Disparities in Cancer Treatment: Factors That Impact Health Equity in Breast, Colon, and Lung Cancer Part 2

FINDINGS ACROSS DISEASES

Table 4.1 summarizes the various factors, grouped into four—(a) patient; (b) provider; (c) health system, and (d) contextual—factors found to have contributed to disparities in cancer treatment. Three cancers—breast, lung, and colorectal—are used here as illustrative examples that can serve as models for disparities in other cancers or other chronic illnesses.

Breast Cancer

Despite a lower incidence of breast cancer among African American women, stage for stage, they are more likely than White women to die from breast cancer (McBride et al., 2007), leaving racial differences in treatment, which have been documented in breast cancer treatment for years, as a likely cause for these persistent disparate outcomes.

Survival outcomes depend on the timeliness, quality, and completeness of treatment. Recent advances in treatment that have the greatest potential to affect breast cancer mortality rates are those involving systemic adjuvant therapy for nonmetastatic patients and may also include benefits from neo-adjuvant chemotherapy. Even in the setting of cooperative group trials, where all women are prescribed identical treatment, disparities exist, despite controlling for tumor characteristics and accesses to care. Disparities in outcome are seen in both pre- and postmenopausal women, and women with and without hormone-sensitive cancer (Albain, Unger, Crowley, Coltman, & Hershman, 2009; Hershman et al., 2010). However, African American women are more likely to miss appointments and discontinue therapy earlier than White women (Hershman et al., 2010).

TABLE 4.1 Factors Potentially Influencing Disparities in Cancer Treatment

¹    "Provider" includes "hospital."

²    Cancer threat appraisal is an attitude reflecting distress about a cancer diagnosis and has demonstrated a major impact on uptake of cancer care.

³    Health literacy maybe inadequate for more than 20% of adult Americans and has a major impact on cancer care communication, recommendations, and adherence.

⁴Perceived discrimination is a belief reflecting general perception of unfair treatment associated with discrimination due to membership in a socially vulnerable group.

A number of studies have demonstrated that under use of adjuvant treatment plays an important role among various subpopulations of minority women (Bickell et al., 2006; Prehn et al., 2002), including data showing that African American women are more likely to discontinue treatment (Griggs, Sorbero, Stark, Heininger, & Dick, 2003; Hershman et al., 2009; Li, Malone, & Daling, 2003; Mandelblatt et al., 2000).

The presence of comorbid conditions (such as obesity, diabetes, or hypertension) may also influence the range of treatments offered to women with breast cancer and contribute to both higher breast cancer mortality rates and higher rates of death from competing causes among Black women (Schairer, Mink, Caroll, & Devesa, 2004; Tammemagi, Nerenz, Neslund-Dudas, Feldkamp, & Nathanson, 2005). As survival rates from breast cancer improve due to advances in screening and treatment, the influence of co-morbid conditions is emerging as a competing cause of death among breast cancer survivors and highlights yet another racial inequity.

While differences in patterns of care in breast cancer treatment have been established, why they occur and how to overcome them has not. In the past few years, much attention has been paid to the dramatic geographic differences in the magnitude of the African American/White breast cancer mortality, with cities such as Chicago and Houston experiencing some of the greatest disparities and New York City exhibiting some of the lowest differences in the United States (McBride et al., 2007). These data suggest that health system and contextual factors, such as the location, accessibility, and quality of screening, diagnostic, and treatment facilities, are potentially some of the largest contributors to racial disparities in breast cancer mortality (Ansell et al., 2009).

Colorectal Cancer

Although colorectal cancer screening and treatment regimens have become increasingly effective, African American patients continue to experience about 20% higher cancer-specific mortality than their White counterparts (American Cancer Society, 2009; Chien, Morimoto, Tom, & Li, 2005; Morris, Wei, Birkmeyer, & Birkmeyer, 2006). This disparity persists even after adjusting for stage of disease and comorbidities, implying that under treatment and poorer quality of care may play a major explanatory role. In fact, most previous research into mechanisms underlying disparate colorectal cancer outcomes has simply measured race-related differences in use of care—and found them abundant. Previous studies have found that Black patients were considerably less likely than Whites to undergo colectomy for colon cancer (Ball & Elixhauser, 1996; Lee, Gehlbach, Hosmer, Reti, & Baker, 1997). National data indicate that Black patients with advanced colorectal cancer are 20%-50% less likely than Whites to undergo adjuvant treatment after surgery (Ayanian et al., 2003; Baldwin et al., 2005; Baxter, Rothenberger, Morris, & Bullard, 2005; Govindarajan, Shah, Erkman, & Hutchins, 2003; Morris, Billingsley, Baxter, & Baldwin, 2004; Morris et al., 2006; Potosky et al., 2002;). Although some have suggested that Black patients are inherently less likely than Whites to benefit from standard colon cancer chemotherapy regimens (Jessup, Stewart, Greene, & Minsky, 2005), the results of rigorously conducted multicenter randomized controlled trials negate such claims (Bach et al., 2002; Dignam et al., 1999, 2003; McCollum et al., 2002). Additionally, among a national cohort of veterans with colorectal cancer, no racial difference in colorectal cancer mortality was identified after controlling for other patient characteristics, stage of disease, and surgical and adjuvant therapy (Dominitz, Samsa, Landsman, & Provenzale, 1998). Taken together, these and similar population-based studies (Bach et al., 2002; Schwartz, Crossley-May, Vigneau, Brown, & Banerjee, 2003) imply that identical treatment of Black and White colorectal cancer patients leads to identical cancer outcomes. Conversely, the unexplained variation or absence of evidence-based treatment among Black colorectal cancer patients results in higher cancer mortality or poorer outcomes.

