the 528
Table 1. Centers for Medicare and Medicaid Services hospital-acquired conditions Hospital-acquired Condition Foreign object retained after surgery Air embolism Blood incompatibility
Pressure ulcers, stage III and IV Falls and trauma Fracture Dislocations Intracranial injury Crushing injury Burn Electric shock Catheter-associated UTI Vascular catheter-associated infection Manifestations of poor glycemic control Diabetic ketoacidosis Nonketotic hyperosmolar coma Hypoglycemic coma Secondary diabetes with ketoacidosis Secondary diabetes with hyperosmolarity Total
2010 ICD-9-CM Codes 998.4 998.7 999.1 999.60 999.61 999.62 999.63 999.69 707.23 707.24
Events (n) 32 (0.05) 5 (0.01) 0 (0.00)
545 (0.89) 62 (0.09)
800-829 830-839 850-854 925-929 940-949 991-994 996.64 999.31
42 (0.06) 765 (1.25) 40 (0.06)
250.10-250.13 250.20-250.23 251.0 249.10-249.11 249.20-249.21 1464 (2.42)
ICD-9-CM, “International Classification of Disease, 9th revision, Clinical Modification”; UTI, urinary tract infection. Data in parentheses are percentages.
Centers for Disease Control to identify the NEs from earlier years.19
Independent Variables The patient demographics included age, sex, race, payer, median household income, and number of comorbidities. We used the comorbidity software included with the NIS to create comorbidity measures, as designed by Elixhauser et al.20 We divided the median household income into 3 groups according to 2 postal code-level income variables provided in the NIS: low ($1-$24,999), medium ($25,000-$34,999), and high ($35,000). The hospital characteristics consisted of the number of hospital beds, teaching status, and cystectomy volume. We calculated the cystectomy volume by averaging the number of cystectomies performed annually for each hospital and then creating low-, medium-, and high-volume tertiles.
Outcomes The 3 main outcomes included in-hospital mortality, LOS, and total hospital costs. A LOS at or greater than the 90th percentile was considered “prolonged” and was found to be >20 days. Similarly, total costs at the 90th percentile were defined as “increased” and were found to be >$52,711. UROLOGY 81 (3), 2013
Statistical Analysis Baseline differences in patient demographics, hospital characteristics, and outcomes were compared using the chi-square test for categorical variables and Student’s t test for continuous variables. Factors predicting the development of an NE were assessed using univariate and multivariate logistic regression models. Using similar regression analyses, we also examined the effect that the presence of any CMS NE had on in-hospital mortality, elongated LOS, and increased total costs. All multivariate regression analyses were adjusted for clustering using the generalized estimating equations method.21-24 To assess the stepwise effect that each variable had on costs and LOS, linear regression models were also developed. All analyses were performed using SAS, version 9.2 (SAS Institute, Cary, NC) with a 2-sided significance level of P <.05.
RESULTS Overall, we identified 12,451 patients (nationally weighted sample of 61,142) who had undergone radical cystectomy for a diagnosis of bladder cancer in the NIS from 2002 to 2009. The frequency of the individual CMS NEs is listed in Table 1. The overall rate of an individual patient developing 1 NE was 2.42%. Of those developing an NE, most patients developed only 1 (2.35%); however, a few developed 2 NEs (0.05%). The most common NEs were vascular-catheter infection (1.25%) and pressure ulcer (0.89%). All other NEs occurred very rarely (<0.10%) and included falls and trauma (0.09%), catheter-associated urinary tract infection (0.06%), manifestations of poor glycemic control (0.06%), foreign object retained after surgery (0.05%), air embolism (0.01%), and blood incompatibility (0.00%). The patient demographics, hospital characteristics, and outcomes between patients who developed no NEs and those who developed 1 NE are listed in Table 2. Most of these patients were white (66%) and male (82%). A greater percentage of black patients developed 1 NE (5.3% vs 3.3%) than no NE. The mean age was 67.9 years for those who did not develop an NE and 69.2 years for those who did. Most patients had Medicare (61%) or private health insurance (31%). Among the patients who developed 1 NE, a greater percentage had Medicare (67% vs 60%). Most patients had 1 comorbidity (84%); however, those who developed 1 NE were more likely to have >3 comorbidities than were those without an NE (46% vs 31%). Most patients underwent cystectomy at an urban, teaching hospital (68%). The median LOS was 9 days, and the median costs were $23,609. Those patients who had 1 NE had greater in-hospital mortality (8.0% vs 2.2%, P <.01), longer mean LOS (26 vs 10 days, P <.01), and greater hospital costs ($66,792 vs $29,792, P <.01). The results of the multivariate analyses predicting the development of any NE adjusted for all covariates are listed in Table 3. We identified several predictors of experiencing 1 NE, including black race (odds ratio [OR] 1.59, 95% confidence interval [CI] 1.27-1.99 vs white rate, P <.01), increasing age (OR 1.02, 95% CI 1.01-1.02, P <.01, continuous variable), 1 comorbidity UROLOGY 81 (3), 2013
Table 2. Baseline characteristics of patients undergoing cystectomy (n ¼ 61,142) Independent Variable Total Age (y) Sex Male Female Race White Black Hispanic Asian or Pacific Islander Native American Other Missing Hospital size (by no. of beds) Small Medium Large Hospital location/ teaching status Rural Urban, nonteaching Urban, teaching Hospital surgical volume (per year) Small (5) Medium (5-17) Large (>17) Comorbidities (n) 0 1 2 >3 Payer Private Medicare Medicaid Self-pay No charge Other Median household income* Quartile 1 (1-34,999) Quartile 2 (35,000-44,999) Quartile 3 (45,000)
Patients With HAC No
Yes
59,677 (97.6) 1464 (2.4) 68.3 70.4 48,949 (82) 10,728 (18) 38,436 1970 1484 669
(64.4) (3.3) (2.5) (1.1)
206 (0.3) 948 (1.6) 15,964 (26.7)
1,208 (82.5) 256 (17.5) 968 78 55 5
P Value <.01 .66 <.01
(66.1) (5.3) (3.7) (0.3)
5 (0.3) 20 (1.4) 333 (22.7) .08
5,294 (8.9) 155 (10.6) 9,985 (16.7) 241 (16.5) 44,398 (74.4) 1,068 (72.9) 3,150 (5.3) 14,923 (25)
<.01
58 (4) 320 (21.8)
41,604 (69.7) 1,086 (74.2) 20,677 (34.7) 19,080 (31.9) 19,920 (33.4)
545 (37.2) 492 (33.6) 427 (29.2)
9,611 15,647 15,427 18,992
(16.1) (26.2) (25.9) (31.8)
123 342 324 675
18,416 36,460 2,406 1,157 158 1,080
(30.9) 280 (19.1) (61.1) 1,071 (73.1) (4) 64 (4.4) (1.9) 36 (2.5) (0.3) 0 (0) (1.8) 13 (0.9)
(8.4) (23.4) (22.1) (46.1)
<.01
<.01
<.01
.04 12,055 (20.2)
265 (18.1)
16,104 (27)
430 (29.3)
31,518 (52.8)
769 (52.6)
HAC, hospital-acquired condition. Data in parentheses are percentages. P value for chi-square of each variable. * Determined by postal code.
