Emergency Medicine · Medical Literature
Learning to manage tracheostomy-related emergencies: a pilot study comparing three teaching strategies for junior doctors in intensive care.
Abstract
BACKGROUND: In intensive care settings, doctors face low-frequency and high-stakes events such as tracheostomy-related emergencies. Junior doctors must learn to manage these situations to ensure patient safety, yet they often lack clinical exposure to such critical events. Clinical teachers should develop methods that provide enough learning opportunities while remaining feasible in terms of workload. This study compared three strategies for teaching junior doctors to manage tracheostomy-related emergencies in intensive care unit (ICU). METHOD: In this pilot study, conducted in the ICU of Lausanne University Hospital, twenty-four junior doctors were randomized into three groups: Group A: Access to a printed algorithm; Group B: E-learning, including the algorithm with access to a toolbox; Group C: Flipped classroom course, including a short low-fidelity simulation. Performance was assessed through pre- and post-intervention sessions consisting of three tracheostomy-related simulation scenarios. A modified Delphi method was used to develop a performance assessment tool consisting of an evaluation scale for each scenario. Knowledge was also assessed before and after the intervention through a multiple-choice questionnaire. The primary outcome was the Global Performance Score (GPS), calculated for each teaching method by summing the performance scale scores across the three simulation scenarios as the total score across all scenarios, after the intervention. RESULTS: The GPS significantly improved after the intervention in the three groups (Group A: M = 141, SD = 22 vs. M = 116, SD = 16.7 p = 0.008; Group B: M = 157, SD = 14 vs. M = 110.1, SD = 17.3 p = 0.008; Group C: M = 171.9, SD = 20.1 vs. M = 99.4, SD = 16 p = 0.008). There was a trend for a higher post-intervention GPS in Group B as compared to Group A (p = 0.105), and in Group C as compared to Group B (p = 0.279). By contrast, post-intervention GPS was significantly higher in Group C as compared to Group A (p = 0.021). CONCLUSION: The flipped classroom strategy, combined with a brief simulation, yielded the highest post-intervention score. This strategy appears promising for junior doctors learning to manage low-frequency critical events, such as tracheostomy-related emergencies. Further research should explore its feasibility at larger scale, confirm the results in a larger population of learners with adequately powered studies and the applicability of such a teaching approach to other critical events.
Clinical relevance
Among junior doctors training in the ICU, does a flipped classroom course with low-fidelity simulation, e-learning with a toolbox, or access to a printed algorithm alone produce superior performance in managing tracheostomy-related emergencies as assessed by simulation-based scoring? __SNAPSHOT__ {"studyDesign":"Pilot randomized controlled educational trial with three parallel arms; pre- and post-intervention simulation-based assessment","setting":"Single tertiary academic center — ICU of Lausanne University Hospital (CHUV), Switzerland","sampleSize":"n=24 (Group A: 8, Group B: 8, Group C: 8; assumed equal allocation — group sizes not explicitly reported in abstract but implied by total n and three groups)","population":"Junior doctors rotating through the ICU; clinical experience level and prior tracheostomy exposure not reported in abstract","intervention":"Group B: E-learning module including a printed algorithm, a toolbox, an online lecture with narrative slides, practical guidelines, and a video. Group C: Flipped classroom — same pre-class e-learning materials followed by a 30–45 minute one-to-one low-fidelity manikin simulation course","comparator":"Group A: Access to a printed tracheostomy emergency management algorithm only (active control)","primaryOutcome":"Global Performance Score (GPS) — sum of performance scale scores across three tracheostomy-related simulation scenarios after intervention; assessment tool developed via modified Delphi method","followUp":"Not reported (assessment conducted immediately post-intervention; no long-term retention data reported)","fundingSource":"Not reported"}
Key findings
All three groups demonstrated statistically significant improvements in Global Performance Score (GPS) after their respective interventions. Group A (algorithm only): pre-intervention M=116, SD=16.7 vs. post-intervention M=141, SD=22 (p=0.008). Group B (e-learning): pre-intervention M=110.1, SD=17.3 vs. post-intervention M=157, SD=14 (p=0.008). Group C (flipped classroom + low-fidelity simulation): pre-intervention M=99.4, SD=16.0 vs. post-intervention M=171.9, SD=20.1 (p=0.008). In between-group comparisons, no statistically significant difference was found between Group B and Group A (p=0.105) or between Group C and Group B (p=0.279). However, Group C's post-intervention GPS was significantly higher than Group A's (p=0.021). NNT is not calculable given the continuous scoring outcome and available data. Pre-post absolute GPS gains were: Group A, +25 points; Group B, +46.9 points; Group C, +72.5 points (all calculated from reported means). Knowledge outcomes via multiple-choice questionnaire were assessed but specific MCQ results are not reported in the available abstract. Adverse event data are not applicable in an educational simulation study.
