key: cord-327661-osx42wdh
authors: Schaefer, E. J.; Geller, A. S.; Diffenderfer, M. R.; Dulipsingh, L.; Wisotzkey, J.; Kleiboeker, S. B.
title: Coronavirus Disease-2019 Case, Death, and Testing Rates in the United States and Worldwide: Primary Data and Review
date: 2020-10-14
journal: nan
DOI: 10.1101/2020.10.13.20172957
sha: 
doc_id: 327661
cord_uid: osx42wdh

ABSTRACT Coronavirus disease-2019 (COVID-19), due to the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), has been associated with a world-wide pandemic, with the United States (US) having the largest total number of cases and deaths (>7 million and >200,000, respectively) at this time. We assessed data as of September 1, 2020 from our combined laboratories and as reported for selected states and countries for case, death, and testing rates per 1 million in the population. Our goal was to elucidate potential causes for the large rate differences observed. SARS-CoV-2 naso-pharyngeal (NP) RNA swab testing in 985,219 US subjects referred to our laboratories by healthcare providers revealed an overall 10.1% positive rate, comparable to the 7.3% rate reported nationwide. In a small subset of 91 subjects, all of whom had been positive for SARS-CoV-2 RNA in NP swabs 2-4 weeks earlier, NP swab testing was twice as likely to be positive (58.6%) as saliva samples (21.5%), based on paired sampling. Our positive rates per state agreed reasonably well with reported Centers for Disease Control and Prevention (CDC) data (r=0.609, P<0.0001) based on 19,898 cases, 593 deaths, and 271,637 tests, all per 1 million in the US population. Louisiana had the highest case rate; New Jersey had the highest death rate; and Rhode Island had the highest testing rate. Of 47 countries, including all countries with populations >50 million, Qatar had the highest case rate; Peru had the highest death rate; and Israel had the highest testing rate for SARS-CoV-2 infection. Correlations between case rates and death rates as well as testing rates were 0.473 and 0.398 for US states and 0.473 and 0.476 for the various countries, respectively (all P<0.0001). In conclusion, outpatient saliva testing is not as sensitive as NP testing for SARS-CoV-2 RNA detection. While testing is important, without adequate public health measures, it is unlikely that we will get this pandemic under adequate control until vaccines become available.

Coronavirus disease-2019 (COVID-19) is caused by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), which is now associated with a world-wide pandemic (>30 million cases and >1 million deaths). The diagnosis is made by SARS-CoV-2 RNA detection in naso-pharyngeal (NP) swabs, nasal swabs, oro-pharyngeal (OP) swabs, or saliva. [1] [2] [3] [4] The greatest number of deaths/1 million in the population has been reported in the densely populated northeastern states of the United States (US).

Up to 50% of SARS-CoV-2 positive patients can remain symptomatic; however, such individuals can spread infections. [5] [6] [7] The average onset of symptoms after infection is about 5 days (range 2-14 days).

COVID-19 fatality is substantially higher in the elderly, and in those with cardiovascular disease, diabetes, obesity, hypertension, and lung disease.

COVID-19 disease symptoms include fever, fatigue, cough, loss of smell and taste, gastrointestinal symptoms, and shortness of breath. The virus spreads between people mainly via respiratory droplets generated by talking, singing, coughing, and/or sneezing. Severe complications include severe acute respiratory distress due to pneumonia (which may require a ventilator), and potentially death from overwhelming infection and inflammation. [1] [2] [3] [4] The virus is highly contagious, with about 50% of infected people being asymptomatic. 5, 6 While testing is critical for diagnosis and documentation of potential immunity, public health measures (e.g. face masks, shields, social distancing, and hand washing) are critical for prevention of new cases until a vaccine becomes available. Subjects that are positive for SARS-CoV-2 RNA based on NP swabs may not have transmissible virus over time, but only viral fragments. 7 Our goals were to assess our own data as well as available US and worldwide data in terms of cases, deaths, and testing per 1 million in the population, in order to examine potential causes for the large rate differences observed between states and countries.

A total of 985,219 subjects (58.2% female; age range 1-101 years; median [IQR] age 49.0 [35.0-6.0] years; 18.2% ≥65 years of age) were assessed in physician offices, clinics, and hospitals. These subjects had NP, OP, or nasal swab samples collected by healthcare providers at various sites throughout the United States, placed in viral transport media, and submitted by overnight express courier service for SARS-CoV-2 RNA detection to Boston Heart Diagnostics (Framingham, MA) beginning on April 17, 2020, Diatherix (Huntsville, AL) beginning on March 16, 2020, and/or Viracor (Lee's Summit, MO) beginning on March 13, 2020. For this analysis, data assessment was ended as of September 1, 2020. Table 1 presents data from hospitals, clinic sites, and healthcare provider offices in 38 states with more than 100 results for samples sent to these laboratories.

