Overview

This project at an academic medical center utilized a clinical decision support tool to reduce inappropriate MRI use in adult patients with low back pain.

Collaborators

Esteban F. Gershanik, MD, MPH, MMSc
Louise I. Schneider, MD
Ali S. Raja, MD, MPH, MBA
Wenhong Mar, MSc
Steven Seltzer, MD
Michael J. Healey, MD
Ramin Khorasani, MD, MPH

Status/Stage of Development

Completed

Funding Sources

This study was funded in part by Grant 1UC4EB012952-01 from the National Institute of Biomedical Imaging and Bioengineering.

Practice Setting

Academic Medical Center

National/Policy Context

  • Advanced medical technology brings the potential to enhance patient safety and improve quality of care, but also often results in unnecessary and high-cost medical tests, such as inappropriate imaging. 
  • ACP/APS guidelines outline situations where imaging is not clinically indicated.

Local/Organizational Context

  • Low back pain affects 70-85% of Americans and continues to be a significant clinical driver of unnecessary imaging orders, despite the evidence that imaging results in the context of simple back pain and a patient’s clinical symptoms are poorly correlated. 
  • Therefore, BWH introduced a clinical decision support intervention based on published ACP/APS guidelines on MRI use for low back pain and aimed to examine its impact on MRI use.

Patient Population Served and Payor Information

  • Patients involved were aged 18 or older with an associated primary or secondary diagnosis of low back pain.

Project Research + Planning

  • Before the intervention began, there was a baseline data-gathering observational period that lasted seven chronological quarters. Data on MRI use for low back pain and other physician MRI ordering habits were collected. 
  • Restructuring of the computerized physician order entry system to implement the clinical decision support (CDS) also occurred.

Tools or Products Developed

  • Clinical decision support (CDS): This is a multi-faceted, real-time, online clinical decision support enabled intervention based on published ACP/APS guidelines for ordering low-back MRIs. It will notify the ordering clinician if an MRI order is likely not indicated.  Based on the clinical history input via the CPOE system, CDS launches, advising the orderer regarding the best diagnostic strategy if evidence is available.

Training

  • The project did not require pre-training for staff; however, when PCPs tried to override a low-back MRI CDS alert that said the MRI was not indicated, they were required to have peer-to-peer phone calls with a radiologist or internist familiar with MRI evidence before doing so.

Tech Involved
  • Electronic medical record
Team Members Involved
  • Physicians
Workflow Steps

  • The institutional billing database was used to identify all primary care visits of patients 18 years and older with an associated primary or secondary diagnosis (within their top two diagnoses) of low back pain.
  • When physicians in the intervention tried to order an MRI for a patient with low back pain, the real-time CDS tool was launched and advised the physician with the best diagnostic strategy for ordering the MRI.  
    • If the request was flagged as “not indicated,” then the clinician could either cancel the MRI request or ignore the CDS alert and proceed with the order.  
  • Two accountability tools were also involved:
    • Mandatory real-time consultation with a radiologist or internist was required before completion of an order that was alerted by CDS to be “not indicated.” 
    • Quarterly practice pattern reports were sent to individual physicians to show their MRI utilization.

Outcomes

  • Rate of lumbar-spine MRI use: Assessed by measuring the rate of lumbar-spine MRIs performed on patients reporting low back pain.
    • 5.3% (n = 443/8437) of pre-intervention low back pain-related primary care visits resulted in lumbar-spine MRI, whereas 3.7% (n = 477/13,008) of visits post-intervention resulted in MRI (p <0.0001).
  • MRI use of any body part: Assessed by number of other organ system-focused MRIs performed on patients.
    • There was a 30.8% relative decrease in the use of MRI of any organ system by the primary care providers in the study cohort. 
  • Lumbar-spine MRIs adherent to guidelines: Assessed by proportion of lumbar-spine MRIs performed that were adherent to evidence-based guidelines for MRIs.
    • Adherence to guidelines for lumbar-spine MRI increased from 78% to 96% in the intervention group (p = 0.0002). 

Future Outcomes

  • The team did not assess the intervention’s impact on patient or provider satisfaction, an important topic for future studies.

 

Benefits

  • This project introduced one way by which clinicians could better understand, monitor, and adapt their own MRI ordering habits.
  • This project could result in improving quality, reducing waste, and enhancing value for patients.
  • The CDS system can potentially be implemented in other medical institutions in the future to further decrease unnecessary MRI use and raise awareness about overuse.
  • Overall, this intervention generates cost savings to healthcare system by preventing unnecessary MRIs.

Intervention-Specific Challenges

  • The team was not able to measure the specific impact of individual components of the intervention (i.e., CDS, quarterly reporting, and peer-to-peer consultation) on ordering behavior.
  • Billing data were used in cohort identification, so the intervention may not have captured all eligible patients.

Sources

  • Ip, I. K., Gershanik, E. F., Schneider, L. I., Raja, A. S., Mar, W., Seltzer, S., Khorasani, R. (2014). Impact of IT-enabled Intervention on MRI Use for Back Pain. Am J Med. 2014 June; 127(6): 512–518.e1. doi:10.1016/j.amjmed.2014.01.024. 
  • Chou R, Qaseem A, Snow V, et al. Diagnosis and treatment of low back pain: a joint clinical practice guideline from the American College of Physicians and the American Pain Society. Ann Intern Med. 2007; 147(7):478–491. [PubMed: 17909209]

Innovators
  • Esteban Gershanik, MD MPH
Talk to the Innovators