Article Text
Abstract
Background The Institute of Medicine called attention to the pervasive differences in treatments and outcomes between ethnic groups. We sought to highlight the geographic and racial disparities in access to treatment for unruptured cerebral aneurysms.
Methods We performed a retrospective cohort study involving patients with unruptured cerebral aneurysms from 2000 to 2010, registered in the National Inpatient Sample (NIS) database. Primary outcomes were those patients receiving treatment and the ratio of untreated to treated aneurysms per state. The purpose of this study was to determine if there were geographic and racial disparities in access to treatment of unruptured cerebral aneurysms based on the NIS. Logistic regression and analysis of variance (ANOVA) techniques were used.
Results There were 57 418 patients diagnosed with unruptured aneurysms (mean age 61.4 years, 70.5% females), with 18 231 undergoing treatment. Males (OR 0.67, 95% CI 0.64 to 0.71, p<0.0001), Asian (OR 0.88, 95% CI 0.81 to 0.96, p=0.003), Hispanic (OR 0.76, 95% CI 0.65 to 0.90, p=0.001), African American (OR 0.57, 95% CI 0.53 to 0.62, p<0.0001), and patients without insurance (OR 0.76, 95% CI 0.67 to 0.87, p<0.0001) were associated with decreased chance of treatment. The opposite was true for lower Charlson Comorbidity Index (OR 3.03, 95% CI 2.71 to 3.39, p<0.0001), coverage by Medicaid (OR 1.12, 95% CI 1.03 to 1.23, p=0.012), or private insurance (OR 1.92, 95% CI 1.80 to 2.04, p<0.0001), and lower income (OR 1.22, 95% CI 1.15 to 1.31, p<0.0001). Significant regional variability was observed among the different states (p=0.006, ANOVA), with Maryland being an outlier.
Conclusions Based on the NIS database, the rate of treatment of unruptured cerebral aneurysms varies according to sex, race, and region.
- Aneurysm
- Intervention
Statistics from Altmetric.com
Introduction
Cerebral aneurysm rupture, causing subarachnoid hemorrhage (SAH), is a devastating event, with significant morbidity and mortality.1 ,2 Treatment of incidentally discovered aneurysms with clipping or coiling, and smoking cessation are the most important preventive measures for SAH.1 ,2 The literature supports treatment decisions based on aneurysm size and morphology.1 ,2 However, there are some indications that racial disparities might exist among groups receiving treatment for ruptured and unruptured cerebral aneurysms.3–5 The Institute of Medicine has demonstrated a lower rate of evidence based practice in the African American population.6 This has been confirmed for cardiovascular pathology by several groups.4 ,7–9
In addition to racial and ethnic disparities, where a patient lives can by itself have a large impact on the level and quality of healthcare they receive.10–13 As black and Hispanic populations tend to live in different areas in comparison with non-Hispanic white populations, location matters in the measurement and interpretation of healthcare disparities.10 ,13 There is wide variation in racial disparities across geographic lines: areas have substantial treatment differences, while others are equal. Furthermore, there is no consistent pattern of disparities across treatment differences.10 ,13 This problem of quality of care across regions and races should remain the target of policy makers in order to map and address barriers to care, and has not been investigated in the setting of access to care for cerebral aneurysms.
The National Inpatient Sample (NIS) is a hospital discharge database that represents approximately 20% of all inpatient admissions to non-federal hospitals in the USA.14 Using this database, we investigated the association of cerebral aneurysm treatment with socioeconomic characteristics (age, gender, race, income). We also hypothesized that significant regional variation exists in regards to access to care, and we quantified that with the use of the ratio of untreated to treated patients with unruptured aneurysms per state per year.
Methods
NIS database
All patients diagnosed with an unruptured cerebral aneurysm in the NIS Database14 (Healthcare Cost and Utilization Project, Agency for Healthcare Research and Quality, Rockville, Maryland, USA) between 2000 and 2010 were included in the analysis. The NIS contains discharge data regarding all discharges from a stratified random 20% sample of non-federal hospitals in several States. More information about the NIS is available at http://www.ahcpr.gov/data/hcup/nisintro.htm.
