A collaborative regional multisite and multi-departmental initiative for improving consensus on clinical outcomes data in spine surgery

Study Design Retrospective chart review with consensus. Objective To establish a consensus-driven multisite collaborative database of individuals undergoing care for spondylolisthesis as a model for identifying barriers to standardized quality data collection in spine care. Summary of Background Data As health care initiatives move toward value-based care implementation, improving standardization of data collection through large databases and registries is increasingly important. However, in the absence of resources, many community practices are not represented, and models for low-cost implementation of such data collection efforts have yet to be established. Methods Data sharing agreements were established across participating private, community, and academic spine centers and data from charts of patients with spondylolisthesis was retrospectively reviewed for prevalence of demographic, surgical, operational, and outcomes variables. A consensus-based needs assessment then identified sources of error, additional data needs, and barriers to collection. Results 115 adult and pediatric charts were reviewed across 10 surgeons and 4 sites. High prevalence (>90%) variables included demographic (age, gender, BMI, smoking status) and operational variables (operating room time, blood loss, length of stay, 90-day readmission rate), and preoperative back pain severity. Low prevalence variables (<25%) included medication use, prior treatments, and skin to skin time for staged surgeries. 90% of surgeons identified inconsistencies between the coded procedure documented and surgery type. Variables identified for future collection included leg pain severity, and a single measure for comorbidity assessment. Barriers to quality data collection included lack of administrative resources, education of research or medical assistants extracting data, and regulatory burden. Conclusion These data provide a model for a low-cost collaborative database infrastructure, with recommendations for standardized data collection. Errors and site variability in surgical coding may have a significant impact on interpretability of study designs involving chart review, along with sparsity of patient reported outcomes.