Personalized bacteriophage therapy uses preparations selected, adapted or engineered against a patient’s bacterial isolate. This review aimed to synthesize human evidence on personalized or strain-matched therapy and distinguish it from fixed-product randomized evidence.

Reporting was informed by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement. Searches ran through 5 February 2026. Screening and extraction were performed in duplicate. Methodological limitations and certainty were appraised descriptively rather than through completed risk-of-bias forms or GRADE profiles. Findings were synthesized narratively. The review was not prospectively registered.

The search flow contained 468 records, 346 after deduplication, 43 assessed full texts and 13 reports mapped to 10 study datasets. Four reports described personalized cohorts or series, six provided case-level or mechanistic evidence, and three were fixed-product randomized trials. In the largest series, clinical improvement occurred in 88/114 infection episodes (77.2%) and pathogen eradication in 65/106 episodes with follow-up (61.3%). Heterogeneity, co-interventions, probable overlap and selective reporting limit attribution.

Personalized phage therapy shows clinical signals in selected rescue settings, but a causal treatment effect remains unproven. Comparative studies with standardized matching, manufacturing, outcome assessment and longitudinal monitoring are needed.