Skip to content
Switch branches/tags
Go to file
Cannot retrieve contributors at this time

Who are the "false-negative" patients of COVID-19?

Originally published in Southern Weekly (南方周末) magazine on Feb. 6th, 2020

On Jan. 24th, In ICU of Zhongnan Hospital of Wuhan University, Medical staff wrote down their names on isolation suits for recognition.

Tan Wei, deputy director of the Hubei Imaging Society of Integrative Medicine, said that 30-40% of the cases whose CT results were very consistent with the new coronavirus pneumonia (i.e. COVID-19) returned negative on the nucleic acid test.

One patient at a hospital in Hangzhou tested negative for the nucleic acid reagents six times and only tested positive the seventh time.

Multiple front-line physicians have expressed a similar view that the most important factor affecting nucleic acid detection is, first and foremost, sample collection.

A married couple who were tested negative for nucleic acid were admitted to a designated hospital in Taizhou. The Hospital dared not let them go. His wife complained every day, "The country is sure we're not sick and you're keeping me here," as she was yelling at a nurse in the isolation ward.

Text | Tang, Yucheng (汤禹成); Tan, Chang (谭畅) reporters at Southern Weekly (南方周末); Yan, Liyuan (闫力元) and Jiang, Zhiyu (蒋芷毓), special contributors at Southern Weekly (南方周末)

Editor | He, Haining (何海宁)

Translation | Genefy and Walrus from Chengji Translation Team (翻译者要求署名)

Auntie Wu was surprised when she got the result of the nucleic acid test.

In January 2020, first Auntie Wu and then her son, Li Zhe (pseudonym), became ill with fever. They had taken CT scans at Wuhan Ninth Hospital before testing for nucleic acid. The result of the son was a double-lung infection and the result of Auntie Wu was a double-lung infection with ground glass opacity. Auntie Wu even took CT scans twice, saying that she had shown CT results to some doctors, who agreed that Auntie Wu was highly suspected to have the new coronavirus infection, as her results were highly consistent with the condition.

The mother and son lived in separate wards for a few days in the observation room. Every time Li Zhe visited his mother, he could see her talking on the phone with a different person, hoping for a nucleic acid test. On January 28th and 29th, they finally got the chance to do the nucleic acid test twice in a row. On the first day the doctor took a swab in Wu's mouth to take samples and the next day in the nasal cavity.

Surprisingly, Li Zhe's nucleic acid test results were "positive". Auntie Wu, however, who had been more severely ill and started earlier, was tested "negative".

This is not an exception. Earlier, getting a reagent box was "winning the lottery" to the suspected patient. And now it's changed. Even if the symptoms are consistent and CT scans suggest a pulmonary viral infection, the patient cannot obtain a "positive" diagnosis.

This would have been good news but now it is becoming a new concern for the patient.

Some patients may be true-negative, but others may change to positive after several tests. They cannot initially be identified as being infected by the new coronavirus (SARS-CoV-2), which means that they may have been turned away from hospitals. The condition may suddenly worsen and there is a risk of infection among those around them.

In February of the new year, there has been a growing number of online posts asking for help from those who tested negative on nucleic acid tests. Tan Wei, deputy director of Hubei Medical Imaging Professional Committee, told Southern Weekly reporters that based on his clinical experience, he believed that 30-40% of the cases whose CT results were very consistent with the new coronavirus pneumonia (i.e., COVID-19) tested negative for nucleic acid.

Li Zhe was hospitalized and started to receive in-patient care, but Wu's case didn't go so well. She hasn't been able to do the third nucleic acid test, nor does she know whether it will turn positive later. As her condition is becoming more severe, she worries about not getting hospitalized day and night.

Positive until the seventh test

"I really hope my test result was positive". Auntie Wu said she was unlucky. If she were positive, there would be a chance to be hospitalized.

According to her description her condition is deteriorating. When the first CT scan was taken, Auntie Wu was able to act as usual. And now she's so dependent on oxygen. Her hospital bed in the observation room is less than three meters from the toilet, but she has to sit for a long time every time she goes to the bathroom, waiting until her breath is steady before she dares to walk slowly across it. Having to exhale through her nose and mouth, she can hardly hold out even for as short a time as a bathroom break.