Cohort definition
In order to establish the cohort of patients, we used ICD-9-CM codes to identify patients in the registry who were diagnosed with an unruptured cerebral aneurysm (ICD-9-CM code 437.3). We used ICD-9-CM codes to identify patients that underwent treatment with clipping (ICD-9-CM code 39.51) or coiling (ICD-9-CM code 39.52 (should also have a code 88.41 and no 39.51 during the same hospitalization), 39.72, and 39.79 between 2000 and 2010) (figure 1). Patients with SAH (ICD-9-CM code 430) and ruptured aneurysms were excluded from the analysis.
Variables
The primary outcome variables were whether a patient received treatment (clipping or coiling) for their unruptured aneurysm, and the average ratio of untreated to treated unruptured aneurysms per state per year. The effect on the outcomes of the pertinent exposure variables was examined in a logistic regression analysis. Age was the only continuous variable. Gender, race, payer, income, and modified Charlson Comorbidity Index (CCI)15 ,16 were categorical variables. Race was divided to the following categories: African American, Hispanic, Asian, or other, with Caucasian being the reference value. Payer status was divided into the following categories: private insurance, Medicaid, and no insurance/self-pay, or other, with Medicare being the reference value. Income was divided into quartiles, based on the mean income by zip code. The CCI predicts the 10 year mortality for a patient who may have a range of comorbid conditions.15 ,16
Statistical analysis
Logistic regression was used to develop a multivariable model for the association of socioeconomic variables with the possibility of a patient receiving treatment (clipping or coiling) for their unruptured aneurysm. Regional differences across states were assessed by examining the ratio of untreated to treated patients with unruptured aneurysms per state per year. Comparison of means was performed using analysis of variance (ANOVA) with post hoc Tukey test. The level of statistical significance was set at 0.05. Sensitivity analyses were performed using different reference values (ie, not Medicare, and white) for the socioeconomic disparities, and further post hoc tests, including Bonferroni modification for the state differences, yielding similar results, and are therefore not presented here. Statistical analyses were performed using XLSTAT V.2011.2.01 (Adinsoft) and SPSS V.20 (IBM).
Results
Demographics and clinical characteristics
In the selected study period there were 57 418 patients diagnosed with unruptured aneurysms (mean age 61.4 years, 70.5% women) that were registered in the NIS, of whom 18 231 underwent treatment with clipping or coiling (figure 1). Patients with untreated aneurysms were more commonly older, men, African Americans, Hispanics, Asians, of higher income, had more comorbidities, and had insurance coverage with Medicare or no insurance (table 1).
Socioeconomic disparities
In a multivariate analysis (figure 2), men (OR 0.67, 95% CI 0.64 to 0.71, p<0.0001), Asian (OR 0.88, 95% CI 0.81 to 0.96, p=0.003), Hispanic (OR 0.76, 95% CI 0.65 to 0.90, p=0.001), African American (OR 0.57, 95% CI 0.53 to 0.62, p<0.0001), and patients without insurance coverage (OR 0.76, 95% CI 0.67 to 0.87, p<0.0001) were associated with a decreased chance of receiving treatment (clipping or coiling) for their unruptured aneurysm. The opposite was true for lower CCI (OR 3.03, 95% CI 2.71 to 3.39, p<0.0001), coverage by Medicaid (OR 1.12, 95% CI 1.03 to 1.23, p=0.012), or private insurance (OR 1.92, 95% CI 1.80 to 2.04, p<0.0001), and lower income (OR 1.22, 95% CI 1.15 to 1.31, p<0.0001). Patients with missing variables were excluded from the multivariate analysis.
Regional variability
There was significant regional variability in the practices of treating unruptured aneurysms (figure 3). ANOVA of the ratio of untreated to treated patients with unruptured aneurysms revealed significant variability among the states (p=0.006). Post hoc Tukey test among the individual states represented in the sample identified Maryland as a significant outlier, with increased numbers of patients with untreated cerebral aneurysms in comparison with the following states: CA, CO, FL, GA, IL, KY, MN, MO, NC, NY, OH, OR, TN, TX, UT, VA, WA, WI, and WV. New Jersey demonstrated a similar trend.
