key: cord-0890328-6g7mijbz
authors: Mirza, Fatima N; Malik, Amyn A.; Omer, Saad B.
title: Influenza-Negative Influenza-Like Illness (fnILI) Z-Score as a Proxy for Incidence and Mortality of COVID-19
date: 2020-04-27
journal: nan
DOI: 10.1101/2020.04.22.20075770
sha: de8f856a881abb12fc231283574be8fe125b9770
doc_id: 890328
cord_uid: 6g7mijbz

Though ideal for determining the burden of disease, SARS-CoV2 test shortages preclude its implementation as a robust surveillance system in the US. We correlated the use of the derivative influenza-negative influenza-like illness (fnILI) z-score from the CDC as a proxy for incident cases and disease-specific deaths. For every unit increase of fnILI z-score, the number of cases increased by 70.2 (95%CI[5.1,135.3]) and number of deaths increased by 2.1 (95%CI[1.0,3.2]). FnILI data may serve as an accurate outcome measurement to track the spread of the and allow for informed and timely decision-making on public health interventions.

CC-BY-NC 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 April 27, 2020 . . https://doi.org/10.1101 2 Abstract 28 29 Though ideal for determining the burden of disease, SARS-CoV2 test shortages preclude its 30 implementation as a robust surveillance system in the US. We correlated the use of the derivative 31 influenza-negative influenza-like illness (fnILI) z-score from the CDC as a proxy for incident 32 cases and disease-specific deaths. For every unit increase of fnILI z-score, the number of cases 33 increased by 70.2 (95%CI[5.1, 135.3] ) and number of deaths increased by 2.1 (95%CI[1.0,3.2]). 34 FnILI data may serve as an accurate outcome measurement to track the spread of the and allow 35 for informed and timely decision-making on public health interventions. CC-BY-NC 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 (which was not certified by peer review)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 April 27, 2020 . . https://doi.org/10.1101 3 Background 52 53 Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has spread exponentially since 54 December 2019, transforming from a localized outbreak in Wuhan, China to a global pandemic.

As of mid-April, roughly 2 million cases have been reported(1), with one-fifth of cases requiring 56 inpatient hospitalization and a case-fatality at an estimated 2%(2).

As the number of individuals infected rapidly climbs, the testing capacity has been outpaced by 59 the need for such tests. In the United States, this has posed a challenge to physicians and public 60 health professionals at large, particularly as it relates to accurately tracking the spread of disease.

Assessing the intensity of the epidemic nationally in a given region is the backbone of allocating 62 resources at the federal and state level, and inform the implementation or relaxing of public 63 health restrictions (e.g. initiating or easing a lockdown).

Given the rapid increase in cases in the previous weeks without parallel expansion in testing 66 capacity and unclear specificity/sensitivity, this problem will only continue to be exacerbated 67 until a nationwide program is made available and further validation studies have been completed 68 (3). In the interim, there is an urgent need to identify proxies for disease incidence that are 69 routinely collected through available infrastructure in the United States in order to guide the 70 evolving public health response in this country(4).

The Centers for Disease Control and Prevention (CDC) centrally collates data using the U.S.

Outpatient Influenza-like Illness Surveillance Network (ILINet) and the National Respiratory 74 and Enteric Virus Surveillance System (NREVSS) (5). We believe that combining both sources 75 . CC-BY-NC 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 April 27, 2020. . https://doi.org/10. 1101 4 of this publicly-available, routinely collected data may serve as a reliable proxy for SARS-CoV-76 2 incidence and mortality. In this study, we used influenza-negative ILI (fnILI) z-scores and 77 compared them against the reported COVID-19 cases and deaths by week to document trends 78 over time. Reich et al (6) reviewed twenty-three seasons of influenza data, beginning in 1997, and ten 97 seasons of statewide data, beginning in 2010 and calculated fnILI from the CDC (5). fnILI was 98 determined using weighted influenza-like illness (wILI) from ILINet -which represents the 99 . CC-BY-NC 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 April 27, 2020. These data included a z-score that represents the degree to which a given fnILI observation was 105 significantly lower or higher than expected based on past trends at similar times during prior 106 years. Z-score was calculated as: We merged this dataset with the CDC-reported SARS-CoV-2 cases and disease-specific deaths.

