What data and guideline recommendations are available regarding use of metformin for patients admitted to a hospital?

Comment by InpharmD Researcher

Evidence supporting the use of metformin in hospitalized patients is scarce. The American Diabetes Association does not provide any guidance for or against metformin use in the inpatient setting, only noting that it should be stopped prior to surgery. Similarly, the Joint British Diabetes Societies guideline provides recommendations for metformin use in special patient populations that are admitted, but not for hospitalized patients overall. In general, it is typically recommended to avoid metformin use in hospitalized patients who are hemodynamically unstable, have renal impairment, or are at risk for lactic acidosis. One retrospective study in patients undergoing percutaneous coronary intervention reported that continued metformin use was not associated with adverse outcomes in patients presenting with STEMI (Table 1); however, there is an overall lack of data evaluating the efficacy and safety of metformin use in hospitalized patients.

Background

The 2023 American Diabetes Association (ADA) Standards of Care in Diabetes guideline discusses recommendations for diabetes care for hospitalized patients, but mainly focuses on insulin therapy. In a non-critical care setting, insulin is mentioned to be the preferred treatment for hyperglycemia in hospitalized patients, though it is noted that oral glucose-lowering medications may be continued under certain conditions. Unfortunately, no guidance was provided for or against metformin use in the inpatient setting; however, when discussing perioperative care, the guidelines recommend that metformin should be held on the day of surgery. [1]

The 2019 Joint British Diabetes Societies document for the management of diabetes in inpatient settings summarizes previous guideline recommendations on metformin use, separated by patient populations. For breastfeeding patients, metformin may be taken after birth, and other oral anti-diabetic treatments should be avoided. For patients on maintenance hemodialysis, metformin should be avoided. For stroke patients who require enteral feeding, resuspended metformin powder may be considered for mild hyperglycemia over other oral hypoglycemic medications (e.g., sulphonylureas). There is a lack of guidance in the document for use of metformin overall in hospitalized patients. [2]

A 2022 study reviewed the evidence for the inpatient management of hyperglycemia in non-critically ill patients with type 2 diabetes, focusing on non-insulin treatments. With regards to metformin, its use has not been recommended in the hospital setting by clinical guidelines due to the risk for lactic acidosis and other side effects; however, a number of retrospective studies have shown no increased risk of adverse events. Additionally, it is common practice to hold metformin in patients undergoing radiological studies with administration of intravenous contrast for 72 hours from the start of the procedure (i.e., percutaneous coronary intervention [PCI]); again several observational studies have reported a low risk of lactic acidosis attributed to metformin after radiological procedures or PCI, especially in patients with glomerular filtration rate > 30-60 mL/min/1.73 m2. In patients at risk for lactic acidosis (i.e., patients with renal failure, sepsis, hypoxia, liver failure, and alcoholism), metformin is not recommended to be used. Overall, there is a lack of evidence to support the use of metformin in the hospital setting; the authors concluded that if patients are relatively stable, without renal impairment, and not at increased risk for lactic acidosis, then continuation of metformin may be safe. [3]

Additionally, a recently published meta-analysis examined metformin therapy in patients hospitalized for COVID-19 to determine whether an association existed between therapy and the reduction of in-hospital mortality. A pooled analysis of 5 studies appears to link inpatient metformin use with a significant reduction of in-hospital death (unadjusted odds ratio [OR] 0.18; 95% confidence interval [CI] 0.1 to 0.31; I2= 0%). However, an adjusted analysis looking at the hazard ratio [HR] of 2 studies suggested no statistically significant association between in-hospital mortality and inpatient metformin use (HR 1.1; 95% CI 0.38 to 3.15; I2= 43%). The authors note that although there is mounting evidence to suggest that using metformin to treat diabetic patients with COVID-19 is not harmful, it does not appear to endow a protective effect, either. [4]

References:

[1] ElSayed NA, Aleppo G, Aroda VR, et al. 16. Diabetes Care in the Hospital: Standards of Care in Diabetes-2023. Diabetes Care. 2023;46(Suppl 1):S267-S278. doi:10.2337/dc23-S016
[2] Sampson M, Jones C; Joint British Diabetes Societies (JBDS) for Inpatient Care. Joint British Diabetes Societies for Inpatient Care: clinical guidelines and improving inpatient diabetes care. Diabet Med. 2018;35(8):988-991. doi:10.1111/dme.13672
[3] Galindo RJ, Dhatariya K, Gomez-Peralta F, Umpierrez GE. Safety and Efficacy of Inpatient Diabetes Management with Non-insulin Agents: an Overview of International Practices. Curr Diab Rep. 2022;22(6):237-246. doi:10.1007/s11892-022-01464-1
[4] Ma Z, Krishnamurthy M. Is metformin use associated with low mortality in patients with type 2 diabetes mellitus hospitalized for COVID-19? a multivariable and propensity score-adjusted meta-analysis. PLoS One. 2023;18(2):e0282210. Published 2023 Feb 23. doi:10.1371/journal.pone.0282210

Literature Review

A search of the published medical literature revealed 2 studies investigating the researchable question:

What data and guideline recommendations are available regarding use of metformin for patients admitted to a hospital?

