- Email: [email protected]
Sickle Cell Disease in Priapism: Disparity in Care?
Male Sexual Dysfunction Sickle Cell Disease in Priapism: Disparity in Care? Gregory A. Joice, Max Kates, Nikolai A. Sopko, Johanna L. Hannan, and Trinity J. Bivalacqua OBJECTIVE MATERIALS AND METHODS
RESULTS
CONCLUSION
To determine the effect of sickle cell disease (SCD) on hospital resource use among patients admitted for priapism. Using the Nationwide Inpatient Sample, a weighted sample of 12,547 patients was selected with a primary diagnosis of priapism from 2002 to 2011. Baseline differences for patient demographics and hospital characteristics were compared between SCD and non-SCD patients. Multivariate analysis was performed to identify the effect of SCD on length of stay, use of penile operations, blood transfusion, and cost. The proportion of SCD patients was 21.5%. SCD patients were younger, more often black, more likely to have Medicaid insurance, and treated more frequently in Southern urban teaching hospitals. SCD was a significant predictor of having a blood transfusion (odds ratio [OR], 16.3; P <.001), and an elongated length of stay (OR, 1.42; P <.001). SCD was associated with less penile operations (OR, 0.40; P <.001). When SCD patients did have an operation, it was performed later in the admission (mean, 0.87 vs 0.47 days; P <.001). SCD was not a significant predictor of increased cost (OR, 1.02; P ¼ .869). SCD patients represent a demographically distinct subgroup of priapism patients with different patterns of resource use manifested by longer hospital stays and more blood transfusions. Moreover, despite evidence that immediate treatment of priapism results in improved erectile function outcomes, SCD patients had less surgical procedures for alleviation of acute priapism events. UROLOGY 86: 72e79, 2015. 2015 Elsevier Inc.
P
riapism represents one of the most common acute emergencies in urology, affecting nearly 10,000 men annually with nearly 30% of these patients going on to be admitted to the hospital.1 However, the hospital course and inpatient costs of treating priapism are unknown. Approximately 1 in 5 patients presenting to an emergency room in the United States with priapism have sickle cell disease (SCD).1,2 Immediate nonsurgical and surgical interventions are necessary to preserve erectile function. Ischemic priapism is a well-known complication of SCD and has been reported to have a lifetime prevalence of approximately 28%-35%.3,4 Historically, SCD patients were treated conservatively with oxygenation, hydration, and blood transfusions; however, we now have a better understanding of the pathophysiology of SCD-associated priapism, and new guidelines recommend immediate interventions to preserve erectile function.5 Patients with
SCD encounter the health care system frequently, with 68 admissions per 100 patients annually and an estimated $1.5 million in annual inpatient charges per 100 patients.6 Prior studies have demonstrated that among patients presenting to emergency departments with priapism, SCD is an independent predictor of hospital admission.2 We hypothesize that SCD and non-SCD patients with priapism are distinct with regard to their hospital resource use and may represent a disparity in care. The present study compares patient demographics, hospital characteristics, and inpatient outcomes between SCD and non-SCD patients. The primary purpose of the study is to assess associations between SCD and inpatient costs and treatment patterns of patients admitted with priapism in the United States.
MATERIALS AND METHODS Data Sources and Patient Selection
Financial Disclosure: The authors declare that they have no relevant financial interests. From the James Buchanan Brady Urological Institute and Department of Urology, Johns Hopkins School of Medicine, Baltimore, MD Address correspondence to: Gregory A. Joice, M.D., James Buchanan Brady Urological Institute and Department of Urology, Johns Hopkins School of Medicine, 1800 Orleans Street, Marburg 134, Baltimore, MD 21287. E-mail: [email protected] Submitted: October 17, 2014, accepted (with revisions): January 23, 2015
72
ª 2015 Elsevier Inc. All Rights Reserved
The Nationwide Inpatient Sample (NIS) is the largest all-payer inpatient database in the United States, consisting of an annual 20% stratified sample of community and academic hospitals. The NIS contains data elements for each discharge including diagnoses and procedures along with patient demographics and hospital characteristics. The recent update of the NIS through http://dx.doi.org/10.1016/j.urology.2015.01.050 0090-4295/15
2011 includes data from approximately 1000 hospitals over 45 states.7 Patients were selected from the 2002-2011 NIS with a primary diagnosis of priapism (“International Classification of Disease, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis code 607.3). A subset of patients was identified who had a concomitant diagnosis of SCD (ICD-9-CM codes 282.6x). The data were weighted using the NIS0 weighting methodology to create national estimates.7
Independent Variables Patient demographics included age, sex, race, payer, median household income, and comorbidities. Patient comorbidities were represented by the Charlson comorbidity index (CCI) and calculated using the Charlson-Deyo adaption for administrative data to create the CCI within the data set.8,9 To account for missing values in the multivariate analysis, the race variable was simplified into white, black, Hispanic, other, and missing. For demographic comparisons, the missing values were excluded when reporting percentages. Hospital characteristics included the number of hospital beds, teaching status, hospital location, and hospital region.
Main Outcome Measures The primary outcomes were penile operation, transfusion, length of stay (LOS), and total hospitalization cost. Penile operation included penile incision, placement of cavernosal shunts, and insertion of prostheses (ICD-9-CM procedure codes 64.92, 64.95, 64.97, and 64.98). Patients who received a transfusion were similarly identified (ICD-9-CM procedure codes 99.0x). The number of days from admission that the patient received either a transfusion or operation was also analyzed. LOS was dichotomized into a new variable, elongated LOS, defined as the top 90th percentile of LOS within the sample. The NIS includes data on total charges for each hospitalization. However, this variable represents the amount the hospitals billed for the service rather than the actual cost spent during hospitalization.10 Supplemental files can be requested that include a cost-to-charge ratio for each hospital allowing conversion of total charges to total cost. The cost-to-charge ratio was used to create a total cost variable for each patient and was further dichotomized to a new variable, increased total cost, defined as the top 90th percentile.
Statistical Analysis Baseline differences in demographics and outcomes were compared using the chi-square test for categorical variables and the Mann-Whitney U test for continuous variables. The effects of independent variables and SCD were assessed using multivariate logistic regression models to produce odds ratios (ORs). Penile operation was included as an independent variable in the transfusion model, and both penile operation and transfusion were included in the LOS and total cost models. All multivariate regression analyses were adjusted for clustering using the generalizing estimating equations method.11-14 All analyses were performed using SAS (version 9.2; SAS Institute, Cary, NC) with a 2-sided significance level of P <.05.
the NIS from 2002 to 2011. The patient demographics and hospital characteristics are listed in Table 1. Overall, among patients admitted to the hospital for priapism, 21.5% had a concomitant diagnosis of SCD. Patients with SCD compared to those without SCD were younger (20 years median age vs 42 years; P <.001), more often black (96.0% vs 36.7%; P <.001), in low median income brackets (47.0% vs 35.8%; P <.001), with Medicaid insurance (49.8% vs 18.0%; P <.001), and healthier (CCI of 0: 83.5% vs 77.0%; P <.001). SCD patients were more likely to receive their care at urban (97.5% vs 92.6%; P <.001), teaching (74.8% vs 58.0%; P <.001) hospitals in the Southern region (51.3% vs 42.5%; P <.001). The outcome variables are listed in Table 2. The overall rate of penile operations in the sample was 68.3%, but SCD patients were significantly less likely to receive operations (42.4% vs 75.4%; P <.001). In addition, SCD patients on average had their operations performed later during their admission (0.87 days vs 0.47 days; P ¼ .01). SCD patients were more likely to receive a blood transfusion during their hospitalization (37.6% vs 2.8%; P <.001) and on average had their transfusions earlier (0.9 days vs 1.77 days; P <.001). In addition, SCD patients had a longer mean LOS (3.38 vs 2.69 days; P <.001), but there was no statistically significant difference between mean total cost ($5899 vs $5698; P ¼ .89). Similar analysis comparing African American patients to noneAfrican American patients (See Supplementary Table) revealed fewer operations (72.1% vs 79.1%; P <.001) and more transfusions (2.2% vs 4.7%; P <.001). However, African American patients did not experience any difference in time to operation, time to transfusion, or increased LOS. On multivariate analysis, SCD patients were less likely to receive a penile operation (OR, 0.40; P <.001) after controlling for independent variables (Table 3). These patients were also more likely to receive a blood transfusion (OR, 16.3; P <.001) and have an elongated LOS (OR, 1.42; P ¼ .001). However, there was no significant difference in increased total cost between SCD and nonSCD patients (OR, 1.02; P ¼ .869). The largest relative differences between these patient populations were penile operation and transfusion rates, where SCD patients were 33% less likely to undergo a penile procedure while being 35% more likely to have a blood transfusion (Fig. 1A). In patients who did not undergo any penile procedure, SCD patients had a higher average total cost by $882, whereas those who underwent a penile procedure had an average increase of $1479 (Fig. 1B). However, neither of these differences was statistically significant in multivariate models including only those patients either with or without a penile operation.
