PATCH trial sets its statistical frame before a single unblinded look at the data
Infants with peripheral intravenous catheters face a predictable and sometimes preventable hazard: infusate escaping the vein and injuring fragile tissue. In the PATCH trial, investigators are testing whether an IV biosensor, ivWatch, added to standard bedside observation can reduce infiltration and extravasation injuries in hospitalized infants compared with routine observation alone. The pre-specified statistical analysis plan matters because it fixes the rules before outcome data can influence interpretation, preserving the trial’s internal validity at a point where enthusiasm, disappointment, or clinical urgency could otherwise distort the readout.
The trial uses a multicentre, type 1 hybrid effectiveness-implementation randomised controlled design across three Queensland hospitals. Infants younger than 1 year, weighing at least 1000 g, are eligible if they have a peripheral intravenous catheter expected to remain in place for more than 24 hours and are receiving intermediate- or high-risk infusates, antimicrobials, or continuous fluids. Randomisation is 1:1 and stratified by site across four strata, a structure intended to keep the treatment groups balanced across hospitals while the trial unfolds in busy pediatric inpatient settings.
Why catheter injury in infants remains a bedside problem
Peripheral intravenous therapy is routine, but the infant vein is not forgiving. Small vessel caliber, limited soft tissue reserve, and difficulty recognizing early signs of infiltration create a narrow margin between a functioning line and tissue damage. An infusate injury may begin quietly, then progress to swelling, erythema, blistering, or deeper harm before visual inspection alone catches the problem. In that setting, a sensor that continuously monitors for deviation from expected flow has a clear biological rationale: earlier detection should shorten exposure to the injurious agent and limit the depth and extent of tissue damage.
The PATCH plan is built around that premise without overreaching beyond the data it will generate. It does not assume the sensor will prevent every complication; it specifies how the trial will determine whether the added technology changes the rate of injury, the severity of injury, and the durability of the catheter itself. The analysis also addresses implementation, a crucial point for a device intended to function alongside already demanding nursing workflows rather than replace clinical judgment.
Primary endpoint locked to harm, not device performance alone
The primary outcome is any infiltration/extravasation injury, from mild through severe harm, classified using the CCHMC Extravasation Harm Scale. A blinded outcome assessment committee will confirm the endpoint, a key safeguard because the intervention cannot be blinded at the bedside and clinical awareness of the device could otherwise influence how injury is recognized or recorded.
The primary analysis will follow intention to treat. That choice keeps every randomized infant in the group to which they were assigned, preserving the practical meaning of the comparison and limiting bias from catheter changes, early discontinuation, or workflow disruptions that are common in hospitalized children. The plan specifies logistic regression adjusted for randomisation stratum, with the risk difference estimated by standardisation and the odds ratio also reported. In practice, this means the trial will not rely on one statistical lens alone; it will quantify relative effect and absolute effect, both of which matter when a device must justify adoption in clinical care.
A narrow statistical plan is a clinical shield against post hoc storytelling.
Secondary outcomes extend beyond a single injury signal
The secondary outcomes widen the lens to include injury severity, time to PIVC failure, dwell time, replacement PIVC placement, skin injuries, and serious adverse events. Each endpoint addresses a different failure mode of peripheral access. A catheter that lasts longer but causes harm would be a poor trade; a device that reduces injury but creates extra replacements or workflow burden would also be difficult to defend in routine pediatric care. By naming these outcomes in advance, the plan recognizes that safety, efficiency, and device durability can move in different directions.
The emphasis on time-to-failure is especially relevant in infants, where one catheter may be expected to support fluids, antimicrobial therapy, or irritant infusions for a prolonged period. Earlier failure can trigger repeated cannulation attempts, distress, procedural pain, and additional opportunities for vascular injury. Dwell time and replacement frequency therefore function as more than operational metrics; they are proxies for how often a vulnerable child is exposed to repeated line placement and the cascade that follows when vascular access breaks down.
A type 1 hybrid design demands clinical and implementation credibility
The trial’s type 1 hybrid label signals that clinical effectiveness sits at the center, while implementation questions remain part of the package. In pediatric hospitals, adoption rarely depends on efficacy alone. Nurses must be able to use the sensor consistently, interpret alerts correctly, and integrate the device into the realities of medication delivery, handoffs, and rounding. A technology that works only under ideal research conditions is unlikely to change care where peripheral access injuries are most common.
The plan was finalized on 10 November 2026, while recruitment was still underway and before database lock or any unblinded analysis. That sequence is methodologically important. It prevents the statistical machinery from shifting after investigators have seen emerging trends, and it signals that the primary readout will be judged against rules written in advance rather than against a desired outcome.
What the trial can and cannot answer yet
Because this is a statistical analysis plan, no efficacy or safety results are presented. The document instead establishes the framework for how the trial will evaluate whether adding ivWatch to standard observation reduces infusate-related injury in infants. The target sample size is 532, and the protocol registration number is ACTRN12623000561684. Ethical approval came from the Human Research Ethics Committee of Children’s Health Queensland Hospital and Health Service, with ratification by The University of Queensland.
The next decisive step is the primary results publication, which will need to show not only whether injuries fall, but whether the effect is large enough, consistent enough, and operationally feasible enough to matter in everyday pediatric infusion care. Until those data arrive, PATCH remains a carefully defined test of whether continuous sensing can outperform the imperfect human eye in one of the most delicate corners of inpatient pediatrics.
- • Clinical Trial Registry (ACTRN12623000561684): https://www.australianclinicaltrials.gov.au/anzctr/trial/ACTRN12623000561684
- • medRxiv DOI 10.64898/2026.11.15.26367580: https://doi.org/10.64898/2026.11.15.26367580