Pooling 15 studies and 117,096 post-heart-attack patients, SGLT2 inhibitors were linked to lower heart failure hospitalization (OR 0.74, 95% CI 0.63-0.88) and lower all-cause death (OR 0.78, 0.67-0.90). The mortality signal came almost entirely from propensity-matched observational studies (OR 0.64); randomized trials showed a favourable but non-significant trend (OR 0.87, 0.76-1.01). Dapagliflozin and empagliflozin performed similarly, early or delayed initiation both reduced hospitalization, and cardiovascular death, recurrent MI and stroke did not differ.
The gap between study designs is the real story. Matched observational cohorts are prone to healthy-user and immortal-time biases, and the dramatic OR of 0.64 is exactly what those biases tend to produce, so the RCT estimate is the more credible one. Because observational studies supply most of the 117,000 patients, the headline pooled mortality figure overstates certainty. The authors also flag small-study effects and possible publication bias, which further tempers the heart failure result. Still, the direction fits what trials in heart failure and chronic kidney disease have established, and the null findings for recurrent MI and stroke suggest the benefit works through haemodynamic and cardiorenal pathways rather than atherothrombotic ones. For adults who have had a heart attack, this supports raising SGLT2 inhibitors with a cardiologist, particularly when diabetes, kidney disease or reduced heart function is present, but it does not settle a mortality benefit. Expect this to be confirmatory rather than paradigm-shifting; larger dedicated post-MI trials will decide whether routine use is warranted.