(OR 1.32, 95% CI 1.25-1.39 vs none, P <.01), Medicare and Medicaid insurance (OR 1.92, 95% CI 1.65-2.24, and OR 1.77, 95% CI 1.37-2.25, P <.01, respectively, vs private). Lower odds of developing a NE were associated with a high hospital cystectomy volume (OR 0.75, 95% CI 0.65-0.87 vs low volume, P <.01). The presence of 1 NE was strongly associated with a prolonged LOS (OR 529
Table 3. Predicting never events CMS Event (Multivariate Analysis) Independent Variable Sex Male Female Race White Black Hispanic Asian or Pacific Islander Native American Other Missing Age (continuous) Hospital size (by no. of beds) Small Medium Large Hospital location/ teaching status Rural Urban, nonteaching Urban, teaching Hospital surgical volume (per year) Small (5) Medium (5-17) Large (>17) Comorbidities (every 1 additional comorbidity) Payer Private Medicare Medicaid Self-pay No charge Other Median household income (by postal code) Quartile 1 (1-34,999) Quartile 2 (35,000-44,999) Quartile 3 (45,000)
OR (95% CI)
P Value .07
Reference 0.88 (0.77-1.01) Reference 1.59 (1.27-1.99) 1.46 (1.12-1.89) 0.69 (0.41-1.17)
<.01 <.01 .17
0.90 (0.37-2.21) 0.91 (0.61-1.36) 0.83 (0.73-0.94) 1.02
.83 .65 <.01 <.01
Reference 0.89 (0.73-1.09) 0.90 (0.76-1.07)
.28 .25
Reference 1.26 (0.94-1.67) 1.57 (1.19-2.07)
.1 <.01
Reference 1.00 (0.88-1.13) 0.75 (0.65-0.87) 1.32 (1.25-1.39)
.97 <.01 <.01
Reference 1.92 (1.65-2.24) 1.77 (1.37-2.25) 1.96 (1.38-2.80) NA 0.74 (0.42-1.29)
<.01 <.01 <.01 NA .29
Reference 1.18 (1.02-1.37) 1.14 (0.99-1.30)
.02 .05
CI, confidence interval; CMS, Centers for Medicare and Medicaid Services; NA, not applicable; OR, odds ratio.
8.76, 95% CI 7.89-9.76, P <.01), increased total costs (OR 7.49, 95% CI 6.65-8.26), and in-hospital mortality (OR 3.48, 95% CI 2.91-4.18; Table 4). The linear relationship of each independent variable to the cost and LOS included a base cost of $18,835. An NE added $31,332 to the cost of care and also added 12.48 extra days to the LOS. Each comorbidity added $3819 and 0.84 day to care for patients after radical cystectomy (data not shown).
COMMENT The present analysis showed that CMS NEs occurred in 2.42% of patients after radical cystectomy and have a significant effect on LOS and mortality. The least common NEs are blood incompatibility (0.00%), air embolism (0.01%), foreign object retained after surgery 530
(0.05%), and manifestations of poor glycemic control (0.05%). All 4 of these events can be directly associated with provider error and are generally considered completely preventable. Alternatively, the most common NEs were vascular catheter-associated infection (1.25%) and pressure ulcers (0.89%). These findings suggest that the CMS might have correctly identified some NEs as true preventable events but might have incorrectly identified others as being truly preventable. Alternatively, improved structural processes might further reduce the frequency of these events. Several patient factors, including increasing age and comorbidities, predicted developing a CMS NE. In this patient population, increasing age is often associated with more complications.8,25 Increasing age, when described by decade, added 20% to the risk of developing a NE. The presence of 1 comorbidity increased the likelihood of developing an NE by 28%. In a surgical population in which 84% have 1 comorbid condition, this is very clinically relevant. Because both age and patient comorbidities are unmodifiable risk factors, the current CMS policy suggests a financial incentive for hospitals to avoid treating patients with bladder cancer who are also elderly and ill. This has particular significance for tertiary and quaternary academic referral hospitals, because they care for many of these patients, and the policy does not explicitly stipulate for any additional resources to be provided to these centers for the prevention of NEs. Thus, the effect of this policy could result in lower reimbursements to these centers simply because they treat patients with more challenging high-risk conditions. Among the hospital characteristics, hospitals with high cystectomy volumes were less likely to have NEs after radical cystectomy. This adds to the evidence demonstrating an inverse relationship between hospital cystectomy volume and negative patient outcomes.26,27 Despite controlling for the clustering of hospital characteristics, including volume, location, and teaching status, patients treated at urban-teaching hospitals were still more likely to develop NEs. Unquantifiable hospital-specific processes of care might account for the complex interaction between hospital case volume and teaching status. The present study has also showed that the development of any NE significantly increased the LOS, costs, and in-hospital mortality. This suggests that the CMS has correctly identified conditions that cause an increased cost burden on the healthcare system. Patient care programs designed to prevent NEs must demonstrate efficacy in terms of event rate reduction and subsequent cost and LOS reduction. The development of an NE affected the LOS more (OR 8.8) than it increased total costs (OR 7.5) or in-hospital mortality (OR 3.5). The development of a NE would naturally lead to a longer LOS and increased costs related to the treatment of the NE. These findings suggest that preventing NEs might have the greatest effect on decreasing the patient LOS with a corresponding decrease in the financial burden of hospitalization. UROLOGY 81 (3), 2013
Table 4. Multivariate regression analysis for outcomes Independent Variable Elongated LOS Any CMS HAC None 1 Increased total costs Any CMS HAC None 1 HAC In-hospital mortality Any CMS HAC None 1 HAC
Patients (n)
Multivariate OR (95% CI)
4,827 (8.1) 675 (46.1)
Reference 8.8 (7.9-9.8)
5,513 (9.24) 663 (45.3)
Reference 7.5 (6.7-8.3)
1,362 (2.28) 117 (8.03)
Reference 3.5 (2.9-4.2)
LOS, length of stay. Data in parentheses are percentages, unless otherwise noted.
