Physician and researcher

Critical reading of research, statistics and computing.

Hi, I’m Stefano. I’m an interventional cardiologist, a PhD candidate and a fan of statistics, code and AI in medicine. I write about the studies I read, the tools I test (and sometimes a few I build) and what I learn in research.

What I post on Instagram and X is collected here, by topic. Read, search and come back whenever you need.

Recent

StatisticsOct 23, 2026Absolute and relative effect: RRR, ARR, and NNTDefinitions of relative risk reduction, absolute risk reduction, and number needed to treat, and why a relative effect alone does not tell the benefit for a patient.ResearchOct 23, 2026Participant adherence: measurement, run-in and effect on the estimateDefinitions of adherence, crossover and withdrawal, the effect of low adherence on power and on the estimate, the run-in period, and methods to measure adherence in clinical trials.StatisticsOct 23, 2026Allocation concealment and baseline imbalanceExplains concealment of the randomization sequence, its empirical link to baseline imbalance, and how to interpret and adjust for differences in a trial's table 1.ResearchOct 23, 2026Crossover, factorial, withdrawal and cluster designsExplains the variations on the parallel trial: crossover, factorial, withdrawal and group allocation, with the assumptions of each and what they require of the reader.ComputingOct 23, 2026claim-check: a Claude skill to find the evidence behind a claimFree Claude skill that searches PubMed for the evidence behind a manuscript sentence, quotes the exact passage and keeps a per-manuscript Excel evidence table.ComputingOct 23, 2026sr-screening: a Claude skill for systematic review screeningFree Claude skill that runs a systematic review from the PICO question to the extraction table and PRISMA 2020 counts, with human approval of every exclusion.Critical reviewSep 17, 2026Fauchier et al.: reduced-dose DOAC versus antiplatelet strategies after left atrial appendage occlusionAfter LAAO, the SAPT advantage over DAPT survives the paper's own sensitivity analyses. The reduced-dose DOAC advantage over DAPT does not.Critical reviewSep 5, 2026SELUTION DeNovo HBR: sirolimus-eluting balloon versus DES in high bleeding riskThe SELUTION DeNovo HBR substudy describes good results with a sirolimus balloon but has too few events to compare it with DES.