Skip to main content
. 2024 Mar 19;26:e53951. doi: 10.2196/53951

Table 2.

Type, features, and results of the CDSSsa for preventing unnecessary imaging.

CDSS name and study CDSS type Features of the CDSS Results
Choosing Wisely recommendation (Stanson Health), Chen et al [30], 2020 Knowledge based

  • Pop-up and soft-stop alert:

    • Provides best practice advise when a CTb scan, x-ray, or MRIc of lumbar spine is ordered for a female patient aged 18 to 49 y or a male patient aged 18 to 69 y

    • Alert suppressed for comorbidities; complicated back pain owing to trauma, cauda equina syndrome, spondylitis, disc herniation, history of back surgery; and disciplines, including neurosurgery, orthopedics, trauma surgery, anesthesiology, rheumatology, physical medicine and rehabilitation, oncology, and neurology

  • Post-CDSS implementation:

    • Overall imaging rate decreased from 5.8% to 5.2% (9.6% decrease; P=.02)

    • MRI imaging rate decreased from 1.8% to 1.5% (16.7% decrease; P<.01)

    • No statistically significant differences in the rates of x-ray (P=.39) or CT (P=.88) orders

  • Rationale for override

    • A total of 64% used preset options: duration >6 weeks (37%), focal neurological deficit (14%), history of trauma (10%), previous spine surgery (1%), unexplained weight loss or insidious onset (1%), and unexplained fever or recent infection (1%)

    • Free-text rationale (n=125, 36%); 56% were inappropriate
NEXUSd clinical decision rule (Medweb), Goergen et al [31], 2006 Knowledge based

  • Guideline questionnaire based on NEXUS criteria and passive alert:

    • Helps physicians to determine which patients to image and which imaging method (eg, plain radiography or helical CT) to use first

  • Compliance with CDSS and imaging guidelines:

    • 40% (141/353) of patients were managed using the CDSS

    • Of the 51 patients for whom the NEXUS guideline did not recommend imaging, 86% (43/51) did not receive any imaging

  • Cervical spine imaging ordered: CDSS intervention group: 63.8% and control group: 78.5% (P=.01)

  • Cervical spine imaging ordered (non-CDSS intervention):

    • Non-CDSS intervention groupe: 72.6% and control group: 78.5% (P=.11)
Combined NEXUS criteria and CCSRf CDSS, Hynes et al [32], 2020 Knowledge based

  • Guideline questionnaire based on NEXUS and CCSR criteria and soft-stop alert

    • Integrated in electronic imaging ordering system

    • Helps physicians follow evidence-based guidelines when ordering cervical spine radiographs for patients who have experienced trauma

    • Physicians asked to check boxes indicating which criteria the patient meets when ordering imaging

  • Cervical spine radiograph orders:

    • Preintervention: 182

    • Postintervention: 126 (P<.001)

  • Proportion of requests meeting NEXUS or CCSR criteria:

    • Preintervention: 76.7%

    • Postintervention: 99.2% (P<.001)
ACPg APSh guideline derived CDSS, Ip et al [33], 2014 Knowledge based

  • Guideline questionnaire based on ACP or APS criteria and soft-stop alert

    • Integrated into the CPOEi system

    • Provides real-time decision support to physicians for imaging orders based on the patient’s clinical history

  • Lumbar spine MRI orders:

    • Preintervention: 5.3% (443/8437)

    • Postintervention: 3.7% (477/13,008; P<.001)

  • Outpatient MRI orders 30 d after:

    • Preintervention: 2.2%

    • Postintervention: 2.7% (P=.03)

  • LBPj-related visits that resulted in an MRI within 30 d of the index visit, accounting for imaging that was ordered by specialists

    • Preintervention: 8.9%

    • Postintervention: 7.8% (P=.002)
ACRk select tool (National Decision Support Company), Mallavarapu and Christiason [34], 2020 Knowledge based

  • Guideline questionnaire based on ACR criteria and soft-stop alert

    • Integrated into the electronic medical health system

    • The free text field, which allowed providers to bypass the ACR select tool within EHRsl, has been removed to increase provider adherence to the tool

  • Preintervention: 13 scans/mo and postintervention: 11.6 scans/mo (P=.54)
Choosing wisely Canada CDSS, Min et al [35], 2017 Knowledge based

  • Guideline questionnaire based on recommendations from the Canadian Association of Emergency Physicians, the College of Family Physicians of Canada, Occupational Medicine Specialists of Canada, the Canadian Association of Radiologists, and the Canadian Spine Society, and soft-stop alert

    • Integrated with the CPOE

    • Physicians must select a suspected diagnosis when ordering an imaging test for LBP

    • If a physician selects other as the suspected diagnosis, they will need to provide an explanation for ordering the imaging test outside the established appropriateness criteria

  • Proportion of LBP patients with imaging order fell significantly compared with preimplementation baseline after CDSS implementation

    • Median: 22% decreased to 17%

    • Mean: 23% decreased to 18%; (P<.001)

  • Imaging ordering patterns

    • A total of 60% (26/43) of the physicians reduced their ordering of imaging tests.

  • Imaging orders placed 1 to 30 d after LBP presentation

    • Preintervention: 2.3% and postintervention: 2.2% (P=.97)

  • EDm revisit: preintervention 8.2% and postintervention 6.9% (P=.17)
ACR appropriateness criteria CDSS (Institute of Clinical Systems Improvement), Solberg et al [36], 2010 Knowledge based

  • Guideline questionnaire based on ACR criteria and soft-stop alert

    • Integrated within the EHR system and requires physicians to enter a reason for every order placed.

    • No safeguards were in place to prevent orders from being placed even if they did not meet certain criteria. Physicians received little feedback on the outcomes of their orders.

  • Volume of spine MRI ordered decreased by 20%

  • Impact of CDSS on patient’s health after spine MRI increased from 14% to 30% (P=.18)
Zafar et al [37], 2019 Knowledge based

  • Guideline based on ACP and APS criteria and soft-stop or passive alert

    • Embedded in CPOE, the CDSn algorithm screen the lumbar spine MRI orders for adherence to the guideline

    • Intervention groups: periodic CDSS report cards vs real-time CDSS alerts vs both

  • Likelihood of placing lumbar spine MRI orders at the time of LBP presentation when compared with baseline

    • CDSS report cards: 38% lower likelihood

    • Real-time CDSS alerts: not associated with any change (P=.59)

aCDSS: clinical decision support system.

bCT: computed tomography.

cMRI: magnetic resonance imaging.

dNEXUS: National Emergency X-Radiography Utilization Study.

eImaging guidelines given in a form of pocket card and posters, with small group teaching sessions.

fCCSR: Canadian Cervical Spine Rule.

gACP: American College of Physicians.

hAPS: American Pain Society.

iCPOE: Computerized provider order entry.

jLBP: low back pain.

kACR: American College of Radiology.

lEHR: electronic health record.

mED: emergency department.

nCDS: clinical decision support.