Objectives: The ongoing COVID-19 pandemic continues imposing a demand for diagnostic screening. In anticipation that the recurrence of outbreaks and the measures for lifting the lockdown worldwide may cause supply chain issues over the coming months, we assessed the sensitivity of a number of one-step retrotranscription and quantitative PCR (RT-qPCR) solutions to detect SARS-CoV-2. Methods: We evaluated six different RT-qPCR alternatives for SARS-CoV-2/COVID-19 diagnosis based on standard RNA extractions. That of best sensitivity was also assessed with direct nasopharyngeal swab viral transmission medium (VTM) heating, overcoming the RNA extraction step. Results: We found a wide variability in the sensitivity of RT-qPCR solutions that associated with a range of false negatives from as low as 2% (0.3-7.9%) to as much as 39.8% (30.2-50.2). Direct preheating of VTM combined with the best solution provided a sensitivity of 72.5% (62.5-81.0), in the range of some of the solutions based on standard RNA extractions. Conclusions: We evidenced sensitivity limitations of currently used RT-qPCR solutions. Our results will help to calibrate the impact of false negative diagnoses of COVID-19, and to detect and control new SARS-CoV-2 outbreaks and community transmissions.