How to avoid being re-hospitalized for patients with COVID-19

COVID-19: 2019 Novel Coronavirus; RT-PCR: ReverseTranscriptase–Polymerase-Chain-Reaction; RNA: Ribonucleic Acid; COVID-19: Corona Virus Disease 2019; NCP: Novel Coronavirus Pneumonia; NCP: Novel Coronavirus Pneumonia; CFDA: the China Food and Drug Administration; CDC: the Centers for Disease Control; ARDS: Acute Respiratory Distress Syndrome; MOF: Multiple Organ Function Failure; ACE-2: Angiotensin-Converting Enzyme-2; BALF: Bronchoalveolar Lavage Fluid; WBC: White Blood Cell; NE: neutrophile Granulocyte; CRP: C-Reactive Protein; PCT: Procalcitonin; CT: Computed Tomography; LYM: Lymphocyte Introduction

In the present study, we used RT-PCR as a reference standard against which we tested the performance of followup chest CT in diagnosing COVID-19. The dynamic conversion of RT-PCR results was studied and compared with the serial follow-up chest CT scans.

Patients
By the time of April 30, 2020, a total of 260 patients who were diagnosed with COVID-19 and were subsequently admitted to and discharged from our hospital, underwent both chest CT and RT-PCR tests for several times between January 21 and March 6, 2020. The repeated RT-PCR tests were conducted up to and including one day after the initial test was performed.

RT-PCR assay
The RT-PCR assays were performed by using TaqMan One-Step RT-PCR Kits, which were approved by the China Food and Drug Administration (CFDA). Using viral nucleic acid extracted from oropharyngeal swab, RT-PCR assay was carried out. To accurately determine the detection limits, we serially tested diluted positive control plasmids with a dynamic range of at least seven orders of magnitude (2 x 10−4-2000 TCID50/reaction) in these assays. In our preliminary trial, reactions with ≥10 copies of this positive control plasmid were consistently positive. The amplifi cation effi ciencies of RT-PCR assay were proven. RT-PCR assay of COVID-19 nucleic acid tested using the oropharyngeal swab interval were carried out ever 24 hours.

CT scan
All chest CT images were obtained on the CT system (Optima 660, GE, America or Somatom Defi nition AS+, Siemens Healthineers, Germany) with patients in the supine position and adopting following parameters: tube voltage of 120 kVp, automatic tube current modulation between 30 and 70 mAs), a pitch between 0.99 and 1.22 mm, matrix with 512 × 512, slice thickness with 10 mm, a fi eld of view with 350 mm × 350 mm.
Images were reconstructed with a slice thickness between 0.625 and 1.250 mm with the same increment. Serial follow-up chest CT scans were carried out ever 1 week interval.

Statistical analysis
All statistical analyses were performed using SPSS 21.0.
Quantitative data were presented as the mean ± standard deviation (minimum-maximum) and the counting data were presented as the percentage of the total.

Result
The cases in our study were confi rmed for COVID-19 by the positive result of epidemiological history, symptoms, and signs (fever, cough, fatigue and/or shortness of breath), chest CT images and/or RT-PCR or gene sequencing of COVID-19 nucleic acid. The positive rates of chest CT imaging in our cohort were 96.99% (252/260) for patients with positive and negative RT-PCR results. The typical characteristics found on chest CT in patients with COVID-19 include ground-glass 4 opacities, multifocal patchy consolidation, and/or interstitial changes with a peripheral distribution. If patients were initially positive or negative for RT-PCR, typical imaging features were observed on the chest CT scan before or within six days of the initial positive RT-PCR results in all cases.

Case 1
The patient suffered from slight nausea and discomfort for 5 days without vomiting, abdominal pain, diarrhea, cold, fever, obvious cough.

The epidemiological history
The patient`s cousin came back from Wuhan on 17 January 2020. His aunt suffered from fever and was diagnosed with COVID-19 following positive test for RT-PCR of COVID-19 nucleic acid using the oropharyngeal swab test.

Past medical history
The patient was diagnosed with colorectal cancer, for which he was receiving chemoradiotherapy treatment.

