ObjectivesThe influence of age on intensive care unit (ICU) decision-making is complex, and it is unclear if it is based on expected subjective or objective patient outcomes. To address recent concerns over age-based ICU decision-making, we explored patient-assessed quality of life (QoL) in ICU survivors before the COVID-19 pandemic.DesignA systematic review and meta-analysis of cohort studies published between January 2000 and April 2020, of elderly patients admitted to ICUs.Primary and secondary outcome measuresWe extracted data on self-reported QoL (EQ-5D composite score), demographic and clinical variables. Using a random-effect meta-analysis, we then compared QoL scores at follow-up to scores either before admission, age-matched population controls or younger ICU survivors. We conducted sensitivity analyses to study heterogeneity and bias and a qualitative synthesis of subscores.ResultsWe identified 2536 studies and included 22 for qualitative synthesis and 18 for meta-analysis (n=2326 elderly survivors). Elderly survivors’ QoL was significantly worse than younger ICU survivors, with a small-to-medium effect size (d=0.35 (−0.53 and −0.16)). Elderly survivors’ QoL was also significantly greater when measured slightly before ICU, compared with follow-up, with a small effect size (d=0.26 (−0.44 and −0.08)). Finally, their QoL was also marginally significantly worse than age-matched community controls, also with a small effect size (d=0.21 (−0.43 and 0.00)). Mortality rates and length of follow-up partly explained heterogeneity. Reductions in QoL seemed primarily due to physical health, rather than mental health items.ConclusionsThe results suggest that the proportionality of age as a determinant of ICU resource allocation should be kept under close review and that subjective QoL outcomes should inform person-centred decision -aking in elderly ICU patients.PROSPERO registration numberCRD42020181181.