The limitations of the present study related to the use of an administrative database. The present analysis relied on coding and reporting of conditions for identification of CMS NEs. ICD-9-CM coding of a large administrative hospital data source has been demonstrated to be accurate approximately 80% of the time; thus, some complications or index subjects might have been missed if miscoded.28,29 Also, the CMS has made recent changes to the ICD-9-CM coding to help identify NEs that were not available in previous years. For example, the CMS NE of a pressure ulcer is meant to only capture the most severe ulcers, grades III and IV. However, this specific ICD-9-CM code did not exist before 2009. Before this addition, the codes were determined by the location, but not the severity, of the ulcers. For the present study, we captured all pressure ulcers within the study period, which likely overestimated the true occurrence of stage III and IV pressure ulcers. Complete adherence to the clinical guidelines for preventing NEs, such as operating room count systems or check lists for vascular catheter maintenance, might significantly reduce their prevalence in this population. However, these data were also not available from the NIS. The NIS database also does not capture any events that developed after the patient was discharged. If a patient developed a complication at home, even if readmitted to the same hospital, the NIS would not capture it. The final limitation was that the NIS database does not capture specific information on the severity of bladder cancer. Thus, cancer grade and stage could not be included in the model.
CONCLUSION CMS NEs after radical cystectomy are rare events. Vascular catheter infection and pressure ulcers were most common and were the only events with rates >0.1%. Several patient- and hospital-level factors can predict the development of these events, including advanced age, increased comorbidities, and hospital radical cystectomy volume. Once a patient has developed a NE, the patient is more likely to have a prolonged LOS, accrue increased total costs, and die in the hospital. The entire system, UROLOGY 81 (3), 2013
including physicians, physician extenders, nurses, pharmacists, and technicians, treats the diseases of hospitalized patients. There are likely many steps in care that can be improved to limit these complications. Ultimately, these findings should be used to inform future healthcare policy regarding “preventable” complications after major oncologic surgery. References 1. SEER. SEER fact sheet, 2008, vol. 2011. Available at: http://seer. cancer.gov/statfacts/html/urinb.html. Accessed October 10, 2011. 2. Piccirillo JF, Tierney RM, Costas I, et al. Prognostic importance of comorbidity in a hospital-based cancer registry. JAMA. 2004;291: 2441-2447. 3. National Comprehensive Cancer Network. 2011 Guidelines for Bladder Cancer. Available at: http://www.nccn.org/professionals/ physician_gls/f_guidelines.asp#site. Accessed October 2, 2011. 4. Madersbacher S, Hochreiter W, Burkhard F, et al. Radical cystectomy for bladder cancer today—a homogeneous series without neoadjuvant therapy. J Clin Oncol. 2003;21:690-696. 5. Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol. 2001;19:666-675. 6. Shabsigh A, Korets R, Vora KC, et al. Defining early morbidity of radical cystectomy for patients with bladder cancer using a standardized reporting methodology. Eur Urol. 2009;55:164-174. 7. Nieuwenhuijzen JA, de Vries RR, Bex A, et al. Urinary diversions after cystectomy: the association of clinical factors, complications and functional results of four different diversions. Eur Urol. 2008;53: 834-842. 8. Konety BR, Allareddy V, Herr H. Complications after radical cystectomy: analysis of population-based data. Urology. 2006;68:58-64. 9. Konety BR, Allareddy V. Influence of post-cystectomy complications on cost and subsequent outcome. J Urol. 2007;177:280-287. 10. Chang SS, Cookson MS, Baumgartner RG, et al. Analysis of early complications after radical cystectomy: results of a collaborative care pathway. J Urol. 2002;167:2012-2016. 11. Deficit Reduction Act of 2005, S. 1932, 103d Congress, 1st Session (2006). 12. Centers for Medicare and Medicaid Services: Fiscal Year 2009 Inpatient Prospective Payment System (IPPS) final rule, in Department of Health and Human Services: Federal Register, Vol. 73, No. 161, 2009. 13. Rosenthal MB. Nonpayment for performance? Medicare’s new reimbursement rule. N Engl J Med. 2007;357:1573-1575. 14. Milstein A. Ending extra payment for “never events”—stronger incentives for patients’ safety. N Engl J Med. 2009;360:2388-2390. 15. Pronovost PJ, Goeschel CA, Wachter RM. The wisdom and justice of not paying for “preventable complications.” JAMA. 2008;299: 2197-2199. 16. 2010 Introduction to the NIS. Healthcare Cost and Utilization Project (HCUP). August 2012. Agency for Healthcare Research and Quality, Rockville, MD. Available at: http://www.hcup-us.ahrq.gov/ db/nation/nis/NIS_Introduction_2010.jsp. accessed October 2, 2011. 17. Creating national estimates. Nationwide Inpatient Sample, Agency for Healthcare Research and Quality, Rockville, MD. Available at: http://www.hcup-us.ahrq.gov/tech_assist/nationalestimates/508_ course/508course.htm. Accessed July 1, 2011. 18. Centers for Medicare and Medicaid Services. Hospital Acquired Conditions Factsheet, 2010. Available at: https://www.cms.gov/ HospitalAcqCond/downloads/HACFactsheet.pdf. Accessed June 1, 2011. 19. Conversion Table of New ICD-9-CM Codes. National Center for Health Statistics, Centers for Medicare and Medicaid Services. Available at: http://www.cdc.gov/nchs/data/icd9/CNVTB12.pdf. Accessed June 3, 2011. 20. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care. 1998;36:8-27.
