COVID-19 : doctors should expect a second wave of coronavirus infections from people with false-negative test results

As COVID-19 testing becomes more widely available, it’s vital that health care providers and public health officials understand its limits and the impact false results can have on efforts to curb the pandemic.

A special article published in Mayo Clinic Proceedings calls attention to the risk posed by overreliance on COVID-19 testing to make clinical and public health decisions.

The sensitivity of reverse transcriptase-polymerase chain reaction (RT-PCR) testing and overall test performance characteristics have not been reported clearly or consistently in medical literature, the article says.

As a result, health care officials should expect a “less visible second wave of infection from people with false-negative test results,” says Priya Sampathkumar, M.D., an infectious diseases specialist at Mayo Clinic and a study co-author.

“RT-PCR testing is most useful when it is positive,” says Dr. Sampathkumar.

“It is less useful in ruling out COVID-19. A negative test often does not mean the person does not have the disease, and test results need to be considered in the context of patient characteristics and exposure.”

Even with test sensitivity values as high as 90%, the magnitude of risk from false test results will be substantial as the number of people tested grows.

“In California, estimates say the rate of COVID-19 infection may exceed 50% by mid-May 2020,” she says.

“With a population of 40 million people, 2 million false-negative results would be expected in California with comprehensive testing. Even if only 1% of the population was tested, 20,000 false-negative results would be expected.”

The authors also cite the effects on health care personnel. If the COVID-19 infection rate among the more than 4 million people providing direct patient care in the U.S. were 10% — far below most predictions – more than 40,000 false-negative results would be expected if every provider were tested.

This poses risks for the health care system at a critical time.

“Currently, CDC (Centers for Disease Control and Prevention) guidelines for asymptomatic health care workers with negative testing could lead to their immediate return to work in routine clinical care, which risks spreading disease,” says Colin West, M.D., Ph.D., a Mayo Clinic physician and the study’s first author. Victor Montori, M.D., a Mayo Clinic endocrinologist, also is a co-author.

While dealing with the enormity of the growing COVID-19 pandemic, it’s important for public health officials to stick to principles of evidence-based reasoning regarding diagnostic test results and false-negatives.

Four recommendations are outlined in the Mayo Clinic article:

1. Continued strict adherence to physical distancing, hand-washing, surface disinfection and other preventive measures, regardless of risk level, symptoms or COVID-19 test results. Universal masking of both health care workers and patients may be necessary.
2. Development of highly sensitive tests or combinations of tests is needed urgently to minimize the risk of false-negative results. Improved RT-PCR testing and serological assays — blood tests that identify antibodies or proteins present when the body is responding to infections such as COVID-19 — are needed.
3. Risk levels must be carefully assessed prior to testing, and negative test results should be viewed cautiously, especially for people in higher-risk groups and in areas where widespread COVID-19 infection has been confirmed.
   Risk-stratified protocols to manage negative COVID-19 test results are needed, and they must evolve as more statistics become available.

“For truly low-risk individuals, negative test results may be sufficiently reassuring,” says Dr. West. “For higher-risk individuals, even those without symptoms, the risk of false-negative test results requires additional measures to protect against the spread of disease, such as extended self-isolation.”

At Mayo Clinic, RT-PCR testing is “one of many factors we take into account in deciding whether the patient meets criteria for COVID-19,” Dr. Sampathkumar says.

If the RT-PCR test is negative but chest X-ray or CT scan results are abnormal, or there has been close contact with a person who has confirmed COVID-19, the recommendation is to continue caring for the patient as if he or she has COVID-19.

“We need to continue to refine protocols for asymptomatic patients and exposed health care workers,” says Dr. Sampathkumar.

---

The spread of new pathogen Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) has caused an expanding pandemic of Corona Virus Disease-2019 (COVID-19) (1, 2, 3, 4).

As of 10^th Mar 2020, the global number of confirmed cases of COVID-19 had surpassed 118,000 with more than 4,292 deaths due to acute respiratory failure or other related complications (5).

Most cases (68.42%) occurred in China; outside China, a total of 37,371 cases of COVID-19 has been confirmed in 114 countries/territories/areas with 1130 deaths (5). T

o allow prompt patient identification and clinical treatment, the Chinese government has released seven successive editions of Guidelines for Diagnosis and Treatment of COVID-19 since Jan 15, 2020.

Laboratory viral nucleic acid testing (NAT) has been recommended as a gold standard for COVID-19 diagnosis, together with serological examination (6).

