Author: Samuel Tan
Randomized controlled trials (RCTs) first rose to prominence in the mid-20th century, heralded as a systematic, rigorous and efficient alternative to case studies and physician testimonials. The initial breakthroughs were numerous – streptomycin for pulmonary tuberculosis, heparin for atherosclerosis – and RCTs rapidly grew to their current position as the pre-eminent study design in medicine. Nonetheless, RCTs are not without their weaknesses, many of which significantly impede their applicability.
The Limits of RCTs
Methodologically, RCTs are well-suited to evaluating pharmacological interventions. However, RCTs evaluating non-drug interventions, such as surgery and psychotherapy, often face challenges with control selection and blinding. Efforts to compensate for this, such as ‘sham’ surgery, can be invasive or unfeasible. In addition, RCTs are costly and protracted enterprises. Lengthy RCTs run the risk of lagging behind innovation, in which new or refined interventions arise during the course of an existing trial. Another significant downfall of RCTs is their tendency to emphasize the ‘average patient’, which can be problematic in heterogeneous study populations. While RCTs may accurately measure population-level responses to interventions, they often do not capture the individual response. This is particularly pertinent in highly multifactorial health conditions, such as insomnia and chronic pain.
Neuropathic pain illustrates this issue well. In a recent meta-analysis of RCTs, Finnerup et al. (2015) identify four gold-standard interventions for neuropathic pain: duloxetine, gabapentin, pregabalin, and topical capsaicin. However, each of these only achieves adequate pain resolution in 10-15% of cases; often, there is no indication as to whether a certain intervention will work or not for a particular person. Thus, despite the large evidence base, many patients have to trial a range of therapies before finding one that provides them with acceptable pain relief.
Focusing on Individuals
Who could talk about healthcare these days without bringing up personalised medicine? As a clinical philosophy, it strives to treat patients as individuals, taking into account their unique genetic, behavioural and environmental profiles. While personalised medicine can be a genetics-driven approach, it also lends well to N-of-1 trials and Single-Case Experimental Designs (SCEDs), which are increasingly being implemented as flexible and robust alternatives to RCTs for making treatment decisions for individual patients.
N-of-1 trials and SCEDs use participants as their own control, and draw comparisons across time periods rather than between participants (as demonstrated in Figure 1). They are able to evaluate the value of a therapeutic intervention for a given participant, as well as identify the best of multiple treatment options. Because each participant receives all treatments, N-of-1 trials avoid participants receiving only placebo, and can more readily maintain therapeutic benefit throughout the study than in RCTs. While individual N-of-1 trials are only relevant to the corresponding participant, data from a series of N-of-1 trials can be aggregated to generate RCT-level evidence for a population.
By definition, N-of-1 trials take into account a patient’s unique disease presentation, as well as their genetic and environmental factors. This makes them especially suited to chronic, stable, and aetiologically complex conditions, such as chronic pain and insomnia. N-of-1 trials offer a particular set of advantages and applications, and are a strong complement to RCTs in many fields. Furthermore, they are relatively quick and cost-effective to execute. Due to these strengths, N-of-1 trials are often able to guide long-term individual treatment decisions more precisely than is possible through RCTs alone.
About the Author
Samuel Tan is a medical student at The University of Queensland and is interested in the application of N-of-1 trials in medicine.