Research & Evidence
Six peer-reviewed studies and a 112-patient randomized trial: the evidence behind Pavaka’s platform, from the bench to people.
Evidence at a glance
Bench (Scientific Reports 2019, 2020)
Engineering (JMEMS 2020)
Veterinary (Vet Surgery 2022)
Murine (PLOS One 2025)
Human feasibility, n = 8 (2019)
Randomized study, n = 112 (NCT04794621)
In a 112-patient randomized study funded by the U.S. Department of Defense (NCT04794621), biofilm was measured during the first three weeks, when patients received Pavaka’s Gen 1 dressing. Biofilm abundance fell 40% from the first visit to week 3 in patients on the Gen 1 dressing plus standard care, while the standard-care group did not improve. Full results are being prepared for peer-reviewed publication.
A research program, not a single experiment
The platform’s history includes laboratory biofilm models, device engineering, animal investigations, early human feasibility work and a completed registered randomized study. These studies are distinct chapters, not evidence that an unchanged commercial device has already demonstrated clinical effectiveness.
The evidence spans several device versions. Each figure names the version studied.
From academic question to printed architecture
The research question was whether a flexible, battery-powered electrode pattern could be studied as an electroceutical wound dressing. The Ohio State program brought engineering, microbiology and wound-care perspectives together.
Pavaka’s investigational dressing generates a low-level electric field at the wound surface. Its platform is protected by an issued U.S. patent and an allowed patent application, exclusively licensed from The Ohio State University.

The early printed dressing: a historical research prototype, not the current Pavaka device. Dusane et al. (2019), Fig. 1 · CC BY 4.0
Bench experiments: activity, viability, structure
Dusane et al. (2019) developed an agar-based model with Pseudomonas aeruginosa lawn biofilms, powered by a 6 V source. This simplified model enabled imaging, culture and microscopy. It does not reproduce perfusion, immune responses or the complexity of a human wound.

Biofilm metabolic activity over time: lower signal develops over the anode. Dusane et al. (2019), Fig. 3 · CC BY 4.0

Recoverable viable bacteria: about a 4-log CFU reduction above the anode at 48 and 72 hours (n = 3). Dusane et al. (2019), Fig. 4 · CC BY 4.0

Architecture under the microscope: damaged cells in treated regions. Dusane et al. (2019), Fig. 5 · CC BY 4.0
Different media and timing questions
The same study examined agar containing human serum or bovine synovial fluid and responses after current was discontinued. These experiments extend the laboratory model. They do not establish clinical wear time, a dressing-change interval or a safe therapeutic regimen.



Protein-containing laboratory media. Dusane et al. (2019), Fig. 7 · CC BY 4.0
Inhibition-zone time course across three media. Dusane et al. (2019), Fig. 8 · CC BY 4.0
After a short current application. Dusane et al. (2019), Fig. 9 · CC BY 4.0
A closer look at bacterial structure (2020)
Lochab et al. compared laboratory conditions and used transmission electron microscopy to assess biofilm ultrastructure. The paper reports about a 4-log decrease in viability over the anode at 48 hours under both room-temperature and incubator conditions. These laboratory findings do not establish a clinical treatment.


Culture-based time courses. Lochab et al. (2020), Fig. 1 · CC BY 4.0
Ultrastructure at 24 hours (TEM). Lochab et al. (2020), Fig. 5 · CC BY 4.0
Preclinical investigation in infected murine wounds
Mack et al. (2025) evaluated an actively powered Ohio State dressing in P. aeruginosa-infected mouse wounds, across four groups of six mice: a standard film dressing alone, a commercial passive dressing, an unpowered Ohio State dressing and a powered Ohio State dressing. The miniaturized device used a 1.5 V source and is not the human-study configuration or the current Pavaka design. The short duration, small groups and attrition limit generalization. Animal findings do not establish clinical benefit in people.



Wound-area and tissue endpoints in mice. Mack et al. (2025), Fig. 1 · CC BY 4.0 · Animal-model data
Reporter signal and recoverable bacteria. Mack et al. (2025), Fig. 2 · CC BY 4.0 · Animal-model data
Exploratory immune measurements. Mack et al. (2025), Fig. 3 · CC BY 4.0 · Animal-model data
