RESUMEN
BACKGROUND AND PURPOSE: The conceptualization of brain death (BD) was pivotal in the shaping of judicial and medical practices. Nonetheless, media reports of alleged recovery from BD reinforced the criticism that this construct is a self-fulfilling prophecy (by treatment withdrawal or organ donation). We meta-analyzed the natural history of BD when somatic support (SS) is maintained. METHODS: Publications on BD were eligible if the following were reported: aggregated data on its natural history with SS; and patient-level data that allowed censoring at the time of treatment withdrawal or organ donation. Endpoints were as follows: rate of somatic expiration after BD with SS; BD misdiagnosis, including "functionally brain-dead" patients (FBD; i.e. after the pronouncement of brain-death, ≥1 findings were incongruent with guidelines for its diagnosis, albeit the lethal prognosis was not altered); and length and predictors of somatic survival. RESULTS: Forty-seven articles were selected (1610 patients, years: 1969-2021). In BD patients with SS, median age was 32.9 years (range = newborn-85 years). Somatic expiration followed BD in 99.9% (95% confidence interval = 89.8-100). Mean somatic survival was 8.0 days (range = 1.6 h-19.5 years). Only age at BD diagnosis was an independent predictor of somatic survival length (coefficient = -11.8, SE = 4, p < 0.01). Nine BD misdiagnoses were detected; eight were FBD, and one newborn fully recovered. No patient ever recovered from chronic BD (≥1 week somatic survival). CONCLUSIONS: BD diagnosis is reliable. Diagnostic criteria should be fine-tuned to avoid the small incidence of misdiagnosis, which nonetheless does not alter the prognosis of FBD patients. Age at BD diagnosis is inversely proportional to somatic survival.