- Email: [email protected]
Trends in Regionalization of Adrenalectomy to Higher Volume Surgical Centers
Adult Urology Oncology: Adrenal/Renal/Upper Tract/Bladder Trends in Regionalization of Adrenalectomy to Higher Volume Surgical Centers Jay Simhan, Marc C. Smaldone, Daniel J. Canter, Fang Zhu, Russell Starkey, Karyn B. Stitzenberg, Robert G. Uzzo and Alexander Kutikov* From the Division of Urologic Oncology, Departments of Surgical Oncology (JS, MCS, DJC, RS, RGU, AK) and Biostatistics (FZ), Fox Chase Cancer Center, Temple University School of Medicine, Philadelphia, Pennsylvania, and Division of Surgical Oncology, Department of Surgery, University of North Carolina at Chapel Hill (KBS), Chapel Hill, North Carolina
Purpose: Although centralization of surgical procedures to high volume centers has been described previously, patterns of care for adrenal surgery are largely unknown. We determined the extent of regionalization of care for adrenal surgery and the extent to which this centralization has evolved with time. Materials and Methods: Using 1996 to 2009 hospital discharge data from New York, New Jersey and Pennsylvania we identified all patients 18 years old or older treated with adrenalectomy. Hospital volume quintiles were created using 1996 hospital volumes. These cutoffs were then applied to subsequent years. Outcome variables were examined by hospital volume status with time using logistic regression models. Results: A total of 8,381 patients underwent adrenalectomy from 1996 to 2009 with a significant 17% to 42% shift toward regionalization to very high volume hospitals, defined as 15 or greater procedures per year (p ⬍0.001). For each successive year the odds of having surgery performed at a very low volume hospital decreased by 13% (OR 0.87, 95% CI 0.84 – 0.89). There were significant differences in patient age, race and payer group for very low volume hospitals, defined as less than 1 procedure per year, compared to very high volume hospitals (p ⬍0.0001). Patients at very high volume hospitals were less likely to be 55 years old or older (OR 0.73, 95% CI 0.61– 0.88), insured through Medicaid (OR 0.60, 95% CI 0.45– 0.79) or uninsured (OR 0.34, 95% CI 0.17– 0.70). When controlling for year treated, patients were less likely to die in the hospital if treated at a very high volume hospital (OR 0.38, 95% CI 0.19 – 0.75). Conclusions: These data reveal the increasing centralization of adrenalectomy to very high volume hospitals since 1996 with improved clinical outcomes. Inequities in access to care to higher volume centers appear to exist and require further investigation.
Abbreviations and Acronyms HLOS ⫽ hospital length of stay HVH ⫽ high volume hospital LVH ⫽ low volume hospital MVH ⫽ medium volume hospital VHVH ⫽ very HVH VLVH ⫽ very LVH Submitted for publication November 20, 2011. Study received Fox Chase Cancer Center institutional review board approval. Supplementary material for this article can be obtained at http://www.fccc.edu/docs/health Professionals/Simhan-Supplementary-Table-1.pdf. * Correspondence and requests for reprints: Urologic Oncology, Fox Chase Cancer Center, Temple University School of Medicine, 333 Cottman Ave., Philadelphia, Pennsylvania 19111 (telephone: 215-728-5342; FAX: 215-214-1734; e-mail: [email protected]).
Key Words: adrenal glands, adrenalectomy, hospitals, delivery of health care, outcome assessment (health care) FIRST reported in 1979,1 the relationship between surgical volume and clinical outcome has been studied extensively. More recently this relationship has been examined in regard to complex cancer procedures such as colorectal surgery, pancreatectomy and
esophagogastrectomy.2 These data reveal a consistent trend that care received at HVHs is associated with decreased HLOS, a lower perioperative complication rate and decreased overall mortality compared to care received at LVHs.3 As a result, experts
0022-5347/12/1882-0377/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
http://dx.doi.org/10.1016/j.juro.2012.03.130 Vol. 188, 377-383, August 2012 RESEARCH, INC. Printed in U.S.A
AND
www.jurology.com
377
378
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
advocate the regionalization of complex surgical procedures as an integral part of continual efforts to improve quality of care. Although similar benefits in patient outcome were described for radical cystectomy, regionalization studies of other urological procedures are limited.4,5 Only a small portion of incidentally discovered adrenal lesions warrants surgical management but the quality of health care delivery to patients with surgical adrenal disease affects a large group of generally healthy individuals.6 The diagnosis, assessment and surgical management of benign or malignant diseases of the adrenal gland require highly specialized training. Currently most surgical management for adrenal disease centers around resection of adrenal masses due to concern for malignancy or benign neoplasms that are metabolically active.7 With the incidence of incidentally discovered adrenal lesions increasing in conjunction with the increased use of routine cross-sectional imaging, the number of adrenalectomies done in the United States has increased.8 –12 Largely based on established institutional and geographic referral patterns, surgical management of adrenal pathology has been distributed among urologists, general surgeons, surgical oncologists and more recently endocrine surgeons.13 As such, to date patterns of care for adrenal surgery are poorly defined. With increased specialization for managing adrenal disorders we hypothesized that centralization of care for adrenal disorders to high volume surgical centers occurred in the last 15 years. Through an analysis of hospital discharge claims data from 3 large northeastern states we evaluated the extent to which regionalization of care currently exists for adrenal surgery and the degree to which this centralization has evolved with time.
