TORONTO (Canadian Press) -- Cancer patients from poor neighbourhoods have a greater chance of dying prematurely than their wealthier counterparts, says a new study, describing a problem that persists despite universal health care in Canada.

Much of the literature on the topic, especially in the United States, has partially explained this disparity by saying it is due to the fact that poorer people are diagnosed when their cancer is at a later stage. But this particular study of Ontario patients didn't find that to be the case.

"I think a lot of us were attributing much more of an impact of the stage at diagnosis, and this study really has highlighted that no, that's not the major factor," Heather Chappell, director of cancer control policy at the Canadian Cancer Society, said in reaction to the findings.

Chappell said it means researchers and the medical community need to be looking at what other reasons might be at play.

"There still is a survival disparity and we need to work to understand why that is the case and to correct that difference in outcome," agreed lead researcher Dr. Christopher Booth of the Queen's University Cancer Research Institute.

The study was published online Monday in Cancer, a journal of the American Cancer Society.

It looked at median household incomes from the 2001 Canadian census and used the Ontario Cancer Registry to identify patients diagnosed from 2003 to 2007.

The chance of a woman from a poor community being alive five years after diagnosis of breast cancer is 77 per cent, compared to 84 per cent for affluent women, Booth said.

Fifty-two per cent of patients with colorectal cancer from the lowest socioeconomic groups are alive five years after diagnosis, compared to 60 per cent of those in the most affluent communities.

"These are important and meaningful differences," Booth said in an interview from Kingston, Ont.

"If we had a form of chemotherapy or cancer treatment that led to an improvement or difference in five-year survival of seven, eight, nine per cent -- the order of magnitude we're seeing with these differences -- it would be a blockbuster home run as far as cancer treatment advances."

As for the fact that there were only modest differences in stage of diagnosis, Booth said it could be because the universal health care system allows Canadians of all economic backgrounds to have access to screening and specialists.

With stage of diagnosis being pretty well discounted as a factor in the disparity, it leaves researchers wondering why the differences in outcome do exist between income levels.

"If it's not stage of disease, what's driving this? And there's a whole host of factors, so it might be at the tumour level. It might be that people with different backgrounds have more aggressive biology in their disease. It might be other aspects of health that are correlated with socioeconomic status," Booth said.

Chappell agreed, noting that people with lower socioeconomic status have more co-morbidities -- other types of disease that may have an impact on their cancer treatment.

"And smoking rates may be another significant fact. And what was interesting was in the study they looked at six specific types of cancers, and every one of them is related to tobacco use," she observed.

In addition, she pointed to a cancer drug access report last year that showed one in 12 Canadians face catastrophic drug costs, and people with lower incomes and seasonal and self-employed workers usually don't have access to affordable insurance for those cancer drugs.

"So we already have identified that as a problem, and we've been advocating the federal and provincial governments to work together to create a national catastrophic drug program," Chappell said.

Mary McBride, senior epidemiologist at the B.C. Cancer Agency, said researchers need to take a more in-depth look at factors that might fit under the broad umbrella of socioeconomic status, such as income and education levels, number of years they've been in Canada and potential language barriers.

"Their health-seeking behaviours and their familiarity with the health system might affect their ongoing care and subsequent survival," she said from Vancouver.

McBride noted that the study only looked at data for cancer patients being treated at eight regional cancer centres in Ontario, and it would be interesting to look at the entire cancer population in the province, and Canada.

Chappell also said more research is needed to see if the findings would be the same in other provinces, and to identify factors that may be having an impact on survival.

"Once we have a better understanding of those factors affecting survival, programs that are effective at addressing those would have a better chance of affecting survival for those patients."

Copyright The Canadian Press, 2010