The evolving model of health-care delivery in the United States has necessitated a closer evaluation of how we deliver musculoskeletal medicine. Population-based medicine will shift reimbursement models to financially reward quality and value. Organizations will be paid a fee for the care of a population of patients. Maintaining a healthy population will result in fewer hospitalizations and the necessity of expensive medical services. Although opponents may argue that patients will receive poorer treatment with this strategy in cost control, maintaining a healthy population through preventative care, lifestyle support, improved outpatient management of patients with chronic diseases, and helping people remain active will be the foundation of keeping the covered population healthy. Ultimately, it is postulated that health-care organizations will be financially successful if they are successful in sustaining a healthy population.

Orthopaedic surgeons will be an integral component of this new model. Our care of people restores function, improves quality of life, and allows individuals to remain active and working. The initial increased cost of operative and nonoperative treatment of musculoskeletal disease meets the vital need of keeping individuals active and healthy. The initial upfront costs of arthroscopic surgery, total joint arthroplasty, or spine surgery provide both quality and long-term value. Unfortunately, our rapidly expanding armamentarium of technologically advanced interventions has often grown more rapidly than the clinical evidence and outcome studies to support their use. The challenge for the future to be able to deliver both quality and value is to identify what interventions work and what interventions fail to improve quality of life and/or function. To achieve this goal, an increased emphasis needs to be placed on high-quality clinical research and outcome studies. Clinical studies performed at single institutions and multiple institutions rarely deliver information with a strong clinical significance because of the inability to recruit enough patients to adequately demonstrate a benefit from a new intervention. Utilization of national databases appeared to be the obvious and exciting solution to meet this challenge.

In this study by Bohl et al., the authors vividly demonstrate the power of large sample sizes and the sobering limitation of national databases. In their study, statistical data were collected for patients undergoing lumbar spine fusion surgery including demographic information, comorbidities, and complications from two widely utilized national databases. The value of large patient sample size is immediately evident with relatively small differences in patient demographic characteristics being significant at p < 0.001 for patient age, residence prior to admission, and number of levels fused. More importantly, the authors have demonstrated that the method of data collection dramatically alters the reported outcomes and they have documented significant differences in comorbidities and complications reported when analyzing these two different databases. When evaluating comorbidities, significant differences (p < 0.05 for all) in rates were observed between the two groups for non-morbid obesity (9.33% in the Nationwide Inpatient Sample and 36.93% in the National Surgical Quality Improvement Program) and peripheral vascular disease (2.35% in the Nationwide Inpatient Sample and 0.60% in the National Surgical Quality Improvement Program). An analysis of inpatient adverse events and complications identified significant differences (p < 0.05 for both) in sepsis rates (0.38% in the Nationwide Inpatient Sample and 0.81% in the National Surgical Quality Improvement Program) and in rates for acute kidney injury (1.79% in the Nationwide Inpatient Sample and 0.21% in the National Surgical Quality Improvement Program).

Although on an initial analysis one may be surprised to see these dramatic differences, a more thorough understanding of the methods of data collection and the goals of the data collection make it evident as to why the differences exist. The Nationwide Inpatient Sample is a database of inpatient International Classification of Diseases, Ninth Revision (ICD-9) codes. The accuracy of this information will be limited by two variables: the documented information on the discharge summary by the physician dictating the discharge summary and a review by hospital coders to identify comorbidities and adverse events that are utilized in hospital billing. Most of us have experienced queries from hospital coders about adding a variety of ICD-9 codes. This method of data collection likely skews the medical information reported to include clinically inactive and insignificant diseases and to possibly not include vital information that was not adequately identified by ICD-9 codes.