- Email: [email protected]
Racial disparities in medicaid patients after brain tumor surgery
Journal of Clinical Neuroscience 20 (2013) 57–61
Contents lists available at SciVerse ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Racial disparities in medicaid patients after brain tumor surgery Debraj Mukherjee a,⇑, , Chirag G. Patil a, , Nathan Todnem b, Beatrice Ugiliweneza b, Miriam Nuño a, Michael Kinsman b, Shivanand P. Lad c, Maxwell Boakye b a
Maxine Dunitz Neurosurgical Institute, Department of Neurosurgery, Cedars-Sinai Medical Center, 8631 West 3rd Street, Suite 800E, Los Angeles, CA 90048, USA Department of Neurosurgery, University of Louisville, Louisville, KY, USA c Division of Neurosurgery, Department of Surgery, Duke University Medical Center, Durham, NC, USA b
a r t i c l e
i n f o
Article history: Received 12 March 2012 Accepted 6 May 2012
Keywords: Brain tumor surgery Complications Medicaid Mortality Racial disparities
a b s t r a c t The presence of healthcare-related disparities is an ongoing, widespread, and well-documented societal and health policy issue. We investigated the presence of racial disparities among post-operative patients either with meningioma or malignant, benign, or metastatic brain tumors. We used the Medicaid component of the Thomson Reuter’s MarketScan database from 2000 to 2009. Univariate and multivariate analysis assessed death, 30-day post-operative risk of complications, length of stay, and total charges. We identified 2321 patients, 73.7% were Caucasian, 57.8% were women; with Charlson comorbidity scores of <3 (56.2%) and treated at low-volume centers (73.4%). Among all, 26.3% of patients were of African-American ethnicity and 22.1% had meningiomas. Mortality was 2.0%, mean length of stay (LOS) was 9 days, mean total charges were US$42,422, an adverse discharge occurred in 22.5% of patients, and overall 30-day complication rate was 23.4%. In a multivariate analysis, African-American patients with meningiomas had higher odds of developing a 30-day complication (p = 0.05) and were significantly more likely to have longer LOS (p < 0.001) and greater total charges (p < 0.001) relative to Caucasian counterparts. The presence of one post-operative complication doubled LOS and nearly doubled total charges, while the presence of two post-operative complications tripled these outcomes. Patients of African-American ethnicity had significantly higher post-operative complications than those of Caucasian ethnicity. This higher rate of complications seems to have driven greater healthcare utilization, including greater LOS and total charges, among African-American patients. Interventions aimed at reducing complications among African-American patients with brain tumor may help reduce post-operative disparities. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Since the Institute of Medicine’s ‘‘Crossing the Quality Chasm’’ was published in 2001, equitable delivery has remained a key aspect in defining quality healthcare.1 However, the presence of healthcare-related disparities between different racial groups within the USA remains an ongoing, widespread, and welldocumented societal and health policy issue.2,3 Disparate access and outcomes after treatment of numerous acute and chronic medical conditions, inclusive of cancer, have been shown consistently over the past several decades, even after accounting for differences in geographic location, patient educational level, and other potential determinants of health and behavior.4,5 With the publication of ‘‘Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care’’ by the Institute of Medicine ⇑ Corresponding author. Tel.: +1 310 729 1247; fax: +1 310 423 0810.
E-mail address: [email protected] (D. Mukherjee). These authors contributed equally to the manuscript.
0967-5868/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jocn.2012.05.014
in 2003, the prevalence of such disparities began to reach a higher level of attention within the realm of health services research.6 As a result, projects within the general medical literature focusing upon disparities grew over the subsequent decade, with newer reports suggesting patient mistrust as well as provider bias may have both played roles in the disparate outcomes observed among otherwise similar groups treated equally.7,8 Until recently, relatively few studies have evaluated disparities in short-term and long-term outcomes within the subspeciality of neurosurgery.9–13 The few reports that have attempted to explore racial disparities in post-operative brain tumor patient outcomes have reported that patients of African-American ethnicity diagnosed with malignant brain tumors have shorter survival,9,10 more severe disease at presentation,13 and are treated more often at lowvolume hospitals11,12 compared to their Caucasian counterparts. However, most of these studies attempt to control for important determinants of health including socioeconomic status using suboptimal regional surrogates, such as median income within the patient’s postal code or county of residence. In an attempt to mitigate
58
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61
potentially confounding variables such as socioeconomic status, we utilized the Medicaid portion of the MarketScan database to evaluate racial disparities in a homogeneous, low-income population of patients undergoing craniotomy for tumor resection.
craniotomy center as a center performing more than 50 craniotomies per year based upon the report by Long et al.16
2. Methods
Primary outcomes studied included inpatient death, length of stay (LOS), total charges, adverse discharge disposition, and incidence of 30-day post-operative complications. The incidence of 30-day complications was assessed at any post-operative hospital admission using ICD-9 diagnosis codes. All other outcomes were inherent data elements contained within the Medicaid dataset. All charges were inflated to 2009 dollars using the medical care component of the consumer price index, accessible from the Bureau of Labor Statistics. Complications included renal, cardiac, neural, venous thromboembolic, pulmonary, infectious and wound complications (Supplementary Table 1) and occurred from the time of initial hospitalization to 30-days following discharge.
