Bacteriophage Research

Urgent note

Phage therapy for infections is a hospital matter. A serious infection with fever, confusion, spreading redness or breathing trouble needs urgent medical care (in the US call 911). Nothing here is advice to treat an infection yourself or to obtain phage products. Evidence labels: preclinical (cells/animals only), promising but limited, mixed, insufficient evidence.

What phages are

Bacteriophages are natural viruses of bacteria. Therapy uses selected phages, sometimes as a cocktail of several, against a specific bacterial infection. [1] A 2022 systematic review in The Lancet Infectious Diseases describes the setting: antibiotic resistance is rising, and the widespread use of phage therapy in routine practice is held back by a scarcity of safety and efficacy data from trials run to current regulatory standards. [2] Phages target particular bacteria, which is both the attraction and the practical challenge.

What randomized trials found

• Burn wounds (PhagoBurn). In a phase 1/2 trial, 27 patients with Pseudomonas aeruginosa burn infections received a 12-phage cocktail or standard silver sulfadiazine cream. The trial was stopped for insufficient efficacy. The time to a sustained reduction in bacteria was longer with phages (median 144 hours versus 47 hours). The phage preparation had lost potency during manufacturing, so patients received much lower doses than planned. [1] Label: mixed: a safety signal without efficacy, and a product-quality problem.
• Urinary tract infection. In a randomized trial in men undergoing prostate surgery, 97 analyzed patients got intravesical phage, placebo or antibiotics. Normalized urine culture occurred in 5 of 28 with phage and 9 of 32 with placebo, with no significant difference. [3] Label: no good evidence of benefit in this trial.
• Cystic fibrosis. A 2026 systematic review of 19 mostly small reports covering 51 patients found lower bacterial load in 35 and no adverse events in 41 of 52 treatment courses, and called the findings supportive of further well-designed studies. The reviewers also noted small samples and different protocols. [4] Label: insufficient evidence (largely case reports).
• Small pooled reviews. A 2026 systematic review found only four eligible studies, two of which showed significant bacterial reductions and two did not. [5]

How to evaluate claims

• Ask whether results come from a randomized trial, a case report, or a lab or animal study. Mouse results are not human evidence.
• Beware wellness sites offering "phage" products for gut health; the evidence above concerns specific infections treated in hospitals.
• A good outcome in one patient (a case report) does not show typical results.

Related guides

Microbiome-Targeted Medicines, Fecal Microbiota-Based Therapies and The Emerging Science Evidence Landscape.