This summary is based on peer-reviewed literature and current clinical guidelines, cited below.
Background
Reading a clinical study’s abstract and conclusion is not the same as being able to act on it. Pharmacists routinely need to go one layer deeper — into the methods section — to judge whether a trial’s design actually supports the claim being made, and whether its reported effect size is one worth changing practice over.
Clinical Considerations
Randomization and Blinding: Why They Matter
Randomization exists to distribute both known and unknown confounders evenly between treatment arms; blinding (of participants, treating clinicians, and outcome assessors) exists to prevent that balance from being undone by differential care, reporting, or assessment after the fact. The Cochrane risk-of-bias framework treats these as distinct, separately assessable domains — random sequence generation, allocation concealment, and blinding are not interchangeable, and a trial can get one right while failing another.1 When reading a trial’s methods, look specifically for how the randomization sequence was generated and whether it was concealed from the people enrolling patients — inadequate allocation concealment is one of the more common ways a “randomized” trial ends up with unbalanced groups anyway.1
Intention-to-Treat vs. Per-Protocol Analysis
These two analytic approaches can produce meaningfully different results from the same dataset. Intention-to-treat (ITT) analysis keeps patients in their originally assigned group regardless of adherence or crossover, preserving the benefit of randomization and generally giving a more conservative, real-world estimate of effect. Per-protocol analysis includes only patients who completed the study as assigned, which can inflate apparent efficacy by excluding non-adherent or early-discontinuation patients. A pharmacist-led randomized trial of a diabetes management clinic illustrates why both matter in practice: the study reported both ITT and per-protocol analyses for its primary endpoint (HbA1c reduction), an approach that lets readers see whether the effect held up under the more conservative ITT assumptions.2 When a trial reports only a per-protocol result, or the two analyses diverge substantially, that is a signal to look more closely at dropout and adherence patterns before trusting the headline effect size.
Absolute vs. Relative Risk Reduction — and Number Needed to Treat
Relative risk reduction (RRR) is the more dramatic-sounding number and the one more often used in marketing, but it can be misleading in isolation. A drug that reduces fracture risk from 2% to 1% and a drug that reduces it from 20% to 10% both report a 50% relative risk reduction, despite very different absolute benefit. Absolute risk reduction (ARR) — the arithmetic difference in event rates — and its inverse, the number needed to treat (NNT), translate a trial’s result into a more clinically usable figure: how many patients must be treated to prevent one additional event.3 A comparative analysis of osteoporosis therapies found NNTs for preventing one vertebral fracture over 3 years ranging from 9 to 21 across different agents, despite each drug’s trial reporting an impressive-sounding relative risk reduction — a concrete illustration of why ARR/NNT, not RRR, should anchor a head-to-head efficacy comparison.3
Bottom Line for Pharmacists
When appraising a clinical study, check how randomization was generated and concealed, note whether results are reported by intention-to-treat or per-protocol (and whether the two diverge), and always convert a relative risk reduction into an absolute risk reduction or NNT before comparing agents or counseling a patient on expected benefit.
References
- Higgins JPT, Altman DG, Gøtzsche PC, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. https://doi.org/10.1136/bmj.d5928
- Halalau A, Sonmez M, Uddin A, Karabon P, Scherzer Z, Keeney S. Efficacy of a pharmacist-managed diabetes clinic in high-risk diabetes patients, a randomized controlled trial – “Pharm-MD”: impact of clinical pharmacists in diabetes care. BMC Endocr Disord. 2022;22(1):69. https://doi.org/10.1186/s12902-022-00983-y
- Ringe JD, Doherty JG. Absolute risk reduction in osteoporosis: assessing treatment efficacy by number needed to treat. Rheumatol Int. 2010;30(7):863-869. https://doi.org/10.1007/s00296-009-1311-y