531
21. Abdollah F, Sun M, Schmitges J, et al. Surgical caseload is an important determinant of continent urinary diversion rate at radical cystectomy: a population-based study. Ann Surg Oncol. 2011;18:2680-2687. 22. Hu JC, Gold KF, Pashos CL, et al. Temporal trends in radical prostatectomy complications from 1991 to 1998. J Urol. 2003;169: 1443-1448. 23. Begg CB, Riedel ER, Bach PB, et al. Variations in morbidity after radical prostatectomy. N Engl J Med. 2002;346:1138-1144. 24. Panageas KS, Schrag D, Riedel E, et al. The effect of clustering of outcomes on the association of procedure volume and surgical outcomes. Ann Intern Med. 2003;139:658-665. 25. Donat SM, Siegrist T, Cronin A, et al. Radical cystectomy in octogenarians—does morbidity outweigh the potential survival benefits? J Urol. 2010;183:2171-2177. 26. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and surgical mortality in the United States. N Engl J Med. 2002;346: 1128-1137. 27. Konety BR, Dhawan V, Allareddy V, et al. Impact of hospital and surgeon volume on in-hospital mortality from radical cystectomy: data from the Health Care Utilization Project. J Urol. 2005;173: 1695-1700. 28. Lawthers AG, McCarthy EP, Davis RB, et al. Identification of inhospital complications from claims data: is it valid? Med Care. 2000;38:785-795. 29. Burns EM, Rigby E, Mamidanna R, et al. Systematic review of discharge coding accuracy. J Pub Health. 2011;34:138-148.
EDITORIAL COMMENT The authors make a valuable contribution to the published data on radical cystectomy. By focusing on “never events,” they bring to urologists’ attention the effect of Medicare changes that will significantly affect the hospitals in which we work. The authors convincingly illustrate in Table 4 why these events are so important, with significantly increased odds of a LOS, increased
532
costs, and increased mortality among patients with these NEs. Fortunately, these events are rare, with 2.4% of cystectomy patients experiencing a NE during their hospitalization. A fundamental concern regarding these “never events” is that CMS will not pay for any care that occurs as a result of the event. This lack of payment could have a significant effect on the hospitals caring for complex cystectomy patients. The authors show that age, race, comorbidities, insurance status, hospital volume, and hospital location/teaching status are all related to having more of the “never events.” If the “never events” are not truly preventable, hospitals and physicians might avoid the treatment of patients with more complex conditions. Are these events always preventable? The finding that highvolume hospitals have significantly lower odds suggests these events can be prevented. The 2 most common events identified by the authors, pressure ulcers and vascular catheter-associated infections, have established processes of care to help prevent the development of the complications. Focusing resources on preventing such complications among the highest risk patients would benefit both patients and hospitals. The CMS will not change their reimbursement policy; the evidence of the adverse effects of the “never events,” as outlined in this report is too strong. Instead, physicians should work as partners with their hospitals to ensure the best care of their high-risk patients and ensure that the processes of care in their hospital are up to the highest standards. Seth A. Strope, M.D., M.P.H., Washington University School of Medicine Department of Surgery, Division of Urologic Surgery, St. Louis, Missouri http://dx.doi.org/10.1016/j.urology.2012.09.057 UROLOGY 81: 532, 2013. 2013 Elsevier Inc.
UROLOGY 81 (3), 2013
RESULTS
CONCLUSION
To describe “never event” (NE) complications after radical cystectomy. The Centers for Medicare and Medicaid Services has denied reimbursement for 10 hospitalacquired conditions that were deemed “reasonably preventable.” These NEs do not take into consideration pre-existing risk factors that make patients more susceptible. Radical cystectomy is a complex surgery that is often necessary in a population with extensive comorbidities. In this setting, the application of an unmodified system to deny reimbursement of adverse outcomes could have a significant effect on healthcare delivery. We hypothesized that measurable patient and hospital characteristics could predict the occurrence of NE complications after radical cystectomy. Using the weighted Nationwide Inpatient Sample database, we identified 61,142 patients with bladder cancer who underwent radical cystectomy from 2002 to 2009. The NE rates were calculated, and their effect on in-hospital mortality, length of stay, and total hospital costs were determined by multivariate regression analysis. The rate for any NE was 2.42%. Vascular-catheter infections (1.25%) were the most common. Black race and comorbidities increased the likelihood of a NE. The presence of any NE increased the average length of stay (by 15 days), total costs (by $37,000), and in-hospital mortality (8.0% vs 2.2%). Centers for Medicare and Medicaid Services NEs are more likely in older patients with comorbidities. NEs strongly predict for negative patient outcomes during hospitalization (length of stay, costs, in-hospital mortality). High-risk populations could benefit from risk adjustment before implementing a significant alteration in hospital or physician reimbursement policies. UROLOGY 81: 527e532, 2013. 2013 Elsevier Inc.
