As cases of COVID-19 increase in number worldwide, clinicians are struggling to diagnose new cases quickly enough to implement appropriate isolation measures. This is particularly difficult given how closely symptoms of COVID-19 match other common viral respiratory infections, including influenza. The aim of this study was to summarize the diagnostic features of suspected cases of COVID-19 in our hospital to help improve differential diagnosis, reduce misdiagnosis in future. Results of our study suggest that pneumonia in COVID-19 patients and pneumonia caused by other pathogens (eg, influenza viruses, adenovirus, and Mycoplasma) are difficult to distinguish based on their clinical manifestations, which included fever, cough, and fatigue in our study. Rarer clinical manifestations, such as expectoration, sore throat, intolerance of cold, shivering, chest tightness, dyspnea, palpitation, and diarrhea, were also common to both COVID-19 and other respiratory pathogens, which was similar to the results from previous studies [8,9].
Routine blood tests of COVID-19-positive patients showed that the WBC count was reduced, inversely correlating with the severity of fever, instead of COVID-19-negative patients. This may contribute to the differential diagnosis of suspected cases. Approximately half (47.1%) of COVID-19-positive patients had reduced lymphocyte count; therefore, a reduced lymphocyte count in suspected cases of COVID-19 suggests the possibility of COVID-19. C-reactive protein of the COVID-19-positive patients was elevated, but was not significant for differential diagnosis. Most COVID-19-positive cases had bilateral pulmonary involvement with GGOs, multiple patchy shadows, and consolidation in their chest upon HRCT imaging, which may be helpful for differential diagnosis. However, these Chest HRCT imaging including patchy shadows, nodular shadows and grid-like also seen in the pneumonia caused by Influenza A virus, influenza B virus and adenovirus, consistent with the previous reports [10,11].
Clustered occurrence is one of the epidemiological criteria for the diagnosis of COVID-19 [12,13]. Our study also shows 6 out of 17 cases were clustered, but clustering is not unique to COVID-19. Mycoplasma pneumonia can also occur in cluster with GGOs in chest CT. In this study, the 7 patients diagnosed with Mycoplasma pneumonia had a mean age of 29.5 years. Three of the patients were younger than 20 years old. Bronchial wall thickening, characteristic change of Mycoplasma pneumonia, in the chest HRCT of young adults may help distinguish Mycoplasma pneumonia from COVID-19.
Additionally, we recommend performing multiplex PCR nucleic acid testing using throat swabs or sputum. It should be noted that these results may be related to factors such as sampling quality, specimen preservation, and different nasopharyngeal virus concentrations at different stages of the disease . Using multiplex PCR, we distinguish influenza A virus, influenza B virus, Adenovirus, Chlamydia pneumoniae, Mycoplasma pneumoniae and so on from suspected cases easily. In this study, no patient with COVID-19 was found co-infection with other respiratory pathogen(s). Recent report showed that rate of co-infection between SARS-CoV-2 and other respiratory pathogens reached 20.7% in northern California, USA. So, testing of SARS-CoV-2 should been done for patients with non–SARS-CoV-2 respiratory pathogens in high incidence of COVID-19 region. The detection of non–SARS-CoV-2 respiratory pathogens by multiplex PCR may help to assess individual the risk of COVID-19 in areas of low transmission.  Because of the highly infectious nature of SARS-CoV2, the suspected COVID-19 cases were all isolated and monitored in a single-person single-room isolation ward. Although communication with healthcare professionals was limited, a detailed medical history should not be neglected. Therefore, it is important to remain open to all causes of lung pathology, including non-infectious causes like pulmonary embolism. For suspected COVID-19 cases, a comprehensive multidisciplinary collaborative diagnosis and treatment (MDT) mechanism should be established. Relevant departments including respiratory medicine, infectious diseases, and radiology should collaborate closely when COVID-19 is suspected to avoid misdiagnosis. Positive result of SARS-CoV-2 nucleic acid testing remains the gold standard for the diagnosis of COVID-19. However, highly suspicious cases with false negative viral nucleic acid testing results should have chest CT and consecutive viral nucleic acid testing in different specimens collected from multiple regions of the body (eg, sputum, throat swabs, blood, urine, and feces) . These patients should also have the tests of serum SARS-CoV-2 specific-IgM and IgG antibodies  to improve the diagnosis rate.
This study has some limitations. Because COVID-19 was managed as Class A infectious disease, this study only performed routine blood tests, C-reactive protein, chest HRCT, throat swab SARS-CoV-2 nucleic acid testing, but not blood biochemical tests in the patients. As a result, we cannot comment on co-morbidities in our population. Additionally, the number of cases in this study was limited by the fact that COVID-19 is an emerging disease, and our findings need to be further verified by a large-scale study.