Analyzing the value of IONM as a complex intervention: The gap between published evidence and clinical practice

• Published in: Clinical neurophysiology 2023-07, Vol.151, p.59-73
• Main Authors: Holdefer, Robert N., Seubert, Christoph N., Skinner, Stanley A., Humbert, Andrew T., Edwards, Mary E., MacDonald, David B.
• Format: Article
• Language: English
• Subjects: Complex intervention; Humans; iCAT_SR; Intraoperative neurophysiological monitoring; Intraoperative Neurophysiological Monitoring - methods; Neurosurgical Procedures - methods; Retrospective Studies; Structural causal model
• ISSN: 1388-2457; 1872-8952
• DOI: 10.1016/j.clinph.2023.03.364

•Intraoperative neurophysiological monitoring (IONM) is a complex intervention as defined by the United Kingdom Medical Research Council.•What is implemented as IONM and how IONM affects postoperative outcomes depends on the specific setting and context.•Causal graphs by representing dependencies on setting and context can increase the internal and external validity of estimated IONM effects. Intraoperative neurophysiological monitoring (IONM) was investigated as a complex intervention (CI) as defined by the United Kingdom Medical Research Council (MRC) in published studies to identify challenges and solutions in estimating IONM’s effects on postoperative outcomes. A scoping review to April 2022 of the influence of setting on what was implemented as IONM and how it influenced postoperative outcomes was performed for studies that compared IONM to no IONM cohorts. IONM complexity was assessed with the iCAT_SR tool. Causal graphs were used to represent this complexity. IONM implementation depended on the surgical procedure, institution and/or surgeon. “How” IONM influenced neurologic outcomes was attributed to surgeon or institutional experience with the surgical procedure, surgeon or institutional experience with IONM, co-interventions in addition to IONM, models of IONM service delivery and individual characteristics of the IONM provider. Indirect effects of IONM mediated by extent of tumor resection, surgical approach, changes in operative procedure, shorter operative time, and duration of aneurysm clipping were also described. There were no quantitative estimates of the relative contribution of these indirect effects to total IONM effects on outcomes. IONM is a complex intervention whose evaluation is more challenging than that of a simple intervention. Its implementation and largely indirect effects depend on specific settings that are usefully represented in causal graphs. IONM evaluation as a complex intervention aided by causal graphs and multivariable analysis could provide a valuable framework for future study design and assessments of IONM effectiveness in different settings.