key: cord-0999137-dxuqkg78 authors: Lascarrou, J. B.; Colin, G.; Le Thuaut, A.; Serck, N.; Ohana, M.; Sauneuf, B.; Geri, G.; Mesland, J. B.; Ribeyre, G.; Hussenet, C.; Boureau, A. S.; Gille, T. title: Predictors of characteristics associated with negative SARS-CoV-2 PCR test despite proven disease and association with treatment and outcomes.The COVID-19 RT-PCR Study. date: 2020-09-15 journal: nan DOI: 10.1101/2020.09.14.20194001 sha: c99e33e53552110f7a4776bc6176ad5e0c43fa7c doc_id: 999137 cord_uid: dxuqkg78 Background: Since December 2019, Coronavirus 2019 (Covid-19) emerged in Wuhan city in China, and rapidly spread throughout China, Asia and worldwide. Recently, concerns emerged about specificity of PCR testing especially sensibility. We hypothesis first that clinical and/or biological and/or radiological characteristics of patients with first false negative COVID19 RT-PCR test despite final diagnosis of COVID-19 are different from patients with first positive COVID19 RT-PCR test. Methods: Case / control study in which patients with first negative COVID19 RT-PCR test were matched to patients with first positive COVID-19 RT-PCR test on age, gender and ward/ICU location at time of RT-PCR test. Results: Between March 30, and June 22, 2020, 82 cases and 80 controls were included. Neither proportion of death at hospital discharge, nor duration of hospital length stay differed between patients Cases and Controls (respectively P=0.53 and P=0.79). In multivariable analysis, fatigue and/or malaise (aOR: 0.16 [0.03 ; 0.81]; P=0.0266), headache (aOR: 0.07 [0.01 ; 0.49]; P=0.0066) were associated with lower risk of false negative whereas platelets upper than 207 per 10.3.mm-3 (aOR: 3.81 [1.10 ; 13.16]; P=0.0344), and CRP>79.8 mg.L-1 (aOR: 4.00 [1.21 ; 13.19]; P=0.0226) were associated with higher risk of false negative. Interpretation: Patients suspected of COVID19 with higher inflammatory biological findings expected higher risk of false negative COVID19 RT-PCR test. Strategy of serial RT-PCR test must be rigorously evaluated before adoption by clinicians. 5 symptoms of COVID-19 infection are not specific: fever, cough, fatigue and lymphopenia (1) , RT-PCR testing and interpretation of results can be a concern for clinicians. We hypothesis first that clinical and/or biological and/or radiological characteristics of patients with first false negative COVID-19 RT-PCR test despite final diagnosis of COVID-19 are different from patients with first positive COVID-19 RT-PCR test; second that patients with first false negative COVID-19 RT-PCR expected better outcome than patients with first positive COVID-19 RT-PCR test. To answer this, we performed a case-control study in which patients with first negative COVID-19 RT-PCR test were matched to patients with first positive COVID-19 RT-PCR test. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . We performed a multicenter retrospective analysis of patients admitted in hospital for suspicion of COVID-19 infection and negative COVID-19 RT-PCR (case) with control: patients hospitalized in the same hospital match on gender, age and ward / intensive care unit (ICU) service with first positive COVID-19 RT-PCR test. Cases were patients with first negative RT-PCR test despite final diagnosis of COVID-19 leading to hospital admission. Controls were patients with first positive RT-PCR test matched on age, gender and ward/ICU in the same hospital. Inclusion criteria were: CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . Our primary objective was to identify factors associated with higher risk of false negative first COVID-19 RT-PCR test (regarding sensibility of RT-PCR testing). Secondary outcomes were treatments delivered, need for mechanical ventilation, duration of mechanical ventilation, occurrence of acute respiratory syndrome and outcome at hospital discharge. All data in the eCase Report Form were anonymized, and no data can be traced back to the patient's identity. Each local investigator filled an eCRF to collect data (Castor EDC, Amsterdam, The Netherlands). Data collected were: characteristics of matching: age, gender, location; baseline demographics (comorbidities); clinical and biological characteristics at hospital admission; history of symptoms; radiological findings; RT-PCR testing results (first and final RT-PCR test if positive for "case" patient); other pathogens testing and result; antiviral treatments; outcomes; modalities of final diagnosis for "case" patient. The study was approved by the appropriate ethics committees (For France: Comité d'éthique de la Société de Réanimation de Langue Française, #20-26; and for Belgium: Comité d'Ethique 045 Clinique Saint Pierre) which waived consent according to data collected. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . https://doi.org/10.1101/2020.09.14.20194001 doi: medRxiv preprint 8 Statistical analysis were performed according to STROBE guidelines (9). Cases were matched with controls on age, gender, and location of hospital admission (ward/intensive care unit) on 1:1 basis. Qualitative variables were described as n (%) and quantitative variables as mean±SD if normally distributed and median [25 th -75 th percentiles] otherwise. Mortality and hospitalization rate were compared between cases and controls using conditional logistic regression to take into account paired data. Conditional logistic regression models were used to identify factors associated with negative RT-PCR test. Step by step backward selection was applied. Predefined factors associated with negative RT-PCR testing at P values Regarding exploratory nature of our study, we did not set sample size but we targeted at least 50 patients and 50 controls (100 patients). . . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . Between March 30, and June 22, 2020, 82 cases and 80 controls were included. Related to non-inclusion of matched controls, 2 cases patients were excluded from analysis. Patients were mainly male (66.25%), were 64.1±16.8 years old and were mainly admitted in ward (71.25%). Of the 80 cases included, a chest radiography was performed for 26 patients (normal (N=1), ground-glass opacities (N=4), local patchy opacities (N=1), bilateral patchy opacities (N=12), interstitial abnormalities (N=7) and a chest CT scan was performed for 75 cases (normal (N=1), ground-glass opacities (N=69), interstitial abnormalities (N=4). Clinical participant's characteristics for the cases and the controls are detailed in Table 1 , and their biological characteristics in Table 2 . On univariable analysis, fatigue/malaise (P=0.0482), headache (P=0.0481), history of fever (P=0.0202), myalgia (0.0239) and elevation of hepatic enzymes (P=0.0239 for ALAT and . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . https://doi.org/10.1101/2020.09.14.20194001 doi: medRxiv preprint P=0.0239 for ASAT) were associated with lower risk of negative PCR test (OR<1);whereas, platelets upper than 207 per 10. 3 .mm -3 (P=0.0015), white blood cells >6.95 per 10. 3 .mm -3 (P=0.0003) and CRP>79.8 mg.L -1 (P=0.279) were associated with higher risk of negative RT-PCR test (OR>1). Because ASAT, ALAT were collinear of platelets count and white blood cells were collinear of CRP, they were not included in multivariable analysis. Result of multivariable analysis is depicted on Figure 1 with an AIC: 54.8 and BIC: 69.1. Proportion of patients "Cases" and "controls" who received at least one treatment (Chloroquine, Corticosteroids, Lopinavir/ritonavir, Macrolids or Tocilizumab) did not differ (P=0.26) ( Table 3) . Mechanical ventilation was required for 10 (12.66%) cases and 14 (17.72%) controls, for duration of 21 [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] days for cases and 15 [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] days for controls. Neither proportion of death at hospital discharge, nor duration of hospital length stay differed between patients "Cases" and "Controls" (respectively P=0.53 and P=0.79). . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . We found that patients with high elevation of CRP, high platelets count, expected high risk of false negative RT-PCR test, whereas patients with non-specific symptoms such as headache or fatigue/malaise expected lower risk. Duration between symptoms onset and time of RT-PCR testing was not associated with false negative result of RT-PCR test in our study. Finally, patients with false negative test did receive neither different treatments nor their expected different outcome according to proportion of patients requiring mechanical ventilation and mortality at hospital discharge. Tree main conclusions can be drawn. First, duration between symptoms onset and time to RT-PCR test was not associated with positivity. We think than such result can be explained by difficulties in the medical history examination especially in older patients (64±17 years) in our cohort as previously described for other disease (11). Additionally, frequent presence of delirium (up to 25%) in the geriatric patients can hypothesized duration reported (12) . Second, association between high level of CRP and higher risk of negative RT-PCR test is of interest because: 1) It is an argument for mortality associated with cytokine storm regardless of viral load (13) 2) It is problematic according to results of RECOVERY trial which indicates than corticosteroids is the only treatment proved effective to reduce mortality for patients with COVID-19 (14) . Small proportion of our patients received corticosteroids but our study took place before evidence of beneficial effects of early short course of corticosteroids. Additional data suggests than corticosteroids benefit most to patients with level of CRP higher than 20 mg/dL (15) striking our results. In other part, RECOVERY did . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted September 15, 2020. . 