Is Race a Real Issue in Colectomy for Ulcerative Colitis?

Abstract

In this issue of Clinical Gastroenterology and Hepatology, Nguyen et al1Nguyen G. LaVeist T. Gerhart S. et al.Racial and geographic variations in colectomy rates among hospitalized ulcerative colitis patients.Clin Gastroenterol Hepatol. 2006; 4: 1507-1513Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar have analyzed national colectomy rates for patients with ulcerative colitis and identified significant racial and regional disparities in the United States. Their report contributes to a body of literature that has identified disparities in several areas of health care related to access, interventions, and outcomes. As in the report from Nguyen et al,1Nguyen G. LaVeist T. Gerhart S. et al.Racial and geographic variations in colectomy rates among hospitalized ulcerative colitis patients.Clin Gastroenterol Hepatol. 2006; 4: 1507-1513Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar most of the previous studies have made use of large governmental or private administrative databases. Although the use of a national administrative database provides reliable global information about the delivery of health care in the United States, they are significantly limited in their ability to explain the precise reasons for any observed difference in care. In their discussion, Nguyen et al1Nguyen G. LaVeist T. Gerhart S. et al.Racial and geographic variations in colectomy rates among hospitalized ulcerative colitis patients.Clin Gastroenterol Hepatol. 2006; 4: 1507-1513Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar clearly cite numerous limitations in the use of the governmental administrative database selected and provide some possible explanations for their findings. Unfortunately, they emphasize racial differences without acknowledging that race may be merely a surrogate maker for other more important factors responsible for their observation. Their use of a single database and focus on race provides an all too simplistic analysis for a problem that is undoubtedly more complex than the authors lead the reader to believe. The authors are to be congratulated for taking on this important question. Although a review of all conflicting literature on the subject of health care disparities is beyond the scope of this editorial, I hope to highlight some of the studies that have evaluated this very important issue. Several reports have found evidence of racial disparities; however, it has been reported frequently that, after correction for other factors, racial disparities assume less importance than in the original analysis. Clearly, there are disparities in our society and they most likely extend to delivery of health care services. However, rigorous analyses are essential to help identify and resolve the real inequities. The US population is becoming increasingly diverse. According to the 2004 census, non-white Hispanics are the largest minority population, representing nearly 14% (41 million) of the population.2US Census Bureau. We the people: Hispanics in the United States in 2004. Available at: www.census.gov. Accessed: September 25, 2006.Google Scholar Other groups such as Asians are growing at a rate 3 times faster than any other ethnic group.3US Census Bureau. We the people: Asians in the United States in 2004. Available at: www.census.gov. Accessed: September 25, 2006.Google Scholar The increasing diversity in our population may present unique challenges to the health care delivery system in relation to cultural differences in the perception of health care and health practices. The US government has made eliminating health care disparities a top priority in the Healthy People 2010 Initiative.4US Department of Health and Human Services. Healthy people 2010. Available at: www.healthypeople.gov. Accessed: September 26, 2006.Google Scholar Importantly, a number of factors likely contribute to health care disparity including limited access, systems of referral into advanced care settings, socioeconomic status, educational level, cultural expectations, population-specific risk factors, and the more recently recognized phenomenon of genetically based factors that influence response to treatment. Although the simplistic view has been that improving access eliminates disparities, the literature provides studies with contradictory information. Many of the early studies evaluating the impact of race on health care use occurred in the Veteran’s Administration (VA) Hospital System. Access to the VA system is based on non–race-based eligibility requirements. Once met, the patient has full and unrestricted use of the VA system. One of the first studies to evaluate the impact of race and care in the VA system highlighted race-based differences in the use of invasive cardiovascular procedures.5Whittle J. Conigliaro J. Good C. et al.Racial differences in the use of invasive cardiovascular procedures in the Department of Veterans Affairs Medical System.N Engl J Med. 1993; 329: 621-627Crossref PubMed Scopus (393) Google Scholar Even after adjusting for confounding factors such as comorbidities, age, treatment center capabilities, and geographic location, African-American veterans had fewer procedures than the number undergone by white veterans. In response to this study, the VA launched an intensive program to evaluate and resolve race-based disparities. Subsequently, studies have shown that there were no significant racial disparities in access, use of services, or outcomes in the VA health system. For example, Petersen et al6Petersen L. Wright S. Petersen E. et al.Impact of race on cardiac and outcomes in veterans with acute myocardial infarction.Med Care. 2002; 40: I-86-I-96Google Scholar found no difference in 30-day, 1-year, or 3-year mortality rates after an acute myocardial infarction or in the use of best-practices guideline medications or diagnostic procedures. On the other hand, Collins et al7Collins T. Johnson M. Henderson W. et al.Lower extremity nontraumatic amputation among veterans with peripheral arterial disease: is race an independent factor?.Med Care. 2002; 40: I-106-I-116Google Scholar did show a difference in rates of amputations between whites and African Americans and Hispanics with peripheral vascular disease using the VA National Surgical Quality Improvement database. However, the significant limitation of this study was the authors’ inability to adjust for disease severity at presentation or prior surgical interventions, which brings into question the conclusions. The findings of the VA system reports are important because they show that a preliminary