Rethinking the appraisal and approval of drugs for type 2 diabetesBMJ 2015; 351 doi: https://doi.org/10.1136/bmj.h5260 (Published 09 October 2015) Cite this as: BMJ 2015;351:h5260
All rapid responses
The recent article by Naci et al. rightly points out that although drugs may be approved on their effectiveness in lowering glucose what we really need are agents that impact upon important outcomes such as the prevention of symptomatic microvascular disease and cardiovascular events (1). The authors suggest that real world evidence of clinical effectiveness should be collected, by randomisation in clinical practice and by using routine healthcare data to monitor outcomes. They suggest this could occur before full drug approval is granted and should be funded in full or part by the pharmaceutical industry (1).
At the University of Surrey we have entered into partnership with Lilly to begin to provide some of these essential real world outcomes data using primary care records. Our preliminary analyses highlight an additional reason to extend trials into the real world; there are significant differences in patient characteristics of people treated with the two newest classes of glucose lowering agents; glucagon-like peptide-1 (GLP1) agonists and sodium-glucose co-transporter-2 (SGLT2) inhibitors. In everyday clinical practice we have found that people with type 2 diabetes treated with these medications have a significantly higher body mass index than those in clinical trials (GLP1 agonists; 37.5kgm-2 in practice, 31.8kgm-2 in aggregated clinical trials, p<0.001. SGLT2 inhibitors; 34.7kgm-2 in practice, 30.6 kgm-2 in trials, p<0.001). This single marker suggests that trial populations are not representative of the populations treated with new glucose lowering therapies, and that the effectiveness of these drugs in everyday clinical practice may be less certain.
If this new era of investigation into real world outcomes is to be funded by the pharmaceutical industry, stringent precautions must be taken to minimise the biases previously identified in industry funded research (2). All trials performed using real world evidence should be registered using existing trial registries. All non-trial research should also be registered in a similar way with demands for high quality research protocols to be submitted before commencement of data extraction and analysis.
We concur that demonstration of drug outcomes should not be confined to glucose lowering but must include outcomes that are relevant to patients. Only by performing well constructed studies in the real world will we establish the effectiveness of new drugs.
1. Naci H, Lehman R, Wouters OJ, Goldacre B, Yudkin JS. Rethinking the appraisal and approval of drugs for type 2 diabetes. BMJ (Clinical research ed). 2015 2015-10-09 10:36:11;351.
2. Lexchin J, Bero LA, Djulbegovic B, Clark O. Pharmaceutical industry sponsorship and research outcome and quality: systematic review. BMJ (Clinical research ed). 2003 2003-05-29 16:48:55;326(7400):1167-70.
Competing interests: AMcG, NM, and SdeL undertake research as part of the newly developed University of Surrey-Lilly Real World Evidence Centre funded by Eli Lilly and Company. RH has no competing interests to declare.
"Do we have clear evidence that this drug for diabetes improves the outcomes that matter to patients?"
So often the answer appears to be - NO.
Thank you so much for this paper, a gleam of hope in the #bigpharma 'feeding frenzy' around people with type 2 diabetes. Which I worry threatens our integrity in the eyes of our patients.
For years I have been disappointed at the rush to accept surrogate measures of success, while accepting side effects such as the loose motions in metformin usage detailed here:
My practice spends £3.936 on drugs for diabetes for every 1000 patients per year, against a CCG average of £7,020 and yet our markers of diabetic care have improved over the last three years using a lower carb higher fat diet described here:
HbA1c is< =59mmol/mol in last 12mths is the national marker. We were average, now improved by 10% to 69.1%, much better than the National average of 61.5%
Obesity: QOF prevalence at Norwood Surgery dropping Was 9.4% now 7.5%, this against a national average of 9.4%
From this it can be deduced that a greater drug spend may not give the best results IF diet is given the priority that its potential for good deserves.
Please, please can we give all patients a three month dietary trial of weight loss and diet BEFORE starting lifelong drugs where the overall outcomes are so uncertain.
Competing interests: No competing interests