Social Class and Cancer Patient Survival in Finland

Abstract

The effect of social class on survival was assessed in a cohort of cancer patients identified from the nationwide population-based Finnish Cancer Registry. The cohort consisted of all reported cases of the 12 most common types of cancer occurring in Finland between 1971 and 1985 among persons born in 1906-1945 (n = 106,661). Social class information based on occupation was obtained individually for each patient from the population census of 1970. Both observed and corrected (i.e., cause of death-specific) 5-year survival rates were used in the analyses. A statistically significant linear effect of social class on age-adjusted relative risk of cancer death was observed in six of 12 cancer types among men and in nine of 12 among women; and the risk was highest for those in the lowest social class. The relative risk of death due to cancer for social class I (highest) relative to social class IV (lowest) was lowest in bladder cancer (relative risk (RR) = 0.46, 95% confidence interval (CI) 0.34-0.61) and kidney cancer (RR = 0.61, 95% CI 0.48-0.78) among men and in corpus uteri (RR = 0.51, 95% CI 0.36-0.72) and rectum cancer (RR = 0.56, 95% CI 0.42-0.74) among women. The differences between results obtained using corrected and observed survival rates were small. These findings indicate that social class is an important determinant of cancer patient survival. Additional research is required to clarify the etiology of the social class differences and to identify factors that could be used for developing strategies to diminish such inequalities.

Social Inequalities and Cancer

“…If all death certificates were based on pathological diagnoses the social class gradient in cancer mortality would be slightly steeper. A smaller proportion of microscopically confirmed diagnosis among lower social classes has also been reported elsewhere (Auvinen et al, 1995).…”

“…No clear social class differences were observed either before or after adjustment for histological type (squamous cell versus others) among lung cancer patients in the study by Stavraky et al (1987). Auvinen et al (1995) reported effect modification by cell type in lung cancer and leukaemia among men in their study based on more than 100 000 cancer patients. Social class differences were not observed among all patients, but they were confined to non-small-cell lung cancer and acute leukaemia.…”

“…In the study by Karjalainen and Pukkala (1990), the differences in results obtained by using relative or corrected rates were small in ages 25-54 years, but increased in older ages. Auvinen et al (1995) observed slightly larger social class differences when using observed rates compared with corrected rates and concluded that the difference was probably due to overestimation of the differences by observed rates.…”

“…They estimated that depending on the type of cancer, 25-50% of the survival difference between the indigent and private patients in their study was due to deaths from causes other than cancer. Also, in the studies by Kogevinas (1990) and Auvinen et al (1995), a social class gradient among cancer patients has been observed in deaths from cancer as well as from other causes. The validity of corrected survival rates depends on the assumption that there are no major social class differences in accuracy of death certificates.…”

“…There is some evidence that diagnosis at autopsy is more common in lower social classes and that in the majority of these cases cancer is the underlying cause of death (Auvinen et al, 1995). This suggests that cases first diagnosed at autopsy (at least those with cancer as a cause of death) should be included in the material with zero survival.…”