Disparities
Examining net cancer survival by certain demographics can help identify which populations experience disproportionate increases to mortality risk from cancer. This chapter breaks net cancer survival rates out by race, neighborhood poverty level and insurance status at time of diagnosis.
Because the PCR does not collect personal income data, this report uses neighborhood poverty level, which is based on the proportion of households in a census tract below the poverty line. See Technical Notes for more details.
Stage, race and poverty
While the physiological effect of race on developing cancer is not well understood, it has been shown as a valuable predictor of outcome. Similarly, socioeconomic status (SES) is associated with other predictors, such as access to care or healthy lifestyles.
Both race and SES affect a patient’s likelihood to have cancer diagnosed in an early stage. Net survival rates were calculated for each combination of race, poverty level and stage to examine how each affects cancer’s deadliness.
Effect of race
When net cancer survival rates are examined by neighborhood poverty and stage at diagnosis, a patient’s race had no statistically significant effect. The rates for corresponding groups of white and black Pennsylvanians were very similar, and there were too few Asians and Pacific Islanders in the study to claim any difference.
Effect of neighborhood poverty
For whites and blacks diagnosed at the same stage, higher neighborhood poverty levels correlated with lower five-year net survival. The effect was only statistically significant among whites, but the same pattern can be seen in the rates for blacks. A more definitive judgement could have been made had there been more black residents in the group.
This relationship of poverty and net cancer survival was not clear among Asians and Pacific Islanders, but this is also attributable to the small number of cases.
Insurance
Michael T. Halpern et al. (2008) found that patients with private insurance or Medicare were more likely to be diagnosed with cancers at an early stage, likely because they have higher screening participation. Again, earlier detection often leads to a better prognosis.
In this report, “insured” includes private insurance, Medicare and insurance provided by the military. Patients are grouped by the primary payer at the time of diagnosis. See the technical notes for more details.
Patients’ insurance statuses varied by age. Among the Pennsylvanians included in the survival analysis, younger adults (those between 15 and 64 years old) were five times more likely to be uninsured, and older adults accounted for 90.0 percent of the Medicare recipients. While the net survival estimates controlled for age, comparing within age groups gives a better picture of cancer’s effect.
For each age group, the net survival rate among Pennsylvanians using Medicaid was lower than among those with other insurance. Again, this could be a case of the life tables used to calculate net survival rates not controlling for some important factors. Medicaid often serves citizens with low income and/or disabilities, both of which increase risk but are not reflected in the life tables.
Still, the differences are dramatic: the 2008-2014 net cancer survival rate among Pennsylvanians aged 55 to 64 was 57.8 percent higher if they had insurance instead of Medicaid (rates of 74.3 and 47.1 percent, respectively).
Among the 65 to 74 age group, patients with no insurance had better net survival than those with Medicaid at diagnosis. This extended to the 75 and older age group, which also had similar rates for the uninsured and those with non-Medicaid insurance.
The differences in net survival between insurance payers can also be examined by years since diagnosis. From the time of diagnosis onward, insured Pennsylvanians had a much higher net survival rate compared to those with Medicaid or no insurance.
During the first six months after diagnosis, patients with Medicaid saw higher net survival rates than those without insurance. By the end of the first year after diagnosis, the rate among those with Medicaid was declining more rapidly than among the uninsured. In fact, the changes in net survival for the uninsured between the second and fifth years mirrored those among insured patients.
Possible reasons this may have been:
- Having no insurance means treatment is delayed, so the net mortality risk spikes in the first year.
- These categories represent the primary payer at diagnosis. A previously healthy patient may not have purchased insurance until after his or her cancer diagnosis. The gap during the first year may reflect the period until the patient was covered.