Learn from the SARS-CoV-2 nucleic acid test to increase the experience of dealing with the “disease X “

Abstract

Background

RT‒PCR is crucial for screening for epidemic diseases such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). However, false positives further plague epidemic prevention and control. This study conducted a stratified study on the initial screening Ct values and false positive ratios, providing experience and reference for addressing future “Disease X”.

Methods

Data from 1,255 positive or suspected positive results were obtained from eleven laboratories with seven different reagents. The proportion of false positives was analyzed on the basis of different Ct values among different reagents and various testing institutions.

Results

When the Ct values of both target genes in the initial detection were < 30, a false positive was considered a small probability event (≤ 1.72%). However, when the 30 ≤ Ct value was < 35, significant differences were noted (0%, 1.41%, 7.69%, and 9.14%, P < 0.001). When the Ct value of any target gene is > 35, 15.58 − 24.22% of positive results may be false positive. Among the suspected positive samples, 53.23% were false positive according to retesting. After separate sampling, 4 tubes (30 people involved) from 19 tubes (133 people involved) were negative.

Conclusions

In summary, different strategies should be adopted according to the different Ct values of primary screening results under pandemic prevention and control conditions, which may provide a better reference for the rapid diagnosis of the next “Disease X”.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-025-10991-7.

Introduction

Real-time reverse transcription PCR (RT‒PCR) analysis of respiratory tract samples in the laboratory, considered the “gold standard” for the detection of some viruses, is characterized by rapid detection, high sensitivity and specificity and has great advantages in detecting SARS-CoV-2 [1–4]. It has been widely adopted by public health organizations in most countries and communities worldwide [5]. During the COVID-19 pandemic, for epidemic prevention and control, many nasopharyngeal swabs from patients or suspected patients need to be sent to laboratories for testing. During this period, RT‒PCR plays a crucial role.

However, during the laboratory testing process, because laboratories operate at full capacity, avoiding nucleic acid contamination to varying degrees is extremely difficult. To rule out false positives caused by nucleic acid contamination, retesting all positive samples places a heavy burden on laboratory work. However, if not retested, once a reported result is a false positive, it might cause trouble for the tested individuals.

This study analyzed large amounts of data and intended to propose reasonable suggestions: (1) Under what circumstances can be directly reported to facilitate the control of infectious sources as soon as possible, and under what situations retesting is required? (2) For the pooled sampling test used in large-scale nucleic acid testing, when should retest be conducted, as well as the separate sampling test to avoid the omission of positive infected samples? When facing a new “Disease X” in the future, this approach can be used. Based on the situation at that time, relevant data should be collected promptly, and corresponding retesting cutoff values should be established. This will better support laboratory testing in playing a critical role in the prevention and control of “Disease X”.

Materials and methods

Laboratory test data collection

Data were collected from laboratories undertook regional new coronavirus nucleic acid testing in China’s western area between August 16 and October 16, 2022. This study collected basic information, including primary screening positive results, retesting results, the results of individual sampling and testing, the results of single-target gene positive or double-target gene amplification, but the cycle threshold (Ct) value exceeded the cutoff to consider the test positive, the names of the corresponding testing reagents and institutions.

Inclusion criteria

(1) Detection institution: The detection institution of COVID-19 nucleic acid should be a laboratory undertaking large-scale nucleic acid testing and engaged in testing for a long time, which is operated and detected by relatively fixed and strictly trained personnel. (2) Analysis reagents and result interpretation: All test reagents should be approved by the National Medical Products Administration. All the data should strictly follow the kit instructions, and the testing quality control also met the relevant standards, including positive quality control, negative quality control, third-party weak positive quality control, and internal control. (3) Reexamination data: Reexamination of the positive (matching the interpretation standard of positive results in the kit manual) and suspected positive samples (those situations that cannot be determined as positive or negative according to the instructions of the test kit, including single-gene amplification and double-gene amplification but with Ct values exceeding the positive determination values specified by the test kit.) with different reagents. (4) Data of positive or suspected positive pooled samples: To improve detection efficiency and reduce laboratory workload, the pooled sampling method has been widely adopted in large-scale nucleic acid testing. For example, in this study, “4 tubes (30 individuals) " indicates that these 4 tubes collectively contain swabs from 30 people. All personnel involved in the initial screening of positive and suspected positive pooled samples were sampled and analyzed separately within 12 h. For example, if the pooled sample (involving 10 people) was positive, the throat swabs of 10 people were collected. Among the samples, at least one is positive, and the combined samples are confirmed as true positive. If all 10 people are negative, then the pooled sample is also negative.