Clinical examination and treatment
On admission, physical examination revealed normal vital signs. Lung auscultation revealed no rhonchi. The patient was negative for infl uenza A and B, parainfl uenza, respiratory syncytial virus, adenovirus, Mycoplasma pneumoniae, Chlamydia pneumoniae, Rickettsia burnetii, and Legionella pneumophila. Chest CT showed double pulmonary ground-glass 4 opacities and multifocal patchy consolidation, which suggested viral pneumonia. He was positive for the RT-PCR assay of COVID-19 nucleic acid that was tested using the oropharyngeal swab by the Centers for Disease Control (CDC).
The patient was diagnosed with COVID-19 on February 6th, 2020 based on the epidemiological history, the clinical character of the respiratory tract, chest CT image, RT-PCR assay of COVID-19 nucleic acid tested using the oropharyngeal swab.
Oseltamivir, arbidol, lopinavir/ritonavir, and moxifl oxacin were given. Two weeks after being admitted to hospital, the symptoms and CT lesion improved and the body temperature was back to normal. RT-PCR of COVID-19 nucleic acid with the oropharyngeal swab tests was repeatedly performed for surveillance, which resulted in negative two times during 24 hours interval. According to the criteria on discharge from hospitals in China (General Offi ce of National Health Commission and General Offi ce of National Administration of Traditional Chinese Medicine, 2020), the patient was discharged from hospital on January 29 th .
After returning home, the patient still complained of a dry cough and fever with the highest body temperature reaching 38.5℃, for which he was admitted to our hospital on February 23, 2020. The RT-PCR of COVID-19 nucleic acid with the oropharyngeal swab test resulted in positive for 2 consecutive times (more than 1 day apart). The patient was admitted to our department for confi rmation of COVID-19. The patient was re-hospitalized for low fever and occasional cough with two consecutive positive RT-PCR tests within 24 hours interval ( Figure 3 roadmap for patients second admitted to hospital).
One day later, his body temperature decreased to 37.3℃. One week after being admitted to hospital, the symptoms and CT lesions improved. During the second admission to our hospital, follow-up chest CT scans showed more consolidation, ground-glass shadow and nodules with absorbing and reducing the area, which indicated that the infl ammation was still progressing. The fi fth scan (2020-02-27) showed consolidations with airbronchogram inside with absorbing and reducing the area. By this time, the infl ammation was not progressing.
The infective marker such as procalcitonin (PCT), white blood cell number (WBC), percent of neutrophils (NE %) increase during re-hospitalization were compared with those during the fi rst admission to our hospital ( Figure 4 Dynamics of an infective marker A the fi rst patient, who was second admitted to hospital)): WBC number, NE percent increased from 5.61 10E9/L to 8.19 10E9/L respectively with the highest body temperature at 38.5℃. PCT(ng/ml) and CRP (mg/L) increased from 0.076 ng/ml, 19.63 mg/L to 0.304 ng/ml, 119.27 mg/L respectively, which further confi rmed that the infl ammation was progressing.
After being admitted to hospital, the symptoms and CT lesion improved and the body temperature was back to normal.
RT-PCR of COVID-19 nucleic acid using the oropharyngeal swab test was repeatedly performed for surveillance, resulting

Case 2
The second patient suffered from fever for 1 day that reached the highest peak of 38.8℃. He also experienced headache,

Figure Explanation
Photo A (2020-02-07 during the fi rst admission to our hospital): there were scattered consolidation, ground-glass shadow and nodules in the upper and lower lobes of the both lungs(Rapid progression stage). Photo B (2020-02-13 during the fi rst admission to our hospital): Compared with those of 2020-02-07, there were more consolidation, ground-glass shadow and nodules in both lungs(Rapid progression stage). Photo C (2020-02-17 during the fi rst admission admitted to our hospital): compared with those of 2020-02-13, there were more consolidation, ground-glass shadow and nodules with increasing density and area(Consolidation stage). Photo D (2020-02-24 during the second admission to our hospital): compared with those of 2020-02-17, there were consolidation, ground-glass shadow and nodules with absorbing and reducing the area(Consolidation stage). photo E (2020-02-27 during the second admission to our hospital): compared with those of 2020-02-24, there were consolidation, ground-glass shadow and nodules with absorbing and reducing the density(Dissipation stage). photo A to D are early stage-progression stage-consolidation stage respectively, which indicated that the infl ammation is still progressing. photo E is dissipation stage. By the time the infl ammation is not progressing. Right Serial follow-up chest CT scans of the second patient. Photo A (2020-01-19 during the fi rst admission to our hospital): there were multiple scattered patchy, light consolidation and agglomerated ground-glass opacities (Rapid progression stage). Photo B (2020-02-04 during the second admission to our hospital): compared with those of 2020-01-19, there were multiplepatchy consolidations in both lungs and grid-like thickness of interlobular septa(Rapid progression stage). Photo C (2020-02-08 during the second admission to our hospital): compared with those of 2020-02-04, there was a lot of consolidation with air-bronchogram inside with absorbing and reducing the area (Consolidation stage). Photo D: (2020-02-12 during the second admission to our hospital): compared with those of 2020-02-08, there was less consolidation, ground-glass shadow and nodules in both lungs with absorbing and reducing the area(Dissipation stage). photo A to C are early stage-progression stage-consolidation stage respectively, which indicated that the infl ammation is still progressing. photo D is dissipation stage. By the time the infl ammation is not progressing.