2.1. Database Thomson Reuter’s MarketScan database from 2000 to 2009 was used for analysis. The MarketScan database is a collection of six different data files, including the: (i) Commercial Claims and Encounters file, (ii) Medicare Supplemental and Coordination of Benefits (COB) file, (iii) Benefit Plan Design file, (iv) Health and Productivity Management file, (v) Medicaid file, and (vi) Lab file. For the current project, we used only the Medicaid database, which is formed by health-care use information for Medicaid recipients. The Medicaid database contains information generated upon inpatient admission as well as data regarding the facility of treatment and all services provided as an inpatient and outpatient, inclusive of pharmaceutical claims and enrollment information. The MarketScan database is a de-identified database that was deemed as exempt from review by the University of Louisville Institutional Review Board. 2.2. Patient selection The MarketScan database was queried for all inpatient admissions with an International Classification of Disease (ICD)-9 primary diagnosis of meningioma or benign, malignant, or metastatic brain tumor as well as a primary procedure code (with either International ICD-9 or Current Procedural Terminology [CPT]-4 coding) consistent with resection of brain tumor. If the primary procedure column was empty, then the secondary procedure column was searched and procedures found in this column were considered to be a primary procedure. All cases in which any combination of ICD-9 or CPT-4 diagnosis and procedural codes shown in Table 1 were achieved were included. All patients under 18 years of age were excluded. 2.3. Covariates Patient race, gender, and age at the time of the initial hospitalization were inherent elements gathered from the database. Race was categorized as either Caucasian or African-American given that an adequate number of other races were not available in the dataset. Gender was a binary variable while age was a continuous variable. Comorbidities were tallied and used to calculate a Charlson comorbidity index score for each patient. The Charlson score is a standardized and validated 10-point weighted measure of a patient’s overall comorbidity.14,15 We defined a high-volume
Table 1 International Classification of Disease (ICD-9) and Current Procedural Terminology (CPT-4) diagnostic and procedural coding Diagnosis
ICD-9-CM diagnosis codes
ICD-9-CM procedure codes
CPT-4 codes
Metastatic tumor Meningioma Benign Malignant Uncertain behavior Malignant brain tumor Benign brain tumor
198.3
01.59 01.51
61510 61512
01.13, 10.14, 01.53, 01.59 01.13, 10.14, 01.53, 01.59
61510
225.2 192.1 237.6 191.0-5, 191.8-9 225.0, 237.5
61510
2.4. Outcomes
2.5. Statistical analysis Data were summarized using means and standard deviations for continuous variables, while using counts and percentages for categorical variables. Chi-squared testing was used to perform unadjusted bivariate analyses for categorical variables, while Mann-Whitney U-testing was used for continuous variables. Inpatient death and 30-day complications were analyzed using multivariate logistic regression. LOS and total charges were analyzed using general linear regression on log-transformed variables. Two-sided tests and p-values <0.05 were considered statistically significant. All statistical analyses were performed using Statistical Analysis Software 9.3 (SAS Institute, Cary, NC, USA). 3. Results Our study identified 2321 patients in the Medicaid database who underwent craniotomy for brain tumor between 2000 and 2009. The mean age of patients was 49 years of age and 57.8% were women. Patients of African-American ethnicity comprised 26.3% (n = 611) and only 26.6% of patients were treated at high-volume hospitals. Charlson comorbidity index was >3 in 43.8% of patients. Overall inpatient mortality was 2.0% with a mean LOS of 9 days and mean total charges of US$42,422. An adverse discharge occurred in 22.5% of patients (that is, they were not discharged home) and the overall 30-day complication rate was 23.4%. The most common tumors were malignant tumors (37.6%), followed by metastases (34.1%), meningiomas (22.2%), and benign tumors (6.1%) (Table 2). The most common inpatient and 30-day complications included pulmonary, wound, thromboembolic, and neurological sequelae (Table 3). 3.1. Bivariate analysis When assessing presenting characteristics in bivariate analysis, age and comorbidities were distributed similarly between Caucasians and African-Americans. However, African-American patients were more likely to be treated at high-volume hospitals than their Caucasian counterparts (33% compared to 24%, p < 0.0001) (Table 2). When assessing the outcomes of African-Americans compared to Caucasian patients, the groups had similar rates of in-hospital mortality and adverse discharge, but those of African-American ethnicity had approximately 4% higher inhospital and 30-day complications, 3-day longer LOS (p < 0.0001), and US$20,000 higher total hospital charges than their Caucasian counterparts (Table 4). Interestingly, in our cohort, the
59
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61 Table 2 Baseline characteristics of patients within the Medicaid Database (2000–2009) with brain tumors who underwent resection
Variable Age, mean (SD) Female gender Charlson index score 0 1 2 P3 High-volume hospital Metastases Meningioma Malignant tumor Benign brain tumor
All (n = 2321) n (%)
White (n = 1710) n (%)
AA (n = 611) n (%)
p-value
49 (14) 1343 (57.86)
49 (14) 950 (55.56)
50 (14) 393 (64.32)
0.1858 0.0002* 0.1936
594 (25.59) 194 (8.36) 516 (22.23) 1017 (43.82) 618 (26.63) 792 (34.12) 515 (22.19) 872 (37.57) 142 (6.12)
429 133 382 766 416 587 313 708 102
165 (27.00) 61 (9.98) 134 (21.93) 251 (41.08) 202 (33.06) 205 (33.55) 202 (33.06) 164 (26.84) 40 (6.55)
(25.09) (7.78) (22.34) (44.80) (24.33) (34.33) (18.30) (41.40) (5.96)
< 0.0001*
AA = African-American, SD = standard deviation. Statistically significant.
*
Table 3 Frequency of inpatient and 30-day complications among patients with brain tumor who underwent resection during 2000 to 2009 Type of complication
Pulmonary Wound DVT/PE Neural Renal Infection Cardiac
15.5%), tripled LOS (up to 21 days), and tripled total charges (up to US$104,649) (Table 6).
Complication rates Inpatient n (%)
30-day n (%)
185 (7.97) 62 (2.67) 58 (2.50) 52 (2.24) 42 (1.81) 26 (1.12) 19 (0.82)
197 (8.49) 93 (4.01) 85 (3.66) 64 (2.76) 44 (1.90) 39 (1.68) 20 (0.86)
DVT = deep vein thrombosis, PE = pulmonary embolism.
African-American patients had no 30-day re-operations, as compared to a 0.7% re-operation rate for Caucasians (Table 4). 3.2. Multivariate analysis Multivariate analyses controlling for age, gender, Charlson score, and hospital-volume status demonstrated that, among all tumor types, African-American patients had 27% higher odds of developing an inpatient post-operative complication compared to Caucasian patients. LOS and total hospital charges were also significantly higher for African-American patients in the multivariate analysis (Table 5). There was no significant association between race and inpatient death or adverse discharges. Analyses demonstrated that, compared to patients with no post-operative complications, one complication increased inpatient mortality by more than six-fold (from 0.8% to 6.0%), LOS by 7 days and total charges by US$26,874. Two post-operative complications increased the mortality rate by more than 15-fold (up to
3.3. Outcomes within the Meningioma subset Patients with meningioma showed the highest level of outcome disparities among the tumor types (other tumor types not shown). In bivariate analyses, African-American patients with meningioma had higher in-hospital mortality compared to Caucasians (3.5% compared to 1.0%), bordering upon statistical significance (p = 0.05). African-American patients with meningioma also had significantly higher 30-day complication rates compared to Caucasians (26.7% compared to 19.1%, p = 0.04). Multivariate analyses demonstrated that African-American patients with meningioma had 58% higher odds of developing a post-operative complication than Caucasians (odds ratio = 1.58, 95% confidence interval) (Table 5). African-American patients with meningioma also had a significantly higher LOS and charges compared to their Caucasian counterparts.
4. Discussion In this study, we demonstrate several significant racial disparities in both inpatient and short-term post-operative outcomes after craniotomy for resection of brain tumors. Among all tumor types, we found in multivariate analyses that African-American patients had significantly higher 30-day complication rates, significantly longer LOS, and accrued significantly greater healthcare related charges. Post-operatively African-American patients with meningioma seemingly had the worst disparities. Increasing LOS and total charges within these groups seemed to be driven primarily by increased complications.