Data processing

The positive results were divided into four grades according to the Ct value: (1) Ct value of the double target gene Ct < 30; (2) double target gene Ct value < 35; (3) any target gene with a Ct value > 35; and (4) suspected positive. The sample data that had neither repeated tests nor individual sampling tests were removed. Since the fluorescence collection of Daan reagent starts after ten cycles, to facilitate statistical analysis with other reagents, the Ct value of the Daan reagent was increased by ten.

Data analysis

GraphPad Prism and SPSS 22.0 Windows were used to analyze the classified data via the chi-square test. When the sample size was not large enough to make the chi-square test valid, Fisher’s test was used for further accurate tests.

Results

Data collection

In this study, the data of 1,255 positive or suspected positive test results from 11 laboratories (GAOXIN, HZHX (ZN), HZHX (YC), HELAN, HYK, JY, JINGKAI, MIDONG, SHAQU, TIANSHAN and DIAN) undertaking large-scale detection of COVID-19 nucleic acid were obtained. A total of seven detection reagents were used (Daan, Daan Gene Co., Ltd., Guangzhou, China, Limitation of detection, LOD: 200 copies/mL; BioGerm, BioGerm Medical Co., Ltd., Shanghai, China, LOD: 150 copies/mL; EasyDiagnosis, Wuhan EasyDiagnosis Biomedicine Co., Ltd., Wuhan, China, LOD: 200 copies/mL; Sansure, Sansure Biotech Inc., Changsha, China, LOD: 200 copies/mL; Bioperfectus, Jiangsu Bioperfectus Technologies Co., Ltd., Jiangsu, China, LOD: 350 copies/mL; Zybio, Zybio Co., Ltd., Chongqing, China, LOD: 200 copies/mL; ZJ, ZJ Biotech Co., Ltd., Shanghai, China, LOD: 200 copies/mL). Among the 1,255 positive or suspected positive test results, 872 were primary screening positive and retesting, 300 were primary screening positive for pooled samples and then individually sampled and tested, 62 were suspected positive and retested, and 21 were primary screening suspected positive for pooled samples and then individually sampled and tested.

The proportion of false positives in the initial positive samples after retesting among the different primary reagents

When the Ct values of the two target genes in the initial detection were all < 30, there was no significant difference in the proportion of false positives among the different reagents (P = 0.947). Although the positive rate of 1.72% (2/116) was still false positive, as a lower probability, the results could be reported immediately to take timely control measures to prevent the spread of the virus (Supplementary Tables 2 and Fig. 1A). When double target genes of the initial detection was 30 ≤ Ct value < 35, with the exception that the false positives detected by Sansu (0/85, 0%) and Daan (4/284, 1.41%) were still small probability events, the other two reagents were prone to a certain proportion of false positives. If reported immediately, it was necessary to consider the potential for false positives for epidemic prevention and control (Supplementary Tables 2 and Fig. 1B). When the Ct value of a target gene was > 35 during the initial test, the proportion of false positives detected by the four reagents increased significantly, but there was no significant difference among the reagents. If the results were reported immediately, at least 15 out of 100 positive cases were false positive. It was recommended to retest the original sample before reporting the results (Supplementary Tables 2 and Fig. 1C).

Fig. 1.

