Effectiveness of Interventions to Decrease Image Ordering for Low Back Pain Presentations in the Emergency Department: A Systematic Review.

BACKGROUND: Low back pain (LBP) is an extremely frequent reason for patients to present to an emergency department (ED). Despite evidence against the utility of imaging, simple and advanced imaging (i.e., computed tomography [CT], magnetic resonance imaging) for patients with LBP has become increasingly frequent in the ED. The objective of this review was to identify and examine the effectiveness of interventions aimed at reducing image ordering in the ED for LBP patients. METHODS: A protocol was developed a priori, following the PRISMA guidelines, and registered with PROSPERO. Six bibliographic databases (including MEDLINE, EMBASE, EBM Reviews, SCOPUS, CINAHL, and Dissertation Abstracts) and the gray literature were searched. Comparative studies assessing interventions that targeted image ordering in the ED for adult patients with LBP were eligible for inclusion. Two reviewers independently screened study eligibility and completed data extraction. Study quality was completed independently by two reviewers using the before-after quality assessment checklist, with a third-party mediator resolving any differences. Due to a limited number of studies and significant heterogeneity, only a descriptive analysis was performed. RESULTS: The search yielded 603 unique citations of which a total of five before-after studies were included. Quality assessment identified potential biases relating to comparability between the pre- and postintervention groups, reliable assessment of outcomes, and an overall lack of information on the intervention (i.e., time point, description, intervention data collection). The type of interventions utilized included clinical decision support tools, clinical practice guidelines, a knowledge translation initiative, and multidisciplinary protocols. Overall, four studies reported a decrease in the relative percentage change in imaging in a specific image modality (22.7%-47.4%) following implementation of the interventions; however, one study reported a 35% increase in patient referrals to radiography, while another study reported a subsequent 15.4% increase in referrals to CT and myelography after implementing an intervention which reduced referrals for simple radiography. DISCUSSION: While imaging of LBP has been identified as a key area of imaging overuse (e.g., Choosing Wisely recommendation), evidence on interventions to reduce image ordering for ED patients with LBP is sparse. There is some evidence to suggest that interventions can reduce the use of simple imaging in LBP in the ED; however, a shift in imaging modality has also been demonstrated. Additional studies employing higher-quality methods and measuring intervention fidelity are strongly recommended to further explore the potential of ED-based interventions to reduce image ordering for this patient population.

Effectiveness of Implementing Evidence-based Interventions to Reduce C-spine Image Ordering in the Emergency Department: A Systematic Review.

OBJECTIVES: Appropriate use of imaging for adult patients with cervical spine (C-spine) injuries in the emergency department (ED) is a longstanding issue. Guidance for C-spine ordering exists; however, the effectiveness of the decision support implementation in the ED is not well studied. This systematic review examines the implementation and effectiveness of evidence-based interventions aimed at reducing C-spine imaging in adults presenting to the ED with neck trauma. METHODS: Six electronic databases and the gray literature were searched. Comparative intervention studies were eligible for inclusion. Two independent reviewers screened for study eligibility, study quality, and extracted data. The change in imaging was reported using individual odds ratios (ORs) with 95% confidence intervals (CIs) using random effects. RESULTS: A total of 990 unique citations were screened for relevance of which six before-after studies and one randomized controlled trial were included. None of the studies were assessed as high quality. Interventions consisted primarily of locally developed guidelines or established clinical decision rules such as the NEXUS or the Canadian C-spine rule. Overall, implementation of interventions aimed at reducing C-spine image ordering resulted in a statistically significant reduction in imaging (OR = 0.69, 95% CI = 0.51-0.93); however, heterogeneity was high (I2  = 82%). Subgroup analysis revealed no differences between studies that specified enrolling alert and stable patients compared to unspecified trauma (p = 0.81) or between studies employing multifaceted versus nonmultifaceted interventions (p = 0.66). While studies generally provided details on implementation strategies (e.g., teaching sessions, pocket cards, posters, computerized decision support) the effectiveness of these implementation strategies were frequently not reported. CONCLUSION: There is moderate evidence regarding the effectiveness of interventions to reduce C-spine image ordering in adult patients seen in the ED with neck trauma. Given the national and international focus on improving appropriateness and reducing unnecessary C-spine imaging through campaigns such as Choosing Wisely, additional interventional research in this field is warranted.