Table 4 Outcomes of all patients with brain tumor within the Medicaid Database (2000–2009) who underwent resection Outcome
White (n = 1710)
AA (n = 611)
p-value
Index hosp. in-hospital death, n (%) 30-day re-operation, n (%) In-hospital complications, n (%) 30-day complications, n (%) Index hosp. discharge home, n (%) Index hosp. LOS, mean (SD) Index hosp. charges (US$), mean (SD)
31 (1.81) 12 (0.70) 243 (14.21) 297 (17.37) 1336 (78.13) 8 (9) $37,853 ($50,769)
16 (2.62) 0 (0.00) 111 (18.17) 129 (21.11) 464 (75.94) 11 (12) $55,209 ($70,636)
0.2248 0.0441* 0.0196* 0.0401* 0.266 <0.0001* <0.0001*
AA = African-American, Hosp. = hospital, LOS = length of stay, SD = standard deviation. * Statistically significant.
60
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61
Table 5 Multivariate analysis of factors associated with increased length of stay, total hospital charges, and 30-day complication rate within the Medicaid Database (2000-2009) Factor
Length of stay combined estimate (p-value)
Race (Ref: White) African1.31 (<0.0001)* American Charlson Index (Ref: 0) 1 1.39 (<0.0001)* 2 1.22 (<0.0001)* 3 1.56 (<0.0001)* Age (per year 1.005 (<0.0001)* increment) Gender (Ref: Male) Female 0.96 (0.2722) Hospital Volume (Ref: Low) High 0.98 (0.5318)
Meningioma estimate (p-value)
Total hospital charges Combined estimate (p-value)
Meningioma estimate (p-value)
Combined estimate (p-value)
Meningioma estimate (p-value)
1.32 (<0.0001)*
1.68 (<0.0001)*
1.85 (<0.0001)*
1.27 (0.0463)*
1.56 (0.0453)*
1.46 (<0.0001)* 1.30 (0.0054)* 1.54 (<0.0001)* 1.003 (0.1188)
1.38 1.39 1.46 0.99
0.91 (0.2422)
0.90 (0.0332)*
0.94 (0.6615)
1.85 (0.1421)
0.56 (0.0168)*
1.07 (0.3751)
1.18 (0.0042)*
1.4 (0.0093)*
0.92 (0.497)
0.89 (0.6194)
(0.0014)* (<0.0001)* (<0.0001)* (0.06)
30-Day complication rate
1.58 1.54 1.40 0.99
(0.0038)* (0.0055)* (0.0565) (0.56)
2.38 1.10 1.45 1.01
(<0.001)* (0.5619) (0.0115)* (0.2108)
2.52 2.01 2.31 1.00
(0.0015)* (0.0186)* (0.0104)* (0.5738)
Ref = reference. * Statistically significant.
Table 6 Bivariate analyses assessing the impact of inpatient complications on Inpatient mortality, length of stay (LOS) and total charges ($US) for brain tumor patients undergoing craniotomy within the Medicaid Database (2000–2009) Outcome
No. complications
All
0 1 2 3+
0.82 6 15.49 15.15
Mortality (%)
Average LOS (days) 0 1 2 3+ Average total charges in 2009 (US$) 0 1 2 3+
7 14 21 41 34,952 61,826 104,649 206,763
While there has been a relative dearth of disparities research in neurosurgical oncology, recent reports by El-Sayed et al.,17 Curry et al.,11 Mukherjee et al.,12 and Dasenbrock et al.18 do help frame the current results. Although our study of 2321 Medicaid patients is smaller in raw number than the studies by Mukherjee et al. and Curry et al. which utilized the National Inpatient Sample, our analysis does represent a more homogenous population of socioeconomically similar patients with equivalent access to care imparted by their Medicaid coverage. Previous authors have attempted to adjust for confounding variables such as socioeconomic status when examining race as a predictor of patient outcomes. However, a majority of these studies have utilized suboptimal regional surrogates, such as median income within the patient’s postal code or county of residence to try to reduce this bias. By including only a homogeneous, low-income population of patients with a relative over-representation of those with African-American ethnicity, we have rigorously analyzed the effect of race on patient outcomes. Among varying covariates and outcomes, the results of our study seem to mirror some basic trends found within previous studies. For instance, a majority of patients (56.18%) had Charlson comorbidity scores of 62, mirroring the relatively healthy population in the Mukherjee et al.12 report, in which the mean Charlson score was 1.47. Furthermore, a 30-day complication rate of 20% was similar to the 30-day complication rate of 25.8% reported by El-Sayed et al.17 The complication rate in El-Sayed et al. held a fur-
ther association with insurance type, with privately insured patients having a 30-day complication rate of only 11.2% while the uninsured had a 30-day complication rate of up to 27.3%. Trends in our data, which appear to demonstrate an association between higher complication rates and worse in-hospital mortality or adverse discharge disposition between Caucasians and AfricanAmerican patients, seem similar to those in other reports. For instance, our data demonstrated a higher inpatient mortality among African-American patients (2.62%) relative to Caucasians (1.81%), which is similar to the results of Curry et al.,10 who showed significantly higher odds of inpatient mortality in those of African-American ethnicity. The degree of outcome disparity was consistently greatest between African-American and Caucasian patients within the meningioma tumor subtype in both our report and that of Curry et al.,11 although the specific mechanism possibly underlying this relationship could not be elucidated with either the Nationwide Inpatient Sample database or in the current MarketScan database study with greater outpatient follow-up data. While research by Claus et al.19 using the Surveillance Epidemiology and End Results (SEER) database may point toward some biological differences between African-American patients with meningioma relative to their Caucasian counterparts, with African-Americans presenting at a significantly younger age, there have been to date no strong scientific data at the genetic or biochemical level to support the hypothesis that the biological characteristics of meningiomas in African-Americans are significantly different to those of their Caucasian counterparts.13 Rather, such outcome differences, particularly post-operatively, may be due to varying social support between these groups. For instance, Wong et al.20 recently identified gaps in supportive care among all patients after meningioma resection. They commented on the lack of a formal support system for patients with benign brain tumors and recommended construction of Internet-based support groups, face-to-face support groups, individual peer support, and extending counseling services for patients with cancer as a means of fostering practical support for such patients post-operatively. Although the Wong et al. report identified no racial disparities within their cohort, the sample included only 29 patients, of which a majority was Caucasian.20 It is possible that these gaps in supportive care may have a larger impact on the African-American population and contribute to the racial disparities seen among patients with meningioma. Our data also indicate that African-American Medicaid patients were more likely to be treated at high-volume centers relative to Caucasian patients. Mukherjee et al.12 reported that AfricanAmerican patients were significantly less likely to be admitted to
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61