In another family, three of the seven members of Wu Xi's family have been diagnosed. Grandpa died the night he was diagnosed. The scarce bed was left for Grandma who was later diagnosed.

Wu Xi's cousin Dou Juan was the only negative patient among those who had had a nucleic acid test in the family, but her double lung imaging had shown viral inflammation. She became one of the family members that Wu Xi was most concerned about. Wu Xi said, "Two negative tests cannot rule (her) out; for example, Li Wenliang wasn't tested positive until the third time.

Li Wenliang, as she mentioned, is one of the first "disinformation" doctors to whistleblow the outbreak. Li Wenliang tested for nucleic acid altogether 3 times. The first result was unknown and the second was negative. Only the third time was positive. This was what he announced on his Weibo account that day, 23 days after he got sick.

It's not just Wuhan. The medical director of a designated hospital in Taizhou, Zhejiang told Southern Weekly reporters that in a video conference on February 3, he heard provincial experts talking about that one patient tested negative for the nucleic acid reagents six times and did not test positive until the seventh time in a hospital in Hangzhou.

The director's hospital had a similar case. A couple who came back from Wuhan with a fever also reported a lung imaging result showing a viral infection, although all three tests for nucleic acid were negative.

Nor is it an exception among patients. Several doctors in the affected areas have confirmed to Southern Weekly reporters that there are now cases where a patient tested negative would turn positive on the nucleic acid test.

A doctor deployed to the front-line in Wuhan said that the clinical symptoms and CT images of many patients in his hospital were both consistent with the new coronavirus pneumonia diagnosis, but among them there were fewer positive cases than negative ones for nucleic acid tests.In a statistics table obtained by the Southern Weekly reporters, many negative patients did not undergo a second test.

This doctor explained that the Wuhan Centers for Disease Control and Prevention (CDC) only gave his hospital a limited quota of 50 tests per day. There's nothing the hospital could do about it. There have also been patients who were tested again. Some of them remained negative, although there was also one person who turned positive.

Many negative patients are applying for a second test, including Lin Jun and his mother. Lin Jun's daughter said that the application should be submitted to the community for approval. "But it's too long to wait". They queued up in different hospitals for another nucleic acid test, and Lin Jun heard there were patients who had been tested negative three times, too. It took a fourth time to show positive results.

Faulty throat swab sampling

On February 3rd, a screenshot of a WeChat Moments posted by Zhang Xiaochun, deputy director of Radiology department at Zhongnan Hospital of Wuhan University, circulated on the Internet. According to her, CT imaging should be the main basis for the current 2019-nCoV pneumonia (i.e., COVID-19) screening in Wuhan outbreak area; at the same time, asymptomatic or nucleic acid test negative but CT imaging positive people should be isolated to prevent familial clustering.

She told the press that one of the reasons she posted the Moments was that some patients who were diagnosed as positive based on CT imaging returned to their homes because of false-negative nucleic acid test results, which eventually caused familial clusters.

Why is there a "false negative" in the test?

Southern Weekly reporters learned that the detection of the new coronavirus included "sample collection", "viral nucleic acid extraction", and "viral nucleic acid detection". Each step has an impact on the outcome.

Multiple front-line doctors have expressed a similar view that the most important factor affecting nucleic acid detection is, first and foremost, sample collection. Currently, oral and nasal swabs are the most common way of collecting samples.

Wang RuiZhi, a front-line medical examiner in Wuhan, said that in the field it's often said that "garbage in, garbage out". He said that swab sampling, which would involve swabbing the arches of the palate, pharynx and palate tonsils for secretion to obtain specimens, would work for the upper respiratory tract, but the new coronavirus was primarily in the lower respiratory tract. . According to the new Interim Clinical Guidance for Management of Patients with COVID-19 (by WHO), both upper and lower respiratory tract specimens should be tested to improve detection rates. But in Wang's observations, because the lower respiratory tract specimens were relatively difficult to obtain, many hospitals did not do simultaneous testing.

Another front-line doctor said that swabs were operable and the easiest to popularize. Phlegm is not always present and extraction of alveolar lavage involves tracheoscopy, which is not always convenient and renders the paramedics more susceptible to infection.