Figure Explanation
A. The fi rst patient, who was second admitted to hospital increase of infective marker (WBC (white blood cell number (10E9/L), NE (percent of Neutrophils (%), LYM (percent of lymphocyte (%) and PCT (procalcitonin (ng/ml)) during the second admission to our hospital, which indicated that the infl ammation was progressing low B. The second patient second, who was second admitted to hospital Dynamics of an infective marker for patients in control group with complete recovery During being admission to our hospital, increase of infective marker(WBC number , NE percent, LYM percent, CRP and PCT) were observed, which indicated that the infl ammation was progressing. After being discharged from hospital, the levels of infective markers returned to normal and never increased, which indicated that the infl ammation was no longer progressing. C. The fi rst patient with complete recovery D. Rhe second patient with complete recovery. chest tightness, chills, cough, stuffy and runny nose, muscle aches without nausea, vomiting, and slight pharyngitis. He came to the fever clinic at our Hospital. He showed no obvious abnormalities during the physical examination with reference to liver function, renal function, myocardial enzymes, electrolyte, and serum procalcitonin. He was negative for the infl uenza A and B, parainfl uenza, respiratory syncytial virus, adenovirus, Mycoplasma pneumoniae, Chlamydia pneumoniae, Rickettsia burnetii, and Legionella pneumophila. CT showed double pulmonary exudative lesions in multiple ground glass samples, suggesting viral pneumonia.

The epidemiological history
The patient came from Wuhan on January 19. He was

Clinical examination and treatemnt
The patient was diagnosed with COVID-19 on February 6th, 2020 based on the epidemiological history, the clinical character of the respiratory tract, chest CT image, and RT-PCR assay of COVID-19 nucleic acid that was tested using the oropharyngeal swab.
Oseltamivir, arbidol, Lopinavir/ritonavir, and moxifl oxacin were given. One day later, the patient's body temperature decreased to 37.0℃u. One week after being admitted to hospital, the symptoms and CT lesions improved. RT-PCR of COVID-19 nucleic acid using the oropharyngeal swab tests were repeatedly performed for surveillance, resulting negative for two consecutive times within 24 hours interval. According to the discharge criteria of hospitals in China, the patient was discharged from hospital on January 29 th .
After returning home, the patient still complained of low fever with an occasional cough. He was re-admitted to our hospital on January 31th with the highest body temperature reaching 38.0℃. The RT-PCR of COVID-19 nucleic acid obtained using the oropharyngeal swab test was positive for 2 times during 24 hours interval. The patient was admitted to our department for confi rmation of COVID-19.
The patient was re-hospitalized for low fever and occasional cough with two consecutive positive RT-PCR tests within 24 hours interval ( During the second admission to our hospital, follow-up chest CT scans showed multiplepatchy consolidations in both lungs and grid-like thickness of interlobular septa(Rapid progression stage), which indicated that the infl ammation was still progressing. The fi fth scan (2020-02-27) showed consolidations with air-bronchogram inside with absorbing and reducing the area. By this time, the infl ammation was not progressing.

Clinical examination
The infective marker such as PCT, WBC number, NE percent were high during re-hospitalization (Figure 4 Dynamics of an infective marker B the second patient, who was second admitted to hospital), which indicated the progressing infl ammation. The isolation was not recommended, and the patient was not allowed to being discharged from hospital with the falsenegative of RT-PCR of COVID-19 nucleic acid.
After being admitted to hospital, the symptoms and CT lesion improved and the body temperature was back to normal.
RT-PCR of COVID-19 nucleic acid using the oropharyngeal swab test was repeatedly performed for surveillance, resulting in negative for two consecutive times (more than one day apart).
The patient, in general, were not allowed to be discharged from hospital for the fi rst time of admission to hospital with the false-negative of RT-PCR of COVID-19 nucleic acid and the progressing infl ammation of chest CT scans and infective marker.

Patients with complete recovery on serial CT scans
In order to further study the relation between the dynamics of the infective marker and serial CT scans, patients with complete recovery on serial CT scans were selected, who were matched for age, gender and disease severity compared with the above patients.

