OBJECTIVE: To systematically review evidence about the relationship between metformin administration and the use of iodinated contrast medium and risk of lactic acidosis (LA) and to assess the quality of five current guidelines for use of contrast medium in patients who are taking metformin.
METHODS: A search strategy was developed by using search termsrelated to metformin, contrast media, and LA. Searches were conducted in MEDLINE (Ovid), all Evidence-based Medicine Reviews (Ovid), EMBASE, and Cochrane library databases and were augmented with searches for evidence-based guidelines on radiology and evidence-based medicine Web sites by using the Google Internet search engine.
Guidelines were appraised by two independent reviewers by using the Appraisal of
Guidelines Research and Evaluation Collaboration Instrument. Other studies were appraised by using structured appraisal checklists.
RESULTS: Five guidelines were identified and five empirical studies met inclusion criteria. All guidelines had poor scores on some Appraisal of Guidelines for Research and Evaluation (AGREE) Collaboration criteria; poorer scores tended to occur in relation to objective assessment of rigor of
guideline development, editorial independence, and applicability of the
guideline to clinical practice. Lack of agreement was observed among guidelines about the need to stop taking metformin after contrast medium is administered, risk of LA in patients with normal renal function before contrast medium injection, recommended method of measuring renal function, and values used to define abnormal function. The evidence that was used as a basis for determining the guidelines for metformin administration, use of contrast medium, and risk of LA consisted of a limited number of observational studies, including case reports, summaries of case reports, and case series (National Health and Medical Research Council of Australia level IV-V evidence). More important, it was not apparent that a systematic search strategy had been used to identify the studies referenced by the guidelines nor that the referenced studies had been appraised. Individual
guideline recommendations were not clearly connected to cited references.
CONCLUSIONS: Substantial inconsistencies exist between the recommendations of the five international guidelines about contrast medium administration in patients who are taking metformin. These are, in part, caused by the low level of evidence underpinning
guideline recommendations.