What you should know about N-of-1 trials

Bigger isn’t always better—at least when it comes to the utility of clinical research results. Medical researchers and clinicians tend to believe that the larger the sample size—and the more powerful the study—the more dependable the results.

Keep in mind, however, that many great breakthroughs have occurred as a result of self-experimentation, with a sample size (N) of 1. In fact, smaller sample sizes help to adjust for confounders and can generate more accurate data.^[]

Definition and design

In clinical medicine, N-of-1 trials involve a single patient who participates in multiple crossover trials. In the trials, the patient is periodically switched from an active treatment to a placebo or between active treatments (ie, withdrawal-reversal designs).

According to the authors of a user guide on the subject, these trials emerged in 1953 but it took 3 decades before they caught on.^[] Over the years, many N-of-1 trial efforts languished due to lack of funding, but various N-of-1 trials have been published in the literature.

For those interested in designing an N-of-1 trial, the authors provide further detail. In short, the characteristic elements include blinding, systematic outcomes measurement, and balanced-sequence assignment.

Benefits

More recently, stakeholder interest in N-of-1 trials has taken off. This blossoming appeal coincides with a shift toward patient-centered care and personalized medicine. Unlike the “gold standard” randomized, controlled trial, N-of-1 trials offer data on the amelioration of health problems and improvements in well-being that occur at the individual patient level.

“There is growing recognition of the wide applicability of N-of-1 trials and single-case designs to a number of diverse health disciplines and the value they can bring to clinical research and practice through the focus on understanding individuals,” wrote the authors of an article published in Healthcare.^[]

N-of-1 trials allow for the testing of new treatments to assess impact on the individual level, as well as to compare different dosages/combinations to optimize the regimen. They may also help to indicate when to discontinue treatments that are ineffective. These trials help in settings where there is a paucity of comprehensive empirical evidence.

Clinicians in various settings can leverage the data from N-of-1 trials. For example, the results of an N-of-1 trial can inform discussions of treatment options with patients and enable the process of shared decision-making.

In pragmatic terms, N-of-1 trials can aid in selecting the most cost-effective treatment. Since the trials provide objective evidence of an individual patient’s response, the cheaper option could be chosen. This could be helpful in deciding between two equally effective treatments, a generic vs a brand-name formulation, or different classes of drugs.

It is often not possible to conduct a randomized, controlled trial for a rare disease because of the small patient numbers involved and the wide clinical and genetic heterogeneity in the condition.

Limitations

Indications for N-of-1 trials are limited to contexts where there is substantial uncertainty regarding the comparative efficacy of treatments.

This uncertainty can result from a paucity of evidence—or conflicting results— from randomized, controlled trials, when the evidence is of limited relevance to the particular patient needing treatment.

For obvious reasons, these trials are not appropriate for acute conditions or those that can rapidly progress to catastrophic outcomes. Nor are they applicable to asymptomatic conditions, unless there is an objective marker to assess (eg, blood pressure, ESR). N-of-1 trials are most useful with chronic and stable disease, as well as with conditions whose pathology progresses slowly.

The treatments used in N-of-1 trials should have rapid onset and washout. If the treatments take too long (eg, methotrexate for rheumatoid arthritis), the patient and clinician could lose patience to continue. Moreover, if the carryover of the effects of the agent linger, the washout period would also take a long time, thus lengthening the trial. Interventions with complex dosing patterns and titration are poor candidates for N-of-1 trials.

Authors of a proposed framework for N-of-1 trials made an important distinction about the data derived from these studies.^[]

“Thus, using an N-of-1 trial design to get reliable data on treatment efficacy in one or multiple single patients (with an individual N-of-1 trial or a series of N-of-1 trials design without aggregation) can, under certain circumstances, be considered as an evidence-based improvement of care and not a medical scientific experiment,” they wrote. “Obtaining any form of statistical generalizable evidence on drug effectiveness on the patient group level using a series of aggregated N-of-1 trials design is, however, categorized as a medical scientific experiment in our framework.”

In other words, a single N-of-1 study could be viewed as an evidence-based improvement, whereas a series of N-of-1 trials—replete with statistical analysis—could culminate into a bonafide scientific experiment.

N-of-1 trials in clinical practice

The authors of the user guide make a good case for incorporating N-of-1 trials in clinical practice. They write that “...making practice more like research will create opportunities for developing the clinical evidence base by enhancing systematic data collection on the comparative effectiveness of treatments by real health care professionals treating real patients.”

They noted that with further integration of N-of-1 trials into practice, physicians could become increasingly engaged in clinical research and evidence generation.