COMMENT RESULTS Overall, a nationally weighted sample of 12,547 patients with a primary diagnosis of priapism was identified within UROLOGY 86 (1), 2015
The current analysis demonstrates that priapism is a costly disease with significant disparity in management of patients with SCD. Patients with SCD experience far 73
Table 1. Patient demographics and hospital characteristics Characteristic
Sickle Cell
Both
P Value*
(78.5)
2694 (21.5)
12,547 (100.0)
N/A
(18.2) (40.5) (41.3)
2247 (83.4) 394 (14.6) 53 (1.9)
4041 (32.2) 4386 (35.0) 4120 (32.8)
<.001
(49.1) (36.7) (10.5) (3.7)
14 2120 35 39
3935 5050 872 331
<.001
(35.8) (28.3) (36.0)
1218 (47.0) 655 (25.3) 717 (27.7)
4587 (38.2) 3319 (27.6) 4108 (34.2)
(18.6) (18.0) (30.6) (32.9)
246 1333 698 398
2066 3099 3693 3619
(77.0) (15.6) (7.4)
2250 (83.5) 358 (13.3) 86 (3.2)
9832 (78.4) 1897 (15.1) 818 (6.5)
(7.4) (92.6)
66 (2.5) 2606 (97.5)
793 (6.4) 11,665 (93.6)
<.001
(42.0) (58.0)
674 (25.2) 1998 (74.8)
4784 (38.4) 7675 (61.6)
<.001
(7.1) (25.7) (67.2)
222 (8.3) 632 (23.6) 1818 (68.0)
919 (7.4) 3147 (25.3) 8392 (67.4)
(21.3) (17.9) (42.5) (18.2)
547 566 1383 198
2649 2333 5575 1990
Nonsickle Cell
All, N (%) 9853 Age, n (%) <30 1794 30-44 3992 >45 4067 Race, n (%)y White 3921 Black 2929 Hispanic 838 Other 291 Median income, n (%) Low 3369 Medium 2664 High 3391 Payer, n (%) Medicare 1819 Medicaid 1766 Private 2995 Other 3221 CCI, n (%) 0 7582 1 1539 2þ 732 Hospital location, n (%) Rural 727 Urban 9059 Hospital academic status, n (%) Nonteaching 4110 Teaching 5676 Hospital bed size, n (%) Small 697 Medium 2515 Large 6574 Hospital region, n (%) Northeast 2102 Midwest 1767 South 4192 West 1792
(0.6) (96.0) (1.6) (1.8)
(9.2) (49.8) (26.1) (14.9)
(20.3) (21.0) (51.3) (7.4)
(38.6) (49.6) (8.6) (3.2)
(16.6) (24.8) (29.6) (29.0)
(21.1) (18.6) (44.4) (15.9)
<.001
<.001
<.001
.023
<.001
CCI, Charlson comorbidity index; N/A, not applicable. * P values are derived from chi-square tests for categorical variables. y Missing values for race were excluded when calculating percentages.
fewer operative interventions (OR, 0.40; P <.001), and when they do receive an operation, it is with a significant delay to surgery when compared to the non-SCD cohort despite the known benefits of immediate treatment of acute priapism on long-term erectile function rates. Whether this constitutes a disparity in care, reluctance to perform shunting penile surgery in a young population, or an unmeasured difference in their disease course remains undetermined. The International Consultation on Sexual Medicine and the European Association of Urology have published clinical practice guidelines for treatment of priapism.5,15,16 Within the guidelines, they give specific treatment considerations for patients with SCD. For the general population, they suggest a stepwise approach beginning with decompression of the corpora cavernosa with either aspiration or injection of sympathomimetic drugs followed by surgical intervention.16 They further suggest that patients with SCD should undergo the same initial treatment plan and that supportive measures 74
including intravenous hydration, oxygen administration, and blood exchange transfusion should be considered but should not be the primary treatment. However, specifically they note that these additional treatments “should not delay initial treatment.”16 Despite these guidelines, the present study suggests that SCD patients are less likely to receive any form of penile operation and thus have a delay in the treatment of their priapic episodes. In addition, we showed that patients with SCD on average receive penile operations 0.4 day later in their hospitalization than patients without SCD. Given that extended delay in treatment of ischemic priapism leads to erectile dysfunction, this finding may represent a potential health care quality disparity for SCD patients.17-19 Our study also revealed an alarmingly high absolute rate of transfusions among SCD patients, who are transfused during 38% of admissions for priapism. This high transfusion rate among the SCD cohort does not appear to be supported by the evidence. Blood transfusion has UROLOGY 86 (1), 2015
Table 2. Inpatient outcomes Outcome
Nonsickle Cell
Penile operation, n (%) No Yes Day to operation, mean (95% CI) Transfusions, n (%) No Yes Day to transfusion, mean (95% CI) LOS, mean (95% CI) Total cost, $, mean (95% CI)
2426 (24.6) 7427 (75.4) 0.47 (0.40-0.52)
1552 (57.6) 1141 (42.4) 0.87 (0.62-1.15)
9575 278 1.77 2.69 5698
1680 1014 0.90 3.38 5899
(97.2) (2.8) (1.22-2.33) (2.56-2.82) (5312-6085)
Both
P Value*
3978 (31.7) 8569 (68.3) 0.52 (0.46-0.59)
<.001
Sickle Cell
(62.4) (37.6) (0.67-1.13) (3.09-3.66) (5315-6484)
11,254 1292 1.09 2.83 5744
(89.7) (10.3) (0.87-1.30) (2.71-2.95) (5417-6070)
.01 <.001 <.001 <.001 .891
CI, confidence interval; LOS, length of stay. * P values are derived from chi-square tests for categorical variables and Mann-Whitney U test for continuous variables.
Table 3. Multivariate outcome models for SCD Outcome Penile operation Transfusion Elongated LOS Increased total cost
Odds Ratio (95% CI)* 0.40 16.3 1.42 1.02
(0.35-0.45) (12.9-20.6) (1.15-1.75) (0.81-1.29)
P Value <.001 <.001 .001 .869
SCD, sickle cell disease; other abbreviations as in Table 2. * Odds ratios from generalized estimating equations multivariate models controlling for patient demographics and hospital characteristics.
been proposed as therapy for priapism in SCD patients because of the mechanism that transfusion can lead to increased oxygen delivery to tissues and prevent tissue damage. As mentioned, the International Consultation on Sexual Medicine and European Association of Urology guidelines recommend blood transfusion only as a supplemental therapy for priapism.16 However, additional studies question the efficacy of blood transfusion in this setting.17 Merritt et al20 published a study of a series of case reports to evaluate the efficacy of blood transfusions in SCD patients with chronic priapism. When compared to patients receiving conventional therapy, the group receiving blood transfusions experienced increased mean time to detumescence (11 vs 8 days). Among the blood transfusion cohort, 9 of 26 patients who received blood transfusions developed neurologic sequelae ranging from mild and temporary to severe and permanent. Association of SCD, priapism, exchange transfusions, and neurologic events (ASPEN) is a known complication of blood transfusions in SCD patients that could account for the neurologic decline observed in this study.21,22 Although there have been no prospective studies looking at the efficacy and risks of blood transfusions in priapism, the current literature appears to lack evidence to support its routine use as a primary therapy. These previous studies highlight the concerning finding of our study that SCD patients with priapism are frequently receiving transfusions. This high rate of transfusions may be exposing SCD patients to a higher risk of complications with no benefit on erectile function preservation. However, within the NIS, we cannot specifically determine whether a blood transfusion was used as a primary or supplemental UROLOGY 86 (1), 2015
therapy. Future studies should look at the timing between transfusion and additional therapies and the incidence of ASPEN syndrome in these patients. Potential disparities in care are exaggerated by the contrasting demographic cohorts of the SCD and nonSCD groups. SCD patients are more likely to have Medicaid insurance and be treated at urban, teaching hospitals in the South. This finding is significant because these hospitals will bear a more significant cost burden of treating these publically insured patients. Health care use and cost is important to consider in SCD patients given the chronicity of the disease and the need for multiple hospitalizations. A recent study estimated that SCD patients average approximately 2.5 acute care visits per year, representing a relatively high health care use subgroup of patients.23 Dinan et al24 showed that among patients undergoing cholecystectomy or hip replacement, SCD was an independent predictor of increased hospital cost. They further go on to propose that this increase in cost is primarily because of the increased LOS in SCD patients. In the present study, we did show an approximate 25% increase in LOS for sickle cell patients but did not show any significant increase in total costs, likely owing to the decreased use of operative intervention. On further analysis, we demonstrated that among priapism patients with nonoperative management, the cost for SCD was 24% higher and for those who did have an operation the increase was 23%. However, neither was a significant difference on multivariate analysis. This finding suggests that the increased LOS in the SCD population causes an increase in cost that is partially offset by the increased rate of penile operations in the general population. In addition, blood transfusion (OR, 5.27; P < .001) was a strong predictor of increased cost in the model, suggesting that the high transfusion rate among SCD patients may also be an important driver of increased costs for patients. As a result of more SCD patients being on Medicaid insurance plans, their increased total cost may have a greater relative effect on hospital reimbursements. In addition, SCD patients are significantly more likely to be African American, which may influence decisions on surgery and transfusions. We indeed did show that within the non-SCD cohort, 75
Figure 1. (A) Differences in outcomes between SCD and non-SCD patients. SCD patients had higher transfusion rates (þ35%), lower rates of penile operations (33%), and longer LOS (þ0.7 days), which were all statistically significant on multivariate analysis. SCD patients had slightly higher cost (þ$201), but this difference was not statistically significant. (B) Graph of total hospital cost by SCD and penile operation. SCD patients had a higher average cost for both patients receiving penile operations (þ$1479) and those not receiving an operation (þ$882). However, these increases were not statistically significant on multivariate analysis within each of the subgroups. LOS, length of stay; SCD, sickle cell disease. (Color version available online.)