12 not mandate positive RT-PCR to be included and randomized in the study (11% of the whole cohort). Interestingly, Hu et al found than headache was associated with intermittent negative COVID-19 RT-PCR status (16) , highlighting our result about headache in our multivariable analysis. Third, low rate of final diagnosis according to positive PCR test is also a concern. Some protocols advocated positive PCR test to include patients (17) and we can hypotheses than physician will be less prone to prescribe treatment to patients without positive PCR test. Global sensitivy of RT-PCR is describe as to 70% (18) but with major impact of duration between symptoms onset and day of RT-PCR testing: between 38% of false negative at day of symptoms onset to 20% at day 8 and then false negative rate which increase again (19) . Long et al, found a positive rate of only 3.5% for patients first initial RT-PCR testing and subsequent retest for the next 7 days. Ai et al, found than chest CT has a high sensitivity for diagnosis of COVID-19 and may be considered as a primary tool for detection in epidemic area (20) . Finally, strategy to perform several tests to document virologic proof of COVID-19 can be debated. Our study took place during first epidemic wave in France and Belgium and was dedicated only to patients requiring hospitalization. It leads to high pre-test probability of COVD-19.We selected carefully patients hospitalized with several strong arguments for COVID-19 and final diagnosis of COVID-19 at hospital discharge. Some limitations of our study must be highlighted. First, negative RT-PCR test can be related to other disease. However, 45 (59.96%) of cases received negative others pathogens research during their hospital length stay and final diagnosis of COVID-19 was performed according to multimodal strategy including chest CT-. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted September 15, 2020. . https://doi.org/10.1101/2020.09.14.20194001 doi: medRxiv preprint 13 scans in 88.75% of case. Second, some issues could occurred during technical perform of RT-PCR but all RT-PCR were performed in hospitals with trained nurses and dedicated protocol to ensure high adherence to methods of RT-PCR realization. Third, our sample size is limited, but we choose to restrained inclusion of patients with robust arguments of COVID-19 according to others methods of diagnosis (especially chest CT-scans) with limited availability during epidemic wave in Europe. Last, we included patients for several centres with different RT-PCR detection kit. However, evidence suggests similar performance of available RT-PCR kits (21, 22) . Patients with first negative RT-PCR test for COVID-19 expected inflammatory markers even at median duration of 6 days after symptoms onset. Decision to perform or to withdraw special treatments such as corticosteroids for patients with COVID-19 cannot be done only on virologic isolation of SARS-CoV-2. Multimodal strategy for diagnosis including radiological findings and clinical history is mandatory for each patient suspected of COVID-19. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . https://doi.org/10.1101/2020.09.14.20194001 doi: medRxiv preprint We thank M. Rouaud, PharmD, for help during administrative process. We thank Mariana Ismael for Castor EDC (Amsterdam, The Netherlands) for technical support to design eCRF. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . Yang . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . https://doi.org/10.1101/2020.09.14.20194001 doi: medRxiv preprint Figure 1 : Forrest plot of multivariable analysis of factors associated with first negative RT-PCR COVID-19 testing . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint The copyright holder for this this version posted September 15, 2020. . https://doi.org/10.1101/2020.09.14.20194001 doi: medRxiv preprint Clinical Characteristics of Coronavirus Disease 2019 in China. The New England journal of medicine A Novel Coronavirus from Patients with Pneumonia in China Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study False-Negative Results of Real-Time Reverse-Transcriptase Polymerase Chain Reaction for Severe Acute Respiratory Syndrome Coronavirus 2: Role of Deep-Learning-Based CT Diagnosis and Insights from Two Cases Performance of Radiologists in Differentiating COVID-19 from Non-COVID-19 Viral Pneumonia at Chest CT Real-time RT-PCR in COVID-19 detection: issues affecting the results Virological assessment of hospitalized patients with COVID Evaluation of Risk Factors in Older Patients with Acute Myocardial Infarction Study Clinical Characteristics and Outcomes of 821 Older Patients with SARS-Cov-2 Infection Admitted to Acute Care Geriatric Wards. The journals of gerontology Series A, Biological sciences and medical sciences COVID-19: consider cytokine storm syndromes and immunosuppression. The Lancet Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report Effect of Systemic Glucocorticoids on Mortality or Mechanical Ventilation in Patients With COVID-19 Factors associated with negative conversion of viral RNA in patients hospitalized with COVID-19. The Science of the total environment High dose dexamethasone treatment for Acute Respiratory Distress Syndrome secondary to COVID-19: a structured summary of a study protocol for a randomised controlled trial False Negative Tests for SARS-CoV-2 Infection -Challenges and Implications Variation in False-Negative Rate of Reverse Transcriptase Polymerase Chain Reaction-Based SARS-CoV-2 Tests by Time Since Exposure Correlation of Chest CT and RT-PCR Testing for Coronavirus Disease 2019 (COVID-19) in China: A Report of 1014 Cases First results of a national external quality assessment scheme for the detection of SARS-CoV-2 genome sequences Performance; Quality Evaluation of Marketed COVID-19 RNA Detection Kits Lymphocyte (cells/microL) , median Neutrophil (cells/microL) , median Haematocrit (%), median PT (seconds), median Total Bilirubin (µmol/L), median Lactate (mmol/L), median Creatininemia (µmol/L), median Sodium (mEq/L) , median Procalcitonin (ng/mL) , median