MATERIALS AND METHODS This study was approved by the institutional review board of the Fox Chase Cancer Center. Data analysis was done using hospital discharge data from New York, Pennsylvania and New Jersey from 1996 to 2009. From 1996 to 1999 data were only available for New York and Pennsylvania. Starting in 2000 procedures from all 3 states were included. Using ICD-9 coding we identified all patients 18 years old or older treated with partial (07.2, 07.21, 07.29), unilateral (07.22) or bilateral (07.3) adrenalectomy. Since CPT coding is not included in inpatient discharge data, information on the surgical approach was not available, ie open vs laparoscopic or robotic. Obtained data included patient age, race (white, black, Asian, Native American, other or unknown), gender (male or female), clinical diagnoses, procedure codes, disposition on hospital discharge (home discharge, hospital facility transfer, discharge to rehabilitation/skilled nursing facility, death during hospitalization or unknown) and payer group (private insur-
ance, Medicaid, Medicare, uninsured or unknown). Mortality was defined as any inpatient death after an adrenal procedure during the same hospitalization. Using the methods described by Birkmeyer et al,14 hospitals where adrenalectomy was done in 1996 were ranked in order of increasing 1996 hospital adrenalectomy volume. Five volume groups were then defined by selecting whole number cutoffs for annual volume that most closely sorted the number of procedures into 5 groups of equal size (quintiles). These volume group quintiles were designated VLVH— 0 or 1, LVH—2 or 3, MVH— 4 to 6, HVH—7 to 14 and VHVH—15 or more adrenalectomies annually. The 1996 cutoffs were applied to each subsequent year to determine centralization trends in the distribution of patients who underwent adrenal surgery among the volume categories. Multiple logistic regressions provided estimates of the odds of surgery at a hospital in the lowest or the highest quintile with time with year as a continuous variable and controlling for the overall number of cases done in the study area annually. Independently of treatment year we further investigated the association between sociodemographic factors and volume status. Finally, using hospital volume by quintile, treatment year and total number of procedures in the study area annually as covariates, we investigated the association between hospital volume, and HLOS and inpatient mortality. For each regression analysis a robust cluster variance estimator was used to estimate the SE to account for the intragroup correlation of observations in the same stage. All hypothesis tests were 2 sided with statistical significance considered at p ⬍0.05.
RESULTS A total of 8,381 patients treated with adrenal surgery met study inclusion criteria. While 2,142 (25.5%) and 2,167 patients (25.9%) were between 18 to 44 and 45 to 54 years old, respectively, the largest proportion (48.6%) was 55 years old or older. Of the patients 6,088 (72.6%) were white and 1,050 (12.5%) were designated black. Of the patients 5,261 (62.8%) had private insurance while 2,103 (25.1%) had Medicare. Although 7,992 patients (95.4%) were discharged home after adrenal surgery, 63 (0.8%) died during the hospital stay, including 43 (68.3%) of malignancy involving the adrenal gland as the primary diagnosis. Other primary diagnoses were hyperactive adrenal nodules in 15 patients (23.8%), primary adrenal malignancy in 27 (42.9%), adrenal metastasis in 16 (25.4%) and other in 5 (7.9%). While gender was similar among the groups, age (p ⫽ 0.0005), race (p ⬍0.0001) and payer group (p ⬍0.0001) varied significantly by volume status category. Although the number of hospitals where adrenalectomy was done remained stable during the study period (p ⫽ 0.11), the mean number of adrenal procedures performed per hospital with time increased (mean 2.9 in 1996 vs 5.5 in 2009, p ⬍0.0001, fig. 1). The proportion of hospitals that met VLVH (39.5% vs 28.3%, p ⫽ 0.09) and VHVH (2.6% vs 8.6%,
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
379
Figure 1. Mean number of adrenalectomies performed per hospital by year during study period, stratified by volume status group. From 1996 to 1999 only New York and Pennsylvania data were available.
p ⫽ 0.09) criteria did not significantly change between 1996 and 2009 but a shift was noted from VLVH to VHVH in the proportion of patients who underwent adrenal surgery during a similar period (fig. 2). Of adrenal procedures 42%were performed at VHVHs in 2009 compared to 17% in 1996 (p ⬍0.001).
In 2009 compared to 1996 significantly fewer patients underwent treatment at VLVHs and LVHs (each p ⬍0.0001) while no significant trends were noted for MVHs or HVHs (fig. 3). Multivariate regression analysis was done to determine associations between patient characteris-
Figure 2. Proportion of procedures performed yearly at VHVHs (purple curve) and VLVHs (blue curve) during study period
380
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
Figure 3. Proportion of patients treated with adrenal surgery by hospital volume category based on defined cutoff quintiles determined from 1996 data. Blue bars indicate VLVHs. Red bars indicate LVHs. Yellow bars indicate MVHs. Light green bars indicate HVHs. Purple bars indicate VHVHs. Asterisk indicates significant change in proportion of patients with adrenal surgery with time (p ⬍0.0001).
tics and treatment hospital volume status. All factors were studied individually while controlling for all other assessed characteristics. Excluding patients with an unknown race designation while controlling for year of surgery, patient age, race, gender and payer group, the odds of undergoing surgery at a VLVH decreased by 13% with each subsequent year from 1996 to 2009 (OR 0.87, 95% CI 0.84 – 0.89,
p ⬍0.001, table 1). During the study period patients treated at a VHVH were more likely to be male (OR 1.09, 95% CI 1.09 –1.10, p ⬍0.001) and less likely to be older than 55 years (OR 0.73, 95% CI 0.61– 0.88, p ⫽ 0.001), have Medicare (OR 0.86, 95% CI 0.79 – 0.93, p ⬍0.001) or Medicaid (OR 0.60, 95% CI 0.45– 0.79, p ⫽ 0.004), or be uninsured (OR 0.34, 95% CI 0.17– 0.70, p ⫽ 0.003). In contrast, patients seeking treatment at VLVHs tended to be older than 55 years (OR 1.43, 95% CI 1.10 –1.84, p ⫽ 0.007), or have Medicaid (OR 1.75, 95% CI 1.06 –2.88, p ⫽ 0.03) or no insurance (OR 2.17, 95% CI 1.47– 3.20, p ⬍0.001). Associations of volume status with inpatient mortality and HLOS were then examined by volume quintile using VLVHs as the referent. For the entire cohort independent of year treated and total procedures per year the risk of inpatient mortality decreased as surgical volume increased for each volume quintile compared to VLVHs (table 2). Patients were significantly less likely to die when treated at VHVHs than at VLVHs (OR 0.38, 95% CI 0.19 – 0.75, p ⫽ 0.006). Likely due to the confounding competing risk profiles of patients treated at the highest volume tertiary care centers, those presenting to HVHs were at lowest risk for mortality compared to those at VLVHs (OR 0.26, 95% CI 0.11– 0.59, p ⫽ 0.001). Similar trends were noted for median HLOS since a shorter HLOS was associated with each increasing volume quintile. When controlling for year treated, for the entire cohort a median difference of –1.75 days (95% CI –1.80 –⫺1.69, p ⬍0.001) of hospital stay was noted in the VHVH cohort compared to the VLVH group.
