As spine surgeons, we frequently advise our patients that they are at risk for medical complications (i.e. myocardial infarction, respiratory difficulty, DVT, etc.) following surgery. However, detailed discussion of those risks and how they relate to the patient’s age and underlying comorbidities and the invasiveness of the planned surgical procedure tend to occur less frequently, probably due to a lack of hard data about how those factors relate to their risks. In the October 1 issue, Dr. Lee and his colleagues from the University of Washington attempted to explicitly define these risk factors. They used a prospectively collected database including 767 patients who underwent lumbar surgery over a two year period to perform univariate and multivariate analyses to define risk factors for a variety of medical complications. Not surprisingly, the strongest predictors of medical complications were the invasiveness of surgery, baseline comorbidities, and age. While the authors did report the rates of different categories of complications (i.e. cardiac, pulmonary, neurological, etc.), the actual rates of these categories of complications are probably less important than identifying the predictors of complications. Inspection of the appendices—which I would recommend given the utility of the data contained there—reveals that the authors cast a wide-net with their definitions of complications, and, thus, the overall rate tended to be relatively high (i.e. 13% of patients had a cardiac complication). While the overall complication rate is directly dependent on the definition of a “complication”, predictors are less sensitive to this, and those identified here are likely very generalizable, regardless of one’s definition of a “complication”.
Most studies of complication rates following spine surgery have focused on overall rates for the “average” patient undergoing an “average” surgery. However, we know that there is no such thing as the “average” patient or the “average” surgery, and the “average” complication rate from these studies can be very misleading to patients considering surgical treatment. Prior studies have attempted to define risk factors for a few major complications (i.e. re-operation, re-admission, infection, death) using a limited number of covariates available from claims data. While these studies are useful, claims data rarely have sufficient information about invasiveness of surgery, baseline comorbidites, and post-operative complications in order to provide data that can be of value in counseling an individual patient. The current study advances the field substantially since the authors prospectively gathered a huge amount of data that begins to allow for individualized consideration of the risks of surgery. Rather than just providing the relative risks, it would have been useful if they had also provided the actual complication rates stratified by the level of risk factor (i.e. the risk of complications for each level of surgical invasiveness). The authors point out how these data could be used as the basis for a calculator that could predict the risks of specific complications for individual patients based on their unique characteristics and the invasiveness of the planned surgery. Such a calculator has been developed and used in cardiac surgery in order to predict peri-operative mortality,1 and a similar predictive model could be helpful in identifying patients at high risk for medical complications following spine surgery and would also allow for patients to make a more informed decision. If surgeons and patients realized that increasing the invasiveness of surgery by one “level” increased their rate of a neurological or hematological complication 15-fold, they might reconsider the aggressiveness of the intervention.
Please read Dr. Lee’s article and accompanying commentary. Will this affect how you counsel patients about their risks of peri-operative medical complications? Let us know by posting a comment on The Spine Blog.
Adam Pearson, MD, MS
Associate Web Editor
1. O'Connor GT, Plume SK, Olmstead EM, et al. Multivariate prediction of in-hospital mortality associated with coronary artery bypass graft surgery. Northern New England Cardiovascular Disease Study Group. Circulation 1992;85:2110-8.