Identifying patients with metformin associated lactic acidosis in the emergency department

Metformin-Associated Lactic Acidosis-Is This on Your Radar?

BACKGROUND

Metformin is a biguanide hyperglycemic agent used to manage non-insulin-dependent diabetes mellitus. Adverse reactions include mainly mild gastrointestinal adverse effects, but severe complications, such as metformin-associated lactic acidosis (MALA) can occur. Metformin is excreted renally and, therefore, not recommended in patients with renal impairment. The reported incidence of MALA is 3 cases per 100,000 patient-years.

CASE REPORT

A 79-year-old woman with a complex medical history, including end-stage renal disease on dialysis and type 2 diabetes, presented to the emergency department (ED) for altered mental status. Prior to arrival, she was found to be hypoglycemic. Her laboratory results were significant for creatinine of 6.56 mg/dL and an anion gap of 52 mmol/L. The venous blood gas revealed a venous pH of 6.857 [reference range (7.32-7.43)], pCO2 of 15.9 mm Hg (40.6-60 mm Hg), HCO3 of 2.7 mmol/L (21-30 mmol/L), lactate of 27 mmol/L (0.5-2 mmol/L), and ammonia of 233 µmol/L. The patient was dialyzed emergently in the ED; repeat laboratory test results showed blood urea nitrogen of 10 mg/dL, creatinine of 1.65 mg/dL, carbon dioxide of 26 mmol/L, and anion gap of 13 mmol/L. The repeat ammonia was 16 µmol/L. The patient's metabolic encephalopathy resolved, and she was discharged home on hospital day 3. WHY SHOULD AN EMERGENCY PHYSICIAN BE AWARE OF THIS?: MALA has a high mortality rate (36%). Laboratory markers have not been found to be a reliable predictor of mortality. Sodium bicarbonate is controversial, but a pH < 7.15 indicates consideration of its use. A pH < 7.1 and a lactate level > 20 mmol/L indicate the need for emergent hemodialysis. Prompt recognition and management in the ED with early hemodialysis can result in good patient outcomes, with a return to their baseline function despite severe laboratory findings.

Risk of Lactic Acidosis in Hospitalized Diabetic Patients Prescribed Biguanides in Japan: A Retrospective Total-Population Cohort Study

Patient data from the National Database of Health Insurance Claims and Specific Health Checkups of Japan (NDB) are used to assess the effect of biguanide administration on rates of lactic acidosis (LA) in hospitalized diabetes mellitus (DM) patients. In this retrospective cohort study (from April 2013 to March 2016), we compare DM inpatients prescribed biguanides to DM inpatients who were not prescribed biguanides to quantify the association between biguanides and incidence of LA. In total, 8,111,848 DM patient records are retrieved from the NDB. Of the 528,768 inpatients prescribed biguanides, 782 develop LA. Of the 1,967,982 inpatients not prescribed biguanides, 1310 develop LA. The rate ratio of inpatients who develop LA and are administered biguanides to those who developed LA without receiving biguanides is 1.44 (95% CI, 1.32-1.58). Incidence rates and rate ratios for both sexes are elevated in the group prescribed biguanides for patients aged 70 years and older, markedly in those 80 years and older: 40.12 and 6.31 (95% CI, 4.75-8.39), respectively, for men and 34.96 and 5.40 (95% CI, 3.91-7.46), respectively, for women. Biguanides should be used conservatively in patients older than 70 years, particularly for those with comorbidities, and with caution in patients 80 years and older.

Utility of support vector machine and decision tree to identify the prognosis of metformin poisoning in the United States: analysis of National Poisoning Data System

BACKGROUND

With diabetes incidence growing globally and metformin still being the first-line for its treatment, metformin's toxicity and overdose have been increasing. Hence, its mortality rate is increasing. For the first time, we aimed to study the efficacy of machine learning algorithms in predicting the outcome of metformin poisoning using two well-known classification methods, including support vector machine (SVM) and decision tree (DT).

METHODS

This study is a retrospective cohort study of National Poison Data System (NPDS) data, the largest data repository of poisoning cases in the United States. The SVM and DT algorithms were developed using training and test datasets. We also used precision-recall and ROC curves and Area Under the Curve value (AUC) for model evaluation.

RESULTS

Our model showed that acidosis, hypoglycemia, electrolyte abnormality, hypotension, elevated anion gap, elevated creatinine, tachycardia, and renal failure are the most important determinants in terms of outcome prediction of metformin poisoning. The average negative predictive value for the decision tree and SVM models was 92.30 and 93.30. The AUC of the ROC curve of the decision tree for major, minor, and moderate outcomes was 0.92, 0.92, and 0.89, respectively. While this figure of SVM model for major, minor, and moderate outcomes was 0.98, 0.90, and 0.82, respectively.

CONCLUSIONS

In order to predict the prognosis of metformin poisoning, machine learning algorithms might help clinicians in the management and follow-up of metformin poisoning cases.