Table 1. Multivariate regression analysis for predictors of surgery at VLVHs and VHVHs, excluding patients with unknown race designation VLVH
Yr Annual procedure vol Age 55 or older Race: White Black Asian Native American Other Gender: F M Payer group: Private Medicare Medicaid Uninsured Unknown
VHVH
OR (95% CI)
p Value
OR (95% CI)
p Value
0.87 (0.84–0.89) 1.00 (1.00–1.00) 1.43 (1.10–1.84)
⬍0.001 0.29 0.007
1.02 (0.90–1.17) 1.00 (1.00–1.00) 0.73 (0.61–0.88)
0.71 0.32 0.001
1.0 1.19 (0.97–1.47) 0.62 (0.50–0.77) 0.55 (0.04–7.69) 0.75 (0.46–1.23)
— 0.10 ⬍0.001 0.65 0.26
1.0 0.69 (0.45–1.06) 1.44 (0.99–2.08) 3.13 (0.52–18.9) 0.90 (0.81–2.07)
— 0.09 0.052 0.21 0.81
1.0 0.89 (0.85–0.93)
— ⬍0.001
1.0 1.09 (1.09–1.10)
— ⬍0.001
1.0 1.03 (0.76–1.40) 1.75 (1.06–2.88) 2.17 (1.47–3.20) 1.06 (0.62–1.80)
— 0.85 0.03 ⬍0.001 0.83
1.0 0.86 (0.79–0.93) 0.60 (0.45–0.79) 0.34 (0.17–0.70) 0.43 (0.10–2.01)
— ⬍0.001 0.004 0.003 0.28
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
381
Table 2. Odds of HLOS difference and inpatient mortality controlled for year treated and annual number of procedures HLOS Difference
Mortality
Hospital Vol Group
OR (95% CI)
p Value
OR (95% CI)
p Value
VLVH (referent) LVH MVH HVH VHVH
0 ⫺0.86 (⫺0.93–0.80) ⫺0.94 (⫺1.01–0.88) ⫺1.68 (⫺1.74–1.61) ⫺1.75 (⫺1.81–1.69)
— ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
1 0.66 (0.43–0.99) 0.44 (0.26–0.74) 0.26 (0.11–0.59) 0.38 (0.19–0.75)
— 0.043 0.002 0.001 0.006
DISCUSSION Adrenal surgery rates have been steadily increasing in conjunction with the increased prevalence of incidental adrenal masses found on routine cross-sectional imaging.8,11 A recent retrospective cohort analysis of 3,144 adrenalectomies identified from the Nationwide Inpatient Sample revealed that surgeon volume and not specialty was an independent predictor of postoperative complications and HLOS.15 These findings mirror recent provider volume-outcome trends in patients undergoing major cancer surgery14,16,17 and they have important implications for quality of care. Despite these observations additional efforts to further define the provider volume-surgical outcomes relationship using population based cohort data have yielded conflicting results.18,19 To date centralization trends for adrenalectomy have been poorly characterized. The association between quality of care and the volume-outcome relationship has led many groups to advocate centralization of complex urological and cancer procedures to high volume centers. Using a state based discharge claims data set for patients treated with extirpative surgery for prostate, colorectal, esophageal or pancreatic cancer Stitzenberg et al reported a significant shift toward high volume centers for esophageal, pancreatic and colon cancer procedures.2,4 Using similar methods our findings demonstrate that in 3 northeastern states adrenal surgery became increasingly centralized to VHVHs during a 15-year period. In our sample the odds of undergoing adrenal surgery at a VLVH decreased by 13% per year and 42% of all adrenal surgeries were done at the highest volume centers in 2009. In our data set only crude outcomes measures were available, including inpatient mortality and HLOS. Results are limited by a lack of confirmatory comorbidity and pathological data. Despite these limitations distinct trends toward decreased length of stay and mortality with increasing volume status were observed in our sample, which should be confirmed using other data sets. Further analysis of social and demographic trends revealed several disparities indicating disparate access to care. In our entire cohort there were significant differences among volume status groups
in patient age, race, gender and payer group. On multivariate analysis controlling for all factors patients 55 years old or older, patients insured through Medicare or Medicaid and uninsured patients were significantly less likely be treated at a VHVH. In contrast, patients treated at a VLVH were more likely to be 55 years old or older and have Medicaid or no insurance while black race approached significance. More characterization of centralization trends and disparities in access to care are challenging but necessary in the health care climate today. With growing subspecialization in surgical disciplines today adrenal procedures are increasingly referred to urologists and general surgeons with fellowship training in laparoscopy and oncology, who are more likely to operate at tertiary care or high volume centers.20,21 Previous analyses showed repeat disparities in access to care at high volume centers for complex surgery, which may be influenced by socioeconomic status, geographic location, travel burden, cultural influences and personal preference.2,22,23 Ayanian et al suggested that the improved outcomes in patients with insurance treated at HVHs may be further confounded by the higher disease severity in patients without insurance treated at LVHs, which may predispose to poor outcomes.24 Furthermore, the regionalization of adrenal surgery to higher volume centers may contribute to the closing of adrenal surgery programs at smaller hospitals, which could potentially exacerbate access disparities in underserved areas. Clear limitations to the use of our data set include the lack of patient specific data, such as pathological stage, comorbid conditions, perioperative complication data and rehospitalization rates. As a result, our outcome measures do not adequately adjust for severity and must be interpreted with caution. Furthermore, New Jersey, New York and Pennsylvania represent a unique urban area with extensive health care resources. As a result, these findings may not be generalizable to less densely populated regions of the country. Nevertheless, despite these limitations all payer administrative data provide the opportunity to accurately assess regionalization trends with
382
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
time and provide some insight into the access to care challenges currently facing our health care system.