a high-volume hospital from 2001 to 2005. However, they reviewed an 18-year period from 1988 to 2005 which demonstrated that, in the late 1980s and early 1990s, African-American patients were more likely to be admitted to high-volume centers, although this was followed by a steady decline over time. The trends in the Mukherjee et al. dataset may reflect the changing healthcare access patterns of a very different patient population, as that study included a majority of privately insured patients who may have chosen to receive their care at less urban, or high volume, centers. One of the strengths of our study was the optimal control for socioeconomic status and level of insurance coverage through the utilization of a Medicaid-specific database. Enrollment in Medicaid is frequently utilized in the literature as a proxy for lowincome because of its usefulness as a socioeconomic status indicator at the individual level.21–24 Excluding data from Medicare, self-pay, non-paying and privately insured patients allowed us to investigate racial disparities in a more socioeconomically homogenous, nationwide, low-income patient population. A primary disadvantage in using Medicaid-specific data is a relatively small total cohort with a larger proportion of African-American patients (26.32%) than in the actual United States population. While we were unable to adjust for acuity of presentation as other studies recently have,18 by including information on patient race within our study, we have addressed a significant weakness of other reports seeking to associate the impact of insurance status upon outcomes within the neurosurgical oncology literature. We demonstrate that racial disparities continue to be a significant public health issue within the neuro-oncology sector despite public and private efforts focused upon curbing these inequalities in care. It has proven to be a challenging task to localize the root cause for ongoing racial disparities due to its probable multifactorial nature. Distinct cultural differences exist between different racial groups, which may influence African-Americans’ pursuit of treatment following a diagnosis of cancer, regardless of physician recommendations.25,26 Others have indicated that these disparities may stem from a distrust of the healthcare system or federal government as well as lack of familial support.7 Other possible important factors may include inadequate health literacy, inadequate access to care, and lack of financial resources within certain racial subgroups. Although the specific mechanisms driving such outcome disparities may not be fully understood, our data demonstrate that African-American Medicaid patients with brain tumors, particularly with meningioma, seem to have higher complication rates that secondarily drive greater healthcare utilization, including greater LOS and total charges. Interventions aimed at reducing complications among this patient population may help reduce post-operative disparities and mitigate excessive healthcare utilization. Although both public and privately funded bodies have spent time and effort over the past decade to bridge the quality chasm and bring parity to the care of racial and ethnic minorities, our data support the theory that such disparities in healthcare continue, affecting individual health and continuing to bring undue economic burden upon our healthcare system. By placing greater emphasis on new alternatives to engaging minorities in appropriate post-operative care and by studying such efforts in rigorous prospective studies, we may see a greater impact of our concerted efforts to reduce disparities and improve care for cancer patients undergoing surgery, including surgery for primary brain tumors. Conflict of interest/disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication.
61
Acknowledgement The authors wish to thank Sherry Brandon for her expert help with copyediting, referencing and administrative support provided on this project. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jocn.2012.05.014. References 1. Committee on Quality of Health Care in America. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: Institute of Medicine; 2001. 2. McGlynn EA, Asch SM, Adams J, et al. The quality of health care delivered to adults in the United States. N Engl J Med 2003;348:2635–45. 3. Asch SM, Kerr EA, Keesey J, et al. Who is at greatest risk for receiving poorquality health care? N Engl J Med 2006;354:1147–56. 4. Morris AM, Rhoads KF, Stain SC, et al. Understanding racial disparities in cancer treatment and outcomes. J Am Coll Surg 2010;211:105–13. 5. Sherwood PR, Dahman BA, Donovan HS, et al. Treatment disparities following the diagnosis of an astrocytoma. J Neurooncol 2011;101:67–74. 6. Smedley BD, Stith AY, Nelson AR, editors. Unequal treatment: confronting racial and ethnic disparities in health care. Washington, DC: Institute of Medicine, The National Academies Press; 2003. 7. Masi CM, Gehlert S. Perceptions of breast cancer treatment among AfricanAmerican women and men: implications for interventions. J Gen Intern Med 2009;24:408–14. 8. Haider AH, Sexton J, Sriram N, et al. Association of unconscious race and social class bias with vignette-based clinical assessments by medical students. JAMA 2011;306:942–51. 9. Barnholtz-Sloan JS, Sloan AE, Schwartz AG. Racial differences in survival after diagnosis with primary malignant brain tumor. Cancer 2003;98:603–9. 10. Robertson JT, Gunter BC, Somes GW. Racial differences in the incidence of gliomas: a retrospective study from Memphis Tennessee. Br J Neurosurg 2002;16:562–6. 11. Curry WT, Carter BS, Barker IInd FG. Racial, ethnic, and socioeconomic disparities in patient outcomes after craniotomy for tumor in adult patients in the United States, 1988-2004. Neurosurgery 2010;66:427–37 [discussion 37– 38]. 12. Mukherjee D, Zaidi HA, Kosztowski T, et al. Disparities in access to neurooncologic care in the United States. Arch Surg 2010;145:247–53. 13. Curry Jr WT, Barker IInd FG. Racial, ethnic and socioeconomic disparities in the treatment of brain tumors. J Neurooncol 2009;93:25–39. 14. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83. 15. 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. 16. Long DM, Gordon T, Bowman H, et al. Outcome and cost of craniotomy performed to treat tumors in regional academic referral centers. Neurosurgery 2003;52:1056–63 [discussion 63–65]. 17. El-Sayed AM, Ziewacz JE, Davis MC, et al. Insurance status and inequalities in outcomes after neurosurgery. World Neurosurg 2011;76:459–66. 18. Dasenbrock HH, Wolinsky JP, Sciubba DM, et al. The impact of insurance status on outcomes after surgery for spinal metastases. Cancer 2012;118:4833–41. 19. Claus EB, Bondy ML, Schildkraut JM, et al. Epidemiology of intracranial meningioma. Neurosurgery 2005;57:1088–95 [discussion 95]. 20. Wong J, Mendelsohn D, Nyhof-Young J, et al. A qualitative assessment of the supportive care and resource needs of patients undergoing craniotomy for benign brain tumours. Support Care Cancer 2011;19:1841–8. 21. Bradley CJ, Given CW, Roberts C. Race, socioeconomic status, and breast cancer treatment and survival. J Natl Cancer Inst 2002;94:490–6. 22. Pittard IIIrd WB, Laditka JN, Laditka SB. Associations between maternal age and infant health outcomes among Medicaid-insured infants in South Carolina: mediating effects of socioeconomic factors. Pediatrics 2008;122:e100–6. 23. Sherwood PR, Stommel M, Murman DL, et al. Primary malignant brain tumor incidence and Medicaid enrollment. Neurology 2004;62:1788–93. 24. Downie DL, Schmid D, Plescia MG, et al. Racial disparities in blood pressure control and treatment differences in a Medicaid population, North Carolina, 2005–2006. Prev Chronic Dis 2011;8:A55. 25. Kim SH, Ferrante J, Won BR, et al. Barriers to adequate follow-up during adjuvant therapy may be important factors in the worse outcome for Black women after breast cancer treatment. World J Surg Oncol 2008;6:26. 26. Morris AM, Billingsley KG, Hayanga AJ. Residual treatment disparities after oncology referral for rectal cancer. J Natl Cancer Inst 2008;100:738–44.
