SARS-CoV-2 Seroprevalence and Drug Use in Trauma Patients from Six Sites in the United States

In comparison to the general patient population, trauma patients show higher level detections of bloodborne infectious diseases, such as Hepatitis and Human Immunodeficiency Virus. In comparison to bloodborne pathogens, the prevalence of respiratory infections such as SARS-CoV-2 and how that relates with other variables, such as drug usage and trauma type, is currently unknown in trauma populations. Here, we evaluated SARS-CoV-2 seropositivity and antibody isotype profile in 2,542 trauma patients from six Level-1 trauma centers between April and October of 2020 during the first wave of the COVID-19 pandemic. We found that the seroprevalence in trauma victims 18–44 years old (9.79%, 95% confidence interval/CI: 8.33 – 11.47) was much higher in comparison to older patients (45–69 years old: 6.03%, 4.59–5.88; 70+ years old: 4.33%, 2.54 – 7.20). Black/African American (9.54%, 7.77 – 11.65) and Hispanic/Latino patients (14.95%, 11.80 – 18.75) also had higher seroprevalence in comparison, respectively, to White (5.72%, 4.62 – 7.05) and Non-Latino patients (6.55%, 5.57 – 7.69). More than half (55.54%) of those tested for drug toxicology had at least one drug present in their system. Those that tested positive for narcotics or sedatives had a significant negative correlation with seropositivity, while those on anti-depressants trended positive. These findings represent an important consideration for both the patients and first responders that treat trauma patients facing potential risk of respiratory infectious diseases like SARS-CoV-2.

pathogens, the prevalence of respiratory infections such as SARS-CoV-2 and how that relates with other 48 variables, such as drug usage and trauma type, is currently unknown in trauma populations. Here, we 49 evaluated SARS-CoV-2 seropositivity and antibody isotype profile in 2,542 trauma patients from six Level-50 1 trauma centers between April and October of 2020 during the first wave of the COVID-19 pandemic. We 51 found that the seroprevalence in trauma victims 18-44 years old (9.79%, 95% confidence interval/CI: 8.33 52 -11.47) was much higher in comparison to older patients (45-69 years old: 6.03%, 4.59-5.88; 70+ years 53 old: 4.33%, 2.54 -7.20). Black/African American (9.54%, 7.77 -11.65) and Hispanic/Latino patients 54 (14.95%, 11.80 -18.75) also had higher seroprevalence in comparison, respectively, to White (5.72%, 4.62 55 -7.05) and Non-Latino patients (6.55%, 5.57 -7.69). More than half (55.54%) of those tested for drug 56 toxicology had at least one drug present in their system. Those that tested positive for narcotics or sedatives 57 had a significant negative correlation with seropositivity, while those on anti-depressants trended positive. 58 These findings represent an important consideration for both the patients and first responders that treat 59 trauma patients facing potential risk of respiratory infectious diseases like SARS-CoV-2. 60 61 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. infection than previously expected when treating trauma patients. As reported, COVID-19 related fatality 88 risks were the single highest cause of officer line-of-duty deaths 14,15 . EMS providers, who have been 89 operating on the far-forward front lines of the pandemic in 2020, had more cases of severe COVID-19 than 90 firefighters (1.2% versus 0.19% respectively) 16 . This risk could be exacerbated by the elevated ability of 91 SARS-CoV-2 to be transmitted by asymptomatic patients. Byambasuren  This study assesses the SARS-CoV-2 seropositivity of 2,542 de-identified serum samples from trauma 103 patients using a standardized enzyme-linked immunosorbent assay (ELISA) protocol that was previously 104 developed for the national serosurvey, conducted May 10 th and July 31 st , 2020 18,19 . The serosurvey ELISA 105 protocol identified IgG, IgM, and IgA antibodies for the SARS-CoV-2 spike protein and its receptor binding 106 domain (RBD). This assay can assess SARS-CoV-2 seropositivity objectively using either IgG or IgM 107 detected levels -for both spike and RBD expression -based on a threshold determined by pre-pandemic 108 control samples. The goal of this study was to evaluate SARS-CoV-2 seroprevalence in trauma patients to 109 offer insightful information on the association between SARS-CoV-2 infection and trauma, which has not 110 been previously reported. This serological study provides an in-depth assessment of SARS-CoV-2 111 seropositivity in trauma patients as well as detects different anti-SARS-CoV-2 antibodies with high 112 sensitivity and specificity for each patient sample. 