Background: In response to the need for safe care for people with diabetes mellitus in the current outbreak of COVID-19, it is critical to evaluate the model, service delivery, feasibility, and efficiency of diabetes mellitus telecoaching. Objective: This study aimed to conduct a systematic review and meta-analysis of the model and efficacy of telecoaching to improve self-care and clinical outcomes. Methods: This study uses the Preferred Reporting Item for Systematic Review and Meta-Analysis (PRISMA). We searched on 22 March 2022, using keywords that matched the MeSH browser in four databases to find relevant studies, namely, PubMed/Medline, Proquest, Scopus, and EBSCOhost. Additionally, we collected randomized controlled trials (RCTs) on Google Scholar using the snowball technique. A quality assessment was performed using the Cochrane Collaboration's Risk of Bias tool (RoB)2. The meta-analysis used the DerSimonian-Laird random-effects model to analyze the pooled mean difference (MD) and its p-value. Results: Thirteen RCT studies were included for the systematic review and meta-analysis with a total number of participants of 3300. The model of telecoaching is a form of using nurses-led telephone and mobile apps, which are relatively cost-effective. The meta-analysis showed a positively improved statistically significance in clinical outcomes, including in HbA1c (a pooled MD of -0.33; 95% CI: -0.51--0.15; p = 0.0003), blood glucose (-18.99; 95% CI: -20.89--17.09; p = 0.00001), systolic blood pressure (-2.66; 95% CI: -3.66--1.66; p = 0.00001), body mass index (-0.79; 95% CI: -1.39--0.18; p = 0.01), and weight (-2.16 kg; 95% CI: -3.95--0.38; p = 0.02). It was not, however, statistically significant in diastolic blood pressure (-0.87; 95% CI: -2.02-0.28; p = 0.14), total cholesterol (-0.07; 95% CI: -0.26-0.12; p = 0.46), low-density lipoprotein (-2.19; 95% CI: -6.70-2.31; p = 0.34), triglycerides (-13.56; 95% CI: -40.46-13.35; p = 0.32) and high-density protein (0.40; 95% CI: -1.12-1.91; p = 0.61). Conclusions: The telecoaching with nurses-led telephone and mobile apps significantly affected clinical outcomes on HbA1c, systolic blood pressure, weight, and BMI. Moreover, there was no significant effect on the total cholesterol, low-density lipoprotein, triglycerides, and high-density lipoprotein. Thus, telecoaching has the potential as a care model in diabetes mellitus during COVID-19 and similar pandemics to improve self-care and clinical outcomes, but all the studies analyzed involved non-COVID-19 patients, limiting the generalizability of the results to COVID-19.