First patient
The fi rst patient was diagnosed with COVID-19 on January 22, 2020 based on epidemiological history, the clinical character of the respiratory tract, chest CT image, RT-PCR assay of COVID-19 nucleic acid obtained via the oropharyngeal swab.
The patient came to our hospital with fever and discomfort for 5 day with the highest temperature reaching 38.1℃, chills, cough, stuffy nose, slight pharyngitis, and muscle aches. Oseltamivir, arbidol, lopinavir/ritonavir, and moxifl oxacin were given. One day later, the body temperature decreased to 37.5℃. After admission to hospital, the symptoms improved.
Accentuations of CT lesion were observed, which indicated pathological changes in the progression stage. Twelve days after being admitted to hospital, the CT lesions improved.
RT-PCR of COVID-19 nucleic acid using the oropharyngeal swab tests was repeatedly performed for surveillance, resulting negative for two times within 24 hours interval. According to the discharge criteria of hospitals in China, the patient was discharged on February 2 th . were observed, which indicated that the infl ammation was progressing ( (Figure 4 Dynamics of an infective marker C the fi rst patient, who was patient with complete recovery).

Clinical examination
After being discharged from hospital, the levels of infective markers returned to normal and never increased, which indicated that the infl ammation was no longer progressing.

Clinical examination
During being admission to our hospital increase of infective marker(WBC number , NE percent, LYM percent, CRP and PCT) were observed, which indicated that the infl ammation was progressing (Figure 4 Dynamics of an infective marker D the second patient with complete recovery).
After being discharged from hospital, the levels of infective markers returned to normal and never increased, which indicated that the infl ammation was no longer progressing. Readmission and isolation were requested. There was a signifi cant progression observed on CT images as well as an increase in infective markers during their second admission to hospital, which indicated the progressing infl ammation. After being discharged from hospital for second time, the lesions on CT image improved signifi cantly and levels of infective markers (WBC count, NE percent, temperature, PCT and CRP) returned to normal level and never increased, which indicated that the infl ammation was not progressing. LYM percent increase to normal level, which indicated that CD4 and CD8 lymphocyte level increase to normal level. Immune recovery occurred, which was impaired by the infection of COVID-19.

Serial
The period between the fi rst discharge from hospital and the second admission to hospital was shorter than the latent incubation period of COVID-19, which indicated recurrence rather than reinfection of COVID-19. The isolation was not an option, and patients could not be discharged from hospital for the false-negative of RT-PCR of COVID-19 nucleic acid obtained using the oropharyngeal swab.
In order to further study the relation between the dynamics of the infective marker and serial CT scans, patients with complete recovery were selected, who were matched for age, gender and disease severity compared with the above patients. There was a signifi cant progression of symptoms, progression on CT image, increase of infective markers (WBC number, NE percent, LYM percent, temperature, CRP and PCT) during admission, which indicated the progressing infl ammation. After a period of therapy, symptoms were relieved; dissipation almost completely disappeared on CT image; infective markers signifi cantly decreased to normal levels, which indicated that the infl ammation was not progressing. LYM percent increase to normal level, which indicated that CD4 and CD8 lymphocyte level increase to normal level. Immune recovery occurred, which was impaired by the infection of COVID-19. After being tested negative for RT-PCR of COVID-19 nucleic acid for two consecutive times (more than one day apart), the patients were subsequently discharged.
We assume that false-negative results of RT-PCR with the oropharyngeal swab were obtained due to the following reasons: the limitations of sample collection, transportation, and kit performance. The total positive rate of RT-PCR with samples obtained via throat swab was reported to be about 30% to 60% at the initial presentation, while the contagious period of COVID-19 has not yet been specifi ed. COVID-19 nucleic acid from respiratory tract specimens may be persistent or recurrently positive during the course of COVID-19, even during recovery. Patients, asymptomatic carriers, and patients in convalescence positive for COVID-19 nucleic acid are all infectious and should remain in isolation until obtaining two consecutive negative results on COVID-19 nucleic acid test [5,6].
In our study, two patients with COVID-19 were discharged from hospital with false-negative results of RT-PCR of COVID-19 nucleic acid testing performed using the oropharyngeal swab. Their CT images revealed signifi cant progression and pulmonary fi brosis structural changes in the lungs, where there may be a lot of hidden COVID-19.
Angiotensin-Converting Enzyme-2 (ACE-2), identifi ed as the cell entry receptor of COVID-19, is highly expressed in the lungs rather than in the upper respiratory tract, such as oropharyngeal or nasopharyngeal tract. RT-PCR of COVID-19 nucleic acid may give false negatives when testing is performed using the oropharyngeal swab, which can be overcome using the Bronchoalveolar Lavage Fluid (BALF) specimen but with a higher exposure risk [6].

Declaration of interests
Ethics approval and consent to participate: The patient in the case report was available to consent. The case has been discussed with the most senior member of staff in charge of the patient's care, who has provided consent for this, and consent was obtained for use of accompanying radiological images from the consultant radiologist. The study was reviewed and approved by the Hainan Provincial People's Hospital Institutional Review Board.