African American patients are less likely to undergo penile operations and more likely to have blood transfusions. However, these differences were far less clinically significant than similar differences observed between SCD and non-SCD patients. Although race may play a role in choice of priapism interventions, SCD still appears to be an independent and clinically significant indicator of different hospital resource use. 76
Few recent population-based studies have analyzed the experience for SCD patients presenting with priapism. Roghmann et al2 looked within the Nationwide Emergency Department Sample for predictors of admission for priapism. They showed that among several patient demographics and hospital characteristics, SCD was an independent predictor of admission for priapism (OR, 2.22; P <.001). Stein et al1 also looked within the UROLOGY 86 (1), 2015
Nationwide Emergency Department Sample to show that 21.1% of patients with a diagnosis of priapism had concurrent SCD, but they did not further analyze what effect this had on their treatment or admission. Chrouser et al25 studied an inpatient population with priapism using the NIS. They showed that SCD patients were less likely to undergo a penile operation compared to non-SCD patients. However, they did not further assess the effect of SCD on other outcomes including LOS, total cost, and transfusion. The limitations of the present study relate to the use of administrative data. These data are inherently susceptible to interinstitution variability in coding practices that could cause inaccuracies in the database. However, despite this limitation, previous studies have shown accuracy rates as high as 80% in correctly identifying conditions with administrative data.26,27 In addition, it is difficult to isolate whether a patient with SCD is only having an isolated case of priapism or if they also have a concomitant painful occlusive crisis. We have attempted to decrease these confounding variables by only including patients with a primary diagnosis of priapism. Finally, the NIS does not contain longitudinal data on individual patients, and thus, tracking outcomes such as erectile dysfunction or readmissions is not possible. In addition, the effect of potential readmissions and recurrent transfusions cannot be included in the total cost model. Future studies should analyze the long-term outcomes and treatment sequelae of patients undergoing different treatment modalities.
CONCLUSION Recent guidelines for the treatment of acute priapism recommend that SCD patients should be treated similarly to the general population with early intervention to achieve detumescence. Despite these guidelines, we show that patients with SCD admitted for priapism are treated less often with operative interventions, and when they are, their surgery is delayed 2-fold compared to others. In addition, SCD patients are transfused more than 1 in every 3 admissions despite a paucity of data supporting its efficacy in priapism. These findings highlight a significant treatment disparity for SCD patients that can potentially lead to more complications and worse erectile function outcomes. References 1. Stein DM, Flum AS, Cashy J, et al. Nationwide emergency department visits for priapism in the United States. J Sex Med. 2013; 10:2418. 2. Roghmann F, Becker A, Sammon JD, et al. Incidence of priapism in emergency departments in the United States. J Urol. 2013;190: 1275. 3. Adeyoju AB, Olujohungbe AB, Morris J, et al. Priapism in sicklecell disease; incidence, risk factors and complications—an international multicentre study. BJU Int. 2002;90:898. 4. Mantadakis E, Cavender JD, Rogers ZR, et al. Prevalence of priapism in children and adolescents with sickle cell anemia. J Pediatr Hematol Oncol. 1999;21:518.
UROLOGY 86 (1), 2015
5. Broderick GA. Priapism and sickle-cell anemia: diagnosis and nonsurgical therapy. J Sex Med. 2012;9:88. 6. Lanzkron S, Carroll CP, Haywood C Jr. The burden of emergency department use for sickle-cell disease: an analysis of the national emergency department sample database. Am J Hematol. 2010;85:797. 7. 2011 Introduction to the NIS. Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality; 2013; Available at: http://www.hcup-us.ahrq.gov/ nisoverview.jsp, 2013; Accessed January 21, 2014. 8. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613. 9. Charlson ME, Ales KL, Pompei P, MacKenzie CR. A new method of classification of prognostic comorbidity for longitudinal studies: development and validation. J Chronic Dis. 1987;40:373. 10. 2011 Cost-to-Charge Ratio. Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality; 2013; Available at: http://www.hcup-us.ahrq.gov/db/state/ costtocharge.jsp, 2013; Accessed January 21, 2014. 11. 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. 12. Hu JC, Gold KF, Pashos CL, et al. Temporal trends in radical prostatectomy complications from 1991 to 1998. J Urol. 2003;169:1443. 13. Begg CB, Riedel ER, Bach PB, et al. Variations in morbidity after radical prostatectomy. N Engl J Med. 2002;346:1138. 14. 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. 15. Albersen M, Bivalacqua TJ. Sexual dysfunction: first EAU priapism treatment guidelines published. Nat Rev Urol. 2014;11:132. 16. Salonia A, Eardley I, Giuliano F, et al. European Association of Urology guidelines on priapism. Eur Urol. 2014;65:480. 17. Olujohungbe A, Burnett AL. How I manage priapism due to sickle cell disease. Br J Haematol. 2013;160:754. 18. Burnett AL, Bivalacqua TJ. Priapism: new concepts in medical and surgical management. Urol Clin North Am. 2011;38:185. 19. Hinman F Jr. Priapism: reasons for failure of therapy. J Urol. 1960; 83:420. 20. Merritt AL, Haiman C, Henderson SO. Myth: blood transfusion is effective for sickle cell anemia-associated priapism. CJEM. 2006;8:119. 21. Siegel JF, Rich MA, Brock WA. Association of sickle cell disease, priapism, exchange transfusion and neurological events: ASPEN syndrome. J Urol. 1993;150:1480. 22. Jones G, Lee BR, Woods GM, et al. Neurologic complications in pediatric sickle cell patients hospitalized for priapism: an analysis of the Pediatric Health Information System (PHIS) database. Blood. 2013;122:2287. 23. Brousseau DC, Owens PL, Mosso AL, et al. Acute care utilization and rehospitalizations for sickle cell disease. JAMA. 2010;303:1288. 24. Dinan MA, Chou CH, Hammill BG, et al. Outcomes of inpatients with and without sickle cell disease after high-volume surgical procedures. Am J Hematol. 2009;84:703. 25. Chrouser KL, Ajiboye OB, Oyetunji TA, et al. Priapism in the United States: the changing role of sickle cell disease. Am J Surg. 2011;201:468. 26. Lawthers AG, McCarthy EP, Davis RB, et al. Identification of inhospital complications from claims data. Is it valid? Med Care. 2000;38:785. 27. Burns EM, Rigby E, Mamidanna R, et al. Systematic review of discharge coding accuracy. J Public Health. 2012;34:138-148.
