Measuring clinical practice: science and methods

Abstract

The routine practice of neurosurgery generates a large amount of clinical data. The structured, systematic capture of this information using clinical registries or other rigorously designed observational studies can yield useful evidence to help improve the care of patients. Registries in particular can be designed to measure outcomes in real-world clinical settings and to study differences in outcomes between subgroups. This information can help clinicians to advise patients regarding their treatment options. To provide valid, generalizable evidence, however, registries and other observational studies must be designed and conducted with a rigor similar to that of randomized clinical trials. Neurosurgeons with a basic understanding of the potential advantages and pitfalls of nonrandomized trials and the methods of statistical analysis will be able to assess the quality of clinical data and to incorporate the findings appropriately into their patients' care.