gms | German Medical Science

Introduction: The current method of choice to address and govern human factor aspects in medical teams seems to be "Crew Resource Management" (CRM) training formats. This is reflected by a strong and steady increase in publications regarding such trainings that can be observed in medical literature since the year 2000 (see figure 1 [Fig. 1]). A lot of evidence is suggesting that CRM trainings are contributing to patient safety and nearly all articles suggest implementing more of such trainings. However, little is known as to what exactly is being taught as "CRM" so that heavy variations between the curricula can be assumed.

Objectives: We wanted to find out what is meant by "CRM trainings" in medicine, i.e., if there is some common standard they follow, a conceptual core they are derived from, and how the evaluation is carried out. Specific interest was put on nine aspects in three areas that would provide the necessary information to reproduce and compare interventions. Intervention design: (1) Description of intervention, setting and/or format. Training conditions: (2) Duration, (3) Target group, (4) Number of participants, (5) Trainer qualification. Evaluation: (6) Kirkpatrick-level of evaluation, (7) Method of evaluation, (8) Sample group, (9) Reported outcome.

Methods: A systematic review of available literature for CRM trainings in medicine was carried out in pubmed, PsychInfo, and ERIC databases and following the PRISMA statements recommendations. From 797 search results 58 studies could be included in a qualitative synthesis of publications both containing a description of a CRM intervention and evaluation results.

Results: The review uncovers that more than 50% (30 of 58) of the publications do not provide a reproducible description of their intervention (see figure 2 [Fig. 2]).Furthermore several descriptive aspects of training conditions and evaluation methods are portrayed.

For a subgroup of 10 publications reporting organisational change management efforts accompanying their trainings, we compared the occurrence of adverse events before and after the intervention and found an average odds ratio of 2,01 (CI95%:1,34-2,99).

Conclusion: Participants perceive CRM trainings as beneficial and positive effects on an organisational level have also been found in several studies. Still, the review indicates that contents as generally enveloped in CRM trainings in medicine might not be comparable to a large extent. It also indicates that there are success factors for CRM interventions that are beyond the scope of trainings.

The results underline the need for development of evaluation and reporting standards for CRM intervention studies to foster completeness, reproducibility, and comparability.