APPENDIX SUPPLEMENTARY DATA
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.urology. 2015.01.050. 77
EDITORIAL COMMENT The Institute of Medicine defines health care disparities as “differences in the quality of health care that are not due to access-related factors or clinical needs, preferences, and appropriateness of intervention.”1 Joice et al2 hypothesize that a preexisting diagnosis of sickle cell disease (SCD) may lead to health care disparities because of biases, prejudices, or stereotypes held by providers that influence recommended clinical care.3 The authors found that SCD patients admitted for priapism are treated less often with surgery, which is often delayed compared to men without SCD with priapism. In addition, the authors identified a 38% transfusion rate for SCD patients who were admitted. The basic issue is whether this represents a health care disparity or rather readily explainable practice patterns. Both American and European guidelines recommend a stepwise approach to the treatment of priapism: First-line treatments are highly recommended before any surgical treatment of ischemic priapism. Surgery should be considered only when conservative management options fail, and surgical treatment is also recommended for priapism events lasting over 72 hours.4,5 Study patients were all hospitalized, and as the authors note, management of priapism occurs 70% of the time in the outpatient setting (eg, the patient presents to the emergency department with a priapism and is managed with intracavernous treatment including aspiration or irrigation with successful detumescence). Therefore, the National Inpatient Sample may lead to biased estimates of resource allocation associated with priapism between the 2 study populations, as most interactions (and interventions) with the health care system are not captured in the Nationwide Inpatient Sample. According to the authors, SCD patients represent a distinct subgroup of priapism when treated in hospitals. This observation may hold true also in outpatient care, and different outpatient treatment of SCD patients may explain differences found in hospitalized patients. For example, because of education about risks of priapism, SCD patients may be more likely to present for care more quickly than non-SCD patients, where embarrassment or lack of awareness of the consequences to erectile function may delay clinical presentation in non-SCD patients. The study seems to assume that conservative treatments had already been tried and failed and therefore, surgical intervention would be needed. Without information on prior outpatient treatment, this assumption cannot be validated. The study did not analyze treatment outcomes, which is the final determinant of quality of care. SCD is an independent predictor of hospital admission for priapism.6 SCD patients who present with priapism alone and without sickle cell crisis are often managed in the emergency room successfully; SCD patients with priapism and additional acute manifestations of SCD typically require inpatient admission to manage SCD clinical sequelae along with priapism. This may explain increased rates of transfusion in hospitalized patients with SCD and priapism. On the other hand, it is possible that SCD patients are admitted to hospital earlier than other patients with priapism, maybe in anticipation of the failure of nonsurgical intracavernous therapy (aspiration, irrigation, or use of adrenergic agonists). Thus delayed or less frequent surgery might be anticipated. The difference of 8-10 hours in the time to surgery in the 2 groups may represent the time needed for clinically appropriate attempts at nonsurgical therapy or medical management of other manifestations of SCD, including
78
exchange transfusion. The authors list reluctance to perform shunting penile surgery in a young population as a possible explanation to differences. Although many factors affect health care equity, disparities in health care based on socioeconomic and racial factors still remain both in the United States and Europe.7,8 In addition, even after controlling for race, discrimination against SCD patients has been shown to exist in some treatment settings.9 In the present study, SCD patients were more often black and more likely to have Medicaid insurance. According to the authors, race may have played a role, but even after controlling for race, SCD was still an independent and clinically significant indicator of different hospital resource use. The fundamental question is whether differences in health care delivery are in fact disparities in care. Analyses such as those performed by these authors are essential in identifying areas for improvement in the care of priapism patients. Sandip Prasad, MD, PhD, MSc, Terhi Hermanson, MD, PhD, MSc, Charles L. Bennett, MD, PhD, and A. Oliver Sartor, MD, South Carolina College of Pharmacy, Columbia, SC
References 1. Siegel R, DeSantis C, Virgo K, et al. Cancer treatment and survivorship statistics, 2012. CA Cancer J Clin. 2012;62:220-241. 2. Joice GA, Kates M, Sopko NA, et al. Sickle cell disease in priapism: disparity in care? Urology. 2015;86:72-79. 3. Miranda J, McGuire TG, Williams DR, et al. Mental health in the context of health disparities. Am J Psychiatry. 2008;165:1102-1108. 4. Salonia A, Eardley I, Giuliano F, et al. European Association of Urology guidelines on priapism. Eur Urol. 2014;65:480. 5. American Urological Association. Guideline on the Management of Priapism. 2003. 6. Roghmann F, Becker A, Sammon JD, et al. Incidence of priapism in emergency departments in the United States. J Urol. 2013;190:1275. 7. National Healthcare Disparities Report 2013. Rockville, MD: U.S. Department of Health and Human Services, Agency for Healthcare Research and Quality; 2014. 8. Health Inequalities in the EU. Final Report of a Consortium. European Commission, Directorate-General for Health and Consumers; 2013. 9. Hassell K, Pace B, Wang W, et al; American Society of Pediatric Hematology Oncology. Sickle cell disease summit: from clinical and research disparity to action. Am J Hematol. 2009;84:39-45.
http://dx.doi.org/10.1016/j.urology.2015.01.051 UROLOGY 86: 78, 2015. 2015 Elsevier Inc.
REPLY We appreciate the authors’1 comments and their contribution to the discussion on sickle cell disease (SCD) and priapism. We would like to take this opportunity to clarify our study and discuss key points regarding current guidelines and the Nationwide Inpatient Sample (NIS). The present study discusses both the European Association of Urology and the International Consultation on Sexual Medicine (ICSM) guidelines, and the American Urological Association guidelines published in 2003 were intentionally not included because they represent outdated nomenclature and do not represent current standard of care. Recent discoveries in the treatment and outcomes of priapism have changed the priority and emphasis of treatments.2 The authors of the editorial note
UROLOGY 86 (1), 2015
that guidelines recommend conservative treatments followed by surgery if conservative measures fail. However, these same guidelines specifically point out that SCD patients should be treated similarly to non-SCD patients. The ICSM guidelines emphasize and clearly point out that “systemic therapy alone is not effective management of SCD priapism.”3 Recent evidence calls into question the efficacy of systemic therapy for priapism in SCD.4 However, what the present study shows is a trend toward more systemic treatments (eg, transfusions) in SCD and more directed therapy (eg, surgery) in non-SCD patients. The authors further suggest that the observed delay in surgery may be due to an appropriate time to attempt nonsurgical therapy or medically manage other aspects of SCD. However, as we discussed in the present study, any delay in definitive treatment can lead to long-term erectile dysfunction.2,5,6 Both the ICSM and European Association of Urology guidelines specifically point out that in SCD patients, systemic therapies can be attempted in conjunction with directed therapy but should not delay more definitive treatments.3,7 Our study highlights that current management of priapism in SCD patients appears to significantly deviate from these guidelines. The authors point out that the present study “does not analyze treatment outcomes, which is the final determinant of quality of care.” As discussed in the article, a limitation of the NIS is the lack of longitudinal patient data preventing this type of analysis. Although we agree that analyzing long-term outcomes is important and would amplify the study, it is important to highlight other strengths of the NIS. As discussed, priapism is generally treated in the outpatient study, and thus, admission and surgery for priapism is a rare event that requires large databases to be studied. We agree with the authors that further analyses are critical to defining the disparity between SCD and non-SCD patients treated for priapism. We hope that by further defining differences in health care delivery between SCD and non-SCD patients,
UROLOGY 86 (1), 2015
future institutional studies can look at the effects on long-term erectile function outcomes. Finally, we should be focusing our basic and clinical science efforts on prevention and education of SCD-associated stuttering priapism that predisposes SCD patients to acute priapic events. We need to move the entire sexual medicine field to prevention of the sequelae of priapism, and SCD patients should be a priority population for these endeavors. Gregory A. Joice, M.D., Max Kates, M.D., and Trinity J. Bivalacqua, M.D., Ph.D., James Buchanan Brady Urological Institute and Department of Urology, Johns Hopkins School of Medicine, Baltimore, MD
References 1. Prasad S, Hermanson T, Bennett CL, Sartor AO. Sickle cell disease in priapism: disparity in care? [Editorial comment]. Urology. 2015;86: 72-79. 2. Burnett AL, Bivalacqua TJ. Priapism: new concepts in medical and surgical management. Urol Clin North Am. 2011;38:185. 3. Broderick GA. Priapism and sickle-cell anemia: diagnosis and nonsurgical therapy. J Sex Med. 2012;9:88. 4. Merritt AL, Haiman C, Henderson SO. Myth: blood transfusion is effective for sickle cell anemia-associated priapism. CJEM. 2006;8: 119. 5. Joice GA, Kates M, Sopko NA, et al. Sickle cell disease in priapism: disparity in care? Urology. 2015;86:72-79. 6. Olujohungbe A, Burnett AL. How I manage priapism due to sickle cell disease. Br J Haematol. 2013;160:754. 7. Salonia A, Eardley I, Giuliano F, et al. European Association of Urology guidelines on priapism. Eur Urol. 2014;65:480.
http://dx.doi.org/10.1016/j.urology.2015.01.052 UROLOGY 86: 78e79, 2015. 2015 Elsevier Inc.
