The aim of current study was to evaluate the outcome of some laboratory tests and accuracy of diagnostic tests for patients with COVID-19. The QUADAS-2 tool was utilized to quality assessment of diagnostic accuracy studies. For data extraction, two reviewers blind and independently extracted data from abstract and full text of studies that included. 95% confidence interval for effect size with fixed effect model and Inver-variance method were calculated. Meta-analysis was performed using Stata/MP v.16 software. 469 studies were reviewed, of which the full text of 52 studies was reviewed and finally twenty-two studies were selected for meta-analysis. Mean differences of Alanine transaminase was 3.40 U/L (MD, 95% CI -2.45, 9.25), correlation between Alanine transaminase and severe COVID-19 was not significantly. Stool, feces, rectal swabs 25% (ES, 95% CI 0.32, 0.82), Urine 1% (ES, 95% CI 0.05, 0.58), overall sensitivity of PCR for detection of COVID-19 was 44% (ES, 95% CI 0.19, 0.68), less sensitive observerd when PCR detection of COVID-19. Evidence revealed that at the time of admission of patients with COVID-19, a specific laboratory model can be used to perform relevant tests and make decisions about patients. PCR using sputum samples was highly sensitive for detecting COVID-19 and after that computed tomography of the chest was identified with high sensitivity.