[Assessment tool's validation study for a simulation-based skin suturing skill]. / Étude de validation d'une grille d'évaluation d'un exercice de suture cutanée en simulation.

CONTEXT: Health simulation is a recognized educational method for teaching and validating surgical procedural skills. The latter requires the development of adapted assessment tools, reaching different validity criteria. The aim of this study was to validate a multimodal assessment tool for a complex skin suturing exercise, combining a manual knot, an intradermal linear suturing and a needle holder tied knot. METHODOLOGY: The suturing exercise was realized on a synthetic skin model by voluntary participants after having obtained their written consent, including 9 postgraduate medical students, 40 surgical residents of different levels of experience, and a group of 9 senior surgeons. The multimodal assessment tool (MAT) combined a checklist, a speed score and a global rating scale. Each exercise was scored by two evaluators. Medical students' performances were filmed anonymously so that they could be scored iteratively. Content validity was tested through a satisfaction questionnaire randomly completed by participants. RESULTS: The MAT was considered relevant or very relevant by 98% of the participants, with a better appreciation for the checklist than for the global rating scale. Internal consistency was strong with a Cronbach α coefficient at 0.78, and a good correlation between the results of the checklist and the global rating scale (r=0.79, P<0.0001). The MAT showed continuous improvement in mean scores from 34.4±3.6 for novices to 47.4±2.5/50 points for experts, passing through three intermediate levels groups, and allowed for significant discrimination between groups. The MAT was reliable, with a coefficient of correlation set at 0.88 for intra-observer reliability, and 0.72 for inter-observer reliability. On sub score analysis, the global rating scale and the speed score better discriminated between groups than the checklist, the latter moreover showing slightly lower reliability than the global rating scale. CONCLUSION: Despite its banality in any surgeon's practice and the fact that it is taught from the 2nd cycle of medical studies, suturing and its technical components have rarely been the subject of publications dedicated to the validation of specific assessment tools. Hence, this work on the MAT and its sub scores made it possible to validate them on many validity and reliability criteria. They can therefore be proposed to surgical teachers for evaluating a complex suturing exercise, with a checklist that is easier to use even for novices and a global rating scale showing better discrimination capacity.

Value of early blood Th-1 cytokine determination in predicting severity of acute pancreatitis.

BACKGROUND: Early evaluation of the severity of acute pancreatitis (AP) requires measurement of many variables within 48 h after admission. Septic complications (SC) are frequent, and preliminary studies have highlighted the value of prophylactic antibiotherapy; however, single and reliable predictive markers of sepsis are not yet available. The aim of this study was to assess the value of determining early blood Th-1 cytokines and their natural antagonists (interleukin-6 (IL-6), IL-1, IL-1ra, and the soluble form of tumor necrosis factor (sTNF) receptors RI and RII) to predict the severity and SC during AP. METHODS: Thirty-seven patients with AP were prospectively included; 25 of them had severe AP, including 8 with SC. Serum cytokines were measured 48 h and 72 h after the onset of AP with an enzyme-linked immunosorbent assay. The optimal severity or SC diagnostic thresholds was determined using receiver operative curves. RESULTS: Severe AP in accordance with the Atlanta criteria were better predicted by C-reactive protein and IL-6 serum determination, albeit these levels could not predict absolutely the death of two patients. In severe AP cases (n = 25) the IL-1 to IL-1-ra ratio was lower in cases further complicated by sepsis ((6+/-4) 10(-3) versus (34+/-13) 10(-3), P < 0.05); moreover, sTNF RI (2497+/-270 pg/ml versus 2133+/-611 pg/ml, P < 0.05) and RII (3751+/-400 pg/ml versus 3045+/-509 pg/ml, P < 0.05) were higher in AP characterized by further SC. The IL-1 to IL-1-ra ratio and IL-1 concentration were dramatically decreased within the first 48 h ((0.4+/-0.4) 10(-3) versus (30+/-11) 10(-3), P < 0.05, and 0.3+/-0.3 versus 15+/-3 ng/l, P < 0.05) in patients with further infection of the pancreatic necrosis (n = 3). The SC diagnosis was better anticipated by an IL-1 to IL-1-ra ratio lower than 5 x 10(-3) or by an sTNF RI higher than 1750 pg/ml and sTNF RII higher than 2750 pg/ml, and the infection of the pancreatic necrosis by an IL-1 concentration <2 ng/l or an IL-1 to IL-1-ra ratio <2 x 10(-3). CONCLUSION: Besides severity markers, IL-1, IL-1-ra, and sTNF RI and RII should be considered in base-line AP assays and, if confirmed by larger studies, could help to screen patients at risk for SC and candidates for prophylactic antibiotherapy with a good negative predictive value.