Co-Q10 has potential to be a clinically effective and cost-effective intervention for HFrEF, if prescribed. At a prescribed cost of < £30 per month, it is relatively inexpensive and appears to have few side effects (although it is more costly than the standard prescribed background medication, which costs around £2 per month). However, given concerns about risk of bias and the applicability of the existing evidence, before considering NHS prescription there is a need to confirm the meta-analysis findings to establish whether or not co-Q10 is genuinely effective in a typical UK population.
Additional primary research could strengthen the evidence base considerably. Given that co-Q10 has not been subject to the rigour and scrutiny of drug-licensing processes, if it were to be offered as an NHS treatment it would seem important that robust independent trial data were available to underpin that decision. Furthermore, although inexpensive and potentially cost saving to the NHS, making co-Q10 routinely available to all HFrEF patients would incur purchase costs of around £113M per year. It would, therefore, be important to understand to what extent these costs might be defrayed by reduced hospitalisation and justified by extending or improving quality of life. Moreover, adding to the pill burden of patients commonly already coping with taking five or so disease-modifying agents per day comes with a responsibility that each has been tested thoroughly and shown to be beneficial.
Should further research confirm the benefits of co-Q10, suggested in the meta-analysis, and its cost-effectiveness, as suggested by our economic model, this would support consideration of NHS prescribing, making co-Q10 available to all relevant CHF patients with a reduced ejection fraction and providing equity of access. Conversely, if further research found benefit to be limited, then this would provide important information for those who currently purchase co-Q10 at their own expense.
Implications for service provision
Currently co-Q10 is not available on NHS prescription. CHF patients who wish to use co-Q10 purchase it at their own expense. Co-Q10 is relatively inexpensive when prescribed and appears to have few side effects. Our analyses have shown that co-Q10 has potential to be a clinically effective and cost-effective intervention for HFrEF patients, from an NHS perspective. However, given concerns about possible bias, applicability and plausibility of meta-analysis effect sizes, stronger evidence may be needed before considering NHS prescription, particularly as co-Q10 has never been through the rigour and scrutiny of drug-licensing processes.
If the variation in the size of benefits on the symptoms observed between the trials included in the meta-analysis reflects differences in patient phenotypes, a targeted approach might be more appropriate, limiting use to those with the most to gain. To the best of our knowledge, there is currently no robust evidence to inform targeting, but this could be generated by a new trial.
Suggested research priorities
Given concerns about possible bias, applicability and plausibility of effect size, and as co-Q10 has never been through the rigour and scrutiny of drug-licensing processes, an adequately powered placebo-controlled RCT of co-Q10 in a typical UK HFrEF population may be warranted. Such a trial would seem to be a necessary step before NHS prescription could be considered. Our analyses suggested that the value of reducing decision uncertainty is highly likely to outweigh the costs of a new trial. The amount of uncertainty that is likely to be resolved by a new trial would need to be assessed against the cost of the trial (based on its design features) to ensure that any further research is an efficient use of NHS resources. However, the cost of a new trial is likely to be much less than the EVPI estimates.
A new trial might be designed to be relevant to primary care, to recruit typical HFrEF patients, based on clinical risk factors, and to resolve questions about whether or not co-Q10 could be particularly beneficial for patients taking statins, stratified by statin use. Outcomes should be patient centred and include mortality, CHF hospitalisation and QoL assessment. Trial duration would need to be 3–5 years to adequately capture long-term outcomes and, ideally, the trial would be linked to routine data sets to allow hospital utilisation to be captured.
A trial of patients with HFpEF could also be considered. HFpEF is a heterogeneous condition with different underlying causes and its response to treatments is likely to vary phenotypically. In the absence of existing evidence, or plausible biological hypotheses about which types of patient might benefit from any particular treatment, trials would need to be powered to detect effectiveness within phenotype subgroups and it would need to be anticipated that an intervention may be ineffective for a proportion of the trial population. Although not eligible for our systematic review, searches identified one trial72 of 30 HFpEF patients, which found no short-term improvement in diastolic function. Results from two current trials122,123 (aiming to recruit 276 and 60 participants, respectively) may provide signals of whether or not co-Q10 is potentially effective in HFpEF and whether or not a trial of co-Q10 in HFpEF would be warranted.
Any new trial should have a clear data-sharing policy in place to enable future independent scrutiny and inclusion in additional research, thereby maximising the utility of the data generated by those patients participating in the trial.