Contents lists available at SciVerse ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Racial disparities in medicaid patients after brain tumor surgery Debraj Mukherjee a,⇑, , Chirag G. Patil a, , Nathan Todnem b, Beatrice Ugiliweneza b, Miriam Nuño a, Michael Kinsman b, Shivanand P. Lad c, Maxwell Boakye b a
Maxine Dunitz Neurosurgical Institute, Department of Neurosurgery, Cedars-Sinai Medical Center, 8631 West 3rd Street, Suite 800E, Los Angeles, CA 90048, USA Department of Neurosurgery, University of Louisville, Louisville, KY, USA c Division of Neurosurgery, Department of Surgery, Duke University Medical Center, Durham, NC, USA b
a r t i c l e
i n f o
Article history: Received 12 March 2012 Accepted 6 May 2012
Keywords: Brain tumor surgery Complications Medicaid Mortality Racial disparities
a b s t r a c t The presence of healthcare-related disparities is an ongoing, widespread, and well-documented societal and health policy issue. We investigated the presence of racial disparities among post-operative patients either with meningioma or malignant, benign, or metastatic brain tumors. We used the Medicaid component of the Thomson Reuter’s MarketScan database from 2000 to 2009. Univariate and multivariate analysis assessed death, 30-day post-operative risk of complications, length of stay, and total charges. We identified 2321 patients, 73.7% were Caucasian, 57.8% were women; with Charlson comorbidity scores of <3 (56.2%) and treated at low-volume centers (73.4%). Among all, 26.3% of patients were of African-American ethnicity and 22.1% had meningiomas. Mortality was 2.0%, mean length of stay (LOS) was 9 days, mean total charges were US$42,422, an adverse discharge occurred in 22.5% of patients, and overall 30-day complication rate was 23.4%. In a multivariate analysis, African-American patients with meningiomas had higher odds of developing a 30-day complication (p = 0.05) and were significantly more likely to have longer LOS (p < 0.001) and greater total charges (p < 0.001) relative to Caucasian counterparts. The presence of one post-operative complication doubled LOS and nearly doubled total charges, while the presence of two post-operative complications tripled these outcomes. Patients of African-American ethnicity had significantly higher post-operative complications than those of Caucasian ethnicity. This higher rate of complications seems to have driven greater healthcare utilization, including greater LOS and total charges, among African-American patients. Interventions aimed at reducing complications among African-American patients with brain tumor may help reduce post-operative disparities. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Since the Institute of Medicine’s ‘‘Crossing the Quality Chasm’’ was published in 2001, equitable delivery has remained a key aspect in defining quality healthcare.1 However, the presence of healthcare-related disparities between different racial groups within the USA remains an ongoing, widespread, and welldocumented societal and health policy issue.2,3 Disparate access and outcomes after treatment of numerous acute and chronic medical conditions, inclusive of cancer, have been shown consistently over the past several decades, even after accounting for differences in geographic location, patient educational level, and other potential determinants of health and behavior.4,5 With the publication of ‘‘Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care’’ by the Institute of Medicine ⇑ Corresponding author. Tel.: +1 310 729 1247; fax: +1 310 423 0810.
E-mail address: [email protected] (D. Mukherjee). These authors contributed equally to the manuscript.
0967-5868/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jocn.2012.05.014
in 2003, the prevalence of such disparities began to reach a higher level of attention within the realm of health services research.6 As a result, projects within the general medical literature focusing upon disparities grew over the subsequent decade, with newer reports suggesting patient mistrust as well as provider bias may have both played roles in the disparate outcomes observed among otherwise similar groups treated equally.7,8 Until recently, relatively few studies have evaluated disparities in short-term and long-term outcomes within the subspeciality of neurosurgery.9–13 The few reports that have attempted to explore racial disparities in post-operative brain tumor patient outcomes have reported that patients of African-American ethnicity diagnosed with malignant brain tumors have shorter survival,9,10 more severe disease at presentation,13 and are treated more often at lowvolume hospitals11,12 compared to their Caucasian counterparts. However, most of these studies attempt to control for important determinants of health including socioeconomic status using suboptimal regional surrogates, such as median income within the patient’s postal code or county of residence. In an attempt to mitigate
58
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61
potentially confounding variables such as socioeconomic status, we utilized the Medicaid portion of the MarketScan database to evaluate racial disparities in a homogeneous, low-income population of patients undergoing craniotomy for tumor resection.
craniotomy center as a center performing more than 50 craniotomies per year based upon the report by Long et al.16
2. Methods
Primary outcomes studied included inpatient death, length of stay (LOS), total charges, adverse discharge disposition, and incidence of 30-day post-operative complications. The incidence of 30-day complications was assessed at any post-operative hospital admission using ICD-9 diagnosis codes. All other outcomes were inherent data elements contained within the Medicaid dataset. All charges were inflated to 2009 dollars using the medical care component of the consumer price index, accessible from the Bureau of Labor Statistics. Complications included renal, cardiac, neural, venous thromboembolic, pulmonary, infectious and wound complications (Supplementary Table 1) and occurred from the time of initial hospitalization to 30-days following discharge.
