Urinary catheter dislodgement studies: what hospital evaluators should compare

A supplier submits a promising securement study. Your committee needs to decide whether its findings support a particular claim in your hospital. Start by writing that claim in a complete sentence: which intervention, compared with what, improves which outcome, in which population and over what period? Then check whether the study actually tested it. A percentage without those qualifiers is an incomplete answer.

This article focuses on interpreting securement and dislodgement studies. For the broader vendor-document process, use the urinary catheter procurement evidence-request matrix. Clinical context on retention and drainage is covered in the review of balloon-free retention and valve drainage.

• What do the selected randomized studies actually compare?
• What counts as dislodgement?
• Patients or catheter-days?
• What can the comparator establish?
• How should withdrawals be counted?
• Which sources add independent evidence?
• A worksheet for the hospital evidence meeting
• Publisher scope and disclosure
• Frequently asked questions

What do the selected randomized studies actually compare?

CDC’s 2009 CAUTI guideline summary, on a page dated March 2024, gives securement recommendation II.E a Category IB designation. It addresses movement and urethral traction. It does not identify a preferred device or resolve every comparison between methods.¹ That distinction allows a committee to support the practice while still examining whether a proposed product’s evidence fits the purchasing question.

Read across a row before reading down a column. The studies do not supply a common league table: their target outcomes and control conditions differ. A favourable result for one question may be irrelevant to another. In particular, an infection finding should not be relabelled as a removal finding, and a pressure-injury finding should not become a promise of fewer infections.

The map also separates uncertainty from contradiction. An imprecise finding can coexist with a later positive study when populations, interventions and outcomes differ. A non-significant result does not establish equivalence. Conversely, a statistically significant result does not establish that every implementation of the intervention will reproduce it. Keep the effect, its uncertainty and the tested setting together.

What counts as dislodgement?

Require an operational definition that another observer could apply. “Dislodgement” might mean complete unintended removal, partial movement requiring clinical intervention, or any recorded movement. “Securement failure” might mean lifting tape while the catheter remains in place. “Traction injury” describes harm, which may occur without removal. If a submission combines these under one failure label, request the component events.

Ask how events were found: scheduled inspection, clinical incident reports, patient reports, device measurements or a combination. An observer who knows the assigned method may assess a subjective outcome differently. Blinding may be impractical, but the paper should explain any independent adjudication and how the same assessment schedule was applied to both groups. Greater surveillance can produce more recorded events without implying worse underlying performance.

Preserve the distinction between absence of an event and absence of measurement. If the study records infection but supplies no displacement definition, write “dislodgement not established by the reported endpoint.” Do not convert that omission into either zero dislodgements or evidence that the intervention failed.

Patients or catheter-days: which denominator answers your question?

CDC’s surveillance summary lists CAUTI per 1,000 catheter-days and calls for CDC/NHSN symptomatic-UTI criteria for numerator data.¹ A patient proportion answers a different question: how many people had the specified outcome during follow-up? A rate describes events relative to exposure time. Neither becomes comparable with another study’s number merely by expressing it as a percentage.

Hypothetical arithmetic example: imagine two cohorts with 100 people each. Each has five people with one defined event. One contributes 500 catheter-days; the other contributes 1,000. Both have an event proportion of 5%, but their rates are 10 and 5 events per 1,000 catheter-days. These invented values illustrate denominator choice; they are not observed catheter outcomes or evidence of an intervention effect.

For real studies, extract the observation window, time at risk, first-event versus recurrent-event rule and treatment of catheter changes. A patient may contribute more than one catheter episode. Ask whether exposure ends at discharge, planned removal, first failure or withdrawal, and whether outcome follow-up continues afterward. Different stopping rules can make a seemingly better rate reflect a different amount of observation.

Do not reconstruct exposure from average dwell time unless the underlying definition and totals support it. Nor should patient-level risk ratios, adjusted model estimates and incidence-rate ratios be placed in one comparison column as interchangeable effects. Request the analysis population, event counts, exposure totals and model specification before calculating a local comparison. Where those details are unavailable, preserve the study’s reported measure and state what cannot be reconciled.