analysis of specific aspects of health care delivery may show racial disparities. However, when adjustments are made for disease severity or social or cultural factors, the race-based difference often lessens or disappears altogether. Another problem with large national administrative databases is that they offer limited or no information regarding the level or type of patient care received before the index admission. Often, disparities exist in the referral systems for ethnic minorities, as pointed out in a study by Shea et al.8Shea D. Stuart B. Vasey J. et al.Medicare physician referral patterns.Health Serv Res. 1999; 34: 331-348PubMed Google Scholar In their study, the investigators found that surgical admissions in New York were 8% for African Americans and 5% for non-white Hispanics, although they represented 15% and 9% of the patients, respectively. To investigate further the reason for the referral disparity, Basu and Clancy9Basu J. Clancy C. Racial disparity, primary care, and specialty referral.Health Serv Res. 2001; 36: 64-77PubMed Google Scholar analyzed the impact of primary care physicians (PCPs) in making referrals for elective referral-sensitive surgeries such as coronary artery bypass graft surgery and total knee and hip replacement. They correlated New York–area hospital discharges with the Area Resource File and the American Hospital Association’s surveys of PCPs. Their hypothesis was that increases in the number of PCPs in a population would lead to a higher rate of ethic minority admissions for referral-sensitive surgeries. The analysis of all admissions of New York residents aged 20–64 years in 1995 revealed a very strong association with PCP density and referral of African Americans for referral-sensitive surgeries. In areas of higher PCP density, there were no racial disparities seen in referrals. In a similar study that examined the influence of having stable access to continuing primary care, Doescher et al10Doescher M.P. Saver B.G. Fiscell K. et al.Racial/ethnic inequities in continuity and site of care: location, location, location.Health Serv Res. 2001; 36: 78-89PubMed Google Scholar reported that racial and ethnic minorities compared with whites are less likely to have regular sites of care after adjustment for income, education, insurance coverage, and overall health status. These studies show that large administrative databases such as the National Inpatient Survey are neither necessarily sensitive nor specific enough to explain accurately the reasons why racial disparities exist for specific referral-sensitive surgical procedures. As one delves deeper into this whole issue of disparity along racial and ethnic lines, it becomes more difficult to assess the relative contribution of other factors. Such an analysis requires the use of multiple data sets that include not only the medical condition, but also data regarding the broader social circumstances of the analyzed population. McGory et al11McGory M. Zingmond D. Sekeris E. et al.A patient’s race/ethnicity does not explain the underuse of appropriate adjuvant therapy in colorectal cancer.Dis Colon Rectum. 2006; 49: 319-329Crossref PubMed Scopus (52) Google Scholar have studied the use of appropriate adjuvant therapy for colorectal cancer in California between 1994 and 2001. In this population-based study, the investigators used 3 different databases: the California Cancer Registry, the California Discharge Database, and the 2000 Census Report to evaluate the variables associated with the underuse of adjuvant therapy. In the cohort of 18,649 patients, 5 different multivariate regression analyses were run using patient characteristics including age, sex, race/ethnicity, comorbidities, payer status, year of diagnosis, and socioeconomic status (SES). In all the models, the only significant association with the underuse of adjuvant therapy for colorectal cancer was the patients’ SES status. Thus, a lower SES, regardless of race or ethnicity, was associated with patients not receiving the appropriate adjuvant therapy. All other variables dropped out as nonsignificant in the multiple regression analysis. Christian et al12Christian C. Niland J. Edge S. et al.A multi-institutional analysis of the socioeconomic determinants of breast reconstruction: a study of the National Comprehensive Cancer Network.Ann Surg. 2006; 243: 241-249Crossref PubMed Scopus (177) Google Scholar found a similar association with SES and not with race or ethnicity in their study of the use of breast reconstruction after treatment for breast cancer. In reference to studies of health care delivery and race or ethnicity, McGory et al11McGory M. Zingmond D. Sekeris E. et al.A patient’s race/ethnicity does not explain the underuse of appropriate adjuvant therapy in colorectal cancer.Dis Colon Rectum. 2006; 49: 319-329Crossref PubMed Scopus (52) Google Scholar made the general recommendation that race and ethnicity in studies that use only a single population or an administrative data set “might be serving as a proxy for inadequately measured economic variables or other unmeasured socioeconomic variables, such as education and acculturation (eg, ability to speak English).” In the present study of a unique patient population, Nguyen et al1Nguyen G. LaVeist T. Gerhart S. et al.Racial and geographic variations in colectomy rates among hospitalized ulcerative colitis patients.Clin Gastroenterol Hepatol. 2006; 4: 1507-1513Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar found that there seems to be a difference in surgical treatment for ulcerative colitis, which may be associated with race. However, the authors acknowledge that the exact cause for this association cannot be established using their study methodology. Therefore, the conclusions of the article must be interpreted with caution because the observation may not be supported either directly or indirectly. Central to this whole issue is that race or ethnicity as a proxy is unreliable, and, more importantly, it assumes that the effect of race or ethnicity is immutable. If we are to move toward a future in which there is no disparity in care, as envisioned in the Healthy People 2010 Initiative, then we need to direct interventions and policies at truly modifiable factors, such as increasing the number of and distribution of PCPs and improving the economic condition of all members of our society. Clearly, there is no simple answer to these difficult and complex problems, but it certainly is important not to stop at simple analyses, which may distract society from understanding the real problems and intervening to correct them.