The trusted source for
healthcare information and
Quality improvement study hampered by IRB concerns
Investigator: Resubmissions added to costs
Pay-for-performance initiatives, which provide bonuses for physicians, hospitals and other providers who meet certain performance standards, are increasingly used by insurers such as Medicare to try to improve quality and efficiency.
Yet the actual effectiveness of such incentives hasn't been well studied, says Laura Petersen, MD, MPH, director of the Veterans Affairs' Health Services Research and Development Service Center of Excellence in Houston, TX.
Petersen's team set out to test whether VA medical centers would do a better job of adhering to clinical guidelines in their care of hypertension patients if the providers received financial incentives. Medical centers would be assigned to one of four study groups, with incentives received by physicians, health care provider groups, both, or no one.1
However, she encountered a significant obstacle in attempting to obtain IRB approval from 17 study sites. Fourteen of the sites eventually approved the study, all determining it minimal risk. But those approvals came only after multiple submissions and a number of concerns raised by various IRBs.1
Additional submissions were required at every site, Petersen says, and the total time spent in the IRB approval process before the trial could commence was more than 27 months. She estimates that the staff time spent obtaining IRB review cost nearly $170,000, not counting her own salary and those of the principal investigators at the sites.
"If you add in all the opportunity costs, what we could have done with the time, the energy, the brainpower and money — it's highly costly," she says.
And Petersen notes that the sites left in the study when it finally launched were more likely to be academic medical centers and specialty hospitals, which treat more complex patients. This affects the generalizability of the results, Petersen says.
"We really wanted to have a study that ended up being generalizable to community, rural non-academic sites, because that's where much of the health care is delivered and we don't have a lot of data about what works in those sites," she says.
Ethics, notification of patients questioned
Concerns raised by IRBs over the study were varied. Some directly questioned the ethics of pay-for-performance incentives. "Offering money to people to do what is expected of them is not ethical," one IRB wrote. "Is this legal for a research study? If legal, it seems to lead to unethical behavior similar to paying finder fees," was another IRB response to Petersen's submission.
Petersen notes that pay for performance is already in use in the Medicare system. "We could just go and do like Medicare's doing and just pay everybody to reward them, but I think it's important to know whether that's an effective method or not, because if it's not effective, we're all just wasting time and money," she says.
Despite the fact that physicians and their institutions were the actual subjects of the study, some IRBs required that their patients be notified about it. The study team offered to put flyers in the clinics describing the study, but one IRB insisted that patients be notified individually. Petersen says investigators were worried this would introduce bias and the potential for breach of confidentiality of the physicians in the study.
That IRB eventually disapproved the study, saying, "The potential risk to hypertensive patients is too great to justify their involvement."
"How would there be risks to patients from following a national guideline?" Petersen asks. "This is not a guideline that I made up — these are guidelines that have been around for a long time."
'A huge variation'
Petersen's own IRB had no problems with the proposal. And of the 14 sites that approved the study, one of them did so through an expedited process.
"There was really a huge variation — everything from an IRB saying it was too risky to justify involvement to someone doing an expedited review. I don't really think there's a good justification for having such variation."
She believes much of the problem lay in IRBs not having enough experience with this type of quality improvement study.
"I think IRBs aren't really used to them," she says. "With all of the patient-centered outcomes, research initiatives and more attention to improving quality, IRBs need to get more expertise on their boards so they can handle these studies without having these kinds of hiccups."
Petersen suggests that IRBs could tap health services researchers, statisticians or epidemiologists to assist them, just as they would seek out experts to consult on other specialized studies.
"There's specific expertise in judging these studies and their methods' strengths and it's not difficult expertise to get," she says.
She also praises proposed changes to the Common Rule that would revise existing review categories to better match the risk levels of different types of research.
"Is the way that we're currently reviewing these studies really justified given this example, where there were actually 14 IRBs that made a determination and they all decided that it was minimal risk?" Petersen says. "Is it really worth putting studies like this through such a rigorous process? Or is there some other sort of streamlined process that they could go though?"