Introduction: The ability to predict risk of perioperative complications based on patient and procedure specific factors would benefit surgeons and patients. Current methods of risk assessment for spine surgery are limited. This study assesses an iPad application that predicts the risk of perioperative complications based on patient and procedure specific unique factors.
Methods: We developed a smooth data entry process using an iPad device that provided appropriate inputs into a previously developed computational model of adverse event occurrence in spine surgery. The model was previously generated using longitudinal prospective data, consisting of 279,145 records, from a claims database and was used to develop a prediction rule incorporating the type and occurrence of complications. The present study applied this model to a group of 200 patients. Patient factors entered into the application's interface included: patient age, patient gender, pre-operative diagnosis, area of the spine affected, comorbidities, fusion status, instrumentation status, number of levels, and use of BMP. Predicted complications were compared to the actual complications for the 30 day postoperative period.
Results: The mean predicted probability of experiencing a complication was .4494 for patients experiencing 1+ complications and .3714 for patients experiencing 0 complications (p=.0436), according to the iPad application. The statistical model utilized by the application previously predicted an adverse event rate of 13.95% for a large-scale administrative claims database. In the current study a group of 70 patients (35%) experienced post-operative complications. The most common comorbidity in patients experiencing complications was hypertension (49%, compared to 36% in patients with no complications).
Conclusions: The iPad application and the statistical model that it utilizes can provide the surgeon with a straightforward and timely means of assessing the risk of perioperative complications based on known patient factors and comorbidities.
Patient Care: By providing the surgeon with a straightforward and timely means of assessing the risk of perioperative complications based on known patient factors and comorbidities. This knowledge could be used to educate the patient on potential risks of a given procedure, which could be weighed against potential benefits.
Learning Objectives: By the conclusion of this session, participants should be able to: 1) Describe the utility of a novel statistical algorithm that predicts adverse event occurrence in spine surgery 2) Discuss, in small groups, how this technology could be implemented in a variety of practice settings.