Proportion of false positives among the initial positive samples screened after retesting with different primary reagents

The difference in the proportion of false positives in primary screening positive samples among different testing institutions

When the Ct value of both target genes in the initial screening was < 30, there was no significant difference in the proportion of false positives among different testing institutions (P = 0.222) (Supplemental Tables 2 and Fig. 2A). When both target genes had a 30 ≤ Ct value < 35, there was a large difference in the proportion of false positives between different testing institutions (P < 0.001) (Supplemental Tables 2 and Fig. 2B).

Fig. 2.

Differences in the proportions of false positives among primary positive samples from different testing institutions

When the Ct value of any target gene was > 35, an obvious difference was noted in the proportion of false positives among different testing institutions (P < 0.001) (Supplemental Tables 2 and Fig. 2C), which stemmed from the influence of personnel operation, equipment and different initial testing reagents.

In general, among the positive results of the preliminary screening, the highest proportion of false positives in the SHAQU could be 30% (36/120), especially when any target gene Ct > 35, which could reach 44.64% (25/56). HZHX (YC) and HYK had the lowest proportions of false positives. (Supplemental Tables 2 and Fig. 2D).

Results of individual sampling of pooled samples with positive initial screening results

When stratifying according to the Ct value, there was an obvious difference between different stratifications in general. The results of the comparison between the two groups were mainly due to the marked difference between A (28/280, 11.11%) and C (1/76, 1.33%) (A: The primary screening and retesting results were positive; C: double target genes Ct < 30 according to the primary screening results).

When the Ct values of the two target genes in the pooled sample initial screening were all < 30, there was no need to retest the original tube, reporting it directly and testing it separately to find and control the source of infection as soon as possible. (Table 1, Supplemental Table 3, Fig. 3 left).

Table 1.

Individual sampling and analysis results of pooled samples that were positive in the primary screening

		Pooled Sample Positive	Resample Individually		Negative Rate	P
			Positive	Negative
Overall	A	280	252	28	11.11%	0.025
	B	186	172	14	8.14%
	C	76	75	1	1.33%
	D	48	47	1	2.13%
Pairwise comparison	A	280	252	28	11.11%	0.229
	B	186	172	14	8.14%
	B	186	172	14	8.14%	0.038
	C	76	75	1	1.33%
	A	280	252	28	11.11%	0.229
	C	76	75	1	1.33%
	C	76	75	1	1.33%	0.626
	D	48	47	1	2.13%
	A	280	252	28	11.11%	0.053
	D	48	47	1	2.13%
	B	186	172	14	8.14%	0.147
	D	48	47	1	2.13%

A: Positive in primary screening and retesting; B: Both target genes were 30 ≤ Ct value < 35 after primary screening; C: Double target genes Ct value < 30 through primary screening; D: Double target gene Ct value < 30 in primary screening and retesting

Fig. 3.

Results of individual sampling of pooled samples with positive initial screening results and the necessity of the original sample retesting when the pooled sample is initially screened as suspected positive

The necessity of an original sample retesting when the pooled sample is initially screened as suspected positive

After 62 samples with suspected positive results were retested, 33 samples (53.23%) were negative, which means that 53.23% of the suspected positive samples would have false positives if they were sent directly as positive reports. Therefore, retesting or resampling is necessary for testing.

No marked difference was found in the proportion of false positives among the positive samples in the initial screening among the different test reagents (P = 0.781) (Supplemental Tables 4 and Fig. 3 right).

The necessity of individual sampling and detection for suspected positive samples

Nineteen tube samples (involving 133 people) were suspected to be positive. After separate collection, 4 test tubes (involving 30 people) were negative, which means that the lack of retesting may lead to unnecessary isolation of 30 false positive “infected people” (Table 2). In addition, when the pooled samples were initially suspected to be positive, the viral load may not have been high enough, but it should still be tested separately to avoid unnecessary isolation and treatment.

Table 2.