113 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Recombinant proteins. The procedure for protein expression and production of the selected spike and 117 RBD in this study has been detailed previously in an established and available protocol 20, 21 . Briefly, 118 recombinant proteins from optimized DNA constructs (Addgene #166010 for Spike, Addgene #166019 for 119 RBD) were produced in an Expi293F mammalian expression system (Thermo Fisher Scientific). After 96 120 hours (Spike protein) or 72 hours (RBD protein) post-transfection, supernatants from transfected cells were 121 harvested by centrifugation, clarified, and subjected to tangential flow filtration (TFF) prior to purification 122 using immobilized metal affinity chromatography (IMAC). Spike proteins were desalted, and RBD proteins 123 were further purified by size exclusion chromatography. Specific details of protein production are described 124 This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. De-identified samples and other data were included in the study under IRB-approved waivers of consent 143 and authorization. All demographic information was obtained from medical records or other secondary 144 sources such as emergency medical services run reports and crash reports. De-identified samples were then 145 sent to NIH for SARS-CoV-2 ELISA testing on dry ice overnight and stored at -80 o C until processing. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. ; https://doi.org/10.1101/2021.08.10.21261849 doi: medRxiv preprint dilution or control was added in technical duplicate into the plates and incubated for 1 hour at room 164 temperature. After sample incubation, plates were washed three times with 300 μL of PBS-T per well. Then, 165 goat anti-human IgA, IgM, and IgG horseradish peroxidase (HRP) secondary antibodies (ThermoFisher) 166 were diluted at 1:4000 in blocking buffer and 100 μL of each secondary antibody solution was added to 167 each well for 1 hour. Plates were again washed three times with PBS-T, then incubated with 100 μL of 1-168 Step™ Ultra TMB-ELISA Substrate Solution (ThermoFisher) for 10 minutes followed by 100 μL of 1 N 169 sulfuric acid STOP Solution (ThermoFisher). Within 30 minutes after adding STOP solution, optical 170 density (OD) was measured at 450 and 650 nm using BioTek Epoch2 plate reader. To remove background, 171 the actual absorbance was calculated as the difference between OD at 450 nm and at 650 nm before further 172 statistical analysis. 173 174 Statistical analysis. Seropositivity was defined as either IgG or IgM OD levels above their respective 175 thresholds for both the spike and RBD expression. Using both spike and RBD expression together increased 176 sensitivity and specificity to 100% for both IgG and IgM based on evaluation with convalescent positive 177 and archival negative controls 18, 19 . The method to determine thresholds was detailed previously using 178 simulations of different samples and control size to model the statistical confidence over a range of disease 179 prevalence and assay specificity. The threshold was determined as previously reported to ensure that the 180 lower 95% confidence limit of specificity is greater than 99%. Exact binomial methods were used to 181 compare seroprevalence between population subgroups. Multiple comparisons were corrected for using the 182 Bonferroni method. To evaluate the association between drug exposure (each drug separately, drug classes, 183 and any drug positivity overall) and SARS-CoV-2 serostatus, multivariable penalized likelihood logistic 184 regression was used, adjusting for age of the trauma patient, sex, race, ethnicity, emergency room admission 185 month, and the admission city among individuals whose samples were tested for the presence of drugs and 186 alcohol. Analysis was done using the "logistf" package version 1.24 in R version 4.0.4 27, 28 . 187 188 for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. population within these study sites (Fig. 1b). When compared to the general US population this trauma 202 population is in general younger and contains more males and more non-white study participants, though 203 this population is specific to the service areas within the bounds of the six trauma centers. Within the sample 204 populations we were able to identify antibodies against the SARS-CoV-2 full spike ectodomain (spike) and 205 spike receptor binding domain (RBD) with IgG, IgM and IgA classes (Fig. 1c-e). Daily measurements of 206 seropositivity and samples collected are displayed in Fig. 1f, with monthly estimates in Fig. 1g. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. ; https://doi.org/10.1101/2021.08.10.21261849 doi: medRxiv preprint were convalescent and most likely outside of the window of when they were most infectious; however, over 240 40% of these participants were pre-convalescent (seronegative) suggesting early stages of disease which is 241 associated with higher viral loads. 7.55 %) (Fig. 4). When calculating seroprevalence point estimates, "other motorized transport injuries" 249 were grouped with MVC's as road traffic injuries (RTI), while drowning and fire/burn injuries were 250 categorized as "other" (Fig. 4a) For further analysis, these drugs were classified in larger groupings as stimulants, narcotics or sedatives, 263 anti-depressants, and others classification of drugs (Fig. 5). 264 265 for use under a CC0 license.