79
RESULTS
CONCLUSION
To determine the effect of sickle cell disease (SCD) on hospital resource use among patients admitted for priapism. Using the Nationwide Inpatient Sample, a weighted sample of 12,547 patients was selected with a primary diagnosis of priapism from 2002 to 2011. Baseline differences for patient demographics and hospital characteristics were compared between SCD and non-SCD patients. Multivariate analysis was performed to identify the effect of SCD on length of stay, use of penile operations, blood transfusion, and cost. The proportion of SCD patients was 21.5%. SCD patients were younger, more often black, more likely to have Medicaid insurance, and treated more frequently in Southern urban teaching hospitals. SCD was a significant predictor of having a blood transfusion (odds ratio [OR], 16.3; P <.001), and an elongated length of stay (OR, 1.42; P <.001). SCD was associated with less penile operations (OR, 0.40; P <.001). When SCD patients did have an operation, it was performed later in the admission (mean, 0.87 vs 0.47 days; P <.001). SCD was not a significant predictor of increased cost (OR, 1.02; P ¼ .869). SCD patients represent a demographically distinct subgroup of priapism patients with different patterns of resource use manifested by longer hospital stays and more blood transfusions. Moreover, despite evidence that immediate treatment of priapism results in improved erectile function outcomes, SCD patients had less surgical procedures for alleviation of acute priapism events. UROLOGY 86: 72e79, 2015. 2015 Elsevier Inc.
P
riapism represents one of the most common acute emergencies in urology, affecting nearly 10,000 men annually with nearly 30% of these patients going on to be admitted to the hospital.1 However, the hospital course and inpatient costs of treating priapism are unknown. Approximately 1 in 5 patients presenting to an emergency room in the United States with priapism have sickle cell disease (SCD).1,2 Immediate nonsurgical and surgical interventions are necessary to preserve erectile function. Ischemic priapism is a well-known complication of SCD and has been reported to have a lifetime prevalence of approximately 28%-35%.3,4 Historically, SCD patients were treated conservatively with oxygenation, hydration, and blood transfusions; however, we now have a better understanding of the pathophysiology of SCD-associated priapism, and new guidelines recommend immediate interventions to preserve erectile function.5 Patients with
SCD encounter the health care system frequently, with 68 admissions per 100 patients annually and an estimated $1.5 million in annual inpatient charges per 100 patients.6 Prior studies have demonstrated that among patients presenting to emergency departments with priapism, SCD is an independent predictor of hospital admission.2 We hypothesize that SCD and non-SCD patients with priapism are distinct with regard to their hospital resource use and may represent a disparity in care. The present study compares patient demographics, hospital characteristics, and inpatient outcomes between SCD and non-SCD patients. The primary purpose of the study is to assess associations between SCD and inpatient costs and treatment patterns of patients admitted with priapism in the United States.
MATERIALS AND METHODS Data Sources and Patient Selection
Financial Disclosure: The authors declare that they have no relevant financial interests. From the James Buchanan Brady Urological Institute and Department of Urology, Johns Hopkins School of Medicine, Baltimore, MD Address correspondence to: Gregory A. Joice, M.D., James Buchanan Brady Urological Institute and Department of Urology, Johns Hopkins School of Medicine, 1800 Orleans Street, Marburg 134, Baltimore, MD 21287. E-mail: [email protected] Submitted: October 17, 2014, accepted (with revisions): January 23, 2015
72
ª 2015 Elsevier Inc. All Rights Reserved
The Nationwide Inpatient Sample (NIS) is the largest all-payer inpatient database in the United States, consisting of an annual 20% stratified sample of community and academic hospitals. The NIS contains data elements for each discharge including diagnoses and procedures along with patient demographics and hospital characteristics. The recent update of the NIS through http://dx.doi.org/10.1016/j.urology.2015.01.050 0090-4295/15
2011 includes data from approximately 1000 hospitals over 45 states.7 Patients were selected from the 2002-2011 NIS with a primary diagnosis of priapism (“International Classification of Disease, Ninth Revision, Clinical Modification [ICD-9-CM] diagnosis code 607.3). A subset of patients was identified who had a concomitant diagnosis of SCD (ICD-9-CM codes 282.6x). The data were weighted using the NIS0 weighting methodology to create national estimates.7
Independent Variables Patient demographics included age, sex, race, payer, median household income, and comorbidities. Patient comorbidities were represented by the Charlson comorbidity index (CCI) and calculated using the Charlson-Deyo adaption for administrative data to create the CCI within the data set.8,9 To account for missing values in the multivariate analysis, the race variable was simplified into white, black, Hispanic, other, and missing. For demographic comparisons, the missing values were excluded when reporting percentages. Hospital characteristics included the number of hospital beds, teaching status, hospital location, and hospital region.
Main Outcome Measures The primary outcomes were penile operation, transfusion, length of stay (LOS), and total hospitalization cost. Penile operation included penile incision, placement of cavernosal shunts, and insertion of prostheses (ICD-9-CM procedure codes 64.92, 64.95, 64.97, and 64.98). Patients who received a transfusion were similarly identified (ICD-9-CM procedure codes 99.0x). The number of days from admission that the patient received either a transfusion or operation was also analyzed. LOS was dichotomized into a new variable, elongated LOS, defined as the top 90th percentile of LOS within the sample. The NIS includes data on total charges for each hospitalization. However, this variable represents the amount the hospitals billed for the service rather than the actual cost spent during hospitalization.10 Supplemental files can be requested that include a cost-to-charge ratio for each hospital allowing conversion of total charges to total cost. The cost-to-charge ratio was used to create a total cost variable for each patient and was further dichotomized to a new variable, increased total cost, defined as the top 90th percentile.
Statistical Analysis Baseline differences in demographics and outcomes were compared using the chi-square test for categorical variables and the Mann-Whitney U test for continuous variables. The effects of independent variables and SCD were assessed using multivariate logistic regression models to produce odds ratios (ORs). Penile operation was included as an independent variable in the transfusion model, and both penile operation and transfusion were included in the LOS and total cost models. All multivariate regression analyses were adjusted for clustering using the generalizing estimating equations method.11-14 All analyses were performed using SAS (version 9.2; SAS Institute, Cary, NC) with a 2-sided significance level of P <.05.
the NIS from 2002 to 2011. The patient demographics and hospital characteristics are listed in Table 1. Overall, among patients admitted to the hospital for priapism, 21.5% had a concomitant diagnosis of SCD. Patients with SCD compared to those without SCD were younger (20 years median age vs 42 years; P <.001), more often black (96.0% vs 36.7%; P <.001), in low median income brackets (47.0% vs 35.8%; P <.001), with Medicaid insurance (49.8% vs 18.0%; P <.001), and healthier (CCI of 0: 83.5% vs 77.0%; P <.001). SCD patients were more likely to receive their care at urban (97.5% vs 92.6%; P <.001), teaching (74.8% vs 58.0%; P <.001) hospitals in the Southern region (51.3% vs 42.5%; P <.001). The outcome variables are listed in Table 2. The overall rate of penile operations in the sample was 68.3%, but SCD patients were significantly less likely to receive operations (42.4% vs 75.4%; P <.001). In addition, SCD patients on average had their operations performed later during their admission (0.87 days vs 0.47 days; P ¼ .01). SCD patients were more likely to receive a blood transfusion during their hospitalization (37.6% vs 2.8%; P <.001) and on average had their transfusions earlier (0.9 days vs 1.77 days; P <.001). In addition, SCD patients had a longer mean LOS (3.38 vs 2.69 days; P <.001), but there was no statistically significant difference between mean total cost ($5899 vs $5698; P ¼ .89). Similar analysis comparing African American patients to noneAfrican American patients (See Supplementary Table) revealed fewer operations (72.1% vs 79.1%; P <.001) and more transfusions (2.2% vs 4.7%; P <.001). However, African American patients did not experience any difference in time to operation, time to transfusion, or increased LOS. On multivariate analysis, SCD patients were less likely to receive a penile operation (OR, 0.40; P <.001) after controlling for independent variables (Table 3). These patients were also more likely to receive a blood transfusion (OR, 16.3; P <.001) and have an elongated LOS (OR, 1.42; P ¼ .001). However, there was no significant difference in increased total cost between SCD and nonSCD patients (OR, 1.02; P ¼ .869). The largest relative differences between these patient populations were penile operation and transfusion rates, where SCD patients were 33% less likely to undergo a penile procedure while being 35% more likely to have a blood transfusion (Fig. 1A). In patients who did not undergo any penile procedure, SCD patients had a higher average total cost by $882, whereas those who underwent a penile procedure had an average increase of $1479 (Fig. 1B). However, neither of these differences was statistically significant in multivariate models including only those patients either with or without a penile operation.