NAT shows better performance than antibody assays as it can identify the viral RNA in the early stage of infection, even during the incubation period.

Since 1988, polymerase chain reaction (PCR) technology has been used extensively in clinical examination with the prominent advantages of high sensitivity, convenience, and economy (7).

Currently, more than 100 enterprises in China have developed real-time PCR (RT-PCR) detection kits for COVID-19, and nine of these kits have been approved by National Medical Products Administration (NMPA) of China (8).

In clinical trials subjected to the NMPA approval procedure, all kits exhibited a high sensitivity of over 90% with samples from confirmed cases.

However, in real clinical settings, the positive diagnosis rates of suspected patients were not as high as previously evaluated(9).

Many suspected patients exhibited typical clinical symptoms or imaging studies consistent with pneumonia, but were not found positive in RT-PCR testing.

This raised the question of what caused these missing positive results in clinical practice?

Given the high quality of NMPA approved kits, false-negative test results were less likely to have arisen from the procedures of nucleic acid extraction and detection.

The types of samples collected clinically appeared to be one of the main causes of the lower performance of RT-PCR tests than expected (10), but, other factors need to be considered, particularly appropriate sample pretreatments before testing.

Considering that SARS-CoV-2 is a single-stranded RNA, isolation of its genome requires cautious handling of samples and good laboratory practices.

Owing to the contagiousness of SARS-CoV-2, many Chinese laboratories have inactivated the virus at 56-60 °C for 30-60 minutes before RNA extraction.

Thermal inactivation at 56 °C has also been recommended to ensure the security of medical inspectors in Chinese expert’s consensus (11).

The consensus also raises the uncertainty of whether heat inactivation may decrease the sensitivity of NAT (11). Although it is suspected that such pretreatment may affect the detection of SARS-CoV-2 in samples with low viral loads, quantitative comparisons on its impact are still unclear.

Herein, we investigated the effects of thermal inactivation on the quantitative RT-PCR results of SARS-CoV-2 and evaluated the false-negative rates due to thermal inactivation.

We further investigated the effects of different specimen types, sample preservation times and chemical inactivation approach on NAT.

Thermal inactivation reduced the detectable amount of SARS-CoV-2 in RT-PCR detection.

To determine the impact of thermal inactivation, parallel NAT was performed using clinical specimens from 4 confirmed COVID-19 patients. Each specimen was serially diluted by a factor of 10^-5 and then determined by RT-PCR.

The line chart presented contrasts the Ct values of each specimen with or without incubation at 56°C for 30 min (Fig. 1).

We classified the original and diluted concentration of 10^-1 and 10^-2 as the high viral load (HVL) group and all remaining dilutions (10^-3 to 10^-5) as the low viral load (LVL) group.

In general, most of the inactivated samples exhibited higher Ct values (mean 33.07 ± SD 5.00) than those with non-inactivated treatment (mean 32.69 ± SD 4.92) with a mean increase of 0.38 (P=0.017, online Supplemental Table 1).

We further calculated the ΔCt of each sample with or without inactivation. The ΔCts of detectable samples in the LVL group (median 1.37, range 0.81 to 2.17) were much higher than those (median 0.14, range -0.38 to 1.57) in the HVL group (P=0.02, online Supplemental Table 1).

These data suggested that in detectable samples of the LVL group, the tendency for increased Ct by thermal inactivation was more substantial than that in the HVL group. Notably, a few diluted samples showed negative results after inactivation.

These results demonstrated thermal inactivation could cause an increased Ct value in the RT-PCR tests of SARS-CoV-2 and might affect the qualitative results of samples carrying low viral loads.

Table 1 Ct value of RT-PCR for detecting SARS-CoV-2 in throat swab specimens

Inactivation                                              Non-inactivation

	Ct 1	Ct 2	Ct 1	Ct 2
Sample A	35.64	35.84	34.65	34.71
Sample B	33.12	33.49	33.34	33.39
Sample C	34.47	34.50	34.55	34.58
Sample D	36.72	NA	36.55	36.62
Sample E	NA	NA	36.75	36.94
Sample F	36.70	36.39	35.51	35.42
Sample G	36.61	36.06	35.41	35.00
Sample H	NA	36.74	36.02	35.92
Sample I	NA	36.61	34.69	34.67
Sample J	34.89	34.98	33.33	33.50
Sample K	35.14	34.97	33.64	33.41
Sample L	36.50	36.93	34.95	35.09
Sample M	NA	NA	36.88	36.90
Sample N	NA	NA	36.80	36.84
Sample O	NA	NA	35.84	36.04

NA, not available, represents undetectable value of Ct > 37.

