Low back pain remains a leading cause of disability. Utilization of healthcare resources has grown without corresponding improvement in outcomes and the use of lumbosacral epidural steroid injections (ESI) remains controversial. Given challenges accurately measuring effectiveness, recent focus has shifted towards patient-centered criteria. Studies show patients wish to have preferences incorporated into treatment algorithm and be involved in decision making. The current quality improvement initiative aims to improve selection and patient outcomes by utilizing a novel clinical analytic model.
Patients undergoing lumbosacral ESI were enrolled from an outpatient spine clinic following evaluation by a single physician. They completed a Oswestry Disability Index(ODI) and Numeric Pain Rating Scale(NPRS) during the initial encounter and at subsequent encounters. Demographic and clinical characteristics were used to create a patient specific mathematical model based on prior patients from our healthcare system, from which an odds ratio for improvement following injection was calculated. This odds ratio was shared with the patient as part of shared decision making.
We propose a patient specific shared decision making tool for evaluating patients who may benefit from ESI. Initial data indicates this is reliable at predicting response to ESI with an accuracy of 75% in our sample population, with additional prospective validation needed. This tool has the potential to help select appropriate patients, improve outcomes, and facilitate dialogue between patient and provider.
Rohit Navlani– Resident Physician, UPMC
Joseph Staszel– Resident Physician, University of Pittsburgh Medical Center, Department of Physical Medicine & Rehabilitation
Michael O'Connell– Fellow, Allegheny Health Network
Suehun Ho– Assistant Professor, University of Pittsburgh Medical Center Department of Physical Medicine & Rehabilitation
Oscar Marroquin– Chief Analytics Officer and Associate Professor of Medicine and Epidemiology, UPMC and University of Pittsburgh, UPMC Health Services Division
Petr Pancoska– Research Associate Professor, University of Pittsburgh Medical Center
Steven Koscumb– Director - Clinical Analytics, University of Pittsburgh Medical Center
William Garrard– Software Engineer, Adjunct Faculty at University of Pittsburgh School of Computing and Information, University of Pittsburgh Medical Center
Kevin Quinn– Director, Clinical Analytics, University of Pittsburgh Medical Center
Gwendolyn Sowa– Professor, Chair of Department of Physical Medicine & Rehabilitation, University of Pittsburgh Medical Center, Department of Physical Medicine & Rehabilitation