Diabetes drugs have been linked to thousands of deaths and hospitalizations over the last decade, an investigation by MedPage Today and the Milwaukee Journal Sentinel found but there is little ability to determine exactly how risky the drugs may be to the public.
That's because the U.S. Food and Drug Administration's reporting system provides only a partial view of the potentially dangerous side effects of the dozens of drugs it approves every year.
This shortcoming is particularly important in the treatment of diabetes, since the agency has approved dozens of drugs based on surrogate measures, such as tests that lower blood sugar levels, rather than proven outcomes such as reducing heart attacks, strokes, blindness or amputations.
Likewise, a Journal Sentinel/MedPage Today examination of 54 cancer drugs approved over the past decade found that 74%were OK'd based on surrogate measures, such as the shrinking of tumors, but not on proof that they extended life. Seldom was there proof of improved quality of life, either.
Independent experts say it is important to have a clear picture of problems attributed to new drugs, so doctors can weigh the side effects against the potential benefits when making decisions on what to prescribe to patients.
"There's gotta be something better than approving drugs based on a lab test and then losing track of them, which is what essentially is happening," said Jerome Avorn, a professor of medicine at Harvard Medical School.
"I think most doctors don't fully understand that the recent drugs approved for diabetes haven't been proven to reduce organ failure and might actually carry risks."
Without proof of real benefit and a growing number of potential harms, the situation puts doctors and patients in a difficult position, said Sandeep Vijan, a physician with the Ann Arbor VA Health System who has studied diabetes drugs.
"Balancing those things out is what we are really concerned about," said Vijan, a professor of internal medicine at the University of Michigan.
One of the few tools available for identifying risks is the FDA's Adverse Events Reporting System, a database that collects case reports from health care providers, pharmaceutical companies and the general public.
A MedPage Today/Journal Sentinel analysis of the case reports from 2004 through March 2014 found about 3,300 deaths and 20,000 hospitalizations in which diabetes drugs were considered the primary suspect.
To account for some of these concerns, the MedPage Today/Journal Sentinel analysis included only reports from health care providers and pharmaceutical companies not the general public. The analysis also excluded case reports that involved diabetes drugs but did not list them as the primary reason for death or a hospitalization.
"The safety and efficacy of these therapies have been evaluated in extensive preclinical and clinical trial programs, which follow widely accepted methodologies of data collection, analysis and independent verification and evaluation," Andrew Davies, spokesman for AstraZeneca, the maker of Byetta, wrote in an email.
Kim Hamilton, a spokeswoman for Merck, the maker of Januvia, wrote in an email that adverse events could be "caused by underlying disease, genetic condition, the medication, concomitant medications or background event that may occur in the population."
Michael Bachner, a spokesman for Novo Nordisk, the maker of Victoza, said the drug has a strong body of supportive evidence from clinical trials and doctors in the field.
"We continue to make our data available and work with all regulatory authorities to provide the most up-to-date medical information for patients and providers," he said.
An FDA spokesman noted the agency does not rely strictly on the database and considers other measures in evaluating a drug's safety.
"It is actually quite difficult to demonstrate causality using these reports alone," FDA spokesman Jeff Ventura wrote in an email.
Nevertheless, the FDA and outside researchers consider reports in the database important signals for issues that need further investigation. But a separate 2009 GAO report on the FDA budget found a system that couldn't keep up.
That report declared: "Although FDA officials told us they received more adverse event reports than staff could review, the agency could not provide data showing the number of adverse event reports staff reviewed during this time period."
The FDA spent roughly 3% of its 2008 budget on tracking adverse events caused by drugs it approved, according to the report.