COMMENT RESULTS Overall, a nationally weighted sample of 12,547 patients with a primary diagnosis of priapism was identified within UROLOGY 86 (1), 2015
The current analysis demonstrates that priapism is a costly disease with significant disparity in management of patients with SCD. Patients with SCD experience far 73
Table 1. Patient demographics and hospital characteristics Characteristic
Sickle Cell
Both
P Value*
(78.5)
2694 (21.5)
12,547 (100.0)
N/A
(18.2) (40.5) (41.3)
2247 (83.4) 394 (14.6) 53 (1.9)
4041 (32.2) 4386 (35.0) 4120 (32.8)
<.001
(49.1) (36.7) (10.5) (3.7)
14 2120 35 39
3935 5050 872 331
<.001
(35.8) (28.3) (36.0)
1218 (47.0) 655 (25.3) 717 (27.7)
4587 (38.2) 3319 (27.6) 4108 (34.2)
(18.6) (18.0) (30.6) (32.9)
246 1333 698 398
2066 3099 3693 3619
(77.0) (15.6) (7.4)
2250 (83.5) 358 (13.3) 86 (3.2)
9832 (78.4) 1897 (15.1) 818 (6.5)
(7.4) (92.6)
66 (2.5) 2606 (97.5)
793 (6.4) 11,665 (93.6)
<.001
(42.0) (58.0)
674 (25.2) 1998 (74.8)
4784 (38.4) 7675 (61.6)
<.001
(7.1) (25.7) (67.2)
222 (8.3) 632 (23.6) 1818 (68.0)
919 (7.4) 3147 (25.3) 8392 (67.4)
(21.3) (17.9) (42.5) (18.2)
547 566 1383 198
2649 2333 5575 1990
Nonsickle Cell
All, N (%) 9853 Age, n (%) <30 1794 30-44 3992 >45 4067 Race, n (%)y White 3921 Black 2929 Hispanic 838 Other 291 Median income, n (%) Low 3369 Medium 2664 High 3391 Payer, n (%) Medicare 1819 Medicaid 1766 Private 2995 Other 3221 CCI, n (%) 0 7582 1 1539 2þ 732 Hospital location, n (%) Rural 727 Urban 9059 Hospital academic status, n (%) Nonteaching 4110 Teaching 5676 Hospital bed size, n (%) Small 697 Medium 2515 Large 6574 Hospital region, n (%) Northeast 2102 Midwest 1767 South 4192 West 1792
(0.6) (96.0) (1.6) (1.8)
(9.2) (49.8) (26.1) (14.9)
(20.3) (21.0) (51.3) (7.4)
(38.6) (49.6) (8.6) (3.2)
(16.6) (24.8) (29.6) (29.0)
(21.1) (18.6) (44.4) (15.9)
<.001
<.001
<.001
.023
<.001
CCI, Charlson comorbidity index; N/A, not applicable. * P values are derived from chi-square tests for categorical variables. y Missing values for race were excluded when calculating percentages.
fewer operative interventions (OR, 0.40; P <.001), and when they do receive an operation, it is with a significant delay to surgery when compared to the non-SCD cohort despite the known benefits of immediate treatment of acute priapism on long-term erectile function rates. Whether this constitutes a disparity in care, reluctance to perform shunting penile surgery in a young population, or an unmeasured difference in their disease course remains undetermined. The International Consultation on Sexual Medicine and the European Association of Urology have published clinical practice guidelines for treatment of priapism.5,15,16 Within the guidelines, they give specific treatment considerations for patients with SCD. For the general population, they suggest a stepwise approach beginning with decompression of the corpora cavernosa with either aspiration or injection of sympathomimetic drugs followed by surgical intervention.16 They further suggest that patients with SCD should undergo the same initial treatment plan and that supportive measures 74
including intravenous hydration, oxygen administration, and blood exchange transfusion should be considered but should not be the primary treatment. However, specifically they note that these additional treatments “should not delay initial treatment.”16 Despite these guidelines, the present study suggests that SCD patients are less likely to receive any form of penile operation and thus have a delay in the treatment of their priapic episodes. In addition, we showed that patients with SCD on average receive penile operations 0.4 day later in their hospitalization than patients without SCD. Given that extended delay in treatment of ischemic priapism leads to erectile dysfunction, this finding may represent a potential health care quality disparity for SCD patients.17-19 Our study also revealed an alarmingly high absolute rate of transfusions among SCD patients, who are transfused during 38% of admissions for priapism. This high transfusion rate among the SCD cohort does not appear to be supported by the evidence. Blood transfusion has UROLOGY 86 (1), 2015
Table 2. Inpatient outcomes Outcome
Nonsickle Cell
Penile operation, n (%) No Yes Day to operation, mean (95% CI) Transfusions, n (%) No Yes Day to transfusion, mean (95% CI) LOS, mean (95% CI) Total cost, $, mean (95% CI)
2426 (24.6) 7427 (75.4) 0.47 (0.40-0.52)
1552 (57.6) 1141 (42.4) 0.87 (0.62-1.15)
9575 278 1.77 2.69 5698
1680 1014 0.90 3.38 5899
(97.2) (2.8) (1.22-2.33) (2.56-2.82) (5312-6085)
Both
P Value*
3978 (31.7) 8569 (68.3) 0.52 (0.46-0.59)
<.001
Sickle Cell
(62.4) (37.6) (0.67-1.13) (3.09-3.66) (5315-6484)
11,254 1292 1.09 2.83 5744
(89.7) (10.3) (0.87-1.30) (2.71-2.95) (5417-6070)
.01 <.001 <.001 <.001 .891
CI, confidence interval; LOS, length of stay. * P values are derived from chi-square tests for categorical variables and Mann-Whitney U test for continuous variables.
Table 3. Multivariate outcome models for SCD Outcome Penile operation Transfusion Elongated LOS Increased total cost
Odds Ratio (95% CI)* 0.40 16.3 1.42 1.02
(0.35-0.45) (12.9-20.6) (1.15-1.75) (0.81-1.29)
P Value <.001 <.001 .001 .869
SCD, sickle cell disease; other abbreviations as in Table 2. * Odds ratios from generalized estimating equations multivariate models controlling for patient demographics and hospital characteristics.
been proposed as therapy for priapism in SCD patients because of the mechanism that transfusion can lead to increased oxygen delivery to tissues and prevent tissue damage. As mentioned, the International Consultation on Sexual Medicine and European Association of Urology guidelines recommend blood transfusion only as a supplemental therapy for priapism.16 However, additional studies question the efficacy of blood transfusion in this setting.17 Merritt et al20 published a study of a series of case reports to evaluate the efficacy of blood transfusions in SCD patients with chronic priapism. When compared to patients receiving conventional therapy, the group receiving blood transfusions experienced increased mean time to detumescence (11 vs 8 days). Among the blood transfusion cohort, 9 of 26 patients who received blood transfusions developed neurologic sequelae ranging from mild and temporary to severe and permanent. Association of SCD, priapism, exchange transfusions, and neurologic events (ASPEN) is a known complication of blood transfusions in SCD patients that could account for the neurologic decline observed in this study.21,22 Although there have been no prospective studies looking at the efficacy and risks of blood transfusions in priapism, the current literature appears to lack evidence to support its routine use as a primary therapy. These previous studies highlight the concerning finding of our study that SCD patients with priapism are frequently receiving transfusions. This high rate of transfusions may be exposing SCD patients to a higher risk of complications with no benefit on erectile function preservation. However, within the NIS, we cannot specifically determine whether a blood transfusion was used as a primary or supplemental UROLOGY 86 (1), 2015
therapy. Future studies should look at the timing between transfusion and additional therapies and the incidence of ASPEN syndrome in these patients. Potential disparities in care are exaggerated by the contrasting demographic cohorts of the SCD and nonSCD groups. SCD patients are more likely to have Medicaid insurance and be treated at urban, teaching hospitals in the South. This finding is significant because these hospitals will bear a more significant cost burden of treating these publically insured patients. Health care use and cost is important to consider in SCD patients given the chronicity of the disease and the need for multiple hospitalizations. A recent study estimated that SCD patients average approximately 2.5 acute care visits per year, representing a relatively high health care use subgroup of patients.23 Dinan et al24 showed that among patients undergoing cholecystectomy or hip replacement, SCD was an independent predictor of increased hospital cost. They further go on to propose that this increase in cost is primarily because of the increased LOS in SCD patients. In the present study, we did show an approximate 25% increase in LOS for sickle cell patients but did not show any significant increase in total costs, likely owing to the decreased use of operative intervention. On further analysis, we demonstrated that among priapism patients with nonoperative management, the cost for SCD was 24% higher and for those who did have an operation the increase was 23%. However, neither was a significant difference on multivariate analysis. This finding suggests that the increased LOS in the SCD population causes an increase in cost that is partially offset by the increased rate of penile operations in the general population. In addition, blood transfusion (OR, 5.27; P < .001) was a strong predictor of increased cost in the model, suggesting that the high transfusion rate among SCD patients may also be an important driver of increased costs for patients. As a result of more SCD patients being on Medicaid insurance plans, their increased total cost may have a greater relative effect on hospital reimbursements. In addition, SCD patients are significantly more likely to be African American, which may influence decisions on surgery and transfusions. We indeed did show that within the non-SCD cohort, 75
Figure 1. (A) Differences in outcomes between SCD and non-SCD patients. SCD patients had higher transfusion rates (þ35%), lower rates of penile operations (33%), and longer LOS (þ0.7 days), which were all statistically significant on multivariate analysis. SCD patients had slightly higher cost (þ$201), but this difference was not statistically significant. (B) Graph of total hospital cost by SCD and penile operation. SCD patients had a higher average cost for both patients receiving penile operations (þ$1479) and those not receiving an operation (þ$882). However, these increases were not statistically significant on multivariate analysis within each of the subgroups. LOS, length of stay; SCD, sickle cell disease. (Color version available online.)
