Low-income was associated with greater impairment prevalence, yet low-income BCS did not report higher PT use. Our study is the first to report on differences of impairments and PT utilization for BCS by race and income. Prior research indicates that AA BCS report higher prevalence of pain, fatigue, and lower overall physical and functional quality of life (Green et al. 2003; Paskett et al. 2008), but among the BCS we interviewed race was inconsistently associated with impairments. Most research on BCS involves white middle-income women with health insurance. Our study has the unique advantage of examining impairments in a large, diverse sample of BCS of all income levels.
Our research supports work that shows those with lower income have significantly lower physical functioning when compared to higher income whites and minorities (Braithwaite et al. 2010). A cross-sectional, observational study of BCS found that 36–59 % had shoulder range of motion restrictions (Cheville et al. 2008). Others found that BCS had significantly more shoulder limitations when compared to controls (Harrington et al. 2011). Our findings highlight the importance of shoulder range of motion limitation and suggest more of a need to address impairments—especially for those with low incomes.
Poorly controlled pain contributes to compensatory patterns of overuse, poor posture, and faulty biomechanics (Norkin and Levangie 1992). Other studies report that AAs may have higher levels of cancer-related pain (Green et al. 2003) yet we found significant differences in pain by income—not race.
Fatigue affects 9–39 % of BC (American Community Survey Design and Methodology 2014; Palmer et al. 2013) and is associated with disability and health care utilization (Servaes et al. 2007). One longitudinal study of 252 BCS reports fatigue affecting 31 % right after treatment declining to 6 % by the end of the first post-treatment year (Green et al. 2003). The majority of our survivors reported fatigue—a substantial finding because our BCS were long-term survivors. Another longitudinal study of 244 long-term BCS found no racial differences in fatigue, pain and other side effects, but they did not make comparisons by income (Gill et al. 2004). Our research shows significant differences by income for pain and utilization of PT to treat pain but not for fatigue.
Our study suggests that PT may be under-utilized for these impairments despite insurance status. The significantly higher use of PT by AAs was unexpected. Since we studied the presence/absence of impairments, not severity, it is possible that severe impairment might be more likely to trigger a referral to PT. In our sample, AAs were more likely to utilize PT if they had complaints of shoulder impairment, muscle weakness or pain. The total number of barriers to PT was not significantly associated with PT utilization—a surprising finding. While not all BCS need PT, our study suggests that impairments persist and the need for referral to PT may also persist.
Recent rehabilitation services documentation for Medicare patients mandate the use of patient reported outcomes such as those examined in this study. It is possible that some BCS may have issues with recall and attribution of impairments to cancer rather than an alternative medical diagnosis. Since our subjects were drawn from many different states, verification of self-reports was logistically impossible because it would have involved hundreds of hospitals, outpatient clinics and internal review boards.
Although this research included high proportions of low-income BCS (an advantage), the relatively few individuals with higher incomes precluded assessing income effects across upper-middle and higher income ranges. While BC is not rare, we studied a large proportion of minority and poor BCS which is rare and has allowed us to consider three key social determinants—race, income, and insurance.
Our results could also be partially explained by residence differences. For example, rural residents are more likely than urban residents to forgo medical and dental care after cancer (Palmer et al. 2013), travel farther for care and have mastectomy (Meilleur et al. 2013). Areas with higher versus lower health care spending also have higher rates of recommended and preferred health care (Keating et al. 2012). Future research can include these comparisons by taking advantage of hierarchical statistical modeling.
Our results may reflect the effects of reduced physical activity among breast cancer survivors. While exercise has many benefits including cancer control and prevention (Courneya et al. 2014) exercise is not part of survivorship care (Phillips et al. 2014).
Our study suggests that disparities exist among BCS for impairments that fall in the scope of practice for physical therapists. The “surveillance” model (Campbell et al. 2012) tracks signs and symptoms indicative of BC-related impairments, but ownership of surveillance is not clear nor tested with population-based samples. However, rehabilitation models exist in all medical systems that can be leveraged to improve BC-related impairment management. Rehabilitation is standard for those with total joint replacement and significantly improves physical and functional ability (Desmeules et al. 2013) self-efficacy, (Lane-Carlson and Kumar 2012) pain, (Niu et al. 2011) and reduces hospital length of stay, (Mertes et al. 2013) post-operative complications, (Husni et al. 2010) and readmission. (Gooch et al. 2012) The inclusion of PT at the beginning of treatment planning could have a similar effect in the BCS population, especially for those with low-incomes who seem to have the most BC-related impairments. Comparative effectiveness trials could compare referral-based versus rehabilitation models of care delivery.