B
ladder cancer is primarily a disease of the elderly, with >70% of patients diagnosed after age 65.1 In addition, nearly 70% of patients with genitourinary malignancies have 1 comorbidity, more than patients with other cancers, placing these patients at high risk of complications.2 Radical cystectomy with lymph node dissection remains the standard surgical treatment of muscle-invasive bladder cancer.3 After radical cystectomy, 5-year recurrence-free survival ranges from 66% to 73% and overall survival from 62% to 66%.4,5 Financial Disclosure: The authors declare that they have no relevant financial interests. Funding Support: This work was supported by a grant from the Doris Duke Charitable Foundation to Columbia University in support of Clinical Research Fellow M. Kates. From the Department of Urology, Columbia University College of Physicians and Surgeons, New York, NY; Department of Epidemiology, Columbia University Mailman School of Public Health, New York, NY; and New York Presbyterian Hospital, New York, NY Reprint requests: James M. McKiernan, M.D., Department of Urology, Columbia University College of Physicians and Surgeons, 161 Fort Washington Avenue, New York, NY 10032. E-mail: [email protected] Submitted: May 17, 2012, accepted (with revisions): September 20, 2012
ª 2013 Elsevier Inc. All Rights Reserved
However, radical cystectomy is a long, complex, and morbid procedure. The 90-day complication rates after radical cystectomy range from 28% to 64%, with inhospital mortality rates of 2%-3%.5-10 Advanced age and comorbidities are associated with an increased rate of complications after radical cystectomy.8 These investigators, and others, have also shown that the occurrence of 1 complications significantly increased the chances of in-hospital mortality (5.36% vs 1.47%), increased the length of stay (LOS) (13 vs 9 days), and increased the total costs ($53,342 vs $38,339).9 However, no study has specifically considered the predictive ability of these characteristics for “preventable” complications in the radical cystectomy population. As a part of the Deficit Reduction Act of 2005, the Secretary of Health and Human Services was instructed by Congress to identify hospital-acquired conditions that were (1) high cost or high volume, or both; (2) resulted in assignment to a higher paying diagnosis-related group when present as a secondary diagnosis; and (3) were reasonably preventable through the application of 0090-4295/12/$36.00 http://dx.doi.org/10.1016/j.urology.2012.09.050
527
evidence-based medicine.11 Conditions were considered reasonably preventable if guidelines were available from professional organizations or academic institutions, with evidence quantifying the extent to which applying these guidelines would result in the prevention of the conditions.12 With this charge, the Centers for Medicare and Medicaid Services (CMS) identified 10 hospital-acquired conditions, also known as “never events” (NEs) (Table 1). As of October 1, 2009, hospitals were no longer reimbursed for the additional care delivered if a patient developed any of these 10 NEs. The policy was intended to motivate hospitals to accelerate programs in patient safety but was met with much resistance.13 Major American healthcare system leaders have suggested that not all NEs are preventable.14 Furthermore, the CMS did not include any exclusions or adjustments for patient demographics or procedures that place patients at greater risk of developing an NE. Pronovost et al15 suggested that this policy might create disincentives for hospitals to treat sicker patients. With the widespread focus on quality of care and the introduction of financial consequences to poor quality, we examined the factors that might predict the development of NEs after radical cystectomy for bladder cancer. If such factors can readily be identified, they could be used to stratify hospitals and procedures according to their risk of patients developing NEs. This risk stratification could be incorporated into current and future CMS reimbursement policies. We hypothesized that patients with more comorbid conditions would experience more NEs, leading to increased in-hospital mortality, LOS, and hospital costs. We also anticipated that higher volume centers would experience fewer NEs, leading to improved outcomes.
MATERIAL AND METHODS Data Source and Patient Selection The Nationwide Inpatient Sample (NIS) is the largest all-payer inpatient database in the United States, consisting of a 20% stratified sample of community and academic hospitals. The NIS contains data elements for each discharge that include the primary and secondary diagnoses and procedures, patient demographics, and hospital characteristics. The recent update of the NIS through 2009 included approximately 1000 hospitals from 44 states.16 The NIS also contains a weighting method based on the hospital stratification to allow for creation of national estimates.17 We queried the 2002-2009 NIS for any adult aged >40 years with a diagnosis of bladder cancer (Clinical Classifications Software diagnosis code 32) who had undergone radical cystectomy (“International Classification of Disease, 9th revision, Clinical Modification” [ICD-9-CM] procedure codes 57.71 and 57.79). We identified CMS NEs by searching each of the diagnosis variables for the relevant ICD-9-CM diagnosis codes according to the most recent documentation provided by the CMS (Table 1).18 Deep vein thrombosis, pulmonary embolism, and surgical site infection are not considered a NE after radical cystectomy, only after certain orthopedic, cardiac, and gastric bypass procedures. Because many of the codes changed between 2002 and 2009, we used the historical codes provided by the 528
Table 1. Centers for Medicare and Medicaid Services hospital-acquired conditions Hospital-acquired Condition Foreign object retained after surgery Air embolism Blood incompatibility
Pressure ulcers, stage III and IV Falls and trauma Fracture Dislocations Intracranial injury Crushing injury Burn Electric shock Catheter-associated UTI Vascular catheter-associated infection Manifestations of poor glycemic control Diabetic ketoacidosis Nonketotic hyperosmolar coma Hypoglycemic coma Secondary diabetes with ketoacidosis Secondary diabetes with hyperosmolarity Total
2010 ICD-9-CM Codes 998.4 998.7 999.1 999.60 999.61 999.62 999.63 999.69 707.23 707.24
Events (n) 32 (0.05) 5 (0.01) 0 (0.00)
545 (0.89) 62 (0.09)
800-829 830-839 850-854 925-929 940-949 991-994 996.64 999.31
42 (0.06) 765 (1.25) 40 (0.06)
250.10-250.13 250.20-250.23 251.0 249.10-249.11 249.20-249.21 1464 (2.42)
ICD-9-CM, “International Classification of Disease, 9th revision, Clinical Modification”; UTI, urinary tract infection. Data in parentheses are percentages.
Centers for Disease Control to identify the NEs from earlier years.19
Independent Variables The patient demographics included age, sex, race, payer, median household income, and number of comorbidities. We used the comorbidity software included with the NIS to create comorbidity measures, as designed by Elixhauser et al.20 We divided the median household income into 3 groups according to 2 postal code-level income variables provided in the NIS: low ($1-$24,999), medium ($25,000-$34,999), and high ($35,000). The hospital characteristics consisted of the number of hospital beds, teaching status, and cystectomy volume. We calculated the cystectomy volume by averaging the number of cystectomies performed annually for each hospital and then creating low-, medium-, and high-volume tertiles.
Outcomes The 3 main outcomes included in-hospital mortality, LOS, and total hospital costs. A LOS at or greater than the 90th percentile was considered “prolonged” and was found to be >20 days. Similarly, total costs at the 90th percentile were defined as “increased” and were found to be >$52,711. UROLOGY 81 (3), 2013
Statistical Analysis Baseline differences in patient demographics, hospital characteristics, and outcomes were compared using the chi-square test for categorical variables and Student’s t test for continuous variables. Factors predicting the development of an NE were assessed using univariate and multivariate logistic regression models. Using similar regression analyses, we also examined the effect that the presence of any CMS NE had on in-hospital mortality, elongated LOS, and increased total costs. All multivariate regression analyses were adjusted for clustering using the generalized estimating equations method.21-24 To assess the stepwise effect that each variable had on costs and LOS, linear regression models were also developed. All analyses were performed using SAS, version 9.2 (SAS Institute, Cary, NC) with a 2-sided significance level of P <.05.