Clinical weak positive specimens were more susceptible to thermal inactivation.

To further verify the effects of thermal inactivation on specimens with low viral loads, 15 specimens with viral loads near the limit of detection were collected from confirmed cases and tested in parallel.

The Ct values of non-inactivated specimens ranged from 33.37 to 36.89. Again, the inactivated group (mean 36.48 ± SD 1.48) showed higher mean Ct values than the non-inactivated group (mean 35.26 ± SD 1.24) with a mean increase of 1.22 (P<0.001, Fig. 2). Moreover, in duplicate RT-PCR tests, positive results in 7 of 15 specimens (46.7%) were converted into undetectable values (false negatives) in at least one parallel testing after thermal inactivation (Table 1). The mean Ct value of the 7 specimens with non-inactivation was 36.25, providing a possible threshold for thermal susceptibility in viral NAT.

Effects of thermal inactivation on different types of samples

Clinical samples for COVID-19 tests included throat swab, sputum, bronchoalveolar lavage fluid, stool, blood, etc. To explore the effect of thermal inactivation in different specimen types, we deployed 4 throat swabs, 2 sputum samples, and 2 stool samples and serially diluted each specimen by a factor of 10^-5. Compared to the non-inactivated group (mean 33.64 ± SD 4.61), thermal inactivation raised the Ct values of most samples (mean 34.08 ± SD 4.68) by a mean increase of

0.44 (P<0.001, Fig. 3, online Supplemental Table 2). We then performed a further analysis within different sample types. While the susceptibility to thermal inactivation in throat swabs has been described above (Fig. 1, online Supplemental Table 3), the

mean Ct values of inactivated stool samples (mean 36.29 ± SD 2.87) were also higher than those of the non-inactivated group (mean 35.55 ± SD 3.24) by a mean increase of 0.74 (P=0.014, online Supplemental Table 3).

However, the difference in mean Ct values in sputum between inactivated (mean 33.90 ± SD 5.06) and non-inactivated groups (mean 33.62 ± SD 4.85) was not found to be significant (P=0.308, online Supplemental Table 3).

Higher temperature and longer time of specimen preservation partially contributed to false-negative results in specimens carrying low viral loads.

To explore the effects of time and temperature on laboratory detection, we diluted four throat swab specimens by factors of 10^-2 and 10^-4. All original and diluted samples were kept at 4°C or room temperature (RT) and harvested at different time points of 12, 24, 36 and 48 hours.

Time and temperature of preservation only exhibited sight impact on the increase of detected Ct value. However, higher Ct values were more prominent with longer time of preservation, especially in diluted samples with lower viral loads. Notably, two positive samples (sample 2 and 4) diluted by a factor of 10^-4 were converted into negative ones with longer storage or preservation at RT (Fig. 4).

Altogether, despite the fact that the influence of storage time and temperature on NAT was slight, an extended time and high temperature of preservation were able to cause false-negative results in low viral load specimens.

Viral inactivation by guanidinium-based lysis exhibited less effects on the detectable amount of SARS-CoV-2 than thermal inactivation.

As genomic RNA of SARS-CoV-2 is unstable and easily degraded by environmental nuclease, the solutions used to preserve specimens are of great importance to protect the viral genomic integrity.

Guanidinium-based buffer, a common solution for specimen preservation has been shown to have the dual function of viral inactivation via chemical destruction of the viral protein (13).

Therefore, we collected 15 throat swab specimens to investigate the effect of chemical inactivation by guanidinium-based solutions on the RT-PCR tests for SARS-CoV-2. Each sample was divided into subgroups based on chemical inactivation by guanidinium-based lysis (GL group), and by thermal inactivation at 56°C for 30 min (VTM group).

VTM was used for specimen preservation before thermal inactivation. The Ct values of most samples in the VTM group (mean 36.48 ± SD 1.48) were higher than those in GL group (mean 35.40 ± SD 1.33) by a mean increase of 1.08 (P<0.001, Fig. 5A, online Supplemental Table 4), demonstrating that GL provided better protection of viral nucleic acid than thermal inactivation.

Moreover, the number of specimens with undetectable results was 2 of 15 in the GL inactivated group but 7 of 15 in the thermal inactivated group. We calculated the ΔCt values of the thermally and chemically inactivated groups versus the non-inactivated group.

