Abstract Background Reverse transcription-polymerase chain reaction (RT-PCR) assays detect severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The number of viruses in the sample varies between patients; it depends on sample location, nasal or throat, and with time infection spreads. Previous studies showed that the viral load of coronavirus disease 2019 (COVID-19) infection is the peak just before symptoms onset. Furthermore, positive and negative results depend on test site, sampling, and timing method; RT-PCR can be 1 to 30% false-negative result. Methods and Materials Within this study, we took RT-PCR test from COVID-19 positive patients who already had the confirmation of the disease either by lung computed tomography (CT)-scan or the symptoms such as dyspnea. The study was explained to all the patients, and they confirmed to take the RT-PCR test. Negative samples from those patients were retested, and if the result came back negative, we included them as negative in the result. Result A total number of 49 patients (25 females) and (24 males) with a mean age of 53.24 years (ranging from 32 to 77) were enrolled. About 32.3% of patients, despite having COVID-19 disease, had a negative RT-PCR test. There is a positive and significant relationship between weight (r = 0.253) and CT at the time of hospitalization of COVID-19 patients and a negative and significant relationship with O2 saturation without oxygen therapy (r = − 0.296), the model can predict 67.7% of the disease due to the beta value, and the share of O2 saturation without oxygen therapy is more than weight. Conclusion We show that a pragmatic model can be designed to predict which patients have a higher chance of getting false-negative result, and should be retested for COVID-19. Among the variables, weight had a negative and significant relationship, and O2 saturation without respiratory support had a negative and significant relationship with COVID-19 disease.