Discussion
There is substantial literature documenting racial and economic disparities in healthcare.10 African Americans and Hispanics experience lower levels of healthcare, even when high quality controls for patient risk adjustment are included in the analysis.10–13 The Institute of Medicine has recognized that many sources, including health systems, healthcare providers, patients, and utilization managers, may contribute to racial and ethnic disparities in healthcare.6 These racial variations in access to evidence based practices, with obvious effects on outcomes, have been extensively investigated for ischemic stroke.8 ,9 In addition to the inherent variation in the presentation and natural history of the disease, shortfalls in the quality of care, particularly across race and ethnicity, have been documented.8 ,9 Similar data are lacking in the context of the treatment of other areas of cerebrovascular pathology, including cerebral aneurysms.
Brinjikji et al3 compared the socioeconomic characteristics of patients undergoing treatment for unruptured and ruptured aneurysms, demonstrating a prevalence for lower socioeconomic status for patients in the SAH category. That was not unexpected given the propensity of these patients for more comorbidities (ie, higher rates of smoking), increasing their risk of SAH. No comparison with untreated patients was performed. In the present analysis, we observed that minority subgroups were least likely to undergo treatment for unruptured cerebral aneurysms. That tendency was more pronounced for African Americans, with Hispanics following closely. On the other hand, decreasing income was associated with increasing chance for treatment. Patients with a higher income are more likely to seek several professional opinions and decide on a more conservative course of action. Patients of lower income are likely to investigate alternative options less, and eventually follow a more aggressive approach.
Lack of insurance was associated with a decreased propensity for treatment whereas private insurance demonstrated the strongest association with intervention. This observation is most likely the combination of patient and systemic factors. Uninsured patients tend to avoid expensive preventive interventions that are done electively whereas healthcare systems and physicians provide barriers to care for these populations. Furthermore, older age and the presence of multiple comorbidities seem to deter physicians and patients from seeking intervention. Further investigation is needed to determine to what degree the latter decision is justified, and how to better select groups who have an unfavorable risk–benefit profile.
Another well recognized contributor to disparities in healthcare is geographic variation.10–13 We identified the effect of differences in where patients live to the chance for receiving treatment for unruptured cerebral aneurysms. Although there was significant variation among all states, Maryland was an outlier, with a significantly elevated ratio of untreated to treated aneurysms. Although this phenomenon has not been described previously in the area of cerebral aneurysms, it has been documented for other pathologies.10–13 Hospitals and regions of the USA vary significantly in the extent to which such disparities are present.10–13 The effect of racial differences is less pronounced when the local characteristics are taken into account. There are healthcare markets that serve large numbers of minorities and do not demonstrate variation in access, or selectively demonstrate disparities for some procedures.10 ,11 ,13 The disadvantaged position of most racial minorities in regards to barriers to care mainly stems from the propensity of these populations to live in areas of the country with low quality hospitals and providers.13 It is a local phenomenon and as such it should be addressed by policy makers.
The present study has several limitations common to administrative databases. Indication bias and residual confounding could account for some of the observed associations. In addition, some coding inaccuracies will undoubtedly occur and can affect our estimates. This is no different than other studies involving the NIS. However, several reports have demonstrated that coding for cerebral aneurysms has shown nearly perfect association with medical record review.17 ,18 The NIS during the years studied did not include hospitals from all states. However, the hospitals included were still diverse with respect to size, region, and academic status, supporting the generalizability of our findings. The NIS does not provide any clinical information on the structure, size, or location of the aneurysms, which are important factors tailoring decision making in cerebrovascular neurosurgery. However, if the decision was based only on the available clinical evidence, we should not be observing any socioeconomic or geographic disparities.
Conclusions
In the US population, the rate of treatment of cerebral aneurysms varies according to sex, race, or ethnic group, and region. In a multivariate analysis, men, Asian, Hispanic, African American, and patients without insurance coverage were associated with a decreased chance of receiving treatment for their unruptured aneurysm. There was significant variation in the rate of untreated to treated aneurysm among all states, with Maryland being a significant outlier. These variations underscore the importance of targeting local factors to improve access to treatment for cerebral aneurysms.
References
Footnotes
-
KB and SM contributed equally and are co-primary authors.
-
Contributors KB: conception and design, data analysis, manuscript preparation, and study supervision. SM: data collection and analysis, and manuscript review. NL: data interpretation, conception, study supervision, and manuscript review.
-
Competing interests None.
-
Ethics approval None. National databases do not require institutional review board approval.
-
Provenance and peer review Not commissioned; externally peer reviewed.