Older age is associated with sustained detection of SARS-CoV-2 in nasopharyngeal swab samples.

. Currently, the standard for diagnosis of Severe Acute Respiratory Syndrome-Coronavirus-2 (SARS-CoV-2) infection is a positive result based on a polymerase chain reaction (PCR) test from nasopharyngeal swab samples. PCR test is also used as a guide for patient discharge from designated hospitals and medical institutions. We recently experienced a case of a 97-year-old female who was diagnosed with coronavirus disease 2019 (COVID-19). Although her clinical symptoms and radiological findings resolved within a few days, PCR results from nasopharyngeal swab samples remained positive for 50 days after the onset. This case prompted us to conduct a retrospective study of the association of age the duration of positive PCR testing. We specifically hypothesized that old age could be a risk for prolonged duration of positive PCR results from nasopharyngeal swab samples.

This study was approved by the ethics committees of National Hospital Organization Hokkaido Medical Center and Hokkaido University Hospital. Nasopharyngeal swab sample was collected and quantitative real-time reverse transcription–PCR (RT-PCR) was conducted as described before. In a RT-PCR assay, cycle threshold (Ct) value is defined as the number of cycles required for the fluorescent signal to cross a baseline threshold. The test results of SARS-CoV-2 were reported as negative in tests in which Ct>45).

We analyzed the records of 66 patients who were diagnosed with COVID-19 between March 1, 2020 and April 30, 2020 at National Hospital Organization, Hokkaido Medical Center. In this hospital, when respiratory physicians (T.H. and M.A.) recognized the peak-out of the COVID-19 infection, PCR tests were performed every day on nasopharyngeal swab samples from each patient. This enabled us to determine the precise day that each COVID-19 patient tested “negative” by PCR. We defined “negative PCR” as confirmed negative results over two sequential days. Characteristics of the 66 patients diagnosed with mild COVID-19 are presented in Table 1 . Forty-two subjects were mild cases, who did not require supplemental oxygen treatment. Eighteen subjects were moderate cases who needed oxygen treatment. Six subjects were severe cases who needed ventilator or/and ECMO.

Table 1

Clinical characteristic of subjects diagnosed as COVID-19 in National Hospital Organization Hokkaido Medical Center.

	Total (n = 66)	Mild (n = 42)	Moderate (n = 18)	Severe (n = 6)	p value*
Male, N (%)	38 (57.6)	23 (54.8)	12 (66.7)	3 (7.9)	0.6421
Age (yr)	61 (22–99)	56 (28–99)	71 (30–96)	59 (22–61)	0.0553
LDH (U/L)	260.5 (127–555)	249.5 (127–452)	330.5 (165–555)	314 (192–452)	0.2927
D-dimer (μg/mL)	0.75 (0.05–20.3)	0.6 (0.05–5)	1.8 (0.4–11.3)	3.25 (0.2–20.3)	0.0076
CRP (mg/dL)	3.495 (0.02–25.82)	2.1 (0.02–17.68)	5.82 (0.28–24.49)	5.54 (1.02–25.82)	0.0122
Ferritin (ng/mL)	293.3 (12.5–1760.3)	268.9 (12.5–1760.3)	402.9 (75.1–156.9)	606.3 (225.5–837.9)	0.1452
HbA1c (%)	5.9 (5.1–11.1)	5.9 (5.1–11.1)	6.15 (5.4–8.0)	6.85 (5.7–9.9)	0.173
WBC (/μL)	4.9 (1.9–14.8)	4.8 (1.9–9.1)	4.8 (2.6–11.6)	6.95 (3.5–14.8)	0.2178
Bas (%)	0.2 (0–3)	0.2 (0–3)	0.15 (0–0.5)	0.2 (0–0.4)	0.7529
Eos (%)	0.05 (0–4.5)	0.05 (0–3.9)	0.1 (0–2.1)	0.1 (0–4.5)	0.9102
Lym (%)	20.85 (5.7–58.0)	24.7 (8.8–58)	15.4 (6–27)	11.2 (5.7–20.7)	0.0009
Mon (%)	6.1 (1.5–15.0)	6.4 (2–11.9)	5.4 (1.5–15)	6.05 (1.9–8)	0.2372
Neu (%)	71.8 (34.0–92.5)	68.65 (34–88)	76.85 (60–92.5)	80.7 (68.8–92)	0.0028
Medication (for COVID-19)	36 (54.5)	17 (40.5)	14 (77.8)	5 (54.6)	0.0097
Cyclesonide, N (%)	32 (48.5)	16 (38.1)	12 (66.7)	4 (66.7)	0.0824
Favipiravir, N (%)	21 (31.8)	5 (11.9)	12 (66.7)	4 (66.7)	<0.0001
Ritonavir/Lopiravir, N (%)	8 (12.1)	2 (4.8)	2 (11.1)	4 (66.7)	<0.0001
Camostat, N (%)	3 (4.6)	1 (2.4)	2 (11.1)	0 (0)	0.2826
Predonisolone, N (%)	2 (3.0)	0 (0)	2 (11.1)	0 (0)	0.0639
Duration of positive PCR (day)	19 (9–45)	17.5 (9–29)	21 (14–45)	24.5 (15–43)	0.0043