African American patients are less likely to undergo penile operations and more likely to have blood transfusions. However, these differences were far less clinically significant than similar differences observed between SCD and non-SCD patients. Although race may play a role in choice of priapism interventions, SCD still appears to be an independent and clinically significant indicator of different hospital resource use. 76
Few recent population-based studies have analyzed the experience for SCD patients presenting with priapism. Roghmann et al2 looked within the Nationwide Emergency Department Sample for predictors of admission for priapism. They showed that among several patient demographics and hospital characteristics, SCD was an independent predictor of admission for priapism (OR, 2.22; P <.001). Stein et al1 also looked within the UROLOGY 86 (1), 2015
Nationwide Emergency Department Sample to show that 21.1% of patients with a diagnosis of priapism had concurrent SCD, but they did not further analyze what effect this had on their treatment or admission. Chrouser et al25 studied an inpatient population with priapism using the NIS. They showed that SCD patients were less likely to undergo a penile operation compared to non-SCD patients. However, they did not further assess the effect of SCD on other outcomes including LOS, total cost, and transfusion. The limitations of the present study relate to the use of administrative data. These data are inherently susceptible to interinstitution variability in coding practices that could cause inaccuracies in the database. However, despite this limitation, previous studies have shown accuracy rates as high as 80% in correctly identifying conditions with administrative data.26,27 In addition, it is difficult to isolate whether a patient with SCD is only having an isolated case of priapism or if they also have a concomitant painful occlusive crisis. We have attempted to decrease these confounding variables by only including patients with a primary diagnosis of priapism. Finally, the NIS does not contain longitudinal data on individual patients, and thus, tracking outcomes such as erectile dysfunction or readmissions is not possible. In addition, the effect of potential readmissions and recurrent transfusions cannot be included in the total cost model. Future studies should analyze the long-term outcomes and treatment sequelae of patients undergoing different treatment modalities.
CONCLUSION Recent guidelines for the treatment of acute priapism recommend that SCD patients should be treated similarly to the general population with early intervention to achieve detumescence. Despite these guidelines, we show that patients with SCD admitted for priapism are treated less often with operative interventions, and when they are, their surgery is delayed 2-fold compared to others. In addition, SCD patients are transfused more than 1 in every 3 admissions despite a paucity of data supporting its efficacy in priapism. These findings highlight a significant treatment disparity for SCD patients that can potentially lead to more complications and worse erectile function outcomes. References 1. Stein DM, Flum AS, Cashy J, et al. Nationwide emergency department visits for priapism in the United States. J Sex Med. 2013; 10:2418. 2. Roghmann F, Becker A, Sammon JD, et al. Incidence of priapism in emergency departments in the United States. J Urol. 2013;190: 1275. 3. Adeyoju AB, Olujohungbe AB, Morris J, et al. Priapism in sicklecell disease; incidence, risk factors and complications—an international multicentre study. BJU Int. 2002;90:898. 4. Mantadakis E, Cavender JD, Rogers ZR, et al. Prevalence of priapism in children and adolescents with sickle cell anemia. J Pediatr Hematol Oncol. 1999;21:518.
UROLOGY 86 (1), 2015
5. Broderick GA. Priapism and sickle-cell anemia: diagnosis and nonsurgical therapy. J Sex Med. 2012;9:88. 6. Lanzkron S, Carroll CP, Haywood C Jr. The burden of emergency department use for sickle-cell disease: an analysis of the national emergency department sample database. Am J Hematol. 2010;85:797. 7. 2011 Introduction to the NIS. Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality; 2013; Available at: http://www.hcup-us.ahrq.gov/ nisoverview.jsp, 2013; Accessed January 21, 2014. 8. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45:613. 9. Charlson ME, Ales KL, Pompei P, MacKenzie CR. A new method of classification of prognostic comorbidity for longitudinal studies: development and validation. J Chronic Dis. 1987;40:373. 10. 2011 Cost-to-Charge Ratio. Healthcare Cost and Utilization Project (HCUP). Rockville, MD: Agency for Healthcare Research and Quality; 2013; Available at: http://www.hcup-us.ahrq.gov/db/state/ costtocharge.jsp, 2013; Accessed January 21, 2014. 11. 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. 12. Hu JC, Gold KF, Pashos CL, et al. Temporal trends in radical prostatectomy complications from 1991 to 1998. J Urol. 2003;169:1443. 13. Begg CB, Riedel ER, Bach PB, et al. Variations in morbidity after radical prostatectomy. N Engl J Med. 2002;346:1138. 14. 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. 15. Albersen M, Bivalacqua TJ. Sexual dysfunction: first EAU priapism treatment guidelines published. Nat Rev Urol. 2014;11:132. 16. Salonia A, Eardley I, Giuliano F, et al. European Association of Urology guidelines on priapism. Eur Urol. 2014;65:480. 17. Olujohungbe A, Burnett AL. How I manage priapism due to sickle cell disease. Br J Haematol. 2013;160:754. 18. Burnett AL, Bivalacqua TJ. Priapism: new concepts in medical and surgical management. Urol Clin North Am. 2011;38:185. 19. Hinman F Jr. Priapism: reasons for failure of therapy. J Urol. 1960; 83:420. 20. Merritt AL, Haiman C, Henderson SO. Myth: blood transfusion is effective for sickle cell anemia-associated priapism. CJEM. 2006;8:119. 21. Siegel JF, Rich MA, Brock WA. Association of sickle cell disease, priapism, exchange transfusion and neurological events: ASPEN syndrome. J Urol. 1993;150:1480. 22. Jones G, Lee BR, Woods GM, et al. Neurologic complications in pediatric sickle cell patients hospitalized for priapism: an analysis of the Pediatric Health Information System (PHIS) database. Blood. 2013;122:2287. 23. Brousseau DC, Owens PL, Mosso AL, et al. Acute care utilization and rehospitalizations for sickle cell disease. JAMA. 2010;303:1288. 24. Dinan MA, Chou CH, Hammill BG, et al. Outcomes of inpatients with and without sickle cell disease after high-volume surgical procedures. Am J Hematol. 2009;84:703. 25. Chrouser KL, Ajiboye OB, Oyetunji TA, et al. Priapism in the United States: the changing role of sickle cell disease. Am J Surg. 2011;201:468. 26. Lawthers AG, McCarthy EP, Davis RB, et al. Identification of inhospital complications from claims data. Is it valid? Med Care. 2000;38:785. 27. Burns EM, Rigby E, Mamidanna R, et al. Systematic review of discharge coding accuracy. J Public Health. 2012;34:138-148.
