A Bayhaven outpatient clinic has paused a pilot program testing direction-based delivery of anti-inflammatory medication through the skin after several participants developed small, painful welts beneath adhesive electrodes — injuries investigators now attribute to polarity selection, electrode reactions and concentrated current in tiny contact areas.

The pause, announced Monday by Bayhaven Clinical Research Partners, follows an internal review of eight reports of redness and blistering in a 42-person study that used low-level direct current to push charged drug molecules across the skin.

“No one was hospitalized, but these were preventable injuries,” said Dr. Ellen Park, the study’s medical director. “The device was doing what we told it to do — we just underestimated what happens at the electrode when you keep driving current through a moist patch of skin.”

The review, shared with the city’s health department, points to a chain of decisions that turned routine sessions into what Park called “chemistry experiments on a postage stamp.”

Polarity did the delivery — and changed the chemistry

The pilot relied on a simple premise: choose the electrode polarity so the drug is repelled from the “like-charged” side and migrates toward the opposite electrode.

That decision had side effects at the skin surface.

According to the review, staff favored one polarity for most participants to improve delivery consistency. But at the electrode interfaces, water and dissolved salts in sweat and gel can be pushed into new reactions.

“When you run current, you don’t just move the medication,” said Priya Nandakumar, a biomedical engineer hired by the clinic to audit the protocol. “You also drive electrode reactions that can generate acid on one side and base on the other. The bulk of the gel might look fine, but right next to the metal, pH can drift hard.”

In practical terms, investigators said, one electrode environment can become more acidic while the other becomes more alkaline over the course of a session, especially if current is maintained without interruption and if the electrode material encourages side reactions.

Nandakumar said the drift may not be obvious until participants report a stinging sensation — often late in a session — because the skin’s surface can tolerate mild changes until localized chemistry overwhelms it.

“That’s why people describe it as a sudden ‘hot’ feeling,” she said. “It’s not heat. It’s a chemical shift happening in a small zone.”

Why Ag/AgCl was recommended — and why it wasn’t used consistently

The clinic’s review singled out electrode choice as a key contributor. Some sessions used generic stainless-steel snaps integrated into disposable patches, while others used silver/silver chloride (Ag/AgCl) inserts.

Ag/AgCl electrodes are widely used in medical sensing and stimulation because their interface tends to support a more predictable exchange of chloride ions rather than forcing more aggressive reactions in the surrounding water.

“In plain language, Ag/AgCl gives the current an easier, steadier ‘reaction pathway,’” said Nandakumar. “That can reduce the push toward extreme pH right at the skin.”

The audit found the study did not mandate Ag/AgCl for every participant because of cost and supply issues.

Clinic administrators said they were trying to keep a single-use patch under $7.

“That decision looks different when you’re treating blisters,” said Park, who added that the clinic is now requiring Ag/AgCl interfaces for any resumed work.

The “hot spot” problem: small areas, big consequences

Even when total current remained within the device’s labeled limits, investigators said irritation clustered in a predictable pattern: the edge of the electrode, a corner of the adhesive, or a small fold where gel was thin.

The logic is straightforward, the report said: if the same current is forced through a smaller effective contact area, current density rises.

Staff were briefed on the concept using a simple comparison. If 2 milliamps is spread across a 10-square-centimeter pad, the “per area” load is modest. If wrinkling or partial lift reduces good contact to 2 square centimeters, the local burden becomes about five times higher.

“That’s when you stop thinking about the pad as one electrode and start thinking about it as a few tiny electrodes,” said Nandakumar. “Those tiny points become the places where pH shifts faster and the skin gets hit harder.”

Participants who reported injury often had patches placed over highly mobile areas — the wrist, ankle and shoulder — where adhesive edges can lift during movement.

One participant, 29-year-old delivery driver Marcus Lee, said the patch “felt fine” for most of the session.

“Then it started to sting like I had spilled something,” Lee said. “When I peeled it off, there was a red ring right where the corner had been rubbing.”

Failure modes investigators say are common — and easy to miss

The internal report listed several recurring problems that can turn routine sessions into high-risk ones:

• Polarity chosen for delivery, not for skin tolerance. Staff prioritized directionality but did not adjust for participants with sensitive skin or low hydration.
• Electrode material mismatch. Mixed use of metal snaps and Ag/AgCl inserts produced inconsistent electrode behavior from session to session.
• Dry-out and uneven gel thickness. As gel thinned near an edge, current concentrated and local chemistry intensified.
• Pressure points under straps or tape. Compression created tiny regions of better contact that drew disproportionate current.
• Long, uninterrupted sessions. Continuous exposure allowed pH drift to build without relief.

The mitigations: safer delivery, slower schedules

Bayhaven Clinical Research Partners said it will redesign the protocol before seeking approval to restart. The changes reflect a shift toward controlling current density and stabilizing electrode reactions while still using polarity to steer molecules.

Among the measures under adoption:

• Increase effective electrode area. Larger pads and designs that maintain full-surface contact reduce localized burden.
• Improve contact uniformity. The clinic plans new skin-prep steps to reduce oil and hair interference, plus placement rules avoiding joints and high-motion skin.
• Standardize Ag/AgCl interfaces. All sessions would use Ag/AgCl contact surfaces to provide a more stable reaction route.
• Use better-buffered gels. Formulations will be selected to resist rapid pH change near the electrode. Investigators cautioned that stronger buffering can complicate drug movement and may require longer sessions or adjusted dosing.
• Set explicit current-density limits. Instead of a single device-wide current cap, staff will calculate allowable current based on pad area and prohibit partial-contact use.
• Duty cycling and shorter sessions. The redesigned plan favors intermittent current and shorter exposures, allowing the skin to recover and reducing the build-up of extreme conditions at the interface.

Park said the changes reflect an uncomfortable tradeoff: “If we lower intensity or break up sessions, delivery can slow. If we push harder for speed, we risk ‘hot spots’ and chemistry shifts. The new protocol is about staying on the safe side of that line.”

Regulators watch, sponsors reconsider timelines

City health officials said they were notified after the third participant reported blistering, and the clinic later provided a written root-cause analysis.

“Any time you’re applying current to skin, you’re managing both physics and chemistry,” said Dr. Luis Herrera, Bayhaven’s deputy health officer, who reviewed a summary of the findings. “The skin doesn’t care that the total current was ‘within spec’ if the contact area is small or the interface chemistry is harsh.”

The device’s manufacturer, Altair Dermal Systems, said in a statement that it is assisting the clinic and that its instructions emphasize full pad contact and recommended electrode materials.

For now, Bayhaven’s study sponsor has agreed to extend the timeline while the clinic revises training and supplies.

Lee, the participant who developed a ring-shaped welt, said he would consider returning if the protocol changes.

“I wasn’t scared off by the idea,” he said. “I just don’t want to be the lesson again.”