Level of evidence

X - No data  Read more→



Please see Tables 1-2 for your response.


 

Effect of Continuous Use of Metformin on Kidney Function in Diabetes Patients with Acute Myocardial Infarction Undergoing Primary Percutaneous Coronary Intervention

Design

Single-center, retrospective study

N= 284

Objective

To assess continuous metformin therapy on kidney function in diabetic patients undergoing coronary intervention

Study Groups

Metformin (n= 119)

No metformin (n= 165)

Inclusion Criteria

Type 2 diabetes mellitus (T2DM); undergoing percutaneous coronary intervention (PCI) for ST-elevation myocardial infarction (STEMI); received metformin therapy before PCI

Exclusion Criteria

Patients without PCI; lack of creatinine data; end-stage kidney disease (estimated glomerular filtration rate [eGFR] < 30 mL/min/1.73 m2); respiratory failure; severe infections; liver disease; history of alcoholism; cardiogenic shock; death ≤ 48 h after hospitalization 

Methods

A single-center retrospective review was conducted on patients on T2DM patients who presented with STEMI and showed an onset of symptoms < 12 h. Continuation of metformin during the peri-angiography period was based on the decision of the treating physicians at that time. Patients in group 1 continued metformin, while group 2 stopped metformin on admission and restarted > 48 h after PCI. Non-ionic, low, or equal osmolality contrast agent was used in all patients and chosen at the discretion of the physician. 

For the primary outcome, the highest serum creatinine level within 2 days after PCI was used to diagnosed contrast-induced acute kideny injury (CI-AKI). 

Duration

January 2008 to December 2018

Outcome Measures

Primary: CI-AKI (elevation in serum creatinine ≥ 27 μmol/L or ≥ 50% over baseline within 48 h after contrast media injection)

Secondary: peak blood glucose and insulin initiation therapy within 2 days after primary PCI; lactic acidosis during hospitalization

Baseline Characteristics

 

Metformin (n= 119)

No metformin (n= 165)

 

Age ≥ 65 years

23.5% 27.9%  

Female

21.8% 28.5%  

Diabetes duration ≥ 5 years

67.2% 60%  

Comorbidities

Hypertension

Hyperlipidemia

Current smoking

Stroke

Prior myocardial infarction

Peripheral arterial disease

 

62.2%

44.5%

67.2%

6.7%

13.4%

2.5%

 

69.1%

37%

60%

8.5%

7.3%

2.4%

 

Metformin dosage

> 0 g to ≤ 0.5 g

> 0.5 g to ≤ 1 g

> 1 g to ≤ 1.5 g

> 1.5 g

 

0.8%

26.9%

71.4%

0.8%

 

4.2%

21.8%

71.5%

2.4%

 

Left ventricular ejection fraction ≤ 40%

5.9%

6.7%

 

Time to PCI, h

4 (6-16)

6 (4-9)  

Single-vessel disease

26.1% 27.3%  

Median HbA1c, % (IQR)

7 (3.6-14.2) 7.8 (6.9-8.8)  

Baseline median creatinine, μmol/L (IQR)

76 (66-86) 73 (61-84)  

Baseline median eGFR, mL/min/1.73(IQR)

89 (73-104) 94 (72-113)  

eGFR: estimated glomerular filtration rate; HbA1c: glycosylated hemoglobin; IQR: interquartile range

Results

Endpoint

Metformin (n= 119)

No metformin (n= 165)

p-value

CI-AKI

12.6% 10.3% 0.545

Post-PCI

Relative creatinine change

≥ 50% increase creatinine

≥ 27 μmol/L creatinine

Relative eGFR change

 

11% (6-20%)

4.2%

12.6%

13.4% (6.4-25.1%)

 

11% (5-22%)

5.5%

9.1%

13.5% (5.2-24.3%)

 

0.858

0.63

0.342

0.449

Initiated insulin ≤ 48 h after primary PCI

Continuous intravenous insulin

Subcutaneous insulin injection

 

1.68%

12.61%

 

7.88%

28.48%

 

0.021

0.001

Blood glucose control ≤ 48 h after primary PCI

Fasting glucose peak, mmol/L

Postprandial glucose peak, mmol/L

n= 102

7.75 (7.1-9.95)

10.65 (9-12.2)

n= 105

9.1 (7.3-11.42)

13.85 (11.9-16.2)

 

< 0.05

0.02

CI-AKI: contrast induced-acute kidney injury; eGFR: estimated glomerular filtration rate; PCI: percutaneous coronary intervention

No case of lactic acidosis was observed during hospitalization for either group. 

Univariate analysis showed that metformin was not associated with increased incidence of CI-AKI after contrast exposure. 