Sampling techniques can also affect detection accuracy. One doctor said that an inexperienced nurse might not be able to obtain a qualified specimen, leading to a false negative. But a more realistic reason lies outside of the lack of experience. Tan Wei told Southern Weekly reporters that the nucleic acid test was first performed by the Hubei CDC. Many of the medical staff were infected because they faced the patients mouth-to-mouth. Tests have now been extended to more hospitals. On the one hand, lots of unprofessional people are performing the test, and the sampling locations they take are not deep enough to reach the pharynx. More importantly, some people are afraid of getting infected if they are too close to the patients, which makes it easier for samples to be collected improperly.

Experimental operation is also a possible factor. Wang Reizhi, who is familiar with in vitro diagnostic reagents, said the new coronavirus test had more operational steps, so the experimenters would be more likely to make mistakes. On top of this, there is a very high number of specimens currently being tested on a daily basis, which increases the probability of errors in the operation of the high workload inspectors.

Hastily prepared reagent kits

Another issue that patients are concerned about is the performance of the reagent kits.

Wang Ruizhi said the outbreak began with a lack of capacity in the production of reagent kits. The National Medical Products Administration (NMPA) had sped up the approval process for reagent kits. Many manufacturers were speeding up research and development to catch up with the market. As a result, the performance of many kits was not adequately tested and evaluated. Generally speaking, It takes 2-3 years from development to certification before such a reagent can be used in clinical applications. And the manufacturing process of the reagent kits will also affect the quality of each batch of the products, which also influences the stability of the detection effectiveness of different batches.

The development of the reagent kits was indeed rushed. A staff member at a company designated to produce the reagent told Southern Weekly that it took less than 20 days for the company to develop the product and get its license. At first, the company often worked overtime till early in the morning, but the workers on the production line never stopped. They had forty or fifty available workers.

"All the companies were racing against the clock and everyone would say they were the first to produce it", said the staff member. On January 26, the reagent kits of four domestic companies passed through the NMPA's Emergency Approval Channel at record speed, which expedited the delivery to local centers and hospitals for disease control. The Southern Weekly reporters learned that the NMPA department had previously assigned two reviewers to each company, providing a 24-hour question-and-answer period to brief the companies on all aspects of the approval process.

Now, at the end of the Spring Festival holiday, with most of the company's workers having returned to work by the third day of the Lunar New Year, the company can reach a production capacity of 200,000 units a day. But daily production is still based on daily orders from hospitals and disease control centers.

When Southern Weekly reporters asked about the sensitivity of the kits, the company responded, "We can only say that our kits are highly sensitive and specific. The exact data is difficult for almost all reagent manufacturers to identify." The staff member said that the company had been very busy lately, and the clinical data had not yet been collected.

Different companies are also producing different reagent kits. "Each company has developed a different process for detecting viral nucleic acid sites, with different sources of raw materials and different production procedures. The kits from different companies will certainly perform differently," Wang Ruizhi told Southern Weekly reporters.

Jiang Rongmeng, a member of the National Advisory Committee on Infectious Diseases, suggested that there was another possibility besides sampling and reagent problems. He believed that the detection of the virus was also related to the time of the patient's onset of symptoms. Generally speaking, one or two days into the onset of symptoms would not test as positive as five or six days. "Even if the reagent and the sampling are not faulty, the time of onset can also have an effect. Low viral load during early onset may cause lower possibility of getting a positive result" Jiang Rongmeng said the same logic held true for chest CT diagnoses in which some patients were tested positive for nucleic acid but lung imaging did not show that. Such patients would also need to be isolated for observation.

How to treat "negative" patients in Wuhan.

On February 2, Wuhan No.9 Hospital informed Auntie Wu and some other patients who had been diagnosed as negative that they were being transferred to another hospital. Auntie Wu said she was too dependent on the Hospital's oxygen supply to go to the lobby for transfer. Then she contacted several hospitals by herself, and for various reasons, she was unable to go.

