Ordering Spine MRI in Emergency Medicine

The 2016 data available from the US National Center for Health Statistics shows the overall Emergency Department visits as follows: 

  • number of visits to emergency departments nationwide was 145.6 million;
  • percent of visits with patient seen in fewer than 15 minutes was 39.0%; and 
  • percent of visits resulting in transfer to a different hospital was 2.7%.

Health care groups rate individual physicians according to how much time they spend evaluating each patient in triage, which also takes into account the time taken to conduct diagnostic tests in investigating clinical cases, such as epidural abscess. These metrics are:

The US emergency department metrics (source:  Emergency Physicians, 11 Benchmarks that should matter to Eps)
Door-to-Doctor Time10 minutes
Door-to-Room Time25 minutes
Admit Decision-to-Depart Time100 minutes
Left Without Being Seen (LWBS)The “Excellent Threshold” is 1% or less

In a community hospital, it is challenging for the physicians to counter delays because of the length of time it takes to receive a radiology report after ordering an urgent spine MRI for the following indications:

  • Spinal cord compressions (herniated disc, burst fracture, tumors)
  • Spinal infection (i.e., abscess)
  • Spinal trauma (epidural hemorrhage)
  • Demyelination with acute neurologic changes

In any of these cases, the physician needs diagnosis as expeditiously as possible to refer the patient to the appropriate specialist, which sometimes requires moving the patient to a different hospital for treatment. In some cases, a quick diagnosis from the radiologists can prevent irreversible disability (i.e., Cauda equina syndrome).

How Long Does It Take to Get an MRI Scan and MRI Read?

Most community hospitals have a shortage of radiologists, usually just one to serve the entire hospital, and it can take up to five hours to get the diagnostic read from the radiology report. Delays may occur due to a variety of factors, including:

How busy the radiologist is?

  • Mondays it can take up to two hours to receive an MRI read because Saturday and Sunday are outside their contractual hours and there is a backlog from Friday.
  • It takes at least two hours of a radiologist absence when performing intervention radiology procedures (e.g. image-guided lumbar epidural corticosteroid injection)

Time of the day

  • There are no in-house radiologists and MRI technologists after contractual hours (18:00).

Time plays a critical role in an emergency situation, and the turnaround time matters to physicians as well as the patients. Receiving a diagnosis within minutes as opposed to hours is beneficial to patients; they can be referred for treatment quickly, which can save lives.

How Can Machine Learning Algorithms Support Radiologists, Physicians, and Patients?

Imagine a radiology AI solution designed to support the work of radiologists as well as clinicians working in urgent/emergency care environments. What if there were a solution that could automatically interpret sequences of musculoskeletal MRI images and provide results back directly into PACS, without an additional interface?

Not only could the AI solution be a helpful partner to emergency physicians, but it could reduce delays and provide a much-needed service to alleviate the stress experienced by overstretched staff juggling multiple tasks (e.g. performing interventional radiology and reading MRI images) and improve their ability to meet deadlines.

Machine learning algorithms can also have a positive impact on patient care by reducing long waiting times for anxious patients and their families due to the lack of human resources to read the MRI images. AI technology has the potential to serve radiologists, physicians, and patients, as well as other key stakeholders. All that is needed is for the clinical community to embrace the technology and be able to trust in the accuracy of the diagnoses from Radiology AI solutions.

About Author

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>