CONCLUSIONS We present a rigorous analysis characterizing the centralization of adrenal surgical care to VHVHs in
a 15-year period. Regionalization of care is associated with improved unadjusted short-term outcomes, which require further evaluation in more robust data sets. Also, the apparent inequities in access to care at high volume centers that may exist for elderly, uninsured and Medicaid patients should be the focus of future investigations.
REFERENCES 1. Luft HS, Bunker JP and Enthoven AC: Should operations be regionalized? The empirical relation between surgical volume and mortality. N Engl J Med 1979; 301: 1364. 2. Stitzenberg KB, Sigurdson ER, Egleston BL et al: Centralization of cancer surgery: implications for patient access to optimal care. J Clin Oncol 2009; 27: 4671. 3. Birkmeyer JD, Stukel TA, Siewers AE et al: Surgeon volume and operative mortality in the United States. N Engl J Med 2003; 349: 2117. 4. Stitzenberg KB, Wong YN, Nielsen ME et al: Trends in radical prostatectomy: centralization, robotics, and access to urologic cancer care. Cancer 2012; 118: 54. 5. Barbieri CE, Lee B, Cookson MS et al: Association of procedure volume with radical cystectomy outcomes in a nationwide database. J Urol 2007; 178: 1418. 6. Kutikov A, Crispen PL, Uzzo RG: Pathophysiology, evaluation, and medical management of adrenal disorders. In: Campbell-Walsh Urology. Edited by AJ Wein, LR Kavoussi, AW Partin et al. Philadelphia: Elsevier 2012; vol 2, chap 57, pp 1685– 1736. 7. Mazzaglia PJ and Vezeridis MP: Laparoscopic adrenalectomy: balancing the operative indications with the technical advances. J Surg Oncol 2010; 101: 739. 8. Gallagher SF, Wahi M, Haines KL et al: Trends in adrenalectomy rates, indications, and physician
volume: a statewide analysis of 1816 adrenalectomies. Surgery 2007; 142: 1011. 9. Saunders BD, Wainess RM, Dimick JB et al: Trends in utilization of adrenalectomy in the United States: have indications changed? World J Surg 2004; 28: 1169. 10. Young WF Jr: Clinical practice. The incidentally discovered adrenal mass. N Engl J Med 2007; 356: 601. 11. Miccoli P, Raffaelli M, Berti P et al: Adrenal surgery before and after the introduction of laparoscopic adrenalectomy. Br J Surg 2002; 89: 779. 12. Kutikov A, Mallin K, Canter D et al: Effects of increased cross-sectional imaging on the diagnosis and prognosis of adrenocortical carcinoma: analysis of the National Cancer Database. J Urol 2011; 186: 805.
17. Dimick JB, Pronovost PJ, Cowan JA Jr et al: Variation in postoperative complication rates after high-risk surgery in the United States. Surgery 2003; 134: 534. 18. Murphy MM, Witkowski ER, Ng SC et al: Trends in adrenalectomy: a recent national review. Surg Endosc 2010; 24: 2518. 19. Stavrakis AI, Ituarte PH, Ko CY et al: Surgeon volume as a predictor of outcomes in inpatient and outpatient endocrine surgery. Surgery 2007; 142: 887. 20. Duchene DA, Moinzadeh A, Gill IS et al: Survey of residency training in laparoscopic and robotic surgery. J Urol 2006; 176: 2158.
13. Sosa JA, Wang TS, Yeo HL et al: The maturation of a specialty: workforce projections for endocrine surgery. Surgery 2007; 142: 876.
21. Le D, Karmali S, Harness JK et al: An update: the operative experience in adrenal, pancreatic, and other less common endocrine diseases of U.S. general surgery residents. World J Surg 2008; 32: 232.
14. Birkmeyer JD, Siewers AE, Finlayson EV et al: Hospital volume and surgical mortality in the United States. N Engl J Med 2002; 346: 1128.
22. Liu JH, Zingmond DS, McGory ML et al: Disparities in the utilization of high-volume hospitals for complex surgery. JAMA 2006; 296: 1973.
15. Park HS, Roman SA and Sosa JA: Outcomes from 3144 adrenalectomies in the United States: which matters more, surgeon volume or specialty? Arch Surg 2009; 144: 1060.
23. Stitzenberg KB and Meropol NJ: Trends in centralization of cancer surgery. Ann Surg Oncol 2010; 17: 2824.
16. Begg CB, Cramer LD, Hoskins WJ et al: Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998; 280: 1747.
24. Ayanian JZ, Weissman JS, Schneider EC et al: Unmet health needs of uninsured adults in the United States. JAMA 2000; 284: 2061.
EDITORIAL COMMENT These authors evaluated trends in the regionalization of adrenalectomy in a claims based administrative data set comprising discharge information on patients in the northeastern United States. They noted regionalization of adrenalectomy to high volume centers during the study period. Patients of increasing age and those with uninsured status or who were insured through the Medicaid program were more likely to be treated at VLVHs. These study findings have significant implications, particularly in regard to the intersection of health care access and quality in impoverished individuals.
This study adds to the growing body of literature on the well documented volume-outcome relationship for major urological1 and nonurological2 surgical procedures. While current data suggest that regionalization to high volume centers improves surgical outcomes, one must consider the potential for variation in access to high quality care with widespread regionalization. Unfortunately the data set in the current study did not allow for the assessment of geographic parameters as they relate to treatment at a high or a low volume center. Given the known inverse association between travel distance and appropriate bladder cancer care,3 it is critical to
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
incorporate geographic data into models identifying predictors of treatment at high volume centers. For regionalization to improve the overall quality of care delivered to our patients we must ensure that the structure and process elements of care allow for referral to high volume centers. This is particularly germane for the patients at risk identified in this
383
study, namely the elderly and under insured populations. Matthew J. Resnick Department of Urologic Surgery Vanderbilt University Medical Center Nashville, Tennessee
REFERENCES 1. Barocas DA, Mitchell R, Chang SS et al: Impact of surgeon and hospital volume on outcomes of radical prostatectomy. Urol Oncol 2010; 28: 243.
