Pharmacoepidemiology research: delivering evidence about drug safety and effectiveness in mental health

Abstract

There is a need for research that provides an evidence base for the pharmacotherapy of people with mental disorders. The abundance of digital data in recent years has facilitated pharmacoepidemiology in the form of observational comparative effectiveness studies at the population level. Advantages are large patient samples, coverage of under-researched sub-populations and naturalistic conditions. Pharmacoepidemiology is also cheaper and quicker to carry out than RCTs, meaning that issues regarding generic medication, stopping medication (deprescribing) and long-term outcomes are more likely to be addressed. Methods can also be extended to pharmacovigilance and drug repurposing. Drawbacks of observational studies come from the non-randomised nature of treatment selection, and the inherent risk of confounding by indication. Potential methods for managing this may include active comparison groups, inter-individual designs, propensity scoring and instrumental variables. Many of the more rigorous pharmacoepidemiology studies have been strengthened through multiple triangulated analytic approaches to improve confidence in inferred causal relationships. With these developments in data resources and analytic techniques, it is encouraging that guidelines are beginning to include evidence from robust pharmacoepidemiogical studies alongside RCTs. Collaboration between guideline-writers and researchers involved in pharmacoepidemiology may help researchers ask the questions that are important to policy-makers and ensure that results get integrated into the evidence-base. Further development of statistical and data science techniques, alongside capacity building in terms of data resources, a wider researcher base and public engagement, will be necessary to take full advantage of future opportunities.

So near yet so far: why won’t the UK prescribe medical cannabis?

“…Indeed, in line with rapid developments in data resources and analytical techniques, many guidelines are now beginning to include evidence from robust observational pharmacoepidemiological studies alongside RCTs. 13 But even more important are the n=1 trials, for these are the core of medical practice since every time a medicine is prescribed an n=1 experiment is being conducted. In some patients the experiment works and in others it fails, the patient either does not respond or the adverse effects outweigh the therapeutic benefit.…”

Evidence-based guidelines for the pharmacological treatment of schizophrenia: Updated recommendations from the British Association for Psychopharmacology

“…Ia: evidence from meta-analysis of randomised controlled trials (RCTs) Ib: evidence from at least one RCT IIa: evidence from at least one controlled study without randomisation IIb: evidence from at least one other type of quasi-experimental study III: evidence from non-experimental descriptive studies, such as comparative studies, correlation studies and case-control studies IV: evidence from expert committee reports or opinions and/ or clinical experience of respected authorities This categorisation is most appropriate for questions of short-term efficacy. Similar taxonomies for other types of research question, including long-term impacts and real-world effectiveness, do not yet exist (Davis et al, 2019) and the following is proposed.…”

Benefits and Harms of Statins in People with Dementia: A Systematic Review and Meta‐Analysis

“…The evidence needed by clinicians and patients with dementia will almost certainly require the inclusion of observational studies in the evidence base, as reflected here. 23 Observational research has inherent problems with confounding; thus a strength of this review is that it specified all observational trials needed to adjust for the main confounder of vascular risk, and it used an established tool to assess the quality of observational studies. 27,43 Even with a wide range of study types allowed, few studies were found, and none directly addressed the question.…”

“…Therefore, it is important at least to consider whether observational research can add to the evidence. 23 This review aims to investigate the wider health effects of statins compared with no pharmacologic lipid management in people older than 65 years with any type of dementia, with the primary outcome the reduction of MACE, by reviewing trials and observational research.…”