At first, there were seventy or eighty patients in the Observation Room with Wu. The positive ones were just a few, a total of eight or nine. Auntie Wu stayed in the Observation Room on the fourth floor for a few days. The patients were mostly elderly and most of them had a fever. Lots of them already had difficulty breathing and were taking oxygen. Other patients close to respiratory failure were still struggling, emitting loud cries. People died almost every day.

In contrast, the son who has been diagnosed as positive is fortunate.

On the night of February 2, the hospital gave Auntie Wu an ultimatum that she must leave. Auntie Wu wouldn't even get an injection if she wouldn't leave.

It was also on February 2, when Wuhan New Coronary Pneumonia Control Command required the city's urban areas to carry out centralized treatment and isolation on "four categories of people". These four categories of people are: confirmed patients, suspected patients, patients with fever that cannot be ruled out of the likelihood of infection, and close contacts of confirmed patients. Many hotels in Wuhan have been converted into New Coronary Pneumonia Isolation Points. During the isolation period, free accommodation, medical observation and treatment are provided in all districts.

But, according to the information that Southern Weekly reporters gathered, the main body of centralized segregation would be mainly grassroots communities. Many hotels have yet to implement the isolation plan, and many do not meet medical conditions. Many patients with negative nucleic acid tests whose condition is yet still developing are either not placed in isolation or unwilling to go there themselves.

Lin Jun and his mother have been isolated at home from their wives and daughters, who are still asymptomatic and remain in the living room. Lin Jun's daughter told Southern Weekly reporters that at her home in the Hongshan District, there was no obvious impact of centralized isolation and no one told them to stay at the hotel.

Wu Xi is also asking the community to coordinate hospital beds for her aunt. Wu Xi said she had told a lot of people about her family and also asked for help online. There were a lot of people who contacted her, and she didn't give up on any of the chances, "So now I face you, and I told it again," Wu Xi said to the Southern Weekly reporters on the evening of February 4.

Another patient, Mr. Lee, had a rather serious lung infection but tested negative twice for nucleic acid. His community wanted to put him in a hotel but they refused because Mr. Lee still needed treatment, which required seven or eight hours a day in a queue for injections at a designated hospital . The community couldn't solve that by just putting them together (in a hotel).

On the front line of treatment, Tan Wei described a sad reality. Some patients had a negative first and second nucleic acid test but CT images showed a viral infection. Followed up with a CT examination over time, they were found to be getting worse. When they rechecked the nucleic acid and waited until the diagnosis was positive, the patients' conditions aggravated and couldn't be saved. The reasons behind, as Tan Wei pointed out, were that the current shortage of the relevant medicines would cause the patient who was not seriously ill to not take such medications. The undiagnosed patients who reached some of the isolation sites did not have relevant medicines either.

On February 4, Southern Weekly reporters called multiple Wuhan community grid workers to ask what to do with the patients whose lung images would show a viral infection but the nucleic acid test would be negative. As a community grid worker said, they did not know how to deal with the situation. Others responded that Wuhan had no designated hospitals for such cases.

Comparing to the SARS treatment back then, according to *The 2004 Edition of the Infectious Atypical Pneumonia Diagnosis and Treatment Program*, people with epidemiological evidence of SARS, corresponding clinical manifestations and lung X-ray imaging changes, and can be ruled out for other disease diagnoses, can make SARS clinical diagnosis. On the basis of the clinical diagnosis, the detection of the pathogen can secure a definite diagnosis. In this Program, clinical diagnosis and definitive diagnosis are classified as one level and the other two levels are medical isolation observation and suspected cases, respectively. Such a hierarchy is more liberal than that of the new coronavirus pneumonia today, where clinically diagnosed patients would receive the same level of isolation as those diagnosed.

Mr. Tan also mentioned his own clinical treatment, saying that they primarily used CT for initial diagnosis. "Because the nucleic acid test takes so long, we have to wait a day for the result of the test. For example, we did a CT scan at first. For a suspected case of mild symptoms, we would advise him to go to the isolation site. We'll treat a case if he's a bit severe. Normally, in five days we'll review the CT scan to see if there's a change in imaging".

