LAUSR

Technological advancements are creating major practice shifts toward minimally invasive techniques with the promise of less pain/morbidity, shorter hospital stays, shorter recoveries and improved cosmesis. These benefits are not necessarily realized in all populations, particularly in small children, in whom these approaches are more challenging and excellent recovery the norm. In the current health care environment where quality is closely scrutinized, less invasive surgery must balance improved outcomes and cost. Outcomes can be broadly categorized as success rates and complications, and while it is reasonably straightforward to compare success rates between traditional and less invasive techniques, direct comparison of complications is more difficult. Minimally invasive surgery may result in a different set of complications or particular complications that are more or less common compared to traditional surgery. As a specialty, we must set minimum outcome reporting standards to provide surgeons and families with appropriate information to assist with management decisions. Proving benefits of minimally invasive surgery requires comparative studies focused on success rates and complications. While cost is another important consideration, this is a more complex issue best addressed by health services researchers and policy makers. Proper outcome reporting begins with a clear research question, clearly defined inclusion criteria, an appropriate control group and a detailed description of how patients are selected or placed into open vs minimally invasive groups. Reportable complications include any adverse events reflecting a deviation from the expected course of the healing process. Unless adverse event profiles are identical for open and minimally invasive surgery, the standard for reporting should include enough detail to consider the types and frequency of specific complications rather than a simple tabulation of rates. In the 1990s a number of classification schemes were proposed by various investigators to standardize complication reporting, although most failed to achieve broad acceptance. Veen et al reviewed results of 8,130 surgical procedures and organized complications by a mix of physical effects of the complication, additional therapy required and whether an error was involved. In 1996 a group of vascular surgeons developed the SCOUT (Surgical Complication Outcome) score by recording, categorizing and assigning a severity score to more than 50 specific complications of major vascular operations. Unfortunately this method was time intensive to develop and specific only to vascular cases. In 1992 Gawande et al focused on complications that cause death, disability at discharge or prolonged hospital stay, and categorized them further according to procedure and preventability. The goal was to identify opportunities for quality improvement measures, not necessarily for use in comparative effectiveness research. Pillai et al used the Otago Audit System to classify complications as none, minor (patient discomfort), intermediate (significant compromise) or severe (major threat to life), leaving ample room for subjectivity. Clavien et al produced the most promising prototype when they proposed a standardized system that categorized complications based on the types of subsequent interventions that were required. The original system had 4 levels of severity, and later revisions added a fifth level as well as a “d” modifier to the level denoting the presence of some disability at the time of hospital discharge. The levels categorize complications as requiring minimal interventions, diagnostic testing/additional medical therapies, procedural interventions (with or without the need for sedation), care in an intensive care unit for significant organ dysfunction/potentially life threatening issues and mortality. Since its introduction, the revised Clavien-Dindo classification has become the most common tool for reporting complications, although not without criticism. It has the benefit of allowing for seemingly simple and reproducible complication grouping. However, the system is bound by certain ordinal assumptions and a broad range of intervention intensities among interventions within each category. Thus, the classification can provide some measure of the severity of