In The Review Trish Greenhalgh1 refers back to the halcyon days of editorials and opinion pieces in medical journals that relied on an author’s bias, speculation, and conjecture that passed for academic debate in the mid-1980s. She decries Cochrane reviews as narrow, boring articles that seldom answer the ‘messy context’ of clinical practice. Sorry, but we don’t agree. Using narrative reviews that do not systematically identify, assess, and synthesise information produces evidence that is biased, misleading, and may harm patients.2
Surely informed clinical practice means telling our patients about what we know is effective, what we know is ineffective, and what we are uncertain about? The Cochrane Collaboration and other useful sources of synthesised evidence such as the BMJ publishing group’s Clinical Evidence demarcate uncertainty and allow us to use regularly updated evidence when helping our patients. What is wrong with this? Every doctor, irrespective of their clinical discipline, cannot possibly keep up with the burgeoning medical literature. In 2009 1.5 million articles were published in 25 400 medical journals. In narrower disciplines than general practice (diagnostic imaging in cardiology) it is estimated a doctor would have to read 95 articles every day to keep up to date.3 One of several solutions lies with clinical summaries that access and synthesise systematic reviews of evidence that address focused clinical problems. Indeed there is now an established hierarchy of pre-appraised synthesised evidence that enables more efficient access to clinically relevant information. The base of this information pyramid is the individual original studies. It then moves to ever-increasing synthesised evidence: synopses of studies, (evidence-based medicine, ACP journal club); synthesis (The Cochrane Library); synopses of synthesis (DynaMed); summaries (National Clinical Guidelines, Clinical Evidence), and ends up with point-of-care decision support, a final ‘systems’ layer.4
The Cochrane Collaboration, and the explicit, systematic, and transparent methods that are used, are prime movers in the information revolution that enables safer patient care. Let us not forget that it is Cochrane authors who have shown how the research agenda can be misled at the inception, funding, publication, and reporting stages.5 We do not have to look far for examples of drugs that had disastrous consequences for some patients (rosiglitazone and rofecoxib); and other drugs that have questionable clinical benefit (gabapentin and oseltamivir).5 While sometimes ‘boring’, Cochrane reviews are seldom irrelevant. A recent example of a Cochrane review that has important clinical and policy implications concerns the use of general health checks in adults, shown not to reduce morbidity and mortality, despite the fact that the UK government began a policy of health checks for adults in 2009.6
We urge readers to take a more critical view of the medical literature and its relevance to their clinical practice. Cochrane reviews remain key components for GPs who wish to remain safe and effective doctors.
Notes
Competing interests
All authors are active members of the Cochrane Primary Health Care Field and Cochrane Collaborative Review Groups.
- © British Journal of General Practice 2013