The authors say the findings coincide with the movement toward consumer-driven health plans.
Those plans were designed to make consumers more cost conscious by requiring them to pay more out of their own pocket for care.
Harvard researchers randomly contacted more than 2,300 U.S. bankruptcy filers in early 2007. Roughly half of those filers agreed to discuss the reasons for their bankruptcy during a follow-up telephone interview.
Nearly two-thirds of all the bankruptcies in the study could be attributed to medical problems, the researchers found. Most of the filers had medical debts of more than $5,000, or 10 percent of pre-tax family income.
The average medical bill for families with private insurance totaled nearly $18,000. The average bill for uninsured families was nearly $27,000.
"In order to protect people, you have to have coverage that's comprehensive and ... that doesn't disappear when you lose your job."
Study co-author Dr. Steffie Woolhandler, an associate professor of medicine at Harvard and a primary care physician, is not surprised by the findings. She said health care costs rose considerably during the study period.
Families also had to pick up a much greater share of those costs as more employers switched to consumer-driven plans that had higher deductibles and co-payments, she said.
"Unfortunately, many people don't have that kind of money. And if the insurance plans require them to pay hundreds or thousands of dollars out of pocket, it can driven them into bankruptcy," Woolhandler said. "So I think that really explains the increase."
The study found that most of the people bankrupted by illness were middle class before their medical crisis occurred. In many cases illness lead to job loss, followed by the loss of insurance.
The results make a strong case that private health insurance is failing people, because it's full of coverage gaps and it's easy to lose health coverage, according to Woolhandler.
"If we have a reform such as that which Mr. Obama is proposing, that just gives people access to private insurance, you will not solve the problem of medical bankruptcy," she said. "In order to protect people you have to have coverage that's comprehensive and that is continuous, that doesn't disappear when you lose your job."
Two of the study authors, including Dr. Woolhandler, are members of the group Physicians for a National Health Program. The organization supports a single-payer national health insurance program.
Woolhander said her personal views did not affect her interpretation of the bankruptcy data, which she said was peer-reviewed and approved for publication in a prestigious medical journal.
Previous medical bankruptcy research conducted by the same researchers has been widely cited by policy leaders, including President Obama.
The Harvard findings are important and should help motivate Americans who want to see lawmakers create a public insurance option, said Joel Albers, a health economics researcher with the Universal Health Care Action Network Minnesota -- a group that advocates for a single-payer health care system.
"If we're going to really get a genuine Medicare For All option it's really going to have to come from the populace," Albers said. "People have to really speak out on this and have a voice on it, and just be an active participant in making that come about."
The insurance industry says the study findings don't tell the whole story, however. Eileen Smith, with the Minnesota Council of Health Plans, said the bankruptcy study places the blame on private health insurance when the main problem is the rising cost of care.
"Health care costs are going to continue to go up regardless of what kind of funding mechanism is in place, whether it's private health insurance or in other parts of the world where you see more government support," Smith said. "We need sustainable coverage for all, and to do that we all need to work on slowing the rising costs to have a long-term solution."
The Harvard study analyzed bankruptcy cases from all 50 states. But researchers did not calculate the percentage of medical bankruptcies among individual states. So it's possible that great variation exists from state to state relative to whether medical costs contributed to the bankruptcies.
For example, Minnesota had 15,473 consumer bankruptcies last year. But it's not known if the study's findings mirror what's happening in Minnesota.
The research is published in the online edition of the American Journal of Medicine. It was paid for by a grant from the Robert Wood Johnson Foundation.