Continuous data were presented as medians and interquartile ranges (IQR). * Comparisons among three groups (mild, moderate, and severe).

We found that older age was significantly associated with prolonged positive PCR tests (P = 0.0053; Fig. 1 A). This relationship remained unchanged when the findings were adjusted for the potential impact of severity of the disease (mild, moderate, severe) and the used of medication (P = 0.026). When we analyzed only mild cases of COVID-19 in order to exclude the influence of disease severity, the result remained significant (P = 0.036; Fig. 1B).

Fig. 1

The association of patient age with prolonged duration of positive PCR results among patients diagnosed with COVID-19 at National Hospital Organization, Hokkaido Medical Center.

At the time of this writing, Hokkaido, most northern island in Japan, has controlled the second wave of COVID-19 and the rate of new cases continues to decrease. However, as per the statements of and guidance from the World Health Organization (WHO), patients are discharged only upon confirmed negative PCR tests of nasopharyngeal swabs taken on two sequential days. This results in prolonged hospital stays as patients remain in beds in the designated infectious disease units designed for acute and ongoing care.

In our analysis, older age is significantly associated with prolonged duration of positive PCR tests from nasopharyngeal swab samples, irrespective of the disease severity and the used of medication (Fig. 1). The reasons underlying these observations remain unclear. Of note, we recently reported that quality of PCR from saliva as a diagnostic measure of SARS-CoV-2 infection was equivalent to that of the samples from nasopharyngeal swabs. We also found that findings from PCR tests reverted from positive to negative much more quickly when using saliva than nasopharyngeal samples. We speculate that in older individuals, cell turnover is less robust and as such, clearance of virus from the nasopharynx is prolonged; these factors may lead to positive PCR tests that persist after acute disease has resolved.

One group from Taiwan has already discussed the possibility that COVID-19 may no longer be contagious at two weeks after the onset of symptoms. It is possible the positive PCR tests reflect the presence of that inactive virions remaining within nose. In a linked study, Zheng and colleagues evaluated viral loads in respiratory samples, stool, serum, and urine using PCR: they found that more than half the respiratory samples remained positive for SARS-CoV-2 as did a full one third of the stool samples at the end of the four week trial period. As such, we propose that there should be a change in the strategy currently in use for determining time of discharge to one that relies on other clinical tests or/and patients’ condition, which could be helpful to inhibit the development of patients’ flail and dementia and to reduce the burden of ongoing and potentially unnecessary prolonged hospitalization.

In summary, we demonstrated that old age is significantly associated with prolonged duration of positive PCR results from nasopharyngeal swab samples; this is the case regardless of disease severity. Further studies will be needed in order to clarify how long these patients are actually contagious.

Declaration of Competing Interest

The authors declare that they have no competing interests.