I’d like to thank the web editor Dr. Pearson for the selection of our manuscript as the focus for The Spine Blog.
Clearly the study and understanding of complications is of paramount importance to the well-being of our patients. This manuscript has provided the reader with extensive risk factor data for complication after lumbar spine surgery. However, it can often be difficult for the reader to interpret these results and relay that information to the patient. While relative risk numbers are of value, our ultimate goal is to devise a predictive model to project a gross likelihood of a certain complication after lumbar spine surgery. It is one thing to tell a patient that they are 2.33 times as likely to have a certain complication because of their risk factor than if they didn’t have that risk factor. It is something else entirely to tell a patient that the likelihood for a particular complication based on their risk factor profile is XX%, or a number they can hang their hat on. Such a predictive model would need to be validated either in existing data registries, or with a future prospective data registry. From the patient and provider perspective, the value of such a predictive model would obviously be tremendous in regard to counseling and the decision making process.
From a medical center perspective, such a predictive model would be of great value as well. In addition to its benefits to patient care, such a predictive model could serve as a risk adjustment modifier in quality assessments. As health care trends toward the emphasis of quality and quality metrics, risk modification for more complex patients becomes essential. Quality metrics should not be absolute thresholds, but rather, a calculated threshold taking into account the patient baseline co-morbidity and extent of treatment. As Dr Pearson points out, there is no such thing as the “average” complication rate from the “average” surgery for the “average” patient. The complication rate for diabetics older than 65 undergoing scoliosis surgery is likely to be higher than that of healthy 25 year olds undergoing microdiscectomy. To use a single absolute threshold for “quality” when examining complications for all patients does not take into account the disparity in baseline health and the invasiveness of treatment. It is essential that future quality metrics account for the spectrum of patient co-morbidity and treatments, and a predictive model would be of great value in that effort.