The necessity of individual sampling and testing for suspected positive pooled samples

Testing Institutions	Sample No	Primary Screen			Resample Individually
		ORF1ab	N	Reagent	P/N#
HZHX (YC)	1,023,186,938	41.98	38.21	Daan	P (1/3) *
HZHX (ZN)	220,324,691,721	37.47	44.11	Daan	P (5/10)
JY	0991952673	39.04	40.82	Daan	P (1/2)
JY	1,015,039,339	39.78	41.55	Daan	P (1/2)
JY	6122000612826SGM	40.02	39.59	Daan	N (0/5)
JY	1,019,899,309	40.19	41.00	Daan	P (1/5)
HYK	1,028,530,576	38.40	37.80	EasyDiagnosis	P (1/3)
HYK	1,034,973,281	38.40	37.90	EasyDiagnosis	P (1/3)
HYK	1,026,801,385	NoCt	39.20	EasyDiagnosis	P (1/10)
HYK	1,039,789,625	38.68	NoCt	EasyDiagnosis	P (1/5)
HYK	1,026,962,157	NoCt	38.90	EasyDiagnosis	N (0/10)
HELAN	6401608071-01	39.14	35.79	Bioperfectus	N (0/5)
DIAN	9,100,000,037,738	NoCt	35.57	ZJ	P (1/10)
DIAN	9,100,000,045,771	NoCt	36.66	ZJ	P (3/10)
DIAN	02211620366471	NoCt	36.86	ZJ	P (1/10)
DIAN	02211620366471	NoCt	36.86	ZJ	P (1/10)
DIAN	9,100,000,041,238	NoCt	38.26	ZJ	P (1/10)
DIAN	1,035,711,068	NoCt	36.71	ZJ	P (1/10)
DIAN	1,026,962,415	NoCt	36.24	Daan	N (0/10)

Note: # P: Positive, N, Negative; NoCt: no Ctvalue; * P (1/3) showed that the sample was a pooled sample involved three people, and one of them was positive after individually sampling and testing, and the following are similar

Discussion

High-precision nucleic acid testing of clinical samples of pathogens plays a crucial role in the diagnosis of infectious diseases and global pandemic control [6]. During the COVID-19 epidemic, RT‒PCR detection plays a vital role in screening high-risk groups and tracking close contacts for better prevention and early control of the epidemic [7]. In the RT‒PCR results, the Ct value is related to the original nucleic acid amount of the virus in the sample [8], and the Ct value represents the number of cycles in which the signal exceeds the positive threshold; thus, generally, a lower Ct value should indicate a greater viral load [9]. However, the Ct value does not mean that there must be viral nucleic acid in the original sample because the diagnostic test may sometimes be inaccurate in false positives.

In this study, the primary screening Ct values were stratified for the first time, and the relationship between different Ct values and the proportion of false positives was analyzed by the results of the original sample retest and resample detection. False SARS CoV-2 PCR test results do occur in clinical settings and can have harmful effects in the case of low-prevalence screening with a low prior probability of positive detection [10]. These include the delay of surgery or other procedures, unnecessary isolation of individuals and close contacts, unnecessary contact tracking and detection, wasteful consumption of personal protective equipment, potential uninfected individuals and infected individuals in hospitals, etc., which have been summarized in detail by BRAUNSTEIN G D [10].

It was found in this study that the number of false positive results could be significantly reduced by retesting, thus greatly lowering the adverse effects of false positive results on epidemic prevention and control, especially for weakly positive samples with any Ct value of double-target genes > 35. Many factors can lead to false positive results [11–14]. Technical issues, such as cross-contamination during sampling (e.g., accidental contact of swabs with contaminated gloves or surfaces [15] ), exposure to infected personnel or contaminated equipment, and viral aerosolization during specimen collection [11], may contribute to false-positive outcomes. Additionally, ambiguous results can arise from low-viral-load samples due to poorly defined detection thresholds [10]. Other contributing factors include suboptimal primer/probe design (e.g., non-specific binding or primer-dimer formation) [16, 17], as well as cross-reactivity with nucleic acids from other pathogens, host tissues, or related coronaviruses [18]. Inexperienced laboratory personnel and inadequate testing conditions (e.g., contaminated workspaces) further exacerbate these risks. Furthermore, RT-PCR may detect non-infectious viral remnants, such as residual SARS-CoV-2 RNA persisting weeks post-infection, which lack clinical relevance but still yield positive results [19].

