Researchers then checked records for those patients for the five years following. They found those who were uninsured were 1.6 times more likely to die in five years than those with private insurance.
More specifically, 35 percent of uninsured patients had died at the end of five years, compared with 23 percent of privately insured patients.
Earlier studies have also shown differences in cancer survival rates of the uninsured and insured, but they were limited to specific cancers and certain geographic areas.
The new findings are consistent across different racial groups. However, the fact that whites have better survival rates cannot be explained by insurance status alone, said Elizabeth Ward, the study's lead author.
The researchers were not able to tell if the numbers were influenced by patients' education levels, or by other illnesses.
Experts said the study also hints at problems with quality of care after diagnosis: such as whether the patient got the appropriate operation from a high-quality surgeon, whether the tumor was thoroughly evaluated by a high-quality pathologist, and whether there was access to needed chemotherapy and radiation.
"The differences that we see in outcomes after people are diagnosed, even among those with early stage disease, suggests that problems with quality of care may be an important reason," said Dr. John Ayanian, professor of medicine and health care policy at Harvard Medical School. He didn't participate in the cancer society study.
The study makes an even stronger statement about the role insurance plays in the timing of screenings and how that can raise the likelihood of a late-stage diagnosis, experts said.
A Kaiser Family Foundation survey last year of 930 households that dealt with cancer found that more than one in four uninsured patients delayed treatment - or decided not to get it - because of the cost.
Such was the case of Edward Hicks.
The retired laborer, had surgery for colorectal cancer in 2005 and was thought to be clear of the disease. Chemotherapy was suggested after the surgery, but he didn't get it.
In February of this year, his wife grew worried when he lost energy and appetite. In April, he told her he felt a lump in his stomach.
Hicks, who lived in Fort Meade, Fla., couldn't get an appointment with a specialist, but a family doctor checked him into a hospital and specialists saw him in late May. They said he was terminal but that chemotherapy might extend his life a little, his wife said.
She was able to get donated chemotherapy drugs from a pharmaceutical company, but it took time to arrange the treatments, which didn't start until mid-June. Meanwhile, her husband's health deteriorated. In July, after just a few treatments, he stopped the chemo, saying it was too hard. He died on Aug. 21, at age 64.
Friends and family told Peggy they believe he would have lived longer had he got chemo earlier, when he was stronger. She doesn't agonize over that, she said, trusting in God's will.
But the devil's in her mailbox - she is facing a $21,000 hospital bill and other costs from his death.