What can the comparator establish?

Translate “standard care” into its actual components. It may mean no securement, a mixture of methods, or an active alternative. The relevant question is whether the tested difference matches the proposed local change. If your hospital already uses securement, a trial against non-securement does not estimate the incremental effect of changing between two securement products.

Also inspect co-interventions: skin protection, training, inspection frequency, replacement rules and catheter configuration. When these differ between arms, the result concerns the tested package. It cannot automatically be assigned to one component. Request enough description to recognize the intervention, without treating a research protocol as instructions for clinical use.

For example, the Brazilian tape trial reported more partial and overall detachment with silicone despite a numerically lower skin-injury proportion.³ That creates a useful appraisal question: are the proposed benefit and the possible failure mode both captured? A committee should decide which outcomes matter before inspecting which method appears to win. It should not select the most flattering endpoint afterward.

The male ICU trial belongs in a urinary evidence file, but its population boundary must remain visible.⁵ Sex-specific findings should not be extended to another population by assumption. Likewise, a study of an external stabilizer does not establish the performance of a different internal retention mechanism. Transferability needs a written explanation rather than a shared use of the word “catheter.”

How should withdrawals and early removals be counted?

Build a participant-flow account alongside the outcome table: screened, enrolled, allocated, exposed, followed, analysed and discontinued, with reasons. Preserve the analysis specified by the study; do not silently move participants between denominators. When a participant stops because of discomfort, detachment or displacement, that reason is directly relevant to evaluating the intervention even if the primary outcome is something else.

Separate planned removal after clinical need ends from unintended removal, withdrawal from data collection, death and transfer out of the unit. These events may end observation for different reasons. Request how the analysis handled each one. A shorter period at risk can reduce the opportunity to observe a complication, while excluding early failures can make the remaining users appear unusually tolerant.

Keep two questions visible: “What happened to everyone allocated?” and “What happened while the intervention was actually used?” They can support different analyses, but the report must identify them clearly. A completer-only satisfaction result should not be presented as the experience of all entrants. Equally, every early removal should not be labelled a dislodgement when the record gives another reason.

The 2006 trial illustrates why the flow account matters: 127 enrolled, but 118 were evaluable.² The publisher’s section excerpt explains that nine were excluded for not being followed beyond ten days. In the intervention arm, two withdrew voluntarily and three developed UTI within ten days; in the control arm, one withdrew and three developed early UTI.¹² Thus, the reported evaluable-group infection result excludes early infections in both arms. Preserve the published analysis, but request the full protocol and participant flow before judging the effect of that rule. These nine exclusions are not a count of device failures.

Which sources add independent evidence?

Count distinct patient cohorts and comparisons, rather than documents. Calpe-Damians and Llauradó-Serra appear in the 2024 trial’s author list and co-author the 2026 ICU securement review.⁴⁶ They are also contacts on NCT07093437.⁷ This identifies an overlapping research group, not a defect in its work. A review discussing an earlier trial does not add another randomized cohort. The other completed studies in Table 1 also show why the entire urinary evidence base should not be described as one investigator lineage.

The retrieved NCT07093437 record describes a randomized, open-label comparison of adhesive Grip-Lok and elastic Conveen securement, with estimated enrolment of 188. It labels the trial Recruiting, but its last posted update is 10 December 2025. Estimated study completion is 30 October 2026; no results are posted in the record retrieved for this article. Listed primary outcomes include meatal injury, pain in communicative patients and CAUTI.⁷ A calendar date reaching its estimate does not establish completion or success. Recheck the registry and subsequent publication before treating planned outcomes as findings.

Broad securement literature requires a second boundary check. Luo and colleagues’ PICC meta-analysis included 13 randomized trials and 1,970 patients, while cautioning about heterogeneity and low-quality evidence for main outcomes.⁸ Those are vascular-access data. Use them to identify appraisal questions, not to supply a urinary effect estimate. Moon and colleagues’ cross-device narrative review likewise acknowledges heterogeneous designs and follow-up.⁹ A broad synthesis should lead you back to the original urinary study for the decision at hand.