2. Finks JF, Osborne NH and Birkmeyer JD: Trends in hospital volume and operative mortality for highrisk surgery. N Engl J Med 2011; 364: 2128.
3. Gore JL, Litwin MS, Lai J et al: Use of radical cystectomy for patients with invasive bladder cancer. J Natl Cancer Inst 2010; 102: 802.
Purpose: Although centralization of surgical procedures to high volume centers has been described previously, patterns of care for adrenal surgery are largely unknown. We determined the extent of regionalization of care for adrenal surgery and the extent to which this centralization has evolved with time. Materials and Methods: Using 1996 to 2009 hospital discharge data from New York, New Jersey and Pennsylvania we identified all patients 18 years old or older treated with adrenalectomy. Hospital volume quintiles were created using 1996 hospital volumes. These cutoffs were then applied to subsequent years. Outcome variables were examined by hospital volume status with time using logistic regression models. Results: A total of 8,381 patients underwent adrenalectomy from 1996 to 2009 with a significant 17% to 42% shift toward regionalization to very high volume hospitals, defined as 15 or greater procedures per year (p ⬍0.001). For each successive year the odds of having surgery performed at a very low volume hospital decreased by 13% (OR 0.87, 95% CI 0.84 – 0.89). There were significant differences in patient age, race and payer group for very low volume hospitals, defined as less than 1 procedure per year, compared to very high volume hospitals (p ⬍0.0001). Patients at very high volume hospitals were less likely to be 55 years old or older (OR 0.73, 95% CI 0.61– 0.88), insured through Medicaid (OR 0.60, 95% CI 0.45– 0.79) or uninsured (OR 0.34, 95% CI 0.17– 0.70). When controlling for year treated, patients were less likely to die in the hospital if treated at a very high volume hospital (OR 0.38, 95% CI 0.19 – 0.75). Conclusions: These data reveal the increasing centralization of adrenalectomy to very high volume hospitals since 1996 with improved clinical outcomes. Inequities in access to care to higher volume centers appear to exist and require further investigation.
Abbreviations and Acronyms HLOS ⫽ hospital length of stay HVH ⫽ high volume hospital LVH ⫽ low volume hospital MVH ⫽ medium volume hospital VHVH ⫽ very HVH VLVH ⫽ very LVH Submitted for publication November 20, 2011. Study received Fox Chase Cancer Center institutional review board approval. Supplementary material for this article can be obtained at http://www.fccc.edu/docs/health Professionals/Simhan-Supplementary-Table-1.pdf. * Correspondence and requests for reprints: Urologic Oncology, Fox Chase Cancer Center, Temple University School of Medicine, 333 Cottman Ave., Philadelphia, Pennsylvania 19111 (telephone: 215-728-5342; FAX: 215-214-1734; e-mail: [email protected]).
Key Words: adrenal glands, adrenalectomy, hospitals, delivery of health care, outcome assessment (health care) FIRST reported in 1979,1 the relationship between surgical volume and clinical outcome has been studied extensively. More recently this relationship has been examined in regard to complex cancer procedures such as colorectal surgery, pancreatectomy and
esophagogastrectomy.2 These data reveal a consistent trend that care received at HVHs is associated with decreased HLOS, a lower perioperative complication rate and decreased overall mortality compared to care received at LVHs.3 As a result, experts
0022-5347/12/1882-0377/0 THE JOURNAL OF UROLOGY® © 2012 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION
http://dx.doi.org/10.1016/j.juro.2012.03.130 Vol. 188, 377-383, August 2012 RESEARCH, INC. Printed in U.S.A
AND
www.jurology.com
377
378
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
advocate the regionalization of complex surgical procedures as an integral part of continual efforts to improve quality of care. Although similar benefits in patient outcome were described for radical cystectomy, regionalization studies of other urological procedures are limited.4,5 Only a small portion of incidentally discovered adrenal lesions warrants surgical management but the quality of health care delivery to patients with surgical adrenal disease affects a large group of generally healthy individuals.6 The diagnosis, assessment and surgical management of benign or malignant diseases of the adrenal gland require highly specialized training. Currently most surgical management for adrenal disease centers around resection of adrenal masses due to concern for malignancy or benign neoplasms that are metabolically active.7 With the incidence of incidentally discovered adrenal lesions increasing in conjunction with the increased use of routine cross-sectional imaging, the number of adrenalectomies done in the United States has increased.8 –12 Largely based on established institutional and geographic referral patterns, surgical management of adrenal pathology has been distributed among urologists, general surgeons, surgical oncologists and more recently endocrine surgeons.13 As such, to date patterns of care for adrenal surgery are poorly defined. With increased specialization for managing adrenal disorders we hypothesized that centralization of care for adrenal disorders to high volume surgical centers occurred in the last 15 years. Through an analysis of hospital discharge claims data from 3 large northeastern states we evaluated the extent to which regionalization of care currently exists for adrenal surgery and the degree to which this centralization has evolved with time.
MATERIALS AND METHODS This study was approved by the institutional review board of the Fox Chase Cancer Center. Data analysis was done using hospital discharge data from New York, Pennsylvania and New Jersey from 1996 to 2009. From 1996 to 1999 data were only available for New York and Pennsylvania. Starting in 2000 procedures from all 3 states were included. Using ICD-9 coding we identified all patients 18 years old or older treated with partial (07.2, 07.21, 07.29), unilateral (07.22) or bilateral (07.3) adrenalectomy. Since CPT coding is not included in inpatient discharge data, information on the surgical approach was not available, ie open vs laparoscopic or robotic. Obtained data included patient age, race (white, black, Asian, Native American, other or unknown), gender (male or female), clinical diagnoses, procedure codes, disposition on hospital discharge (home discharge, hospital facility transfer, discharge to rehabilitation/skilled nursing facility, death during hospitalization or unknown) and payer group (private insur-
ance, Medicaid, Medicare, uninsured or unknown). Mortality was defined as any inpatient death after an adrenal procedure during the same hospitalization. Using the methods described by Birkmeyer et al,14 hospitals where adrenalectomy was done in 1996 were ranked in order of increasing 1996 hospital adrenalectomy volume. Five volume groups were then defined by selecting whole number cutoffs for annual volume that most closely sorted the number of procedures into 5 groups of equal size (quintiles). These volume group quintiles were designated VLVH— 0 or 1, LVH—2 or 3, MVH— 4 to 6, HVH—7 to 14 and VHVH—15 or more adrenalectomies annually. The 1996 cutoffs were applied to each subsequent year to determine centralization trends in the distribution of patients who underwent adrenal surgery among the volume categories. Multiple logistic regressions provided estimates of the odds of surgery at a hospital in the lowest or the highest quintile with time with year as a continuous variable and controlling for the overall number of cases done in the study area annually. Independently of treatment year we further investigated the association between sociodemographic factors and volume status. Finally, using hospital volume by quintile, treatment year and total number of procedures in the study area annually as covariates, we investigated the association between hospital volume, and HLOS and inpatient mortality. For each regression analysis a robust cluster variance estimator was used to estimate the SE to account for the intragroup correlation of observations in the same stage. All hypothesis tests were 2 sided with statistical significance considered at p ⬍0.05.