On the basis of the actual situation of laboratories, a heavy laboratory workload was also an important factor for false positive results. When a regional pandemic occurs, high-frequency and heavy testing work, resulting in physical and mental fatigue of the testing personnel, sometimes results in errors in the sample and even false positives. Moreover, given the 24-hour nonstop working situation, there was no time to clean the laboratory effectively, which would also result in nucleic acid contamination and false positives in the laboratory. Studies have also shown that in the nucleic acid testing laboratory for SARS-CoV-2, the leakage and improper handling of plasmid products have led to persistent nucleic acid contamination in the laboratory [20]. In addition to the above factors, the potential for cross-contamination of samples is large, both at the analytical and preanalytical stages of testing [21].

Notably, in addition to false positives in this study, for the suspected positive samples, some positive results were still obtained after retesting or resampling the original samples and detecting them separately. Therefore, corresponding measures should be taken for such samples, including timely retesting of the original samples or resampling and testing separately, to avoid virus transmission caused by infected persons. This strategy was also consistent with the reported view that repeat testing drastically decreases the chances of failing to identify infected individuals [22]. Sometimes, the viral load may be relatively low in the initial test and may reach the lower limit of reagent detection in the second test [4].

In response to COVID-19 infection emergencies, a series of test kits have been approved by the US FDA Emergency Use Authorization and China’s NMPA [23]. In clinical applications, other factors, such as differences in their composition of reagents, nucleic acid extraction methods, real-time RT‒PCR processes, interpretation of results, personnel, and equipment, lead to variations in testing results among different laboratories [23]. Therefore, correct sampling procedures, strict laboratory standards, and the use of high-quality extraction and real-time RT‒PCR kits could improve test quality and reduce inaccurate results [4]. As demonstrated by our findings, significant differences exist in false-positive rates across detection systems. When the Ct values of both target genes were < 30 (indicating strong positivity), no significant differences in false-positive rates were observed among reagents (P = 0.947) or testing institutions (P = 0.222), suggesting consistent performance across systems for high-viral-load samples. For samples with moderate positivity (30 ≤ Ct < 35), Sansu (0/85, 0%) and Daan (4/284, 1.41%) exhibited minimal false positives, while other reagents showed higher variability. Notably, significant inter-institutional differences emerged in this Ct range (P < 0.001), likely reflecting variations in operational protocols. In weakly positive samples (Ct > 35), all four reagents displayed markedly elevated false-positive rates, though inter-reagent differences remained non-significant. Striking inter-institutional variability (P < 0.001) was observed in this category, potentially attributable to differences in personnel technique, equipment calibration, or reagent batch effects. For instance, institution SHAQU reported an overall false-positive rate of 30% (36/120), escalating to 44.64% (25/56) for Ct > 35, whereas institutions HZHX (YC) and HYK demonstrated the lowest proportions. These findings underscore the importance of context-specific quality control measures, particularly for low-viral-load samples.

Consequently, the Ct values should not be interpreted in isolation. Instead, they should be considered within the operational context of individual laboratories. Differences in equipment, reagent sensitivity, and operator skills can result in different Ct value assignments for the same sample across different institutions. Thus, laboratories need to develop internal validation protocols to standardize Ct value interpretations and maintain consistency in retesting strategies.

On the basis of this study, the following original sample retest strategies are recommended for labs to adopt for primary screening of positive or suspected COVID-19 nucleic acid tests.

In the state of regional control, once the pooled sample is positive or suspiciously positive in the preliminary screening, it should be retested before being reported. In the individual sampling test, if the result is strongly positive with two targets (Ct value < 30), the original sample retest may not be carried out, and the positive result may be directly reported. However, in this case, there is still the possibility of individual false positives, and attention should be given to screening in the follow-up process. If the single target gene is positive or the double target gene is weakly positive (the Ct value of either of the two targets is > 30), one or two more sensitive or equally sensitive reagents should be reexamined and reported.

