Objectives To investigate what can be done with the well-documented problem of low-value imaging.

Method Systematic review searching Medline-Ovid, Embase-Ovid, Scopus, and Cochrane Library for studies published from 2010 to 2020. Medical subject headings (Mesh) for Diagnostic imaging/Radiology, Health service misuse or medical overuse, and Health planning were used for building the search strategy, and articles in English, German, Dutch, Swedish, Danish, or Norwegian were included. To appraise the quality of the included articles, the Mixed Methods Appraisal Tool was applied. The included articles were subject to a narrative synthesis.

Results Of 15,659 identified records, 95 studies were included in the final evidence synthesis, of which 45 studies were found through hand-searching techniques. Included studies had different designs, such as controlled and uncontrolled before-and-after studies, time series, chart reviews, and cohort studies. Interventions were mostly targeted at referring physicians. The most commonly evaluated interventions were education (n = 28) and clinical practice guidelines (n = 28), either alone or in combination with other components. More effective than single-component interventions were multi-component interventions showing a reduction in the use of low-value imaging in 74 and 94% of the studies, respectively. Musculoskeletal (n = 26), neurological (n = 23), and vascular (n = 16) imaging were the most addressed types of imaging.

Conclusion Single-component interventions were generally less effective than multi-component interventions that include education. Successful reduction of low-value imaging strongly depended on contextual and cultural factors in the health care systems. There is no quick fix for reducing low-value imaging as changing practice in complex organizations is challenging in general, and in healthcare systems in particular.