RESULTS A total of 8,381 patients treated with adrenal surgery met study inclusion criteria. While 2,142 (25.5%) and 2,167 patients (25.9%) were between 18 to 44 and 45 to 54 years old, respectively, the largest proportion (48.6%) was 55 years old or older. Of the patients 6,088 (72.6%) were white and 1,050 (12.5%) were designated black. Of the patients 5,261 (62.8%) had private insurance while 2,103 (25.1%) had Medicare. Although 7,992 patients (95.4%) were discharged home after adrenal surgery, 63 (0.8%) died during the hospital stay, including 43 (68.3%) of malignancy involving the adrenal gland as the primary diagnosis. Other primary diagnoses were hyperactive adrenal nodules in 15 patients (23.8%), primary adrenal malignancy in 27 (42.9%), adrenal metastasis in 16 (25.4%) and other in 5 (7.9%). While gender was similar among the groups, age (p ⫽ 0.0005), race (p ⬍0.0001) and payer group (p ⬍0.0001) varied significantly by volume status category. Although the number of hospitals where adrenalectomy was done remained stable during the study period (p ⫽ 0.11), the mean number of adrenal procedures performed per hospital with time increased (mean 2.9 in 1996 vs 5.5 in 2009, p ⬍0.0001, fig. 1). The proportion of hospitals that met VLVH (39.5% vs 28.3%, p ⫽ 0.09) and VHVH (2.6% vs 8.6%,
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
379
Figure 1. Mean number of adrenalectomies performed per hospital by year during study period, stratified by volume status group. From 1996 to 1999 only New York and Pennsylvania data were available.
p ⫽ 0.09) criteria did not significantly change between 1996 and 2009 but a shift was noted from VLVH to VHVH in the proportion of patients who underwent adrenal surgery during a similar period (fig. 2). Of adrenal procedures 42%were performed at VHVHs in 2009 compared to 17% in 1996 (p ⬍0.001).
In 2009 compared to 1996 significantly fewer patients underwent treatment at VLVHs and LVHs (each p ⬍0.0001) while no significant trends were noted for MVHs or HVHs (fig. 3). Multivariate regression analysis was done to determine associations between patient characteris-
Figure 2. Proportion of procedures performed yearly at VHVHs (purple curve) and VLVHs (blue curve) during study period
380
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
Figure 3. Proportion of patients treated with adrenal surgery by hospital volume category based on defined cutoff quintiles determined from 1996 data. Blue bars indicate VLVHs. Red bars indicate LVHs. Yellow bars indicate MVHs. Light green bars indicate HVHs. Purple bars indicate VHVHs. Asterisk indicates significant change in proportion of patients with adrenal surgery with time (p ⬍0.0001).
tics and treatment hospital volume status. All factors were studied individually while controlling for all other assessed characteristics. Excluding patients with an unknown race designation while controlling for year of surgery, patient age, race, gender and payer group, the odds of undergoing surgery at a VLVH decreased by 13% with each subsequent year from 1996 to 2009 (OR 0.87, 95% CI 0.84 – 0.89,
p ⬍0.001, table 1). During the study period patients treated at a VHVH were more likely to be male (OR 1.09, 95% CI 1.09 –1.10, p ⬍0.001) and less likely to be older than 55 years (OR 0.73, 95% CI 0.61– 0.88, p ⫽ 0.001), have Medicare (OR 0.86, 95% CI 0.79 – 0.93, p ⬍0.001) or Medicaid (OR 0.60, 95% CI 0.45– 0.79, p ⫽ 0.004), or be uninsured (OR 0.34, 95% CI 0.17– 0.70, p ⫽ 0.003). In contrast, patients seeking treatment at VLVHs tended to be older than 55 years (OR 1.43, 95% CI 1.10 –1.84, p ⫽ 0.007), or have Medicaid (OR 1.75, 95% CI 1.06 –2.88, p ⫽ 0.03) or no insurance (OR 2.17, 95% CI 1.47– 3.20, p ⬍0.001). Associations of volume status with inpatient mortality and HLOS were then examined by volume quintile using VLVHs as the referent. For the entire cohort independent of year treated and total procedures per year the risk of inpatient mortality decreased as surgical volume increased for each volume quintile compared to VLVHs (table 2). Patients were significantly less likely to die when treated at VHVHs than at VLVHs (OR 0.38, 95% CI 0.19 – 0.75, p ⫽ 0.006). Likely due to the confounding competing risk profiles of patients treated at the highest volume tertiary care centers, those presenting to HVHs were at lowest risk for mortality compared to those at VLVHs (OR 0.26, 95% CI 0.11– 0.59, p ⫽ 0.001). Similar trends were noted for median HLOS since a shorter HLOS was associated with each increasing volume quintile. When controlling for year treated, for the entire cohort a median difference of –1.75 days (95% CI –1.80 –⫺1.69, p ⬍0.001) of hospital stay was noted in the VHVH cohort compared to the VLVH group.
