Potential Impact of Prescribing Metformin According to eGFR Rather Than Serum Creatinine

Delphine S. Tuot; Feng Lin; Michael G. Shlipak; Vanessa Grubbs; Chi-yuan Hsu; Jerry Yee; Vahakn Shahinian; Rajiv Saran; Sharon Saydah; Desmond E. Williams; Neil R. Powe


Diabetes Care. 2015;38(11):2059-2067. 

In This Article

Abstract and Introduction


Objective Many societies recommend using estimated glomerular filtration rate (eGFR) rather than serum creatinine (sCr) to determine metformin eligibility. We examined the potential impact of these recommendations on metformin eligibility among U.S. adults.

Research Design and Methods Metformin eligibility was assessed among 3,902 adults with diabetes who participated in the 1999–2010 National Health and Nutrition Examination Surveys and reported routine access to health care, using conventional sCr thresholds (eligible if <1.4 mg/dL for women and <1.5 mg/dL for men) and eGFR categories: likely safe, ≥45 mL/min/1.73 m2; contraindicated, <30 mL/min/1.73 m2; and indeterminate, 30–44 mL/min/1.73 m2). Different eGFR equations were used: four-variable MDRD, Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine (CKD-EPIcr), and CKD-EPI cystatin C, as well as Cockcroft-Gault (CG) to estimate creatinine clearance (CrCl). Diabetes was defined by self-report or A1C ≥6.5% (48 mmol/mol). We used logistic regression to identify populations for whom metformin was likely safe adjusted for age, race/ethnicity, and sex. Results were weighted to the U.S. adult population.

Results Among adults with sCr above conventional cutoffs, MDRD eGFR ≥45 mL/min/1.73 m2 was most common among men (adjusted odds ratio [aOR] 33.3 [95% CI 7.4–151.5] vs. women) and non-Hispanic Blacks (aOR vs. whites 14.8 [4.27–51.7]). No individuals with sCr below conventional cutoffs had an MDRD eGFR <30 mL/min/1.73 m2. All estimating equations expanded the population of individuals for whom metformin is likely safe, ranging from 86,900 (CKD-EPIcr) to 834,800 (CG). All equations identified larger populations with eGFR 30–44 mL/min/1.73 m2, for whom metformin safety is indeterminate, ranging from 784,700 (CKD-EPIcr) to 1,636,000 (CG).

Conclusions The use of eGFR or CrCl to determine metformin eligibility instead of sCr can expand the adult population with diabetes for whom metformin is likely safe, particularly among non-Hispanic blacks and men.


Healthy People 2020 goals include developing strategies for safe and effective glycemic control.[1] One key strategy to attain this goal is to promote the use of metformin. Compared with other antidiabetes drugs, metformin is associated with decreased risk of cardiovascular events, progression of chronic kidney disease (CKD), and death.[2,3] Also, it is well recognized that metformin has a better safety profile than other medications; in particular, it does not cause hypoglycemia, a common and potentially dangerous adverse effect of insulin secretagogues.[4]

There is considerable reluctance, however, in using metformin among patients with CKD. Early pharmacokinetic studies demonstrated a prolonged half-life of metformin among individuals with severely impaired kidney function, placing them at heightened risk of lactic acidosis, a very rare (3.3–4.3 cases/100,000 patient-years) but serious metabolic complication that can occur in the setting of metformin accumulation.[5] Thus, the U.S. Food and Drug Administration (FDA) has stated that metformin is contraindicated among individuals with kidney disease, "suggested by serum creatinine (sCr) ≥1.4 mg/dL for women and ≥1.5 mg/dL for men, or abnormal creatinine clearance (CrCl), which may result from conditions such as cardiovascular collapse (shock), acute myocardial infarction, and septicemia".[6]

As the benefits of metformin have become more widely appreciated, there has been an ongoing debate as to whether these sCr thresholds are too restrictive and whether the benefits of metformin outweigh potential harm among individuals with mild-to-moderate CKD. At the same time, evidence has accumulated that sCr leads to substantial misclassification in identifying individuals with CKD and that estimated glomerular filtration rate (eGFR) is a more accurate estimation of an individual's kidney function. In 2012, the American Diabetes Association, Kidney Disease Improving Global Outcomes, European Association for the Study of Diabetes, and U.K. National Institute for Health and Care Excellence all recommended metformin as a first-line agent for diabetes treatment among individuals with mild CKD, defined by eGFR ≥45 mL/min/1.73 m2, and stated not to use metformin among individuals with severe CKD, defined by eGFR <30 mL/min/1.73 m2. [7–9] Because robust safety data are lacking for individuals with moderate CKD, defined by an eGFR 30–44 mL/min/1.73 m2, these societies recommended cautious use of metformin for individuals within this range, with frequent review and monitoring of kidney function.

Safely expanding metformin use among individuals with mild CKD may help improve outcomes among U.S. adults with diabetes. Our goals with this study were as follows: 1) to determine prevalence and trends of metformin use among U.S. adults with diabetes from 1999 to 2010, 2) to identify subpopulations of U.S. adults with diabetes for whom metformin is likely safe when implementing eGFR rather than conventional sCr thresholds, and 3) to determine whether different GFR- or CrCl-estimating equations could have substantial impact on the number of individuals who would be considered safe candidates for metformin use.