Practice implications
This study establishes that all three teaching strategies — including the low-resource printed algorithm alone — produced measurable, statistically significant improvements in simulated performance of tracheostomy emergency management. The flipped classroom with brief low-fidelity simulation outperformed the algorithm-only approach on the GPS (p=0.021), suggesting additive benefit from interactive, hands-on components. However, the difference between e-learning and algorithm-only, and between flipped classroom and e-learning, did not reach statistical significance, possibly due to the very small sample size rather than a true absence of effect. Clinical importance cannot be fully determined as the GPS is a study-specific tool without established minimal clinically important difference (MCID), and the study did not measure real patient outcomes, knowledge retention, or skill durability. Findings apply to junior ICU doctors in a Swiss tertiary academic center and may not generalize to other training levels, specialties, resource-limited settings, or non-Swiss health systems. Patients who were not represented include clinicians outside the ICU context, senior physicians, nurses, or those in low-resource environments. Given this is an abstract-only analysis, key details regarding randomization methods, allocation concealment, blinding of assessors, and knowledge outcome data remain unavailable — these limitations are important for interpreting the validity of comparisons. __LIMITATIONS__ Very small sample size (n=24; approximately 8 per group), severely underpowering between-group comparisons and preventing definitive conclusions about differential efficacy.,Single-center design at a Swiss academic tertiary hospital limits external generalizability.,Performance outcome (GPS) is study-specific and lacks a validated MCID; statistical significance cannot be directly translated into clinical importance.,No long-term follow-up — whether performance gains are retained over time (e.g., weeks or months after training) is unknown.,Only simulation-based performance was measured; actual patient outcomes, clinical error rates, or real-world tracheostomy emergency management were not assessed.,Knowledge outcomes (MCQ scores) are mentioned but specific results are not reported in the available abstract.,Assessor blinding to group allocation and inter-rater reliability of the Delphi-derived performance scale are not described in the abstract.,The flipped classroom simulation was delivered by the first author (MG), introducing potential instructor-effect bias.,Groups A and B differ in content richness (algorithm alone vs. algorithm + toolbox + online lecture + video), making it difficult to isolate which component of Group B's intervention drives any additional benefit.,Pilot designation means the study was not powered for inferential hypothesis testing; p-values should be interpreted cautiously and as hypothesis-generating.,Funding source not reported, precluding assessment of potential conflicts of interest.,Only the abstract was available for this analysis; full-text review may reveal additional methodological details that affect interpretation. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ In this small (n=24) single-center pilot RCT, all three teaching strategies improved ICU junior doctors' simulated tracheostomy emergency performance, but the flipped classroom with low-fidelity simulation produced the highest GPS and was the only arm to significantly outperform the algorithm-only control (p=0.021). Given the pilot design, very small group sizes, single-center setting, and lack of long-term or patient outcome data, these findings are hypothesis-generating and should inform — but not yet change — institutional teaching practice; adequately powered multi-center trials are needed to confirm the superiority and durability of the flipped classroom approach.