2.1. Database Thomson Reuter’s MarketScan database from 2000 to 2009 was used for analysis. The MarketScan database is a collection of six different data files, including the: (i) Commercial Claims and Encounters file, (ii) Medicare Supplemental and Coordination of Benefits (COB) file, (iii) Benefit Plan Design file, (iv) Health and Productivity Management file, (v) Medicaid file, and (vi) Lab file. For the current project, we used only the Medicaid database, which is formed by health-care use information for Medicaid recipients. The Medicaid database contains information generated upon inpatient admission as well as data regarding the facility of treatment and all services provided as an inpatient and outpatient, inclusive of pharmaceutical claims and enrollment information. The MarketScan database is a de-identified database that was deemed as exempt from review by the University of Louisville Institutional Review Board. 2.2. Patient selection The MarketScan database was queried for all inpatient admissions with an International Classification of Disease (ICD)-9 primary diagnosis of meningioma or benign, malignant, or metastatic brain tumor as well as a primary procedure code (with either International ICD-9 or Current Procedural Terminology [CPT]-4 coding) consistent with resection of brain tumor. If the primary procedure column was empty, then the secondary procedure column was searched and procedures found in this column were considered to be a primary procedure. All cases in which any combination of ICD-9 or CPT-4 diagnosis and procedural codes shown in Table 1 were achieved were included. All patients under 18 years of age were excluded. 2.3. Covariates Patient race, gender, and age at the time of the initial hospitalization were inherent elements gathered from the database. Race was categorized as either Caucasian or African-American given that an adequate number of other races were not available in the dataset. Gender was a binary variable while age was a continuous variable. Comorbidities were tallied and used to calculate a Charlson comorbidity index score for each patient. The Charlson score is a standardized and validated 10-point weighted measure of a patient’s overall comorbidity.14,15 We defined a high-volume
Table 1 International Classification of Disease (ICD-9) and Current Procedural Terminology (CPT-4) diagnostic and procedural coding Diagnosis
ICD-9-CM diagnosis codes
ICD-9-CM procedure codes
CPT-4 codes
Metastatic tumor Meningioma Benign Malignant Uncertain behavior Malignant brain tumor Benign brain tumor
198.3
01.59 01.51
61510 61512
01.13, 10.14, 01.53, 01.59 01.13, 10.14, 01.53, 01.59
61510
225.2 192.1 237.6 191.0-5, 191.8-9 225.0, 237.5
61510
2.4. Outcomes
2.5. Statistical analysis Data were summarized using means and standard deviations for continuous variables, while using counts and percentages for categorical variables. Chi-squared testing was used to perform unadjusted bivariate analyses for categorical variables, while Mann-Whitney U-testing was used for continuous variables. Inpatient death and 30-day complications were analyzed using multivariate logistic regression. LOS and total charges were analyzed using general linear regression on log-transformed variables. Two-sided tests and p-values <0.05 were considered statistically significant. All statistical analyses were performed using Statistical Analysis Software 9.3 (SAS Institute, Cary, NC, USA). 3. Results Our study identified 2321 patients in the Medicaid database who underwent craniotomy for brain tumor between 2000 and 2009. The mean age of patients was 49 years of age and 57.8% were women. Patients of African-American ethnicity comprised 26.3% (n = 611) and only 26.6% of patients were treated at high-volume hospitals. Charlson comorbidity index was >3 in 43.8% of patients. Overall inpatient mortality was 2.0% with a mean LOS of 9 days and mean total charges of US$42,422. An adverse discharge occurred in 22.5% of patients (that is, they were not discharged home) and the overall 30-day complication rate was 23.4%. The most common tumors were malignant tumors (37.6%), followed by metastases (34.1%), meningiomas (22.2%), and benign tumors (6.1%) (Table 2). The most common inpatient and 30-day complications included pulmonary, wound, thromboembolic, and neurological sequelae (Table 3). 3.1. Bivariate analysis When assessing presenting characteristics in bivariate analysis, age and comorbidities were distributed similarly between Caucasians and African-Americans. However, African-American patients were more likely to be treated at high-volume hospitals than their Caucasian counterparts (33% compared to 24%, p < 0.0001) (Table 2). When assessing the outcomes of African-Americans compared to Caucasian patients, the groups had similar rates of in-hospital mortality and adverse discharge, but those of African-American ethnicity had approximately 4% higher inhospital and 30-day complications, 3-day longer LOS (p < 0.0001), and US$20,000 higher total hospital charges than their Caucasian counterparts (Table 4). Interestingly, in our cohort, the
59
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61 Table 2 Baseline characteristics of patients within the Medicaid Database (2000–2009) with brain tumors who underwent resection
Variable Age, mean (SD) Female gender Charlson index score 0 1 2 P3 High-volume hospital Metastases Meningioma Malignant tumor Benign brain tumor
All (n = 2321) n (%)
White (n = 1710) n (%)
AA (n = 611) n (%)
p-value
49 (14) 1343 (57.86)
49 (14) 950 (55.56)
50 (14) 393 (64.32)
0.1858 0.0002* 0.1936
594 (25.59) 194 (8.36) 516 (22.23) 1017 (43.82) 618 (26.63) 792 (34.12) 515 (22.19) 872 (37.57) 142 (6.12)
429 133 382 766 416 587 313 708 102
165 (27.00) 61 (9.98) 134 (21.93) 251 (41.08) 202 (33.06) 205 (33.55) 202 (33.06) 164 (26.84) 40 (6.55)
(25.09) (7.78) (22.34) (44.80) (24.33) (34.33) (18.30) (41.40) (5.96)
< 0.0001*
AA = African-American, SD = standard deviation. Statistically significant.
*
Table 3 Frequency of inpatient and 30-day complications among patients with brain tumor who underwent resection during 2000 to 2009 Type of complication
Pulmonary Wound DVT/PE Neural Renal Infection Cardiac
15.5%), tripled LOS (up to 21 days), and tripled total charges (up to US$104,649) (Table 6).
Complication rates Inpatient n (%)
30-day n (%)
185 (7.97) 62 (2.67) 58 (2.50) 52 (2.24) 42 (1.81) 26 (1.12) 19 (0.82)
197 (8.49) 93 (4.01) 85 (3.66) 64 (2.76) 44 (1.90) 39 (1.68) 20 (0.86)
DVT = deep vein thrombosis, PE = pulmonary embolism.
African-American patients had no 30-day re-operations, as compared to a 0.7% re-operation rate for Caucasians (Table 4). 3.2. Multivariate analysis Multivariate analyses controlling for age, gender, Charlson score, and hospital-volume status demonstrated that, among all tumor types, African-American patients had 27% higher odds of developing an inpatient post-operative complication compared to Caucasian patients. LOS and total hospital charges were also significantly higher for African-American patients in the multivariate analysis (Table 5). There was no significant association between race and inpatient death or adverse discharges. Analyses demonstrated that, compared to patients with no post-operative complications, one complication increased inpatient mortality by more than six-fold (from 0.8% to 6.0%), LOS by 7 days and total charges by US$26,874. Two post-operative complications increased the mortality rate by more than 15-fold (up to
3.3. Outcomes within the Meningioma subset Patients with meningioma showed the highest level of outcome disparities among the tumor types (other tumor types not shown). In bivariate analyses, African-American patients with meningioma had higher in-hospital mortality compared to Caucasians (3.5% compared to 1.0%), bordering upon statistical significance (p = 0.05). African-American patients with meningioma also had significantly higher 30-day complication rates compared to Caucasians (26.7% compared to 19.1%, p = 0.04). Multivariate analyses demonstrated that African-American patients with meningioma had 58% higher odds of developing a post-operative complication than Caucasians (odds ratio = 1.58, 95% confidence interval) (Table 5). African-American patients with meningioma also had a significantly higher LOS and charges compared to their Caucasian counterparts.
4. Discussion In this study, we demonstrate several significant racial disparities in both inpatient and short-term post-operative outcomes after craniotomy for resection of brain tumors. Among all tumor types, we found in multivariate analyses that African-American patients had significantly higher 30-day complication rates, significantly longer LOS, and accrued significantly greater healthcare related charges. Post-operatively African-American patients with meningioma seemingly had the worst disparities. Increasing LOS and total charges within these groups seemed to be driven primarily by increased complications.