The ΔCt values of thermally inactivated specimens were much higher than those in GL inactivated group (P<0.001, Fig. 5B). These results suggested that GL preservation could attenuate the increased Ct value compared to thermal inactivation.

*-*-*-*-*-*-*

Figure legends

Figure 1. Effects of thermal inactivation on the RT-PCR tests of SARS-CoV-2 in throat swabs samples. Each sample (n= 4) was detected in raw solution or after a series of dilutions by virus transport media from 10-1 to 10-5, followed by treatment with or without incubation at 56°C for 30 minutes. The black dot represents the mean Ct value of the duplicate quantitative RT-PCR experiments. The dotted line represents Ct value 37 and Ct values of undetectable results were determined as 38.

Figure 2. Effects of thermal inactivation on RT-PCR tests of SARS-CoV-2 in clinical throat swabs carrying low viral loads. Each sample (n= 15) was detected with or without incubation at 56°C for 30 minutes. Each dot represents the mean Ct value of duplicate quantitative RT-PCR experiments. The dotted line represents Ct value 37 and Ct values of undetectable results were determined as 38. The comparison between the inactivation and non-inactivation groups was tested by paired two-tailed Student’s t-test. *** represents P<0.001.

Figure 3. Effects of thermal inactivation on the quantitative RT-PCR tests of SARS-CoV-2 in different sample types. Specimens of sputum (n= 2), stool (n= 2) and throat swabs (n= 4) were incorporated. Each sample was detected twice in raw solution or dilution by virus transport media from 10-1 to 10-5. Each symbol represents Ct value from one test. The dotted line represents equal Ct value between the inactivated and non-inactivated group. Shaded areas represent detectable values in both two groups. Ct values of undetectable results were determined as 38.

Figure 4. Effects of temperature and time of specimen preservation on the RT-PCR tests of SARS-CoV-2 in throat swabs samples. Each sample (n= 4) was detected in raw solution or diluted by virus transport media by factors of 10-2 and 10-4 and then preserved at 4°C or room temperature (RT). The samples were detected every 12 hours in the next two days. Each symbol represents mean Ct of duplicate quantitative RT-PCR experiments. The dotted line represents Ct value 37 and Ct values of undetectable results were determined as 38.

Figure 5. Effects of guanidinium-based lysis on the quantitative RT-PCR tests of SARS-CoV-2. All throat swabs samples (n= 15) were placed into guanidinium-based lysis (GL) or virus transport media (VTM). Each sample in the VTM group was divided into subgroups with or without incubation at 56°C for 30 minutes. (A) Comparative analysis of Ct values in specimens between chemical inactivation (preserved in GL) or thermal inactivation (preserved in VTM). Shaded areas represent detectable values in both inactivation methods. (B) Comparative analysis of ΔCt values in specimens preserved in GL or VTM. ΔCt value was calculated by the difference of Ct values between inactivation and non-inactivation specimens. Each symbol or dot represents the mean Ct of duplicate quantitative RT-PCR experiments. Ct values of undetectable results were determined as 38. The comparison of ΔCt values between GL and VTM group was tested by paired two-tailed Student’s t-test. *** represents P<0.001.

---

Source:
Mayo Clinic

References:

