Improving Surgical Triage in Spine Clinic: Predicting Likelihood of Surgery Using Machine Learning.

BACKGROUND: Correctly triaging patients to a surgeon or a nonoperative provider is an important part of the referral process. Clinics typically triage new patients based on simple intake questions. This is time-consuming and does not incorporate objective data. Our goal was to use machine learning to more accurately screen surgical candidates seen in a spine clinic. METHODS: Using questionnaire data and magnetic resonance imaging reports, a set of artificial neural networks was trained to predict whether a patient would be recommended for spine surgery. Questionnaire responses included demographics, chief complaint, and pain characteristics. The primary end point was the surgeon's determination of whether a patient was an operative candidate. Model accuracy in predicting this end point was assessed using a separate subset of patients. RESULTS: The retrospective dataset included 1663 patients in cervical and lumbar cohorts. Questionnaire data were available for all participants, and magnetic resonance imaging reports were available for 242 patients. Within 6 months of initial evaluation, 717 (43.1%) patients were deemed surgical candidates by the surgeon. Our models predicted surgeons' recommendations with area under the curve scores of 0.686 for lumbar (positive predictive value 66%, negative predictive value 80%) and 0.821 for cervical (positive predictive value 83%, negative predictive value 85%) patients. CONCLUSIONS: Our models used patient data to accurately predict whether patients will receive a surgical recommendation. The high negative predictive value demonstrates that this approach can reduce the burden of nonsurgical patients in surgery clinic without losing many surgical candidates. This could reduce unnecessary visits for patients, increase the proportion of operative candidates seen by surgeons, and improve quality of patient care.

Comparison of Minimally Invasive Versus Open Transforaminal Interbody Lumbar Fusion.

STUDY DESIGN: Narrative review. OBJECTIVES: In this review, we address the question of whether the literature supports the notion that minimally invasive transforaminal interbody fusion (MIS-TLIF) improves outcome as compared with open TLIF (open-TLIF). Short and long-term outcomes, fusion rate, and cost-effectiveness were reviewed. METHODS: This is a narrative review using various databases. Open-TLIF and MIS-TLIF studies were included and posterior lumbar interbody fusion studies were excluded. A description of paramedian incision in surgical technique was essential to the definition of MIS-TLIF. The present review included 14 prospective observational studies and 6 randomized controlled trials. RESULTS: With short-term outcomes, some studies indicate a better outcome with MIS-TLIF regarding intraoperative bleeding, hospital stay, time to ambulation, postoperative narcotic use, and time to resume work. Both MIS-TLIF and open-TLIF surgeries improved Oswestry Disability Index, back pain, and leg pain. Some studies show that MIS-TLIF resulted in lower back pain than open-TLIF. Radiation exposure was higher with MIS-TLIF. In the longer term, clinical outcomes were improved in both MIS and open TLIF groups. Fusion rates were more than 90% in both MIS-TLIF and open-TLIF. Cost-effectiveness and length of surgery had mixed results. CONCLUSIONS: The potential benefits of MIS-TLIF might be present in the early recovery period after surgery. Long-term outcomes were similar with both MIS-TLIF and open-TLIF.

Novel Method Using Baseline Normalization and Area Under the Curve to Evaluate Differences in Outcome Between Treatment Groups and Application to Patients With Cervical Spondylotic Myelopathy Undergoing Anterior Versus Posterior Surgery.

STUDY DESIGN: Retrospective review of a prospective database. OBJECTIVE: To describe a novel method that uses baseline normalization and area under the curve (AUC) to compare surgical outcomes between patients surgically treated anteriorly versus posteriorly for cervical spondylotic myelopathy (CSM). SUMMARY AND BACKGROUND DATA: It is important to control for baseline characteristics, especially disease severity, when evaluating differences in outcomes between 2 treatment groups. However, current methods of reporting outcomes are limited perhaps diminish the health impact of the entire postoperative recovery experience. METHODS: In the prospective, multicenter AO Spine North America CSM database, 147 patients had complete modified Japanese Orthopaedic Association (mJOA) data at baseline and at 6-, 12-, and 24-months postoperatively and were either treated anteriorly (n = 94) or posteriorly (n = 53). Each patient's follow-up mJOA scores were normalized by dividing them by the patient's baseline value. A graph was then plotted with the time point on the x-axis and the normalized score or "recovery index" on the y-axis. The AUC was calculated and then compared between the anterior and posterior surgical approach groups. RESULTS: The non-normalized recovery profile of the anterior group was better than that of the posterior group, as the patients treated anteriorly had less functional impairment at baseline. After normalization, patients in the anterior and posterior group had similar recovery indices and AUCs at 6-months following surgery. At 24-months, patients treated posteriorly had a significantly higher recovery index (1.32) and a larger AUC (16.3) than those treated anteriorly (1.11, 14.5, P = 0.004 and P = 0.006, respectively). CONCLUSION: This is the first study to apply AUC analysis to patients with CSM. In surgical patients with CSM, those treated anteriorly achieved a higher mJOA score at all time points than those treated posteriorly. The recovery indices, however, were not significantly different between approach groups at 6 months. LEVEL OF EVIDENCE: 3.