Meta-Analysis Comparing Risk for Adverse Outcomes After Left Ventricular Assist Device Implantation in Patients With Versus Without Diabetes Mellitus

• Published in: The American journal of cardiology 2019-12, Vol.124 (12), p.1918-1923
• Main Authors: Lateef, Noman, Usman, Muhammad Shariq, Colombo, Paolo C., Ahmed, Saba, Khan, Muhammad Shahzeb, Khan, Safi U., Garan, Arthur R., Yuzefpolskaya, Melana, Demmer, Ryan T.
• Format: Article
• Language: English
• Subjects: Bleeding; Cause of Death; Cohort Studies; Confidence intervals; Congestive heart failure; Diabetes; Diabetes mellitus; Diabetes Mellitus - diagnosis; Diabetes Mellitus - drug therapy; Diabetes Mellitus - epidemiology; Female; Heart failure; Heart Failure - diagnosis; Heart Failure - epidemiology; Heart Failure - surgery; Heart-Assist Devices; Hemorrhage; Humans; Hypertension; Implantation; Infections; Ischemia; Life expectancy; Life span; Male; Meta-analysis; Morbidity; Mortality; Prognosis; Retrospective Studies; Risk analysis; Risk Assessment; Stroke; Survival Analysis; Systematic review; Thromboembolism; Transplants & implants; Treatment Outcome; Ventricle
• ISSN: 0002-9149; 1879-1913
• DOI: 10.1016/j.amjcard.2019.09.021

Patients with diabetes mellitus (DM) are known to have reduced life expectancy and be at increased risk for multiple morbidities including serious infection. However, published data on DM outcomes after left ventricular assist device (LVAD) implantation are sparse, inconsistent and individual studies are small with limited power. We conducted a systematic review and meta-analysis to compare survival and adverse events post-LVAD in DM versus non-DM (NDM) patients. Medline, Scopus, and Cochrane Central databases were searched for studies comparing outcomes in DM and NDM patients who underwent LVAD implantation for advanced heart failure. Results were reported as random effect risk ratios (RR) with 95% confidence intervals. We identified 5 retrospective cohort studies, at low risk of bias, reporting on 1,351 patients (n = 488 DM). There was increased 30-day mortality (RR 1.57 [1.00, 2.47]; p = 0.05; I2 = 0%) among DM versus NDM. The DM and NDM groups did not differ significantly in terms of 1-year mortality (RR 1.15 [0.98, 1.35]; p = 0.08; I2 = 39%), device-related infection (RR 1.05 [0.92, 1.19]; p = 0.88; I2 = 0%), ischemic stroke (RR 1.29 [0.91, 1.83]; p = 0.69; I2 = 0%), hemorrhagic stroke (RR 1.10 [0.42, 2.83]; p = 0.85; I2 = 69%), or bleeding (RR 1.06 [0.80, 1.40]; p = 0.70; I2 = 27%). After LVAD implantation, patients with DM, versus patients without, have a modestly elevated 30-day mortality rate. However, 1-year mortality and morbidity did not differ.