Table 1. Multivariate regression analysis for predictors of surgery at VLVHs and VHVHs, excluding patients with unknown race designation VLVH
Yr Annual procedure vol Age 55 or older Race: White Black Asian Native American Other Gender: F M Payer group: Private Medicare Medicaid Uninsured Unknown
VHVH
OR (95% CI)
p Value
OR (95% CI)
p Value
0.87 (0.84–0.89) 1.00 (1.00–1.00) 1.43 (1.10–1.84)
⬍0.001 0.29 0.007
1.02 (0.90–1.17) 1.00 (1.00–1.00) 0.73 (0.61–0.88)
0.71 0.32 0.001
1.0 1.19 (0.97–1.47) 0.62 (0.50–0.77) 0.55 (0.04–7.69) 0.75 (0.46–1.23)
— 0.10 ⬍0.001 0.65 0.26
1.0 0.69 (0.45–1.06) 1.44 (0.99–2.08) 3.13 (0.52–18.9) 0.90 (0.81–2.07)
— 0.09 0.052 0.21 0.81
1.0 0.89 (0.85–0.93)
— ⬍0.001
1.0 1.09 (1.09–1.10)
— ⬍0.001
1.0 1.03 (0.76–1.40) 1.75 (1.06–2.88) 2.17 (1.47–3.20) 1.06 (0.62–1.80)
— 0.85 0.03 ⬍0.001 0.83
1.0 0.86 (0.79–0.93) 0.60 (0.45–0.79) 0.34 (0.17–0.70) 0.43 (0.10–2.01)
— ⬍0.001 0.004 0.003 0.28
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
381
Table 2. Odds of HLOS difference and inpatient mortality controlled for year treated and annual number of procedures HLOS Difference
Mortality
Hospital Vol Group
OR (95% CI)
p Value
OR (95% CI)
p Value
VLVH (referent) LVH MVH HVH VHVH
0 ⫺0.86 (⫺0.93–0.80) ⫺0.94 (⫺1.01–0.88) ⫺1.68 (⫺1.74–1.61) ⫺1.75 (⫺1.81–1.69)
— ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001
1 0.66 (0.43–0.99) 0.44 (0.26–0.74) 0.26 (0.11–0.59) 0.38 (0.19–0.75)
— 0.043 0.002 0.001 0.006
DISCUSSION Adrenal surgery rates have been steadily increasing in conjunction with the increased prevalence of incidental adrenal masses found on routine cross-sectional imaging.8,11 A recent retrospective cohort analysis of 3,144 adrenalectomies identified from the Nationwide Inpatient Sample revealed that surgeon volume and not specialty was an independent predictor of postoperative complications and HLOS.15 These findings mirror recent provider volume-outcome trends in patients undergoing major cancer surgery14,16,17 and they have important implications for quality of care. Despite these observations additional efforts to further define the provider volume-surgical outcomes relationship using population based cohort data have yielded conflicting results.18,19 To date centralization trends for adrenalectomy have been poorly characterized. The association between quality of care and the volume-outcome relationship has led many groups to advocate centralization of complex urological and cancer procedures to high volume centers. Using a state based discharge claims data set for patients treated with extirpative surgery for prostate, colorectal, esophageal or pancreatic cancer Stitzenberg et al reported a significant shift toward high volume centers for esophageal, pancreatic and colon cancer procedures.2,4 Using similar methods our findings demonstrate that in 3 northeastern states adrenal surgery became increasingly centralized to VHVHs during a 15-year period. In our sample the odds of undergoing adrenal surgery at a VLVH decreased by 13% per year and 42% of all adrenal surgeries were done at the highest volume centers in 2009. In our data set only crude outcomes measures were available, including inpatient mortality and HLOS. Results are limited by a lack of confirmatory comorbidity and pathological data. Despite these limitations distinct trends toward decreased length of stay and mortality with increasing volume status were observed in our sample, which should be confirmed using other data sets. Further analysis of social and demographic trends revealed several disparities indicating disparate access to care. In our entire cohort there were significant differences among volume status groups
in patient age, race, gender and payer group. On multivariate analysis controlling for all factors patients 55 years old or older, patients insured through Medicare or Medicaid and uninsured patients were significantly less likely be treated at a VHVH. In contrast, patients treated at a VLVH were more likely to be 55 years old or older and have Medicaid or no insurance while black race approached significance. More characterization of centralization trends and disparities in access to care are challenging but necessary in the health care climate today. With growing subspecialization in surgical disciplines today adrenal procedures are increasingly referred to urologists and general surgeons with fellowship training in laparoscopy and oncology, who are more likely to operate at tertiary care or high volume centers.20,21 Previous analyses showed repeat disparities in access to care at high volume centers for complex surgery, which may be influenced by socioeconomic status, geographic location, travel burden, cultural influences and personal preference.2,22,23 Ayanian et al suggested that the improved outcomes in patients with insurance treated at HVHs may be further confounded by the higher disease severity in patients without insurance treated at LVHs, which may predispose to poor outcomes.24 Furthermore, the regionalization of adrenal surgery to higher volume centers may contribute to the closing of adrenal surgery programs at smaller hospitals, which could potentially exacerbate access disparities in underserved areas. Clear limitations to the use of our data set include the lack of patient specific data, such as pathological stage, comorbid conditions, perioperative complication data and rehospitalization rates. As a result, our outcome measures do not adequately adjust for severity and must be interpreted with caution. Furthermore, New Jersey, New York and Pennsylvania represent a unique urban area with extensive health care resources. As a result, these findings may not be generalizable to less densely populated regions of the country. Nevertheless, despite these limitations all payer administrative data provide the opportunity to accurately assess regionalization trends with
382
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
time and provide some insight into the access to care challenges currently facing our health care system.
CONCLUSIONS We present a rigorous analysis characterizing the centralization of adrenal surgical care to VHVHs in
a 15-year period. Regionalization of care is associated with improved unadjusted short-term outcomes, which require further evaluation in more robust data sets. Also, the apparent inequities in access to care at high volume centers that may exist for elderly, uninsured and Medicaid patients should be the focus of future investigations.