Multivariate analysis found the presence of contrast volume (p= 0.002) and eGFR ≤ 60 mL/min/1.73m2 (p< 0.025) were associated with CI-AKI. 

Adverse Events

See results section.

Study Author Conclusions

The study indicated that the metformin continuation after primary PCI from STEMI in diabetic patients with eGFR > 30 mL/min/1.73m2 did not increase the risk of CI-AKI. 

InpharmD Researcher Critique

The study is limited by its retrospective and single-center nature, limiting ability to control for confounding variables and generalizability of the results.



References:

Yu Q, Zhu JJ, Liu WX. Effect of continuous use of metformin on kidney function in diabetes patients with acute myocardial infarction undergoing primary percutaneous coronary intervention. BMC Cardiovasc Disord. 2020;20(1):187. Published 2020 Apr 21. doi:10.1186/s12872-020-01474-5

 

Association of Metformin Use During Hospitalization and Mortality in Critically Ill Adults With Type 2 Diabetes Mellitus and Sepsis

Design

Retrospective, multicenter, cohort study

N= 14,847

Objective

To compare 90-day mortality in diabetic patients with sepsis with and without exposure to metformin during hospitalization

Study Groups

Metformin exposure (n= 682)

No metformin exposure (n= 14,165)

Inclusion Criteria

Age ≥ 18 years, critically ill with type 2 diabetes and sepsis

Exclusion Criteria

Not specified

Methods

Patient data were compiled via retrospective chart review of 16 different hospitals within a single university medical center. Metformin exposure was defined as administration of at least 1 dose during index hospitalization regardless of timing of sepsis diagnosis. 

Duration

Patients admitted between October 2008 and December 2014

Outcome Measures

Primary: all-cause mortality at 90 days from sepsis diagnosis

Secondary: development of severe acute kidney injury (AKI) between 24 hours prior to diagnosis or dischage/death (whichever comes first), AKI recovery, major adverse kidney events (MAKE) at 30, 60, and 90 days from AKI diagnosis

Baseline Characteristics

 

Original cohort (N= 14,847)

 

 

Age, years

69    

Female

50.6%    

Body mass index, kg/m2

30.31    

Race

White

Black

Other

 

80%

10%

10%

   

Creatinine (interquartile range)

1 (0.8 to 1.4)    

Median time to sepsis diagnosis, days

0.05    

Median ICU length of stay, days

3.8    

Comorbidities

Hypertension

Chronic liver disease

COPD

CKD

Cardiovascular disease

 

16.7%

91.6%

62.8%

71.9%

39.3%

   
 

Metformin cohort (n= 682)

   

Timing of first dose of metformin in relation to ICU admission

Prior to

Within 3 days

> 3 days after

Missing information

 

 

14.3%

26.4%

43.9%

15.2%

   

Timing of first dose of metformin in relation to sepsis diagnosis

Prior to

Within 3 days

> 3 days after

Missing information

 


7.9%

29.6%

47.2%

15.2%

   

Timing of first dose of metformin in relation to AKI diagnosis

Prior to

Within 3 days

> 3 days after

Missing information

 


16.2%

29.8%

37.5%

16.4%

   

Baseline characteristics of patients were similar across groups after propensity matching. 

Results

Endpoint

Metformin exposure (n= 682)

No metformin exposure (n= 14,165)

Adjusted OR* (95% CI); p-value

Mortality at 90 days

82 (12%) 4,887 (34.5%) 0.46 (0.35 to 0.6); < 0.001

Severe AKI

339 (49.7%) 9,116 (64.4%)  0.55 (0.47 to 0.64); < 0.001 

AKI recovery (propensity score matched cohort)

276 of 291 (94.8%) 858 of 993 (86.4%) 4.2 (2.1 to 8.4); < 0.001

MAKE 

30 days

60 days

90 days

 

59 (8.7%)

85 (12.5%)

102 (15%)

 

4,018 (28.4%)

5,098 (36%)

5,606 (39.6%)

Unadjusted OR (95% CI); p-value

0.24 (0.18 to 0.31); < 0.0001

0.25 (0.2 to 0.32); < 0.0001

0.27 (0.22 to 0.68); < 0.0001

*Odds ratio (OR) and 95% confidence interval (CI) adjusted based on 1:4 propensity score matched cohort.

Adverse Events

N/A

Study Author Conclusions

Metformin exposure during hospitalization was associated with decreased risk-adjusted 90-day mortality and severe AKI in adults with type 2 diabetes mellitus and sepsis.

InpharmD Researcher Critique

The current study included patients with any exposure to metformin, as little as one dose. As such, results of this study may not be readily generalizable to patients on chronic treatment with metformin. 



References:

Gómez H, Del Rio-Pertuz G, Priyanka P, et al. Association of Metformin Use During Hospitalization and Mortality in Critically Ill Adults With Type 2 Diabetes Mellitus and Sepsis. Crit Care Med. 2022;50(6):935-944. doi:10.1097/CCM.0000000000005468