RESULTS Overall, we identified 12,451 patients (nationally weighted sample of 61,142) who had undergone radical cystectomy for a diagnosis of bladder cancer in the NIS from 2002 to 2009. The frequency of the individual CMS NEs is listed in Table 1. The overall rate of an individual patient developing 1 NE was 2.42%. Of those developing an NE, most patients developed only 1 (2.35%); however, a few developed 2 NEs (0.05%). The most common NEs were vascular-catheter infection (1.25%) and pressure ulcer (0.89%). All other NEs occurred very rarely (<0.10%) and included falls and trauma (0.09%), catheter-associated urinary tract infection (0.06%), manifestations of poor glycemic control (0.06%), foreign object retained after surgery (0.05%), air embolism (0.01%), and blood incompatibility (0.00%). The patient demographics, hospital characteristics, and outcomes between patients who developed no NEs and those who developed 1 NE are listed in Table 2. Most of these patients were white (66%) and male (82%). A greater percentage of black patients developed 1 NE (5.3% vs 3.3%) than no NE. The mean age was 67.9 years for those who did not develop an NE and 69.2 years for those who did. Most patients had Medicare (61%) or private health insurance (31%). Among the patients who developed 1 NE, a greater percentage had Medicare (67% vs 60%). Most patients had 1 comorbidity (84%); however, those who developed 1 NE were more likely to have >3 comorbidities than were those without an NE (46% vs 31%). Most patients underwent cystectomy at an urban, teaching hospital (68%). The median LOS was 9 days, and the median costs were $23,609. Those patients who had 1 NE had greater in-hospital mortality (8.0% vs 2.2%, P <.01), longer mean LOS (26 vs 10 days, P <.01), and greater hospital costs ($66,792 vs $29,792, P <.01). The results of the multivariate analyses predicting the development of any NE adjusted for all covariates are listed in Table 3. We identified several predictors of experiencing 1 NE, including black race (odds ratio [OR] 1.59, 95% confidence interval [CI] 1.27-1.99 vs white rate, P <.01), increasing age (OR 1.02, 95% CI 1.01-1.02, P <.01, continuous variable), 1 comorbidity UROLOGY 81 (3), 2013
Table 2. Baseline characteristics of patients undergoing cystectomy (n ¼ 61,142) Independent Variable Total Age (y) Sex Male Female Race White Black Hispanic Asian or Pacific Islander Native American Other Missing Hospital size (by no. of beds) Small Medium Large Hospital location/ teaching status Rural Urban, nonteaching Urban, teaching Hospital surgical volume (per year) Small (5) Medium (5-17) Large (>17) Comorbidities (n) 0 1 2 >3 Payer Private Medicare Medicaid Self-pay No charge Other Median household income* Quartile 1 (1-34,999) Quartile 2 (35,000-44,999) Quartile 3 (45,000)
Patients With HAC No
Yes
59,677 (97.6) 1464 (2.4) 68.3 70.4 48,949 (82) 10,728 (18) 38,436 1970 1484 669
(64.4) (3.3) (2.5) (1.1)
206 (0.3) 948 (1.6) 15,964 (26.7)
1,208 (82.5) 256 (17.5) 968 78 55 5
P Value <.01 .66 <.01
(66.1) (5.3) (3.7) (0.3)
5 (0.3) 20 (1.4) 333 (22.7) .08
5,294 (8.9) 155 (10.6) 9,985 (16.7) 241 (16.5) 44,398 (74.4) 1,068 (72.9) 3,150 (5.3) 14,923 (25)
<.01
58 (4) 320 (21.8)
41,604 (69.7) 1,086 (74.2) 20,677 (34.7) 19,080 (31.9) 19,920 (33.4)
545 (37.2) 492 (33.6) 427 (29.2)
9,611 15,647 15,427 18,992
(16.1) (26.2) (25.9) (31.8)
123 342 324 675
18,416 36,460 2,406 1,157 158 1,080
(30.9) 280 (19.1) (61.1) 1,071 (73.1) (4) 64 (4.4) (1.9) 36 (2.5) (0.3) 0 (0) (1.8) 13 (0.9)
(8.4) (23.4) (22.1) (46.1)
<.01
<.01
<.01
.04 12,055 (20.2)
265 (18.1)
16,104 (27)
430 (29.3)
31,518 (52.8)
769 (52.6)
HAC, hospital-acquired condition. Data in parentheses are percentages. P value for chi-square of each variable. * Determined by postal code.
(OR 1.32, 95% CI 1.25-1.39 vs none, P <.01), Medicare and Medicaid insurance (OR 1.92, 95% CI 1.65-2.24, and OR 1.77, 95% CI 1.37-2.25, P <.01, respectively, vs private). Lower odds of developing a NE were associated with a high hospital cystectomy volume (OR 0.75, 95% CI 0.65-0.87 vs low volume, P <.01). The presence of 1 NE was strongly associated with a prolonged LOS (OR 529
Table 3. Predicting never events CMS Event (Multivariate Analysis) Independent Variable Sex Male Female Race White Black Hispanic Asian or Pacific Islander Native American Other Missing Age (continuous) Hospital size (by no. of beds) Small Medium Large Hospital location/ teaching status Rural Urban, nonteaching Urban, teaching Hospital surgical volume (per year) Small (5) Medium (5-17) Large (>17) Comorbidities (every 1 additional comorbidity) Payer Private Medicare Medicaid Self-pay No charge Other Median household income (by postal code) Quartile 1 (1-34,999) Quartile 2 (35,000-44,999) Quartile 3 (45,000)
OR (95% CI)
P Value .07
Reference 0.88 (0.77-1.01) Reference 1.59 (1.27-1.99) 1.46 (1.12-1.89) 0.69 (0.41-1.17)
<.01 <.01 .17
0.90 (0.37-2.21) 0.91 (0.61-1.36) 0.83 (0.73-0.94) 1.02
.83 .65 <.01 <.01
Reference 0.89 (0.73-1.09) 0.90 (0.76-1.07)
.28 .25
Reference 1.26 (0.94-1.67) 1.57 (1.19-2.07)
.1 <.01
Reference 1.00 (0.88-1.13) 0.75 (0.65-0.87) 1.32 (1.25-1.39)
.97 <.01 <.01
Reference 1.92 (1.65-2.24) 1.77 (1.37-2.25) 1.96 (1.38-2.80) NA 0.74 (0.42-1.29)
<.01 <.01 <.01 NA .29
Reference 1.18 (1.02-1.37) 1.14 (0.99-1.30)
.02 .05
CI, confidence interval; CMS, Centers for Medicare and Medicaid Services; NA, not applicable; OR, odds ratio.
