262 - Application of Personalized Medicine Analytics in the Decision to Use...

#262 Application of Personalized Medicine Analytics in the Decision to Use Pulsed Electromagnetic Field Therapy (PEMF) after Anterior Cervical Discectomy and Fusion (ACDF)

Value and Outcomes in Spine Surgery

Poster Presented by: K. Radcliff


K. Radcliff (1)
J. Ryaby (2)
R. Hahn (3)
J. Mackowiak (4)

(1) Thomas Jefferson University, Department of Orthopedic Surgery, Egg Harbor, NJ, United States
(2) Orthofix, Lewsiville, TX, United States
(3) Orthofix, Lewisville, TX, United States
(4) Center for Outcomes Research, Cedar Point, NC, United States


Background: Foley reported the average risk of pseudoarthrosis at 6 months is lowered 14.8 percentage points (16.4% with PEMF vs. 31.2% in controls, p=.0065) and 6.1 percentage points at 12 months (7.2% with PEMF vs.13.3% in controls, p=0.1129). However, authors report risk is a function of patient-specific factors (age, smoking, etc.). No published tool is available to predict a patient's chance of non-fusion with or without a treatment to increase fusion rate.

Purpose: The purpose of this study was to determine if patient characteristics and use of PEMF treatment can predict cervical non-fusion rates at 6 and 12 months, and to report the incremental change in fusion rates associated with use of PEMF in a specific patient. Design: A retrospective analysis of pooled data from a randomized clinical trial (RCT) and an observational study, with both studies focused on pulsed electromagnetic field therapy (PEMF) use in ACDF.

Patient Sample: All patients with complete data from the RCT and the observational study were included in the analysis. Outcome measure: The outcome measures were cervical fusion at 6 months and 12 months.

Methods: Two mathematical models (6 and 12 month outcomes) were developed using logistic regression to predict the risk of non-fusion with and without PEMF therapy following ACDF for a patient with specific characteristics. A variable was included in the models if inclusion resulted in a 10% or greater increase in the magnitude of the regression coefficient. Interaction with treatment terms were included for all variables in the model.

Results: The RCT enrolled 323 patients, and the observational study enrolled 274, for a total of 597. The following variables were included in the models: age, smoking, disability, workman's compensation, and number of cervical levels fused. The numbers of patients with complete data in the 6 and 12 month models were 440 and 454 respectfully. Both 6 and 12 month models were statistically significant with the ability to predict chance of fusion. In both the 6 and 12 month models, chance of fusion is improved with PEMF use (p< 05). The figure shows the predicted benefit of PEMF for a low risk and enhanced risk patient with the specified characteristics. Figure 1. Comparison of the average non-fusion rate risk reduction with PEMF use reported by Foley to 2 different personalized risk reductions with PEMF use calculated using the mathematical models developed in this research. Low risk ACDF patient: 41-year-old smoker on worker's comp, 1 cervical level fused. Enhanced risk ACDF patient: 56-year-old smoker, 2 cervical levels fused.

Conclusions: A mathematical model was developed to predict a patient's probability of non-fusion following ACDF. Use of PEMF therapy improved the chance of fusion at both 6 and 12 months in these models. The magnitude of a patient's baseline risk of non-fusion, and the benefit of PEMF is a function of the patient-specific characteristics.

6 & 12 month risk reduction