We graphically represented the fnILI z-score, cumulative cases, and cumulative deaths for the ke . CC-BY-NC 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 April 27, 2020. . https://doi.org/10.1101/2020.04.22.20075770 doi: medRxiv preprint 6 plotted against total nationwide cases and/or deaths per week. All data was analyzed using Stata There is an apparent tracking between fnILI z-score and cases or deaths by state over the course 132 of the study period. This phenomenon is particularly pronounced when comparing these indices 133 during the month of March ( Figure 1A ).

When assessing the correlation over time between fnILI z-score and either new cases or deaths, 136 we observed a z-score peak prior to an increase in cases or deaths. Therefore, we used a lag 137 variable of two weeks for incidence and one week for mortality to better fit the model (Table 1) .

On the mixed effects linear model accounting for clustering at the state level using random 139 effects, we found that for every unit increase of fnILI z-score two weeks prior, the number of 140 cases increased by 70.2 (95% CI [5.1, 135.3] ). Similarly, we found that for every unit increase of 141 fnILI z-score one week prior, the number of deaths increased by 2.1 (95% CI [1.0, 3.2]), also 142 when correcting for regional effects. When plotting the median nation-wide z-score two week or 143 . CC-BY-NC 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 April 27, 2020. There is increasing evidence that current testing may significantly underestimate the burden of 150 disease (8), necessitating alternative methods to accurately assess these trends over time. Our 151 results suggest that the fnILI z-score data can be used as a proxy for the trajectory of disease 152 incidence and mortality in the United States. In the context of limited resources in a rapidly 153 changing field, it becomes increasingly necessary to innovatively utilize available infrastructure 154 to tackle the apparent gaps in knowledge quickly. To our knowledge, this is the first academic 155 study to use fnILI z-scores from ILINet and NVRESS data in order to model and potentially 156 predict the burden of COVID-19 over time.

This report demonstrates the important potential of such a proxy, and validates its correlation 159 with incidence and mortality. Importantly, we present the optimal model for such a prediction by 160 building in a lag term. This two week lag term is likely necessary for incidence due to the known 161 incubation period of this disease and because of a delay in testing (2). For mortality, our one 162 week lag term likely represents the rapid escalation to fatality (9). These lag terms also may 163 allow our model to function as an early warning system for rise in cases, similar to ILINet.

As there is already a robust infrastructure in place to collect these data, validating the use of such 166 data is extremely valuable, especially in the setting of limited availability and capacity of testing 167 . CC-BY-NC 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 April 27, 2020. CC-BY-NC 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 April 27, 2020 . . https://doi.org/10.1101 Disclosures: The authors have no conflicts of interests to disclose.

Funding/Support: There is no funding for this study. CC-BY-NC 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 April 27, 2020 . . https://doi.org/10.1101 

An interactive web-based dashboard to track COVID-19 in real 217 time. The Lancet Infectious Diseases

Clinical Characteristics of Coronavirus 219 Disease 2019 in China

COVID-19) Current Status and Future Perspectives: A Narrative 222

On the responsible use of digital data to tackle the COVID-19 225 pandemic

Influenza Surveillance System: Purpose and Methods | CDC

COVID-19 in the United States from influenza-like illness data. Github. 2020

Information criteria and statistical modeling

Estimation of the Risk of Death from Novel Coronavirus (COVID-19) Infection: 237 Inference Using Exported Cases. J Clin Med. 2020 Feb 14;9(2):523.

. CC-BY-NC 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 April 27, 2020. 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 April 27, 2020. . CC-BY-NC 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 April 27, 2020.