1. Sun K, Chen J, Viboud C. Early epidemiological analysis of the coronavirus disease 2019 outbreak based on crowdsourced data: a population-level observational study. The Lancet Digit Health. 2020 Feb 20 as DOI: 10.1016/S2589-7500(20)30026-1
2. Leung GM, Leung K. Crowdsourcing data to mitigate epidemics. The Lancet Digit Health. 2020 Feb 20 as DOI: 10.1016/S2589-7500(20)30055-8.
3. Yang XB, Yu Y, Xu J, Shu H, Xia J, Liu H, et al. Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med. 2020 Feb 24 as DOI: 10.1016/S2213-2600(20)30079-5.
4. Zhou P, Yang XL, Wang XG, Hu B, Zhang L, Zhang W, et al. A pneumonia outbreak associated with a new coronavirus of probable bat origin. Nature 2020;579: 270–3.
5. World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report –46. World Health Organization. 2020.
   https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200306-sitrep-46-c ovid-19.pdf?sfvrsn=96b04adf_4. (accessed Mar 12, 2020).
6. National Health Commission (NHC) of the PRC and National Administration of Traditional Chinese Medicine of the PRC. Guidance for Corona Virus Disease 2019: Prevention, Control, Diagnosis and Management (Seventh Edition). 2020.
   http://www.nhc.gov.cn/yzygj/s7653p/202003/46c9294a7dfe4cef80dc7f5912eb1989.shtml (Accessed Mar 13, 2020). [Chinese]
7. Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, et al. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science. 1988;239:487-91
8. National Medical Products Administration. Emergent approval of novel coronavirus testing products by National Medical Products Administration. National Medical Products Administration, 2020. http://www.nmpa.gov.cn/WS04/CL2056/375802.html (accessed Mar 16, 2020) [Chinese]
9. Zhang Q, Zhao Q. Inactivating porcine coronavirus before nuclei acid isolation with the temperature higher than 56 °C damages its genome integrity seriously. Preprint at https://www.biorxiv.org/content/10.1101/2020.02.20.958785v1 (2020).
10. Zou LR, Ruan F, Huang MX, Liang LJ, Huang HT, Hong ZS, et al. SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients. N Engl J Med. 2020 Feb 19 as DOI: 10.1056/NEJMc2001737.
11. Chinese Society of Laboratory Medicine. The Experts’ Consensus on Nucleic Acid Detection of 2019-nCoV. Zhonghua Yi Xue Za Zhi. 2020;100(00):E003-E003. DOI: 10.3760/cma.j.issn.0376-2491.2020.0003 [Chinese]
12. National Health Commission (NHC) of the PRC and National Administration of Traditional Chinese Medicine of the PRC. Guidance for Corona Virus Disease 2019: Prevention, Control, Diagnosis and Management (Sixth Edition). 2020.
    http://www.nhc.gov.cn/yzygj/s7653p/202002/8334a8326dd94d329df351d7da8aefc2.shtml [Chinese]
13. Burton JE, Easterbrook L, Pitman J, Anderson D, Roddy S, Bailey D, et al. The effect of a non-denaturing detergent and a guanidinium-based inactivation agent on the viability of Ebola virus in mock clinical serum samples. J Virol Methods. 2017;250:34-40.

14. Wu F, Zhao S, Yu B, Chen YM, Wang W, Song ZG, et al. A new coronavirus associated with human respiratory disease in China. Nature. 2020;579:265–9.
15. Hoffmann M, Kleine-Weber H, Krüger N, Müller M, Drosten C, Pöhlmann S. The novel coronavirus 2019 (2019-nCoV) uses the SARS-coronavirus receptor ACE2 and the cellular protease TMPRSS2 for entry into target cells. Preprint at https://www.biorxiv.org/content/10.1101/2020.01.31.929042v1(2020)
16. Bermingham A, Heinen P, Iturriza-Gomara M, Gray J, Appleton H, Zambon MC. Laboratory diagnosis of SARS. Philos Trans R Soc Lond B Biol Sci 2004;359:1083-9.
17. Mackay IM, Arden KE. MERS coronavirus: diagnostics, epidemiology, and transmission. Virol J. 2015;12:222.
18. Darnell ME, Subbarao K, Feinstone SM, Taylor DR. Inactivation of the coronavirus that induces severe acute respiratory syndrome, SARS-CoV. J Virol Methods 2004;121:85-91.
19. Leclercq I, Batéjat C, Burguière AM, Manuguerra J-C. Heat inactivation of the Middle East respiratory syndrome coronavirus. Influenza Other Resp 2014;8:585-6.
20. Chen PS, He YT, Huang YL, Chen YL, Huang H, Yu XG, et al. Inactivation of 2019 new coronary virus before quantitative real-time PCR testing. Chin J Lab Med. 2020; 43(00): E004-E004. DOI: 10.3760/cma.j.issn.1009-9158.2020.0004 [Chinese]
21. Pan Y, Zhang D, Yang P, Poon LLM, Wang Q. Viral load of SARS-CoV-2 in clinical samples. Lancet Infect Dis 2020 Feb 24. pii: S1473-3099(20)30113-4.
22. Rabenau HF, Cinatl J, Morgenstern B, Bauer G, Preiser W, Doerr HW. Stability and inactivation of SARS coronavirus. Med Microbiol Immun. 2005;194:1-6.
23. Yunoki M, Urayama T, Yamamoto I, Abe S, Ikuta K. Heat sensitivity of a SARS-associated coronavirus introduced into plasma products. Vox Sang 2004;87:302-3
24. World Health Organization. Laboratory biosafety guidance related to coronavirus disease 2019 (COVID-19). Geneva. https://apps.who.int/iris/bitstream/handle/10665/331138/WHO- WPE-GIH-2020.1-eng.pdf. Published 12 Feb 2020. (Accessed 7 Mar 2020).