8.76, 95% CI 7.89-9.76, P <.01), increased total costs (OR 7.49, 95% CI 6.65-8.26), and in-hospital mortality (OR 3.48, 95% CI 2.91-4.18; Table 4). The linear relationship of each independent variable to the cost and LOS included a base cost of $18,835. An NE added $31,332 to the cost of care and also added 12.48 extra days to the LOS. Each comorbidity added $3819 and 0.84 day to care for patients after radical cystectomy (data not shown).
COMMENT The present analysis showed that CMS NEs occurred in 2.42% of patients after radical cystectomy and have a significant effect on LOS and mortality. The least common NEs are blood incompatibility (0.00%), air embolism (0.01%), foreign object retained after surgery 530
(0.05%), and manifestations of poor glycemic control (0.05%). All 4 of these events can be directly associated with provider error and are generally considered completely preventable. Alternatively, the most common NEs were vascular catheter-associated infection (1.25%) and pressure ulcers (0.89%). These findings suggest that the CMS might have correctly identified some NEs as true preventable events but might have incorrectly identified others as being truly preventable. Alternatively, improved structural processes might further reduce the frequency of these events. Several patient factors, including increasing age and comorbidities, predicted developing a CMS NE. In this patient population, increasing age is often associated with more complications.8,25 Increasing age, when described by decade, added 20% to the risk of developing a NE. The presence of 1 comorbidity increased the likelihood of developing an NE by 28%. In a surgical population in which 84% have 1 comorbid condition, this is very clinically relevant. Because both age and patient comorbidities are unmodifiable risk factors, the current CMS policy suggests a financial incentive for hospitals to avoid treating patients with bladder cancer who are also elderly and ill. This has particular significance for tertiary and quaternary academic referral hospitals, because they care for many of these patients, and the policy does not explicitly stipulate for any additional resources to be provided to these centers for the prevention of NEs. Thus, the effect of this policy could result in lower reimbursements to these centers simply because they treat patients with more challenging high-risk conditions. Among the hospital characteristics, hospitals with high cystectomy volumes were less likely to have NEs after radical cystectomy. This adds to the evidence demonstrating an inverse relationship between hospital cystectomy volume and negative patient outcomes.26,27 Despite controlling for the clustering of hospital characteristics, including volume, location, and teaching status, patients treated at urban-teaching hospitals were still more likely to develop NEs. Unquantifiable hospital-specific processes of care might account for the complex interaction between hospital case volume and teaching status. The present study has also showed that the development of any NE significantly increased the LOS, costs, and in-hospital mortality. This suggests that the CMS has correctly identified conditions that cause an increased cost burden on the healthcare system. Patient care programs designed to prevent NEs must demonstrate efficacy in terms of event rate reduction and subsequent cost and LOS reduction. The development of an NE affected the LOS more (OR 8.8) than it increased total costs (OR 7.5) or in-hospital mortality (OR 3.5). The development of a NE would naturally lead to a longer LOS and increased costs related to the treatment of the NE. These findings suggest that preventing NEs might have the greatest effect on decreasing the patient LOS with a corresponding decrease in the financial burden of hospitalization. UROLOGY 81 (3), 2013
Table 4. Multivariate regression analysis for outcomes Independent Variable Elongated LOS Any CMS HAC None 1 Increased total costs Any CMS HAC None 1 HAC In-hospital mortality Any CMS HAC None 1 HAC
Patients (n)
Multivariate OR (95% CI)
4,827 (8.1) 675 (46.1)
Reference 8.8 (7.9-9.8)
5,513 (9.24) 663 (45.3)
Reference 7.5 (6.7-8.3)
1,362 (2.28) 117 (8.03)
Reference 3.5 (2.9-4.2)
LOS, length of stay. Data in parentheses are percentages, unless otherwise noted.
The limitations of the present study related to the use of an administrative database. The present analysis relied on coding and reporting of conditions for identification of CMS NEs. ICD-9-CM coding of a large administrative hospital data source has been demonstrated to be accurate approximately 80% of the time; thus, some complications or index subjects might have been missed if miscoded.28,29 Also, the CMS has made recent changes to the ICD-9-CM coding to help identify NEs that were not available in previous years. For example, the CMS NE of a pressure ulcer is meant to only capture the most severe ulcers, grades III and IV. However, this specific ICD-9-CM code did not exist before 2009. Before this addition, the codes were determined by the location, but not the severity, of the ulcers. For the present study, we captured all pressure ulcers within the study period, which likely overestimated the true occurrence of stage III and IV pressure ulcers. Complete adherence to the clinical guidelines for preventing NEs, such as operating room count systems or check lists for vascular catheter maintenance, might significantly reduce their prevalence in this population. However, these data were also not available from the NIS. The NIS database also does not capture any events that developed after the patient was discharged. If a patient developed a complication at home, even if readmitted to the same hospital, the NIS would not capture it. The final limitation was that the NIS database does not capture specific information on the severity of bladder cancer. Thus, cancer grade and stage could not be included in the model.