Bench summaries need their own evidence column. FDA’s nonbinding US premarket bench-report guidance covers methods, sample rationale, predefined criteria, results and deviations; clinical studies are outside its scope.¹⁰ These are useful completeness prompts for reading submitted adhesion or force data. They are not hospital regulatory obligations or a bridge from bench strength to fewer clinical removals. The broader bench-submission discussion addresses that document workflow.

A worksheet for the hospital evidence meeting

Complete one record per study and one conclusion per claimed benefit. Keep source location and reviewer comments with the record so that another committee member can reproduce the reading. The worksheet below is an editorial aid: it has no validated score, automatic pass mark or universal adoption threshold.

Close the meeting with a claim-level statement, not a winner. For example: “The submitted study concerns skin tolerance, whereas our purchasing claim concerns fewer unintended removals; the required removal definition and event counts are absent.” That identifies an exact evidence gap. It avoids both accepting an untested promise and dismissing useful findings for a different outcome.

Where the evidence is relevant but incomplete, identify the document or clarification that would change the decision. Where it is precise but inapplicable, record the population or comparator mismatch. Any proposed local evaluation still needs the hospital’s clinical, quality and governance review; a worksheet cannot authorize patient use or create market eligibility.

Publisher scope and disclosure

Silicon Life publishes this appraisal while developing and supplying a catheter in the same broad field. Its Disposable Sterile Controlled Catheter is a Class II device registered in China only (鄂械注准20252145807), for female patients with voiding dysfunction; it is not indicated for men or authorized for sale in other markets. The company-reported evidence is a single-centre, single-arm study: 23 enrolled, 19 completed, with four withdrawals after dislodgement or removal for discomfort. It does not establish superiority.¹¹

Those withdrawals belong in an evidence appraisal alongside the completer findings. The external securement studies discussed here do not establish compatibility with, or improved retention of, Silicon Life’s balloon-free design. Healthcare professionals can request the full clinical study report and Instructions for Use for that China-only, female-indication product. The purpose is to inspect the evidence and its limits, not infer eligibility in another jurisdiction.

Frequently asked questions

Does evidence of fewer infections prove fewer dislodgements?

No. Both could be relevant outcomes, but each needs its own definition, measurement and denominator. An infection endpoint cannot tell you how many catheters migrated or were unintentionally removed unless the study reports those events. Ask for the event table rather than inferring a mechanical benefit from a different clinical result.

Is there one urinary-catheter dislodgement rate we can use as a benchmark?

This appraisal does not establish one. A transferable benchmark would need matching event definitions, catheter configuration, population, observation window and denominator. A rate quoted through a narrative review must be traced to its original study before being used for a local comparison. An impressive number from another device class does not fill that gap.

Can a positive securement study select our hospital’s preferred device?

It can contribute to the decision when the tested comparison matches the proposed change. It cannot rank untested alternatives. Review the full methods, participant flow, safety outcomes, uncertainty and applicability, then combine that assessment with the hospital’s authorized procurement process. A vendor’s product name in a paper is not evidence for every configuration sold under that family.

Scope, source access and update plan

Sources were reviewed on 30 October 2026. The study map uses selected primary abstracts; a publisher section excerpt additionally supplies the 2006 trial’s exclusion reasons. Complete clinical-study methods, protocols and participant-flow tables were not verified. The 2026 ICU review is used only for bibliographic authorship, not unverified content. The trial-registry fields were checked through the official structured record. This is a narrative appraisal with worked examples, not a systematic search, pooled analysis or exhaustive count of trials.

Revisit the appraisal when NCT07093437 posts a status update or results, another relevant urinary study is published, or the CDC recommendation changes. Recheck product scope before any company documentation enquiry. Source access dates describe this review; they do not make an old registry entry or an older clinical study newly updated.