This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. ; https://doi.org/10.1101/2021.08.10.21261849 doi: medRxiv preprint Only two of the individual drugs tested were significantly associated with SARS-CoV-2 seropositivity after 266 controlling for potential confounders. Samples tested positive for Lorazepam -belongs to a class of drugs 267 known as benzodiazepines -were associated with an increased likelihood of being SARS-CoV-2 268 seropositive (Odds Ratio (OR): 8.14, 95% CI: 1.21 -45.0, p = 0.03). Meanwhile, samples tested positive 269 for fentanyl, a synthetic opioid class, were associated with a decreased likelihood of being SARS-CoV-2 270 seropositive (OR: 0.25, 95% CI: 0.03 -0.95, p = 0.04). Narcotics or sedatives as a category were also 271 negatively associated with SARS-CoV-2 seropositivity (OR: 0.56, 95% CI: 0.34 -0.90) (Fig. 6) This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. While it is difficult to compare the relative risk of SARS-CoV-2 seropositivity between the general 298 population and the trauma population due to differences in donor recruitment and study design, we did note 299 that in comparison to a national study conducted by our group using the same seroassays over the same 300 time period, cities in the south/central region of the United States had higher SARS-CoV-2 prevalence in 301 trauma patients comparison to the general population. As the samples represented this trauma population 302 were obtained from only six trauma centers and not all Level-1 trauma facilities throughout the United 303 States, the study cannot be used to infer seroprevalence in the overall trauma population of the United 304 States. In addition, one of six sites, Iowa City, had a slightly delayed collection timeframe in comparison 305 to the other five. Therefore, further investigation is necessary to understand the seroprevalence at each 306 trauma center in comparison to its region. 307 308 Published seroprevalence estimates of the general population of Massachusetts during the same time period 309 as this study showed a lower prevalence (4.0%) in comparison to the trauma population sampled from 310 Worcester, MA (7.72%) 39 . Another study in Miami found that during the spring/summer of 2020 the general 311 population had a lower seroprevalence of 6.0% compared to the trauma population at 12.17% 41 . A study in 312 central North Carolina found increasing seroprevalence in the general community from 2.9% to 9.1% from 313 April through October. Our estimate from Charlotte, NC which was gathered in July at 7.28%, could 314 suggest a higher than state-average rate of SARS-CoV-2 seroprevalence in the trauma population, given 315 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Among motor vehicle crash victims specifically, a large proportion of trauma victims were positive for 319 drugs or alcohol. There was a lower likelihood that these individuals had a prior SARS-CoV-2 infection if 320 they were positive for narcotics or sedatives (including marijuana). Interestingly, there was a positive 321 correlation with the depressant Lorazepam and SARS-CoV-2 seropositivity; this medication induces 322 anxiolysis and sedation. In addition lorazepam can worsen obstructive pulmonary disease and lead to 323 respiratory compromise 42 . Whether seropositivity among specific drug exposed individuals is due to 324 chemical activity of the drug or alterations in behavior of those using these drugs remains to be determined, 325 as do the implications of these relationships for first responders and other medical professionals needing to 326 treat trauma patients under the influence of certain drugs. As the stimulant drug class trends higher 327 seropositivity than THC, alcohol, benzodiazepines, or narcotics, and methamphetamine use is correlated 328 with increased risk-taking behavior, this may explain a higher trending SARS-CoV-2 exposure. 329 Additionally, patients using stimulants such as methamphetamine often present in the emergency room with 330 excited delirium, spitting, and physical aggression can lead to breakdown in PPE protocols for healthcare 331 providers. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. ; https://doi.org/10.1101/2021.08.10.21261849 doi: medRxiv preprint In this study, we have shown that differences in SARS-CoV-2 seroprevalence among trauma patients are, 342 as with the general population, correlated with region, race, ethnicity, and age. There are also correlations 343 This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. ; https://doi.org/10.1101/2021.08.10.21261849 doi: medRxiv preprint under contract number HHSN261200800001E. The content of this publication does not necessarily reflect 366 the views or policies of the Department of Health and Human Services, nor does mention of trade names, 367 commercial products, or organizations imply endorsement by the U.S. Government. Disclaimer: The NIH, 368 its officers, and employees do not recommend or endorse any company, product, or service. 369 for use under a CC0 license. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint this version posted August 11, 2021.  This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.