Saturday, 1 June 2013

From Skullduggery to Scurvy: Patient Centred Evidence Based Medicine

James Lind (1716-1794)

Good clinical practice depends on the following: the individual expertise of the clinician, an understanding of the best available external evidence and informed patient choice. Education of patients is increasingly important so that they can make informed choices about their own treatments. The importance of clinical evidence must be emphasised to patients as well as all clinical staff.
Clinical evidence typically comes from clinical trials. A comparison of a new treatment with a control, ie. an existing treatment or placebo (a fake treatment with no plausible biological effect), is known as a controlled trial. The first documented clinical trial was performed by James Lind (pictured above), a naval surgeon who proved in 1747 that citrus fruits (now known to contain vitamin C) cured scurvy (which at the time killed a huge proportion of sailors). He took 12 sailors with scurvy and divided them into six pairs. Each pair was given a different diet which varied from seawater, a mixture of garlic, mustard and horseradish, and citrus fruits. The sailors given citrus fruits made a striking recovery.
Until this point, treatments were largely unproven and often quackery (skullduggery even). An example of this was the popular practise of blood-letting which most famously killed the first US president, George Washington in 1799. Over 5 litres of blood were drained over a day to treat his throat infection, resulting in his death.
Now clinical trials often involve hundreds if not thousands of patients. These patients must be randomised, in other words selection of patients for both (or more) treatment arms of the trial must be completely random. This ensures that there is no difference between the treatment group and the control group at the beginning of the trial. For example, randomisation ensures that healthier patients are not selected by chance to receive the new treatment. After receiving informed consent, patients must not know whether they have received the new or control treatment (single blinded trial). Where possible, physicians must also not know which treatment they have given (double blinded trial). The best possible trial therefore is a randomised, controlled, double blinded trial.
The best available evidence for a treatment does not end there. One positive trial should stimulate other trials to confirm the evidence. A single positive result can be the result of sheer luck (despite statistical significance). It is important that ALL available evidence is considered by the treating clinician. These are best obtained from rigorous systematic reviews of treatments, such as those provided by the international non-profit Cochrane Collaboration.
Negative results MUST also be included in any analysis of the best available evidence. Evidence that a treatment does not work is just as important as evidence that it might. However, it has been estimated that only half of all registered and completed clinical trials are published. Positive trials are twice as likely to be published than negative ones. The reasons for these are complex and being addressed by the AllTrials campaign. One example of this is the failure of the pharmaceutical company Roche to disclose all trial data concerning Tamiflu to the Cochrane Collaboration. The UK government has spent £500 million stockpiling Tamiflu on the basis of data that is not complete.

All clinical staff should be trained in interpreting clinical trial data and placing this in the context of best available clinical evidence. As patients play a more central role in their treatment, we should also make a bigger effort to educate them on how to interpret the evidence about their own treatment.
May 20th marked International Clinical Trials Day, commemorating the anniversary of the very first clinical trial by James Lind