CONCLUSION CMS NEs after radical cystectomy are rare events. Vascular catheter infection and pressure ulcers were most common and were the only events with rates >0.1%. Several patient- and hospital-level factors can predict the development of these events, including advanced age, increased comorbidities, and hospital radical cystectomy volume. Once a patient has developed a NE, the patient is more likely to have a prolonged LOS, accrue increased total costs, and die in the hospital. The entire system, UROLOGY 81 (3), 2013
including physicians, physician extenders, nurses, pharmacists, and technicians, treats the diseases of hospitalized patients. There are likely many steps in care that can be improved to limit these complications. Ultimately, these findings should be used to inform future healthcare policy regarding “preventable” complications after major oncologic surgery. References 1. SEER. SEER fact sheet, 2008, vol. 2011. Available at: http://seer. cancer.gov/statfacts/html/urinb.html. Accessed October 10, 2011. 2. Piccirillo JF, Tierney RM, Costas I, et al. Prognostic importance of comorbidity in a hospital-based cancer registry. JAMA. 2004;291: 2441-2447. 3. National Comprehensive Cancer Network. 2011 Guidelines for Bladder Cancer. Available at: http://www.nccn.org/professionals/ physician_gls/f_guidelines.asp#site. Accessed October 2, 2011. 4. Madersbacher S, Hochreiter W, Burkhard F, et al. Radical cystectomy for bladder cancer today—a homogeneous series without neoadjuvant therapy. J Clin Oncol. 2003;21:690-696. 5. Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol. 2001;19:666-675. 6. Shabsigh A, Korets R, Vora KC, et al. Defining early morbidity of radical cystectomy for patients with bladder cancer using a standardized reporting methodology. Eur Urol. 2009;55:164-174. 7. Nieuwenhuijzen JA, de Vries RR, Bex A, et al. Urinary diversions after cystectomy: the association of clinical factors, complications and functional results of four different diversions. Eur Urol. 2008;53: 834-842. 8. Konety BR, Allareddy V, Herr H. Complications after radical cystectomy: analysis of population-based data. Urology. 2006;68:58-64. 9. Konety BR, Allareddy V. Influence of post-cystectomy complications on cost and subsequent outcome. J Urol. 2007;177:280-287. 10. Chang SS, Cookson MS, Baumgartner RG, et al. Analysis of early complications after radical cystectomy: results of a collaborative care pathway. J Urol. 2002;167:2012-2016. 11. Deficit Reduction Act of 2005, S. 1932, 103d Congress, 1st Session (2006). 12. Centers for Medicare and Medicaid Services: Fiscal Year 2009 Inpatient Prospective Payment System (IPPS) final rule, in Department of Health and Human Services: Federal Register, Vol. 73, No. 161, 2009. 13. Rosenthal MB. Nonpayment for performance? Medicare’s new reimbursement rule. N Engl J Med. 2007;357:1573-1575. 14. Milstein A. Ending extra payment for “never events”—stronger incentives for patients’ safety. N Engl J Med. 2009;360:2388-2390. 15. Pronovost PJ, Goeschel CA, Wachter RM. The wisdom and justice of not paying for “preventable complications.” JAMA. 2008;299: 2197-2199. 16. 2010 Introduction to the NIS. Healthcare Cost and Utilization Project (HCUP). August 2012. Agency for Healthcare Research and Quality, Rockville, MD. Available at: http://www.hcup-us.ahrq.gov/ db/nation/nis/NIS_Introduction_2010.jsp. accessed October 2, 2011. 17. Creating national estimates. Nationwide Inpatient Sample, Agency for Healthcare Research and Quality, Rockville, MD. Available at: http://www.hcup-us.ahrq.gov/tech_assist/nationalestimates/508_ course/508course.htm. Accessed July 1, 2011. 18. Centers for Medicare and Medicaid Services. Hospital Acquired Conditions Factsheet, 2010. Available at: https://www.cms.gov/ HospitalAcqCond/downloads/HACFactsheet.pdf. Accessed June 1, 2011. 19. Conversion Table of New ICD-9-CM Codes. National Center for Health Statistics, Centers for Medicare and Medicaid Services. Available at: http://www.cdc.gov/nchs/data/icd9/CNVTB12.pdf. Accessed June 3, 2011. 20. Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care. 1998;36:8-27.
531
21. Abdollah F, Sun M, Schmitges J, et al. Surgical caseload is an important determinant of continent urinary diversion rate at radical cystectomy: a population-based study. Ann Surg Oncol. 2011;18:2680-2687. 22. Hu JC, Gold KF, Pashos CL, et al. Temporal trends in radical prostatectomy complications from 1991 to 1998. J Urol. 2003;169: 1443-1448. 23. Begg CB, Riedel ER, Bach PB, et al. Variations in morbidity after radical prostatectomy. N Engl J Med. 2002;346:1138-1144. 24. Panageas KS, Schrag D, Riedel E, et al. The effect of clustering of outcomes on the association of procedure volume and surgical outcomes. Ann Intern Med. 2003;139:658-665. 25. Donat SM, Siegrist T, Cronin A, et al. Radical cystectomy in octogenarians—does morbidity outweigh the potential survival benefits? J Urol. 2010;183:2171-2177. 26. Birkmeyer JD, Siewers AE, Finlayson EV, et al. Hospital volume and surgical mortality in the United States. N Engl J Med. 2002;346: 1128-1137. 27. Konety BR, Dhawan V, Allareddy V, et al. Impact of hospital and surgeon volume on in-hospital mortality from radical cystectomy: data from the Health Care Utilization Project. J Urol. 2005;173: 1695-1700. 28. Lawthers AG, McCarthy EP, Davis RB, et al. Identification of inhospital complications from claims data: is it valid? Med Care. 2000;38:785-795. 29. Burns EM, Rigby E, Mamidanna R, et al. Systematic review of discharge coding accuracy. J Pub Health. 2011;34:138-148.
EDITORIAL COMMENT The authors make a valuable contribution to the published data on radical cystectomy. By focusing on “never events,” they bring to urologists’ attention the effect of Medicare changes that will significantly affect the hospitals in which we work. The authors convincingly illustrate in Table 4 why these events are so important, with significantly increased odds of a LOS, increased
532
costs, and increased mortality among patients with these NEs. Fortunately, these events are rare, with 2.4% of cystectomy patients experiencing a NE during their hospitalization. A fundamental concern regarding these “never events” is that CMS will not pay for any care that occurs as a result of the event. This lack of payment could have a significant effect on the hospitals caring for complex cystectomy patients. The authors show that age, race, comorbidities, insurance status, hospital volume, and hospital location/teaching status are all related to having more of the “never events.” If the “never events” are not truly preventable, hospitals and physicians might avoid the treatment of patients with more complex conditions. Are these events always preventable? The finding that highvolume hospitals have significantly lower odds suggests these events can be prevented. The 2 most common events identified by the authors, pressure ulcers and vascular catheter-associated infections, have established processes of care to help prevent the development of the complications. Focusing resources on preventing such complications among the highest risk patients would benefit both patients and hospitals. The CMS will not change their reimbursement policy; the evidence of the adverse effects of the “never events,” as outlined in this report is too strong. Instead, physicians should work as partners with their hospitals to ensure the best care of their high-risk patients and ensure that the processes of care in their hospital are up to the highest standards. Seth A. Strope, M.D., M.P.H., Washington University School of Medicine Department of Surgery, Division of Urologic Surgery, St. Louis, Missouri http://dx.doi.org/10.1016/j.urology.2012.09.057 UROLOGY 81: 532, 2013. 2013 Elsevier Inc.
UROLOGY 81 (3), 2013