NURS 225 Week 8 Research Critique Part 2 Example

Research Critique Part 2: Design, Sample, Ethics, Measurement, Analysis and Conclusions in a Trial of Clinically Indicated IV Catheter Replacement

Research Design

Kang et al. (2026) used an open-label, non-inferiority, cluster-randomized crossover design. Eight wards, four medical and four surgical, were randomized by an independent statistician using stratified block randomization, and after the first period each ward crossed over to the other strategy. Randomizing wards rather than patients suits a policy-level intervention, since nurses on one ward cannot easily follow two replacement rules at once, and the crossover lets each ward serve as its own comparison. Boswell and Cannon (2017) identify randomization and a control condition as the features that most protect internal validity, and both are present. The main weakness is that neither patients nor staff could be blinded, which matters because phlebitis is graded by observation; assessors who knew the strategy might have graded borderline sites differently. The authors acknowledge this and note standardized training to reduce it. With only eight clusters, period and carryover effects are also possible, which the authors address by stratifying analyses by period.

Sample and Setting

Enrollment was limited to adults on the trial wards expected to need a peripheral line for four days or longer, with exclusions for bloodstream infection, immunosuppression, skin problems at the site and central lines. The authors calculated that 1,230 participants were needed for 80% power, assumed some clustering, added 10% for withdrawal and enrolled 1,324 in the intention-to-treat analysis, 675 and 649 in the two groups. A transparent power calculation is a clear strength. The single 2,700-bed academic hospital with a dedicated IV team, which placed more than 92% of catheters, limits generalizability to community hospitals and units where bedside nurses insert their own lines.

Ethical Considerations

The protocol was approved by the hospital's institutional review board and registered before the trial. Because both strategies fell within accepted standard care and allocation occurred at the ward level, the board waived individual written consent; patients and families were informed through routine communication, could decline study procedures or request replacement and had data de-identified. Boswell and Cannon (2017) emphasize informed consent as a central protection, so the waiver deserves scrutiny. In this case it appears justified: neither strategy exceeded usual care, the risk was minimal and individual consent would have been impractical when the policy applied to every patient on a ward.

Data Collection and Measurement

Phlebitis, the primary outcome, was measured with the Infusion Nurses Society scale, a published, graded tool, and bloodstream infection followed CDC criteria. Ward nurses and the IV team received standardized training before the trial, and each catheter was tracked on a case report form. These are strengths. Interrater reliability for phlebitis grading was not reported, a weakness given the open-label design.

Data Analysis and Results

The analysis treated the ward as the unit of clustering, using generalized estimating equations for the primary outcome and Cox models for time-to-event outcomes, with both intention-to-treat and per-protocol analyses. Phlebitis rates were 12.6% with removal on clinical grounds against 11.7% with scheduled removal; the risk difference was 1.44 percentage points with a 95% confidence interval from -1.47 to 4.35. Because the upper limit, 4.35, stayed below the 5-point margin, non-inferiority was established, and the per-protocol result agreed. No bloodstream infections occurred. The clinically indicated group needed fewer insertions per patient (1.77 versus 2.16) and kept catheters longer (112.0 versus 89.6 hours). The analysis is appropriate for the design, but the upper limit sits close to the margin, so the result is reassuring rather than definitive. Complete case analysis was used for missing data, which is acceptable here because few participants were lost, but the article does not report exactly how many observations were missing.

Discussion, Limitations and Implications

The authors interpret their findings cautiously and list appropriate limitations: single site, open-label assessment, possible period effects and no power to detect differences in rare infections. Their findings agree with the 2026 Cochrane update, whose 14 pooled trials showed no clear rise in bloodstream infection when catheters stayed until a clinical reason arose (Charles et al., 2026), and with evidence of longer dwell times in older adults (Jiang et al., 2026). Phlebitis rates in both groups exceeded the Infusion Nurses Society benchmark of under 5%, which the authors attribute partly to large-bore catheters; this is an important practice point in its own right. The trial shows that stopping routine replacement is safe enough; it does not show that IV sites can be watched less closely.

Overall Appraisal

This is a well-designed pragmatic trial with clear outcome definitions, a transparent sample size, appropriate analysis and honest limitations. Its main weaknesses, open-label assessment and a single specialized setting, are understandable for a policy trial. For nursing practice, it supports replacing peripheral catheters when clinically indicated, combined with careful daily site assessment and attention to catheter size.

References

Boswell, C., & Cannon, S. (2017). Introduction to nursing research: Incorporating evidence-based practice (4th ed.). Jones & Bartlett Learning.

Charles, K. R., Schults, J. A., Marsh, N., Osborne, S. R., Rees, S., Mihala, G., Smith, S., & Rickard, C. M. (2026). Clinically indicated removal versus routine removal of peripheral venous catheters. Cochrane Database of Systematic Reviews, 2026(1), CD007798. https://doi.org/10.1002/14651858.CD007798.pub6

Jiang, X. Y., Wang, L. J., Wei, T., Li, X. Y., Tan, Y., Lin, Q., Yuan, Z., & Wang, T. Y. (2026). Clinically indicated versus routine replacement of peripheral intravenous cannulas in elderly patients: Secondary analysis of a multi-center randomized controlled trial in mainland China. Geriatric Nursing, 72, 104100. https://doi.org/10.1016/j.gerinurse.2026.104100

Kang, N., Kim, H. L., Choi, H. R., & Kim, J. (2026). Routine versus clinically indicated replacement of peripheral intravenous catheters in adults: A non-inferiority cluster-randomised crossover trial. Healthcare, 14(13), 1825. https://doi.org/10.3390/healthcare14131825