Table 4 Outcomes of all patients with brain tumor within the Medicaid Database (2000–2009) who underwent resection Outcome
White (n = 1710)
AA (n = 611)
p-value
Index hosp. in-hospital death, n (%) 30-day re-operation, n (%) In-hospital complications, n (%) 30-day complications, n (%) Index hosp. discharge home, n (%) Index hosp. LOS, mean (SD) Index hosp. charges (US$), mean (SD)
31 (1.81) 12 (0.70) 243 (14.21) 297 (17.37) 1336 (78.13) 8 (9) $37,853 ($50,769)
16 (2.62) 0 (0.00) 111 (18.17) 129 (21.11) 464 (75.94) 11 (12) $55,209 ($70,636)
0.2248 0.0441* 0.0196* 0.0401* 0.266 <0.0001* <0.0001*
AA = African-American, Hosp. = hospital, LOS = length of stay, SD = standard deviation. * Statistically significant.
60
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61
Table 5 Multivariate analysis of factors associated with increased length of stay, total hospital charges, and 30-day complication rate within the Medicaid Database (2000-2009) Factor
Length of stay combined estimate (p-value)
Race (Ref: White) African1.31 (<0.0001)* American Charlson Index (Ref: 0) 1 1.39 (<0.0001)* 2 1.22 (<0.0001)* 3 1.56 (<0.0001)* Age (per year 1.005 (<0.0001)* increment) Gender (Ref: Male) Female 0.96 (0.2722) Hospital Volume (Ref: Low) High 0.98 (0.5318)
Meningioma estimate (p-value)
Total hospital charges Combined estimate (p-value)
Meningioma estimate (p-value)
Combined estimate (p-value)
Meningioma estimate (p-value)
1.32 (<0.0001)*
1.68 (<0.0001)*
1.85 (<0.0001)*
1.27 (0.0463)*
1.56 (0.0453)*
1.46 (<0.0001)* 1.30 (0.0054)* 1.54 (<0.0001)* 1.003 (0.1188)
1.38 1.39 1.46 0.99
0.91 (0.2422)
0.90 (0.0332)*
0.94 (0.6615)
1.85 (0.1421)
0.56 (0.0168)*
1.07 (0.3751)
1.18 (0.0042)*
1.4 (0.0093)*
0.92 (0.497)
0.89 (0.6194)
(0.0014)* (<0.0001)* (<0.0001)* (0.06)
30-Day complication rate
1.58 1.54 1.40 0.99
(0.0038)* (0.0055)* (0.0565) (0.56)
2.38 1.10 1.45 1.01
(<0.001)* (0.5619) (0.0115)* (0.2108)
2.52 2.01 2.31 1.00
(0.0015)* (0.0186)* (0.0104)* (0.5738)
Ref = reference. * Statistically significant.
Table 6 Bivariate analyses assessing the impact of inpatient complications on Inpatient mortality, length of stay (LOS) and total charges ($US) for brain tumor patients undergoing craniotomy within the Medicaid Database (2000–2009) Outcome
No. complications
All
0 1 2 3+
0.82 6 15.49 15.15
Mortality (%)
Average LOS (days) 0 1 2 3+ Average total charges in 2009 (US$) 0 1 2 3+
7 14 21 41 34,952 61,826 104,649 206,763
While there has been a relative dearth of disparities research in neurosurgical oncology, recent reports by El-Sayed et al.,17 Curry et al.,11 Mukherjee et al.,12 and Dasenbrock et al.18 do help frame the current results. Although our study of 2321 Medicaid patients is smaller in raw number than the studies by Mukherjee et al. and Curry et al. which utilized the National Inpatient Sample, our analysis does represent a more homogenous population of socioeconomically similar patients with equivalent access to care imparted by their Medicaid coverage. Previous authors have attempted to adjust for confounding variables such as socioeconomic status when examining race as a predictor of patient outcomes. However, a majority of these studies have utilized suboptimal regional surrogates, such as median income within the patient’s postal code or county of residence to try to reduce this bias. By including only a homogeneous, low-income population of patients with a relative over-representation of those with African-American ethnicity, we have rigorously analyzed the effect of race on patient outcomes. Among varying covariates and outcomes, the results of our study seem to mirror some basic trends found within previous studies. For instance, a majority of patients (56.18%) had Charlson comorbidity scores of 62, mirroring the relatively healthy population in the Mukherjee et al.12 report, in which the mean Charlson score was 1.47. Furthermore, a 30-day complication rate of 20% was similar to the 30-day complication rate of 25.8% reported by El-Sayed et al.17 The complication rate in El-Sayed et al. held a fur-
ther association with insurance type, with privately insured patients having a 30-day complication rate of only 11.2% while the uninsured had a 30-day complication rate of up to 27.3%. Trends in our data, which appear to demonstrate an association between higher complication rates and worse in-hospital mortality or adverse discharge disposition between Caucasians and AfricanAmerican patients, seem similar to those in other reports. For instance, our data demonstrated a higher inpatient mortality among African-American patients (2.62%) relative to Caucasians (1.81%), which is similar to the results of Curry et al.,10 who showed significantly higher odds of inpatient mortality in those of African-American ethnicity. The degree of outcome disparity was consistently greatest between African-American and Caucasian patients within the meningioma tumor subtype in both our report and that of Curry et al.,11 although the specific mechanism possibly underlying this relationship could not be elucidated with either the Nationwide Inpatient Sample database or in the current MarketScan database study with greater outpatient follow-up data. While research by Claus et al.19 using the Surveillance Epidemiology and End Results (SEER) database may point toward some biological differences between African-American patients with meningioma relative to their Caucasian counterparts, with African-Americans presenting at a significantly younger age, there have been to date no strong scientific data at the genetic or biochemical level to support the hypothesis that the biological characteristics of meningiomas in African-Americans are significantly different to those of their Caucasian counterparts.13 Rather, such outcome differences, particularly post-operatively, may be due to varying social support between these groups. For instance, Wong et al.20 recently identified gaps in supportive care among all patients after meningioma resection. They commented on the lack of a formal support system for patients with benign brain tumors and recommended construction of Internet-based support groups, face-to-face support groups, individual peer support, and extending counseling services for patients with cancer as a means of fostering practical support for such patients post-operatively. Although the Wong et al. report identified no racial disparities within their cohort, the sample included only 29 patients, of which a majority was Caucasian.20 It is possible that these gaps in supportive care may have a larger impact on the African-American population and contribute to the racial disparities seen among patients with meningioma. Our data also indicate that African-American Medicaid patients were more likely to be treated at high-volume centers relative to Caucasian patients. Mukherjee et al.12 reported that AfricanAmerican patients were significantly less likely to be admitted to
D. Mukherjee et al. / Journal of Clinical Neuroscience 20 (2013) 57–61