REFERENCES 1. Luft HS, Bunker JP and Enthoven AC: Should operations be regionalized? The empirical relation between surgical volume and mortality. N Engl J Med 1979; 301: 1364. 2. Stitzenberg KB, Sigurdson ER, Egleston BL et al: Centralization of cancer surgery: implications for patient access to optimal care. J Clin Oncol 2009; 27: 4671. 3. Birkmeyer JD, Stukel TA, Siewers AE et al: Surgeon volume and operative mortality in the United States. N Engl J Med 2003; 349: 2117. 4. Stitzenberg KB, Wong YN, Nielsen ME et al: Trends in radical prostatectomy: centralization, robotics, and access to urologic cancer care. Cancer 2012; 118: 54. 5. Barbieri CE, Lee B, Cookson MS et al: Association of procedure volume with radical cystectomy outcomes in a nationwide database. J Urol 2007; 178: 1418. 6. Kutikov A, Crispen PL, Uzzo RG: Pathophysiology, evaluation, and medical management of adrenal disorders. In: Campbell-Walsh Urology. Edited by AJ Wein, LR Kavoussi, AW Partin et al. Philadelphia: Elsevier 2012; vol 2, chap 57, pp 1685– 1736. 7. Mazzaglia PJ and Vezeridis MP: Laparoscopic adrenalectomy: balancing the operative indications with the technical advances. J Surg Oncol 2010; 101: 739. 8. Gallagher SF, Wahi M, Haines KL et al: Trends in adrenalectomy rates, indications, and physician
volume: a statewide analysis of 1816 adrenalectomies. Surgery 2007; 142: 1011. 9. Saunders BD, Wainess RM, Dimick JB et al: Trends in utilization of adrenalectomy in the United States: have indications changed? World J Surg 2004; 28: 1169. 10. Young WF Jr: Clinical practice. The incidentally discovered adrenal mass. N Engl J Med 2007; 356: 601. 11. Miccoli P, Raffaelli M, Berti P et al: Adrenal surgery before and after the introduction of laparoscopic adrenalectomy. Br J Surg 2002; 89: 779. 12. Kutikov A, Mallin K, Canter D et al: Effects of increased cross-sectional imaging on the diagnosis and prognosis of adrenocortical carcinoma: analysis of the National Cancer Database. J Urol 2011; 186: 805.
17. Dimick JB, Pronovost PJ, Cowan JA Jr et al: Variation in postoperative complication rates after high-risk surgery in the United States. Surgery 2003; 134: 534. 18. Murphy MM, Witkowski ER, Ng SC et al: Trends in adrenalectomy: a recent national review. Surg Endosc 2010; 24: 2518. 19. Stavrakis AI, Ituarte PH, Ko CY et al: Surgeon volume as a predictor of outcomes in inpatient and outpatient endocrine surgery. Surgery 2007; 142: 887. 20. Duchene DA, Moinzadeh A, Gill IS et al: Survey of residency training in laparoscopic and robotic surgery. J Urol 2006; 176: 2158.
13. Sosa JA, Wang TS, Yeo HL et al: The maturation of a specialty: workforce projections for endocrine surgery. Surgery 2007; 142: 876.
21. Le D, Karmali S, Harness JK et al: An update: the operative experience in adrenal, pancreatic, and other less common endocrine diseases of U.S. general surgery residents. World J Surg 2008; 32: 232.
14. Birkmeyer JD, Siewers AE, Finlayson EV et al: Hospital volume and surgical mortality in the United States. N Engl J Med 2002; 346: 1128.
22. Liu JH, Zingmond DS, McGory ML et al: Disparities in the utilization of high-volume hospitals for complex surgery. JAMA 2006; 296: 1973.
15. Park HS, Roman SA and Sosa JA: Outcomes from 3144 adrenalectomies in the United States: which matters more, surgeon volume or specialty? Arch Surg 2009; 144: 1060.
23. Stitzenberg KB and Meropol NJ: Trends in centralization of cancer surgery. Ann Surg Oncol 2010; 17: 2824.
16. Begg CB, Cramer LD, Hoskins WJ et al: Impact of hospital volume on operative mortality for major cancer surgery. JAMA 1998; 280: 1747.
24. Ayanian JZ, Weissman JS, Schneider EC et al: Unmet health needs of uninsured adults in the United States. JAMA 2000; 284: 2061.
EDITORIAL COMMENT These authors evaluated trends in the regionalization of adrenalectomy in a claims based administrative data set comprising discharge information on patients in the northeastern United States. They noted regionalization of adrenalectomy to high volume centers during the study period. Patients of increasing age and those with uninsured status or who were insured through the Medicaid program were more likely to be treated at VLVHs. These study findings have significant implications, particularly in regard to the intersection of health care access and quality in impoverished individuals.
This study adds to the growing body of literature on the well documented volume-outcome relationship for major urological1 and nonurological2 surgical procedures. While current data suggest that regionalization to high volume centers improves surgical outcomes, one must consider the potential for variation in access to high quality care with widespread regionalization. Unfortunately the data set in the current study did not allow for the assessment of geographic parameters as they relate to treatment at a high or a low volume center. Given the known inverse association between travel distance and appropriate bladder cancer care,3 it is critical to
REGIONALIZATION OF ADRENALECTOMY TO HIGHER VOLUME SURGICAL CENTERS
incorporate geographic data into models identifying predictors of treatment at high volume centers. For regionalization to improve the overall quality of care delivered to our patients we must ensure that the structure and process elements of care allow for referral to high volume centers. This is particularly germane for the patients at risk identified in this
383
study, namely the elderly and under insured populations. Matthew J. Resnick Department of Urologic Surgery Vanderbilt University Medical Center Nashville, Tennessee
REFERENCES 1. Barocas DA, Mitchell R, Chang SS et al: Impact of surgeon and hospital volume on outcomes of radical prostatectomy. Urol Oncol 2010; 28: 243.
2. Finks JF, Osborne NH and Birkmeyer JD: Trends in hospital volume and operative mortality for highrisk surgery. N Engl J Med 2011; 364: 2128.
3. Gore JL, Litwin MS, Lai J et al: Use of radical cystectomy for patients with invasive bladder cancer. J Natl Cancer Inst 2010; 102: 802.