APPENDIX SUPPLEMENTARY DATA
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.urology. 2015.01.050. 77
EDITORIAL COMMENT The Institute of Medicine defines health care disparities as “differences in the quality of health care that are not due to access-related factors or clinical needs, preferences, and appropriateness of intervention.”1 Joice et al2 hypothesize that a preexisting diagnosis of sickle cell disease (SCD) may lead to health care disparities because of biases, prejudices, or stereotypes held by providers that influence recommended clinical care.3 The authors found that SCD patients admitted for priapism are treated less often with surgery, which is often delayed compared to men without SCD with priapism. In addition, the authors identified a 38% transfusion rate for SCD patients who were admitted. The basic issue is whether this represents a health care disparity or rather readily explainable practice patterns. Both American and European guidelines recommend a stepwise approach to the treatment of priapism: First-line treatments are highly recommended before any surgical treatment of ischemic priapism. Surgery should be considered only when conservative management options fail, and surgical treatment is also recommended for priapism events lasting over 72 hours.4,5 Study patients were all hospitalized, and as the authors note, management of priapism occurs 70% of the time in the outpatient setting (eg, the patient presents to the emergency department with a priapism and is managed with intracavernous treatment including aspiration or irrigation with successful detumescence). Therefore, the National Inpatient Sample may lead to biased estimates of resource allocation associated with priapism between the 2 study populations, as most interactions (and interventions) with the health care system are not captured in the Nationwide Inpatient Sample. According to the authors, SCD patients represent a distinct subgroup of priapism when treated in hospitals. This observation may hold true also in outpatient care, and different outpatient treatment of SCD patients may explain differences found in hospitalized patients. For example, because of education about risks of priapism, SCD patients may be more likely to present for care more quickly than non-SCD patients, where embarrassment or lack of awareness of the consequences to erectile function may delay clinical presentation in non-SCD patients. The study seems to assume that conservative treatments had already been tried and failed and therefore, surgical intervention would be needed. Without information on prior outpatient treatment, this assumption cannot be validated. The study did not analyze treatment outcomes, which is the final determinant of quality of care. SCD is an independent predictor of hospital admission for priapism.6 SCD patients who present with priapism alone and without sickle cell crisis are often managed in the emergency room successfully; SCD patients with priapism and additional acute manifestations of SCD typically require inpatient admission to manage SCD clinical sequelae along with priapism. This may explain increased rates of transfusion in hospitalized patients with SCD and priapism. On the other hand, it is possible that SCD patients are admitted to hospital earlier than other patients with priapism, maybe in anticipation of the failure of nonsurgical intracavernous therapy (aspiration, irrigation, or use of adrenergic agonists). Thus delayed or less frequent surgery might be anticipated. The difference of 8-10 hours in the time to surgery in the 2 groups may represent the time needed for clinically appropriate attempts at nonsurgical therapy or medical management of other manifestations of SCD, including
78
exchange transfusion. The authors list reluctance to perform shunting penile surgery in a young population as a possible explanation to differences. Although many factors affect health care equity, disparities in health care based on socioeconomic and racial factors still remain both in the United States and Europe.7,8 In addition, even after controlling for race, discrimination against SCD patients has been shown to exist in some treatment settings.9 In the present study, SCD patients were more often black and more likely to have Medicaid insurance. According to the authors, race may have played a role, but even after controlling for race, SCD was still an independent and clinically significant indicator of different hospital resource use. The fundamental question is whether differences in health care delivery are in fact disparities in care. Analyses such as those performed by these authors are essential in identifying areas for improvement in the care of priapism patients. Sandip Prasad, MD, PhD, MSc, Terhi Hermanson, MD, PhD, MSc, Charles L. Bennett, MD, PhD, and A. Oliver Sartor, MD, South Carolina College of Pharmacy, Columbia, SC
References 1. Siegel R, DeSantis C, Virgo K, et al. Cancer treatment and survivorship statistics, 2012. CA Cancer J Clin. 2012;62:220-241. 2. Joice GA, Kates M, Sopko NA, et al. Sickle cell disease in priapism: disparity in care? Urology. 2015;86:72-79. 3. Miranda J, McGuire TG, Williams DR, et al. Mental health in the context of health disparities. Am J Psychiatry. 2008;165:1102-1108. 4. Salonia A, Eardley I, Giuliano F, et al. European Association of Urology guidelines on priapism. Eur Urol. 2014;65:480. 5. American Urological Association. Guideline on the Management of Priapism. 2003. 6. Roghmann F, Becker A, Sammon JD, et al. Incidence of priapism in emergency departments in the United States. J Urol. 2013;190:1275. 7. National Healthcare Disparities Report 2013. Rockville, MD: U.S. Department of Health and Human Services, Agency for Healthcare Research and Quality; 2014. 8. Health Inequalities in the EU. Final Report of a Consortium. European Commission, Directorate-General for Health and Consumers; 2013. 9. Hassell K, Pace B, Wang W, et al; American Society of Pediatric Hematology Oncology. Sickle cell disease summit: from clinical and research disparity to action. Am J Hematol. 2009;84:39-45.
http://dx.doi.org/10.1016/j.urology.2015.01.051 UROLOGY 86: 78, 2015. 2015 Elsevier Inc.
REPLY We appreciate the authors’1 comments and their contribution to the discussion on sickle cell disease (SCD) and priapism. We would like to take this opportunity to clarify our study and discuss key points regarding current guidelines and the Nationwide Inpatient Sample (NIS). The present study discusses both the European Association of Urology and the International Consultation on Sexual Medicine (ICSM) guidelines, and the American Urological Association guidelines published in 2003 were intentionally not included because they represent outdated nomenclature and do not represent current standard of care. Recent discoveries in the treatment and outcomes of priapism have changed the priority and emphasis of treatments.2 The authors of the editorial note
UROLOGY 86 (1), 2015
that guidelines recommend conservative treatments followed by surgery if conservative measures fail. However, these same guidelines specifically point out that SCD patients should be treated similarly to non-SCD patients. The ICSM guidelines emphasize and clearly point out that “systemic therapy alone is not effective management of SCD priapism.”3 Recent evidence calls into question the efficacy of systemic therapy for priapism in SCD.4 However, what the present study shows is a trend toward more systemic treatments (eg, transfusions) in SCD and more directed therapy (eg, surgery) in non-SCD patients. The authors further suggest that the observed delay in surgery may be due to an appropriate time to attempt nonsurgical therapy or medically manage other aspects of SCD. However, as we discussed in the present study, any delay in definitive treatment can lead to long-term erectile dysfunction.2,5,6 Both the ICSM and European Association of Urology guidelines specifically point out that in SCD patients, systemic therapies can be attempted in conjunction with directed therapy but should not delay more definitive treatments.3,7 Our study highlights that current management of priapism in SCD patients appears to significantly deviate from these guidelines. The authors point out that the present study “does not analyze treatment outcomes, which is the final determinant of quality of care.” As discussed in the article, a limitation of the NIS is the lack of longitudinal patient data preventing this type of analysis. Although we agree that analyzing long-term outcomes is important and would amplify the study, it is important to highlight other strengths of the NIS. As discussed, priapism is generally treated in the outpatient study, and thus, admission and surgery for priapism is a rare event that requires large databases to be studied. We agree with the authors that further analyses are critical to defining the disparity between SCD and non-SCD patients treated for priapism. We hope that by further defining differences in health care delivery between SCD and non-SCD patients,
UROLOGY 86 (1), 2015
future institutional studies can look at the effects on long-term erectile function outcomes. Finally, we should be focusing our basic and clinical science efforts on prevention and education of SCD-associated stuttering priapism that predisposes SCD patients to acute priapic events. We need to move the entire sexual medicine field to prevention of the sequelae of priapism, and SCD patients should be a priority population for these endeavors. Gregory A. Joice, M.D., Max Kates, M.D., and Trinity J. Bivalacqua, M.D., Ph.D., James Buchanan Brady Urological Institute and Department of Urology, Johns Hopkins School of Medicine, Baltimore, MD
References 1. Prasad S, Hermanson T, Bennett CL, Sartor AO. Sickle cell disease in priapism: disparity in care? [Editorial comment]. Urology. 2015;86: 72-79. 2. Burnett AL, Bivalacqua TJ. Priapism: new concepts in medical and surgical management. Urol Clin North Am. 2011;38:185. 3. Broderick GA. Priapism and sickle-cell anemia: diagnosis and nonsurgical therapy. J Sex Med. 2012;9:88. 4. Merritt AL, Haiman C, Henderson SO. Myth: blood transfusion is effective for sickle cell anemia-associated priapism. CJEM. 2006;8: 119. 5. Joice GA, Kates M, Sopko NA, et al. Sickle cell disease in priapism: disparity in care? Urology. 2015;86:72-79. 6. Olujohungbe A, Burnett AL. How I manage priapism due to sickle cell disease. Br J Haematol. 2013;160:754. 7. Salonia A, Eardley I, Giuliano F, et al. European Association of Urology guidelines on priapism. Eur Urol. 2014;65:480.
http://dx.doi.org/10.1016/j.urology.2015.01.052 UROLOGY 86: 78e79, 2015. 2015 Elsevier Inc.
79