In the nonregional control state, if the pooled sample is positive in the preliminary screening, to isolate and control the potential risk personnel as soon as possible, the results should be reported immediately, and individual sampling and testing work should be carried out (all the people involved in the pooled samples should stay in place and be isolated from others until single sampling and testing results are obtained). In the individual sampling test, the processing strategy is the same as that described above. For samples that are suspected to be positive after initial screening, retesting or resampling should be carried out in time to avoid missing positive infected persons. The retest of the original sample is only for the samples that are initially positive or suspected positive, and the negative samples are not retested, mainly to avoid false positives but cannot avoid false negatives.

The sample data of suspected positive pooled samples are not sufficient to analyze the relationships among the Ct value, different reagents, various test institutions and individual sampling results. In the future, data related to this situation should be added and analyzed to provide evidence for laboratories to decide whether to conduct separate collections in time.

By analyzing the relationship between Ct values and the proportion of false positives, this study analyzed the possibility of false positives if positive results were directly reported with different Ct values. On the basis of the stratification of different Ct values and the pandemic prevention and control situation at that time, the strategy of retesting the original tube was proposed to be an important reference for epidemic prevention and control.

Of course, this study’s definition of false positives based solely on retesting results may oversimplify diagnostic challenges. Future studies should incorporate multi-center validation and quantitative viral load analysis to further refine diagnostic accuracy. Additionally, initial negative results were not retested during large-scale screening, precluding the collection of false-negative data. Although this study focuses on false positives, controlling false negatives remains critical for preventing undetected transmission.

It is worth mentioning that our recommendations are tailored to the detection of RNA viruses via RT-PCR, based on the assumption that the pathogen shares similar characteristics with SARS-CoV-2. If ‘Disease X’ is caused by bacteria or parasites, entirely different diagnostic methods may be required.

Conclusions

The detection method of RT‒PCR for SARS-CoV-2 has withstood large-scale nucleic acid detection during the epidemic and has also led to many valuable experiences and lessons, which provides a better reference for the rapid diagnosis of the next “Disease X “. RT‒PCR is prone to false positives caused by various factors during high-intensity and continuous testing. It is important to avoid these factors and establish a cutoff value in a timely manner (different from the cutoff value determined by the reagent kit) as a criterion for direct reporting, retesting, or resampling. In the early stages of epidemic control, it is necessary to balance the advantages and disadvantages of rapid reporting and accurate results. During the epidemic response period, speed is sometimes more important than accuracy.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

RT‒PCR

Real-time reverse transcription‒PCR

Ct value

Cycle threshold value

Author contributions

Xin Yang：Data curation, Investigation, Methodology, Validation, Writing the original draft, Writing review and editing; Jiming Li: Conceptualization, Project administration, Supervision, Writing review and editing.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

All the data generated or analyzed during this study are included in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

The study conformed to the ethical guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of the Yantai Yuhuangding Hospital. The Ethics Committee of the Yantai Yuhuangding Hospital approved our study also waived informed consent because this study was retrospective, all information was anonymous, and there was no risk to the subjects.

Ethical approval

Unavailable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Clinical trial

Not applicable.

Clinical perspectives

False positives may occur when viral infections are detected through RT‒PCR, and it is impossible to distinguish them according to the criteria for positive interpretation in the kit instructions. Retesting all positive samples greatly increases the laboratory testing burden, and conversely, false positive results seriously interfere with clinical diagnosis and treatment. On the basis of many test results, this study established a cutoff value for sample retesting and elaborated on what kind of results can be directly reported and what kind of results need to be retested. In this way, the reporting of false positive results is minimized as much as possible, the laboratory testing burden is reduced, and timely reporting of the results of true positive samples is ensured. When humans respond to future “Disease X “, if the RT‒PCR technique is used for detection, the experience provided by this study can be directly adopted, providing a more effective guarantee for clinical testing.