Patient n plus 1

The nth patient is the last patient to enter a randomized controlled trial, having satisfied the restrictive entry criteria — a prescribed age range (rarely old or young), sex (often male), lacking specified comorbidities, taking few or no other medications, alert, oriented and cooperative, eager to participate, compliant, English-speaking and living not too far from the participating clinic or hospital.

Patient n plus 1 is in your office: he or she is older, somewhat confused, on multiple medications (some of which you know about and others, you're guessing, you don't), imperfectly compliant, with most or all of the trial exclusion comorbidities and some others you haven't yet been able to define, and whose clinic visits depend on the personal schedule and driving abilities of an equally aged neighbour.

Patient n plus 1, not being in the clinical trial, is also not in the systematic review, meta-analysis or clinical practice guideline. None of the diagnostic and treatment algorithms you have seen provide a precise pathway for n plus 1. Caring for n plus 1 requires a careful history, reliance on the physical exam and a few easily available diagnostic procedures, a pharmacologic acquaintance with several classes of molecules and an ability to instantly recall the latest warnings about adverse effects and drug interactions — along with an estimate of the likelihood that n plus 1 is actually taking the pills you have prescribed.

You're aware of the clinical guidelines you're supposed to be following, but they don't seem to be written for patient n plus 1. Not to mention that there are several other reasons for your guideline-ennui, including your increasing suspicion that evidence is distorted by publication bias, that adverse event data are underreported and that pharmaceutical companies exert a broad influence on the trials they fund and on the development of practice guidelines, and your awareness that, in practice, one drug is measured against available alternatives, not against the placebo required in a drug-approval trial.

This disconnect between clinical trials and guidelines on the one hand and clinical practice on the other will likely get worse with every major advance in genomics, especially pharmacogenomics, at least in the short term. For example, although there is no doubt that β-adrenergic receptor antagonists (β-blockers) improve the prognosis of patients with congestive heart failure, it is also true that adverse events from β-blocker therapy occur in up to 43% of patients with heart failure.¹ Even small test doses may not be tolerated. Recently, however, genetic variants in cytochrome P450 2D6 (controlling metabolism of β-blockers) and in the β₁-adrenergic receptor (controlling response to β-blocker therapy) have been identified, raising the possibility of identifying patients with congestive heart failure who are more (or less) likely to benefit from therapy or to experience adverse events.²

But to apply this new pharmacogenomic information to patient n plus 1, we will need, as Muszkat and Stein conclude in their recent review, “additional larger studies” that take into account these identified genotypes.³ In the short and medium term, it is likely that ongoing trials targeting even more restricted groups of patients based on their genotypes will make the algorithmic fit of n plus 1 even more exasperating. The trials will be conducted on patients with specific genotypes, while clinical practice will continue with unspecified phenotypes. But eventually, we may hope, the genomification of medicine will reduce the mismatch between patients in trials and n plus 1. Especially when n plus 1 is genotyped.

Without specific RCT evidence relevant to the particularity of n plus 1, how are we to proceed in advising patients? Perhaps the way physicians have always done, with a thorough understanding of the patient's dilemmas and a sound knowledge of basic science. And, most difficult of all, by finding sufficient time to think each problem through. — CMAJ