Clinical trials are too seldom directed at the real interests and concerns of patients and healthcare professionals, a meeting heard this week.
One example given by Dr David Tovey, editorial director for BMJ Knowledge, at the conference in London organised by the James Lind Alliance and The Lancet was the proliferation of clinical trials comparing Pfizer’s erectile dysfunction drug, Viagra (sildenafil citrate), with placebo.
An analysis published by BMJ Clinical Evidence in March 2006 found one systematic review of sildenafil versus placebo (2000), which included 27 randomised controlled trials (RCTs), as well as 23 further RCTs on the same subject. The performance of Viagra against placebo was “not a question that anyone is seriously asking anymore,” Tovey commented.
There were many more instances of misdirected clinical research in less specialised therapy areas than erectile dysfunction. The James Lind Alliance, which promotes public and professional awareness of uncertainties about treatment effects, has been consulting with patients and clinicians with the aim of identifying treatment uncertainties and setting research priorities for asthma.
As Sir Ian Chalmers, co-ordinator of the Database of Uncertainties about the Effects of Treatment, told the conference that, out of more than 200 questions submitted to the James Lind Alliance by asthma patients and their carers, more than 100 addressed the long-term effects of steroids and other treatments. Yet there were no robust clinical data available on this topic, he pointed out.
One fundamental problem with asthma trials was that they tended to be conducted in the wrong setting, Sir Ian noted. Most studies took place in hospitals and not in primary care, where the vast majority of patients were treated.
The pitfalls of extrapolating from one setting to another can also cloud clinical judgment on a pan-national scale. Tovey cited a systematic review by the BMJ last year of HIV/AIDS management in resource-poor countries.
It found that treatment models were being exported directly from resource-rich nations such as the United States, where HIV/AIDS largely afflicted gay men or intravenous drug users and there were few complications with co-morbidities such as malaria. In poorer countries, by contrast, AIDS/HIV was largely a heterosexual condition that was exacerbated by local factors such as endemic malaria and lack of a health infrastructure.
Other hinderances to real-world clinical research included setting parameters too broadly. For example, the BMJ had conducted an as yet unpublished review of heart failure and diabetes, and had found hardly any studies addressing heart failure in diabetics. There was, however, plenty of evidence on heart failure in sub-sets of diabetics.
Another problem was too stringent exclusion criteria. In a study of ‘wait-and-see” prescribing for the treatment of otitis media in the Journal of the American Medical Association, 776 people were found to be eligible for the trial yet nearly 500 of these (notably children) were left out of the analysis.
Some interventions were simply under-researched, such as the role of antidepressants in post-natal depression, Tovey told the conference. Or the interventions were poorly defined: what, for example, constituted “exercise” in the context of low back pain? There were also too few trials comparing drug with non-drug interventions. And in most trials the harm caused by treatment interventions was “really poorly” represented compared with the corresponding benefits. In practice, though, a prescribing decision was always a trade off between benefit and harm.
The key question for physicians, Tovey commented, was how much the patient in the study resembled the patient in front of them. This was about defining the right (representative) patient populations, identifying the right questions to be addressed, conducting systematic reviews and targeting outcomes that were important and meaningful to patients, such as fractures rather than bone mineral density.
The pharmaceutical industry contributed to the perception and communication gaps between researchers, clinicians and patients by selecting patients for clinical trials who would derive the maximum benefit from their therapy, argued Stephen Holgate, MRC Clinical Professor of Immunopharmacology at the University of Southampton.
Clinical trial patients accounted for around 2% of the real-world asthma population and the selection criteria were so strict that they left only a “fraction” of asthma patients eligible for study participation, he said.