Other more cautious cities

The patient diagnosis has become a heavy pressure of the Wuhan outbreak area. All involved parties is looking for a solution. Later Zhang Xiaochun also stressed to the medical media, Ding Xiang Yuan, that "Wuhan as an outbreak area cannot completely rely on nucleic acid testing to screen patients at this stage and achieve the prevention and control effect of cutting off the source of infection... The suggestion I made is only suitable for our outbreak area, not for ordinary and sporadic cases screening. Screening for ordinary and/or sporadic cases requires differential diagnosis, which CT imaging cannot achieve."

In other cities, diagnosing seems more cautious.

Shanghai, thousands of miles away, has far fewer infections than Wuhan, where everything seems manageable. The nucleic acid test is still performed by the CDC and can be booked daily at the hospital. A respiratory physician at a Shanghai 3A Hospital told Southern Weekly reporters that now there were strict controls on the detection of nucleic acid. Only those with an indication on CT Imaging will be permitted to test. If the image matches the characteristics of viral pneumonia, and there's an epidemiological history and the leukocytes and lymphocytes are found to be low, "This type of patient cannot be released even if the nucleic acid test is negative for the first time. Instead, they should be observed in the isolation ward and even treated with medication. A few days after they should be tested the nucleic acid again. Very few patients might need to test more times."

Two cities in Zhejiang, Wenzhou and Taizhou, have gradually become serious outbreak areas. A married couple who were tested negative for nucleic acid were admitted to a designated hospital in Taizhou. The Hospital dared not let them go. "Even tested negative, we dare not let them go." Said the director of the aforementioned medical section. The wife complained every day, "The country is sure we're not sick and you're keeping me here," as she was yelling at a nurse in the isolation ward. The nurse felt aggrieved and asked the director to persuade him, and the director talked to the patient for half an hour to soothe her emotions.

The director told Southern Weekly reporters that although the negative results of the two tests met the criteria for exclusion, they were not clinically excluded. "We usually treat him until he basically would show no symptoms on scans or his indicators in lymphocytes would have returned to normal. Then he'd test for nucleic acids. If there is no problem then he can go home".

Several doctors have mentioned that the condition is evolving. The director quoted above analyzed the condition of a woman patient, who, tested negative on the nucleic acid test, had some of the lesions receding while another part growing, such that not all lesions would come out at once.

He also felt the pace of nucleic acid testing quickening. Earlier, the nucleic acid test was performed by the CDC in the local prefecture-level city. As the number of patients being tested increased, it became the responsibility of the county's centers for disease control. And now the county centers for disease control can confirm the diagnosis once the patients are tested positive.

Facing a positive result, doctors are more decisive. A patient's lung CT imaging did not have any abnormalities. Just because he had a slight cough after returning from Wuhan, he tested for nucleic acid which turned out positive. The director mentioned above judged him to be an upper respiratory infection of the coronavirus. Although there were no abnormalities in the lung CT imaging, he must be isolated.

In fact, experts in disease prevention and control are also debating how to confirm the diagnosis. In an interview with Southern Weekly reporters, most experts agreed that diagnosis would be best to have a pathological result. Some experts, however, thought that this would be more difficult in practice in Wuhan. The complexity of the new coronavirus infection requires doctors to synthesize more tests and make more careful judgments. Jiang Rongmeng pointed out to Southern Weekly reporters that imaging diagnosis could be used as an auxiliary way. "CT imaging with no indicators cannot necessarily rule out (a case), but the presence of indicators cannot necessarily be attributed to this disease either. From the point of view of clinical diagnosis, imaging is always a pathway for auxiliary diagnosis , and pathogenic diagnosis should always be the focus".

On February 5, the National Health Commission released the "New Coronavirus Infected Pneumonia Diagnosis and Treatment Scheme (Trial Version V)". The criteria for "suspected cases" have relaxed. Whether there is a history of epidemiology or not, as long as a case has satisfied two clinical manifestations, the "fever and/or respiratory symptoms" and "in early stages total leukocytes at normal or decreased levels, or decreased lymphocyte count ", it can be considered as a suspected case. More importantly, it has added a new "clinically diagnosed cases" besides the "suspected cases" and "confirmed cases", limiting it to the Hubei Province. This means that the results of CT imaging in Hubei Province can be used as the basis for the diagnosis of "clinical cases".