Reasons for under use of adjuvant therapy among Black patients are largely unknown. To determine whether access to care was part of the problem, Morris et al. (2008) reviewed SEER-Medicare data linked to American Medical Association and American Hospital Association data for incident rectal cancer cases over a 6-year interval, and found that Black and White patients were equally likely to see a medical oncologist postoperatively (73.1% vs. 74.9%). However, among patients who saw a medical oncologist, African American patients were substantially less likely than Whites to use chemotherapy (54.1% vs. 70.2%). Examination of baseline patient characteristics that might account for these discrepancies in treatment revealed that individual demographic and clinical variables had minimal influence. For example, as age at diagnosis and comorbidities increased, use of chemotherapy by White patients declined as expected. By contrast, among Black patients in the cohort, those with the lowest comorbidity scores were the least likely to receive treatment. A similar study recently evaluated the use of chemotherapy among stage III colon cancer patients (Baldwin et al., 2005). Again, although Black and White patients were equally likely to be referred to an oncologist, the youngest Black patients in the cohort were the least likely to undergo chemotherapy. Taken together, these data imply that Black patients who are younger and healthier—that is, those who have the most to gain from chemotherapy—are the least likely to receive it.

In exploring reasons for why these disparities were present, studies have found poorer adherence to medical recommendations among healthier patients in general, due to decreased experience with medical encounters (Carlos, Fendrick, Patterson, & Bernstein, 2005; Carlos, Underwood, Fendrick, & Bernstein, 2005) or increased family caregiver responsibilities (Ahmed et al., 2001; Kim, Kabeto, Wallace, & Langa, 2004; Ortiz et al., 2009; Schulz & Beach, 1999). In addition, in spite of access to an initial consultation, Black patients in the study may have been less likely to undergo adjuvant treatment because of difficulty or unfamiliarity with navigating the medical system (Freeman, 2006), lack of transportation (Guidry, Aday, Zhang, & Winn, 1997; Lutgendorf et al., 2010), or lack of available specialist providers (Bach, Pham, Schrag, Tate, & Hargreaves, 2004). Finally, Black patients may be more likely than White patients to refuse adjuvant therapy for social or cultural reasons, such as reduced risk tolerance, lack of trust in medical treatment, or resignation to a disease state. Previous studies of colorectal and prostate cancer have indicated that a greater proportion of Black than White patients may delay or refuse surgical care until the disease has become relatively acute or advanced (Demissie et al., 2004; Shavers et al., 2004). It follows that Black patients are more likely to undergo urgent or emergent surgery, a well-defined risk factor for early recurrence and mortality (Ball & Elixhauser, 1996). In another study, the perceptions of African Americans who considered their physicians uninterested and less engaging led to a lack of understanding of their treatment options and less adherence to physician recommendations (Gordon, Street, Sharf, Kelly, & Souchek, 2006). Thus, poorer colorectal cancer outcomes are primarily influenced by how well (and if at all) each aspect of care is delivered, which in turn is influenced by patient characteristics and by the characteristics of providers and hospitals.

Lung Cancer

Decisions against lung cancer surgery in early stage, non-small-cell disease serves as nearly a guarantee of death within 4 years, with half these deaths occurring within a year of diagnosis (Bach et al., 1999). Patients’ perceptions of progressive lung cancer are that the prognosis for this disease is exceedingly poor and the disease is thus very much feared (Cykert & Phifer, 2003). Only very few patients with early-stage disease have absolute contraindications for surgery, so why are decisions against surgery so common and strongly associated with African Americans? Several reasons are possible. African Americans often hold beliefs that could potentially delay diagnosis or impede treatment of lung cancer (Margolis et al., 2003). Lack of trust, superstitions, religiosity, and fatalism have all been proposed (Cykert, Dilworth-Anderson, et al., 2010; Gordon et al., 2006) as cultural barriers to recommended care. However, although surgical refusal by African Americans has been reported as equal (Cykert, Dilworth-Anderson, et al., 2010) or marginally higher than for Whites (Lathan, Neville, & Earle, 2006), frank refusal only explains a tiny fraction of the treatment gap.

A recent prospective study (Cykert, Dilworth-Anderson, et al., 2010) found that when patients perceived their communication with their physician, and perceived their prognosis to be poor, they were more likely to decide against surgery, regardless of race. Another complicating factor in physician-patient communication about lung cancer surgery is that nearly one-fifth of physicians who provide lung cancer care underestimate survival relative to treatment (Schroen, Detterbeck, Crawford, Rivera, & Socinksi, 2000).

System factors have also been associated with sub-optimal care for African Americans. For instance, low hospital volume for lung cancer procedures predicts poor lung cancer outcomes (Strand et al., 2007). African Americans received proportionately more care at such facilities (Epstein, Gray, & Schlesinger, 2010; Lucas, Stukel, Morris, Siewers, & Birkmeyer, 2006). Yet another contributing factor is lack of regular care. Even when insured, African American patients have less access to primary care (Corbie-Smith, Flagg, Doyle, & O’Brien, 2002) and, in early-stage lung cancer, having this access to continuing care seems to provide the possibility of reevaluation and the opportunity to reengage in cancer treatment.