For this research, patient data were extracted from medical records without name or identification number and were analyzed as anonymized data. In our view, this research is exempted from requirement for human institutional review board approval as per exemption 4, as listed at https://grants.nih.gov/policy/ humansubjects.htm and at the open education resource (OER) website for research involving human subjects. This exemption "involves the collection or study of data or specimens if publicly available or recorded such that subjects cannot be identified". At the request of the preprint server medRxiv, we had this designation and our research reviewed by the Advarra Institutional Review Board (Columbia, MD) and their determination on September 26, 2020 was that "had the request for exempt determination been submitted prior to initiation of research activities, the research would have met the criteria for exemption from institutional review board review under 45 CFR 46.104(d) (4). Therefore, they agreed that this research did not require institutional review board approval.

The paired NP swab and saliva samples were collected from previously positive convalescent plasma donors (n=91, mean age 53 years, 53% female) using a protocol and written informed consent form approved by the human institutional review board of Trinity Health of New England (Hartford, CT).

These subjects had all been positive 2-4 weeks earlier for SARS-CoV-2 RNA based on NP swabs. The . 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)

The copyright holder for this preprint this version posted October 14, 2020. . 6 saliva was collected in viral transport media kits obtained from Strategic Laboratory Partners (Nashville, TN).

We also examined Centers for Disease Control and Prevention (CDC) data from various states for comparison. We included Rhode Island because of the high case and death rates observed for this state.

To obtain information on COVID-19 data for the states in the US and world-wide data, we accessed the CDC website on September 1, 2020, https://www.cdc.gov/covid-data-tracker/index.html#testing, as well as the website https://www.worldometers.info/coronavirus/? and the World Health Organization website https://www.covid19.who.int/.

Detection of SARS-CoV-2 RNA in NP, OP, nasal swabs or saliva was performed using reverse transcriptase polymerase chain reaction methods as previously described for the Viracor assay. 8 Viracor was among the first US laboratories to offer SARS-CoV-2 RNA testing as of March 13, 2020. Viral detection and quantification in plasma was performed using a reverse transcriptase polymerase chain reaction (RT-PCR) method which targets two regions of the SARS-CoV-2 nucleo-capsid (N) gene using TaqMan chemistry. This assay is a modification of an assay approved by the Food and Drug Administration (FDA) under emergency use authorization (EUA) guidelines for qualitative detection of SARS-CoV-2 in respiratory samples. Nucleic acid extraction was performed using the ThermoFisher KingFisher FLEX instrument with MagMAX Viral/Pathogen Nucleic Acid Isolation reagents.

Amplification was performed using the Applied Biosystems™ 7500 Fast Real-Time PCR Systems (SDS v1.5.1) with TaqPath 1-step RT-qPCR master mix CG reagents. A total of 15 µL nucleic acid was amplified in a 30 µL reaction volume. The thermocycling protocol was as follows: 25˚C for 2:00 (1 cycle), 50˚C for 15:00 (1 cycle), 95˚C for 2:00 (1 cycle), 95˚C for 0:03, 60C for 0:30 (45 cycles). For quantification, a standard curve was used by amplifying 10-fold serial dilutions (75 copies/mL to 7.5 x 10 7 copies/mL) of in vitro transcribed RNA prepared from the full SARS-CoV-2 N gene. The copies/mL value for each sample which generated a cycle threshold value was calculated using the slope and y-. 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. . 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 preprint this version posted October 14, 2020. .

Fluorescence signals are analyzed on the SensoSpot Fluorescence Low Density Microarray Analyzer (Sensovation AG, Radolfzell, Germany. The RNA assays in all three laboratories have received EUA approval from the FDA and have excellent sensitivity, reproducibility, and reliability.

All statistical analyses were performed using R software, version 3.6.0 (R Foundation, Vienna, Austria) for comparisons between rates, and the statistical significance of differences between groups was assessed using the nonparametric Kruskal-Wallis method. Pearson correlation analysis was also performed using R.

. 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 preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20172957 doi: medRxiv preprint

Of 985,219 subjects having NP, OP, or nasal swabs done between March 13 and September 1, 2020 at various sites in 43 states, 10.1% were positive for SARS-CoV-2 RNA. The data from 38 states in which at least 100 swabs were submitted for testing are presented in Table 1 . When we first tabulated these data on June 1, 2020, New York State had by far the highest percentage of positive subjects (43.5%); this rate decreased to 4.6%, when we included 104,195 tests done for health screening and for nursing home employees who had much lower positive rates of <4.0%. When a subset of 91 subjects, who served as convalescent plasma donors and had been positive for SARS-CoV-2 RNA based on NP swabs 2-4 weeks earlier, had repeat NP swabs and saliva collected, 58.6% were still positive based on NP swabs; but only 21.5% were positive based on saliva collection. These differences were statistically significant P<0.01). In our total study population, 18.2% were ≥65 years of age, of whom 6.9% were positive compared to 10.5% in the <65-year age group. Therefore, in the population we tested, older people did not have a higher positivity rate; in fact, it was lower (P<0.0001), even though it has been well documented that elderly subjects have a significantly higher case fatality rate than younger subjects.

