Issue 15 / 23 April 2012

MORE support is needed for clinical trials comparing existing therapies, including an easing of regulatory barriers, according to a leading research director.

Professor John Simes, director of the NHMRC’s Clinical Trials Centre, said clinical trials comparing existing therapies should not be subject to the same regulatory standards that apply to randomised controlled trials (RCTs) of new therapies.

He said current clinical trials research regulatory requirements included the need to develop detailed research protocols of 50‒60 pages, and issues around clinical trials insurance.

Professor Simes agreed with comments in a JAMA article on proposed changes to the regulation of research in the US, which said that many types of comparative clinical effectiveness research “pose no or only minimal additional risks or burdens to patients over what patients would experience in clinical care”. (1)

He said if patient and doctor agreed that both treatment options were viable, then the risks were minimal and consent and follow-up protocols should be the same as for routine care.

His comments come in response to a themed-issue of JAMA focused on comparative effectiveness research. (2)

However, Professor Simes and other experts raised concerns that many of the studies were observational rather than RCTs.

For instance, a paper looking at survival outcomes among trauma patients receiving different transport options illustrated many of the difficulties of observational comparative effectiveness research, according to a trauma expert. (3)

The study found that helicopter transport was associated with improved survival compared with ground transport, but Professor Peter Cameron, head of the critical care division in the School of Public Health at Monash University, said the study was flawed.

“Everyone loves a helicopter but to actually prove that transport makes a difference is a very difficult thing. The difficulty of comparative effectiveness research is that as soon as you bring in a helicopter, everything changes”, he said.

The data selection process and coding of injuries in the study was also problematic.

“They say they can adjust for potential bias using propensity scores, but that assumes that their data is of sufficient quality in the first place”, he said.

Professor Cameron questioned whether the difference in survival was due to the training of the staff in the helicopter, the short response time, or the fact that helicopters were usually only used in sophisticated integrated response systems. “What is it [that caused the survival advantage]? We don’t know.”

Professor Simes said a retrospective cohort study in the same issue of JAMA, which found that kidney cancer patients had better survival outcomes following partial rather than radical nephrectomy, was also problematic. (4)

“The worry here is that there’s still a major selection bias going on, in that the people getting the radical nephrectomy may have in fact been sicker patients than those getting the partial”, Professor Simes said.

The retrospective study’s results also contradicted the results of a previous RCT, creating a “dilemma” for clinicians, he said.

The JAMA themed issue comes ahead of an MJA clinical trials research summit on 18 May, which will discuss how to increase support for clinical trials groups.

Professor Simes, who is on the summit organising committee, said more support was needed for research that may not have any commercial value, which was the case for much comparative effectiveness research.

“I think there should be more research of what we can call comparative effectiveness studies trying to better define how to use and optimise existing treatments and existing care”.

– Sophie McNamara

1. JAMA 2012; 307: 1589-1590
2. JAMA 2012; 307: 1555
3. JAMA 2012; 307: 1602-1610
4. JAMA 2012; 307: 1629-1635

Posted 23 April 2012