a high-volume hospital from 2001 to 2005. However, they reviewed an 18-year period from 1988 to 2005 which demonstrated that, in the late 1980s and early 1990s, African-American patients were more likely to be admitted to high-volume centers, although this was followed by a steady decline over time. The trends in the Mukherjee et al. dataset may reflect the changing healthcare access patterns of a very different patient population, as that study included a majority of privately insured patients who may have chosen to receive their care at less urban, or high volume, centers. One of the strengths of our study was the optimal control for socioeconomic status and level of insurance coverage through the utilization of a Medicaid-specific database. Enrollment in Medicaid is frequently utilized in the literature as a proxy for lowincome because of its usefulness as a socioeconomic status indicator at the individual level.21–24 Excluding data from Medicare, self-pay, non-paying and privately insured patients allowed us to investigate racial disparities in a more socioeconomically homogenous, nationwide, low-income patient population. A primary disadvantage in using Medicaid-specific data is a relatively small total cohort with a larger proportion of African-American patients (26.32%) than in the actual United States population. While we were unable to adjust for acuity of presentation as other studies recently have,18 by including information on patient race within our study, we have addressed a significant weakness of other reports seeking to associate the impact of insurance status upon outcomes within the neurosurgical oncology literature. We demonstrate that racial disparities continue to be a significant public health issue within the neuro-oncology sector despite public and private efforts focused upon curbing these inequalities in care. It has proven to be a challenging task to localize the root cause for ongoing racial disparities due to its probable multifactorial nature. Distinct cultural differences exist between different racial groups, which may influence African-Americans’ pursuit of treatment following a diagnosis of cancer, regardless of physician recommendations.25,26 Others have indicated that these disparities may stem from a distrust of the healthcare system or federal government as well as lack of familial support.7 Other possible important factors may include inadequate health literacy, inadequate access to care, and lack of financial resources within certain racial subgroups. Although the specific mechanisms driving such outcome disparities may not be fully understood, our data demonstrate that African-American Medicaid patients with brain tumors, particularly with meningioma, seem to have higher complication rates that secondarily drive greater healthcare utilization, including greater LOS and total charges. Interventions aimed at reducing complications among this patient population may help reduce post-operative disparities and mitigate excessive healthcare utilization. Although both public and privately funded bodies have spent time and effort over the past decade to bridge the quality chasm and bring parity to the care of racial and ethnic minorities, our data support the theory that such disparities in healthcare continue, affecting individual health and continuing to bring undue economic burden upon our healthcare system. By placing greater emphasis on new alternatives to engaging minorities in appropriate post-operative care and by studying such efforts in rigorous prospective studies, we may see a greater impact of our concerted efforts to reduce disparities and improve care for cancer patients undergoing surgery, including surgery for primary brain tumors. Conflict of interest/disclosures The authors declare that they have no financial or other conflicts of interest in relation to this research and its publication.
61
Acknowledgement The authors wish to thank Sherry Brandon for her expert help with copyediting, referencing and administrative support provided on this project. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jocn.2012.05.014. References 1. Committee on Quality of Health Care in America. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC: Institute of Medicine; 2001. 2. McGlynn EA, Asch SM, Adams J, et al. The quality of health care delivered to adults in the United States. N Engl J Med 2003;348:2635–45. 3. Asch SM, Kerr EA, Keesey J, et al. Who is at greatest risk for receiving poorquality health care? N Engl J Med 2006;354:1147–56. 4. Morris AM, Rhoads KF, Stain SC, et al. Understanding racial disparities in cancer treatment and outcomes. J Am Coll Surg 2010;211:105–13. 5. Sherwood PR, Dahman BA, Donovan HS, et al. Treatment disparities following the diagnosis of an astrocytoma. J Neurooncol 2011;101:67–74. 6. Smedley BD, Stith AY, Nelson AR, editors. Unequal treatment: confronting racial and ethnic disparities in health care. Washington, DC: Institute of Medicine, The National Academies Press; 2003. 7. Masi CM, Gehlert S. Perceptions of breast cancer treatment among AfricanAmerican women and men: implications for interventions. J Gen Intern Med 2009;24:408–14. 8. Haider AH, Sexton J, Sriram N, et al. Association of unconscious race and social class bias with vignette-based clinical assessments by medical students. JAMA 2011;306:942–51. 9. Barnholtz-Sloan JS, Sloan AE, Schwartz AG. Racial differences in survival after diagnosis with primary malignant brain tumor. Cancer 2003;98:603–9. 10. Robertson JT, Gunter BC, Somes GW. Racial differences in the incidence of gliomas: a retrospective study from Memphis Tennessee. Br J Neurosurg 2002;16:562–6. 11. Curry WT, Carter BS, Barker IInd FG. Racial, ethnic, and socioeconomic disparities in patient outcomes after craniotomy for tumor in adult patients in the United States, 1988-2004. Neurosurgery 2010;66:427–37 [discussion 37– 38]. 12. Mukherjee D, Zaidi HA, Kosztowski T, et al. Disparities in access to neurooncologic care in the United States. Arch Surg 2010;145:247–53. 13. Curry Jr WT, Barker IInd FG. Racial, ethnic and socioeconomic disparities in the treatment of brain tumors. J Neurooncol 2009;93:25–39. 14. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–83. 15. 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. 16. Long DM, Gordon T, Bowman H, et al. Outcome and cost of craniotomy performed to treat tumors in regional academic referral centers. Neurosurgery 2003;52:1056–63 [discussion 63–65]. 17. El-Sayed AM, Ziewacz JE, Davis MC, et al. Insurance status and inequalities in outcomes after neurosurgery. World Neurosurg 2011;76:459–66. 18. Dasenbrock HH, Wolinsky JP, Sciubba DM, et al. The impact of insurance status on outcomes after surgery for spinal metastases. Cancer 2012;118:4833–41. 19. Claus EB, Bondy ML, Schildkraut JM, et al. Epidemiology of intracranial meningioma. Neurosurgery 2005;57:1088–95 [discussion 95]. 20. Wong J, Mendelsohn D, Nyhof-Young J, et al. A qualitative assessment of the supportive care and resource needs of patients undergoing craniotomy for benign brain tumours. Support Care Cancer 2011;19:1841–8. 21. Bradley CJ, Given CW, Roberts C. Race, socioeconomic status, and breast cancer treatment and survival. J Natl Cancer Inst 2002;94:490–6. 22. Pittard IIIrd WB, Laditka JN, Laditka SB. Associations between maternal age and infant health outcomes among Medicaid-insured infants in South Carolina: mediating effects of socioeconomic factors. Pediatrics 2008;122:e100–6. 23. Sherwood PR, Stommel M, Murman DL, et al. Primary malignant brain tumor incidence and Medicaid enrollment. Neurology 2004;62:1788–93. 24. Downie DL, Schmid D, Plescia MG, et al. Racial disparities in blood pressure control and treatment differences in a Medicaid population, North Carolina, 2005–2006. Prev Chronic Dis 2011;8:A55. 25. Kim SH, Ferrante J, Won BR, et al. Barriers to adequate follow-up during adjuvant therapy may be important factors in the worse outcome for Black women after breast cancer treatment. World J Surg Oncol 2008;6:26. 26. Morris AM, Billingsley KG, Hayanga AJ. Residual treatment disparities after oncology referral for rectal cancer. J Natl Cancer Inst 2008;100:738–44.