When we compared our data with CDC state-wide data, for the 38 states where we had >100 cases/state, we noted a significant correlation for the percentage of positives (r=0.609, P<0.0001). We added Rhode Island to the data set because it is in the very high mortality, densely-populated northeast corridor of the US. As of September 1. 2020, the top ten states in the United States for cases/1 million were Louisiana, Mississippi, Arizona, Alabama, Georgia, Tennessee, South Carolina, Texas, New York, and Iowa (Table 1) . However, the top ten US states in terms of deaths/1 million were New Jersey, New York, Massachusetts, Connecticut, Louisiana, Rhode Island, Mississippi, Arizona, Michigan, and Illinois.

In terms of testing per 1 million, the top ten states were Rhode Island, New York, Louisiana, Connecticut, Illinois, Massachusetts, New Jersey, Tennessee, Michigan, and California.

In Figure 1 , we have plotted the relationship between death and case rates per million by state (Panel A). These data are based on 19,898 cases/1 million and 593 deaths/1million in the US as of September 1, 2020, as well as 271,637 tests/1 million in the US population. As can be clearly seen in . 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 preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20172957 doi: medRxiv preprint Figure 1 , the northeastern states of New Jersey, New York, Massachusetts, Connecticut, and Rhode Island (red circles) had the highest death rates per case, while the southern states (orange circles) of Louisiana, Mississippi, Alabama, and Georgia, and one western state Arizona had higher case rates, but generally much lower mortality rates. Western states (blue circles), especially Oregon, Wyoming, Washington, Colorado, Utah, and California, had the lowest mortality rates and also relatively low case rates. The overall correlation between deaths/1 million and cases/1million was 0.473 (P<0.0001). In Panel B, we plotted tests/1 million versus cases/1 million, and here we see a similar positive relationship with a correlation coefficient of 0.398 (P<0.0001). Clearly higher testing rates are not associated with lower case rates, but in fact, with higher case rates. Louisiana had the highest case rate as well as a high testing rate, while the converse was true for Oregon. Overall, the highest testing rates were observed in Rhode Island, New York, and Louisiana, all states with among the highest case rates.

We carried out a similar analysis for all countries in the world with populations >50 million, as well as other selected countries. This analysis included 47 countries. Of these, 3 countries were in Asia and Oceania; 11 countries were in Europe; 5 countries were in North America; 7 countries were in South America; 7 countries were in the Middle East; and 4 countries were in Africa ( Table 2) . As for US states, we also plotted these relationships in Panel A and B of Figure 2 . The correlations between cases/1 million and deaths/1 million were similar to US states at 0.488 (P<0.0001), as were the correlations between cases/1 million and tests/1 million at 0.395 (P<0.0001). However, these relationships had to be plotted on log scales instead of linear scales because of the very large variability in rates between countries.

Countries with the highest death rates generally also had the highest case rates and testing rates. These 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 October 14, 2020. .

populous Asian countries (India, Bangladesh, Pakistan, and the Philippines) had intermediate case rates, as well as fairly low death and testing rates. Similar relationships between case rates and testing rates were observed. As for US states, high testing rates were generally associated with high case rates.

. 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 preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20172957 doi: medRxiv preprint

The rapid spread of SARS-CoV-2 viral infection has been in large part due to its very contagious nature and the fact that many infected people are asymptomatic. Although the pandemic started in China, its spread there as well as in other countries in Asia, such as Thailand, Vietnam, Myanmar, South Korea, and Japan, has been quite well controlled. In these countries case rates have been <100/1 million and death rates have been <10/1 million. One cannot attribute this excellent infection control to testing, but rather to excellent public health measures including isolation of cases, contact tracing, social distancing, and the wearing of face masks. It is interesting that even in countries such as India, Bangladesh, the Philippines, and Australia, all with >1,000 cases/1 million, death rates have been <60/1 million. In contrast, subjects in Europe, North America, and South America have fared far worse with case rates generally >5,000/1 million and death rates >500/1 million, despite a large amount of testing. In the northeastern and southern US states, case rates have generally exceeded 15,000/1 million; and death rates have been >1,000/1 million, despite a lot of testing.

