More than 11 million people in the United States (U.S.) are cancer survivors. Of the more than one million Americans newly diagnosed with cancer each year, about 40% are working-age adults (Horner et al., 2009). Significant advances in cancer treatment have translated into improved survival and quality of life (Hubbard, 2010) and up to 85% of persons diagnosed continue working during treatment or return to work shortly after treatment (Hoffman, 2005; Short et al., 2005). This percentage varies by cancer type and stage of diagnosis.

Employed cancer survivors tend to be as productive and to have a comparable number of absentee days as employees without a cancer history (NCI, 2009). Nevertheless, due to differences in work load, stress and accommodation (Brown et al., 2006), certain job sectors may facilitate a higher rate of return of cancer patients than others. With earlier diagnoses among younger working-aged persons, job-related accommodations could have far reaching social and economic effects. Information on the occupational distribution and sociodemographic composition of the cancer survivors who continue to work after diagnosis is integral to understanding the potential health needs and burdens on this subpopulation.

Although some studies have investigated the functional limitations of adult survivors (Dellapasqua et al., 2006; Sehl, 2009), few have focused on the cancer survivors who continue and/or return to work (Frazier et al., 2009). Studies aimed at identifying which groups of cancer survivors return to work and if there is a significant disparity between these groups are needed to develop workplace-support programs. With most existing literature focusing on return to work among specific types of cancer, this study satisfies a gap in cancer literature by investigating health disparities among all survivors who continue to work post diagnosis.

We analyzed pooled cross-sectional data from the 1997–2009 U.S. National Health Interview Survey (NHIS). The study sample included persons ≥18 years (n=381,447). In this study, “cancer survivors” were individuals who reported being diagnosed with cancer. Employment status was determined by whether or not respondents were working during the week prior to the NHIS interview. Employed participants were stratified by occupational sector (Krieger et al., 2005).

Participants’ self-reported health status was dichotomized to “poor-fair” and “good-excellent” health. Chronic conditions (other than cancer) and functional limitations were dichotomously re-coded into presence or absence of ≥2 chronic conditions and ≥2 functional limitations, respectively. Finally, we assessed whether or not respondents spent >7 days in bed due to illness or injury in the last 12 months.

We examined socio-demographic characteristics including age (18–39, 40– 64, and ≥ 65 years), gender, race (White, Black, or other), ethnicity (Hispanic or non-Hispanic), education (<12, 12, and >12 years), as well as employment and insurance status at time of interview. Data management and analyses were conducted using SAS version 9.2 which allows the analysis of weighted, complex survey data. We used SAS SURVEYLOGISTIC to perform a multivariable logistic regression with contrasts between occupational sectors and employment status of persons with and without cancer for each health indicator. Analyses were adjusted for cancer type and socio-demographic variables.

Table 1 describes the socio demographics of the study sample. There were 22,952 persons who reported a cancer diagnosis. Among adult cancer survivors, workers were significantly less likely to report poor health and multiple disabilities when compared to the unemployed.

Table 2. illustrates that irrespective of occupational sector, cancer survivor workers were significantly less likely than unemployed survivors, but more likely than workers without cancer, to report “poor-fair” health. Among cancer survivors, blue-collar workers were significantly more likely than white-collar workers (Odd Ratio=1.98; 95% Confidence Interval=1.53–2.56), and farm-workers less likely than blue-collar workers (0.21; 0.09–0.52) to report poor-fair health. Service workers were less likely than blue-collar (0.60; 0.43–0.83) but more likely than farm-workers (2.82; 1.16–6.86) to report a “poor-fair” health status.

Among employed survivors, blue-collar workers were significantly more likely (1.28;1.04–1.59) to report having ≥2 functional limitations compared to white-collar workers. Working cancer survivors were less likely than unemployed cancer survivors, but more likely than employed persons without cancer, to report having ≥2 chronic conditions. Finally, among working cancer survivors, blue-collar workers were twice as likely (2.03;1.35–3.05) as white-collar workers to report having ≥2 chronic conditions.

These nationally representative data reflect differences in the health of working cancer survivors according to occupational sector. As one may expect, the major differences are found between white- and blue-collar workers. The former were more likely to report low prevalence of functional limitations as well as good-to-excellent health. In contrast, blue-collar workers reported 30% increased risk for ≥2 functional limitations, and twice the risk of ≥2 chronic conditions and poor-fair overall health status. This may reflect that financial need outweighs health limitations, especially among survivors in blue-collar jobs, which supports previous studies (Thorpe et al., 2003; Zwerling et al., 2003). However, although this white-blue collar differential is not limited to cancer survivors (Krieger et al. 2005; Fleming et al. 2007), the need for employment is likely to be more relevant in a population with a potentially recurrent and deadly chronic disease. A review of the impact of job loss and retirement on health concluded that job loss and threatened job loss have adverse effects regarding morbidity, all-cause mortality, and psychological health and that social support moderated some of these associations. The interpretation of these effects is complex, however, due to the significant impact of the worker’s years from retirement age and occupational sector. For example, whereas unemployment among middle-age blue-collar workers is likely to have a greater negative impact than for their white-collar peers, being forced to retire early (with adequate benefits), may be welcome by the former but not the latter, especially among highly specialized professionals (doctors, judges) (Kasl and Jones, 2000). Owing to the employment-based U.S. health insurance system, cancer survivors who are under retirement age risk losing not only income and health-enhancing social support (Pransky, 2010; Soler-Vila et al., 2003) but vital health insurance coverage as well. This is highly significant because their pre-existing condition (cancer) will make new coverage difficult to qualify for through employment or to afford as an individual (Cohen et al., 2005; Kaiser, 2006).

NHIS data are self-reported and cross-sectional which limits any causal inferences from findings. Also, clinical variables such as stage at diagnosis and type of treatment are not available. Major strengths of this study include the large sample size and the use of the NHIS data to examine occupational differences in health profiles among persons with a history of cancer. The data are representative of the U.S. population and can be used as a guide to inform public health policies related to work related issues among cancer survivors.

This study highlights the need for specific job accommodations for cancer survivors. With more reports of poor-fair health and increasing time before retirement, it is important to understand and attempt to alleviate some of the work-related difficulties affecting the health of working cancer survivors. Although there is continued support through organizations such as the American Cancer Society and a myriad of non-profit organizations, there is still no official venue providing job-related support for U.S. cancer survivors returning to work.