Rates in the United States (19,898 cases and 593 deaths per million) and Brazil (19,919 cases and 609 deaths per million) were comparable in cases and deaths, due to limited public health measures. In contrast, Japan (579 cases and 11 deaths per million), South Korea (430 cases and 7 deaths per million), China (59 cases and 3 deaths per million), and Thailand (50 cases and < 1 death per million) had much lower rates, presumably due to significantly better public health measures.

Many European countries had case rates lower than the United States, but comparable death rates. The importance of public health measures may best be exemplified by comparing Sweden, which did not introduce such measures, to its neighbors Norway and Finland, which did introduce such measures.

Sweden had case and death rates of 8,555/1 million and 578/1 million, respectively, while Norway and Finland had case and death rates of <2,500/1 million and <65/1 million, respectively. These great differences occurred despite the fact that these countries had fairly equivalent testing rates.

There may be some other potential causes of the large variability in case and death rates between countries. One such possibility is mutations in the virus. The D614G amino acid substitution in the S . 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 preprint this version posted October 14, 2020. .

glycoprotein encoded by the SARS-CoV-2 S gene has been reported to result in a form of the virus that is more infective and virulent than the original Wuhan strain. [9] [10] [11] [12] [13] This variant has been found in over 90% of United States strains. Another possibility is human genetic variation. Genome-wide association studies have identified a 3p21.31 DNA locus as being associated with a significant 1.77-fold increased risk for respiratory failure in hospitalized COVID-19 patients. 14 This genetic variant, apparently inherited from Neanderthals, is not present in subjects indigenous to China, Japan, or Sub-Saharan Africa, and has a frequency of ~5-8% in North America and western Europe and ~20% in India. 15 Therefore, both viral and human genetic variation may play a role in country differences with regard to SARS-CoV-2 cases and mortality rates.

Recently, there has been a significant increase in SARS-CoV-2 RNA testing in the United States in an effort to control the pandemic in the absence of vaccines. There have also been efforts to find easier ways to carry out such testing. Despite data from Yale New Haven Medical Center indicating that selfcollected saliva samples yielded similar or better results than did NP swabs in terms of detecting SARS-CoV-2 positive hospitalized patients based on 44 paired samples, our data in previously positive outpatients did not support these findings. Our data indicated that saliva analysis only found about half as many positive cases than did NP swab analysis. Moreover, data has indicated that rapid antigen testing for SARS-CoV-2 protein is not nearly as sensitive or accurate as RNA testing. Our overall data strongly supports the benefits of public health measures in preventing spread of SARS-CoV-2 infection. In our view the major reason for the very high cases and mortality rates in the United States has been the consistent lack of such measures, due to failures of government and public agency leadership.

. 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 preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20172957 doi: medRxiv preprint

Our data indicate that 1) outpatient saliva testing is not as sensitive as NP testing in detecting SARS-CoV-2 RNA; 2) the marked variability in the case and death rates between states and countries is due mainly to differences in public health measures; 3) variation in SARS-CoV-2 genetics and human genetics may also play a role in such differences; and 4) these difference are least likely to be due to lack of adequate 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.

The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20172957 doi: medRxiv preprint

We thank the medical personnel and technical staff at each institution for their effort and commitment to SARS-CoV-2 testing. We also thank Shannon Foster of Viracor-Eurofins Clinical Diagnostics and Scientific Network for compiling all RNA data. All information was anonymized prior to data analysis.

All relevant ethical guidelines were followed in carrying out this research. . 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 preprint this version posted October 14, 2020. . . 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 preprint this version posted October 14, 2020. . . 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 preprint this version posted October 14, 2020. . . 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 preprint this version posted October 14, 2020. 1M, 1 million people.

. 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 preprint this version posted October 14, 2020. . . 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 preprint this version posted October 14, 2020. . Data as reported on the following websites: https://www.who.org and www.worldometersinfo/coronavirus.

. 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 preprint this version posted October 14, 2020. . . 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)

The copyright holder for this preprint this version posted October 14, 2020. . https://doi.org/10.1101/2020.10.13.20172957 doi: medRxiv preprint

A pneumonia outbreak associated with a new coronavirus of probable bat origin

A Novel Coronavirus from Patients with Pneumonia in China

Saliva is more sensitive for SARS-CoV-2 detection in COVID-19 patients than nasopharyngeal swabs

Pathophysiology, transmission, diagnosis, and treatment of Coronavirus Disease 2019 (COVID-19): A Review

Transmission of 2019-nCoV Infection from an Asymptomatic Contact in Germany

Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship

Temporal profiles of viral load in posterior oropharyngeal saliva samples and serum antibody responses during infection by SARS-CoV-2: an observational cohort study

SARS-CoV-2 viral load assessment in respiratory samples

Emergence of drift variants that may affect COVID-19 vaccine development and antibody treatment