K. Miller, M. Gahan, C. Lam, M. Gallivan, V. Serhiyenko, N. Stephenson, N. MadhavMetabiota, Inc., San Francisco, United States

Correspondence: kmiller@metabiota.com

Simulating worker absenteeism by economic sector during influenza pandemics and implications for economic impact

Indirect costs of infectious disease epidemics are often far greater than financial losses associated with direct costs, such as treatment and response costs

Worker absenteeism:

Large driver of outbreak-related economic lossReduces productivity and precipitates business failuresOften neglected due to quantification uncertainties and limited observational data

Enterprise Risk Management should include preparations for this potential loss

Figure 1. Simulated absenteeism and incidence by week

Methods

Tool developed to estimate risk of worker absenteeism rates during influenza pandemics

Statistical model constructed to predict absenteeism rates using CDC monthly lab-confirmed flu incidence/mortality and national survey data on worker absence (BLS)

Industry-specific data were analysed to estimate the relative contribution to absenteeism by economic sector Absenteeism estimated as a function of Incidence, Mortality, and Sentiment

Fitted absenteeism model is applied to global outbreak simulations to estimate temporal dynamics of pandemic-induced absenteeism rates

Figure 2. Employed people who missed work due to illness, injury, or medical problem (BLS)

Introduction

Results & Conclusions

Simulation based on 2009 H1N1 pandemic produces 4% absenteeism (US) during peak epi-week, comparable to historical reports

Simulation based on 1918-like pandemic produces 27% absenteeism (US) during peak epi-week, though comparable historical reports unavailable

Absenteeism trends observed in reported data by economic sector

Education and health sectors demonstrate highest absenteeism rates

Methods can be expanded to additional locations across the globe with cultural trends in absenteeism and sentiment

Absenteeism models can help quantification of potential indirect economic costs of epidemics

Insurance mechanisms can be created based on such probabilistic models which inform likelihood of reaching an established threshold of absenteeism prior to payout

Figure 3. Simulated absenteeism by sector

References

Ahn & Yelowitz (2016) Paid Sick Leave and Absenteeism: The First Evidence from the U.S. https://mpra.ub.uni-muenchen.de/69794/1/MPRA_paper_69794.pdf An, et al. (2007) Method and system for estimating dynamics of workforce absenteeism using information on pandemic spread and mitigation actions https://www.google.com/patents/US20080177614 De Blasio, et al (2012) Estimating influenza-related sick leave in Norway: Was work absenteeism higher during the 2009 A(H1N1) pandemic compared to seasonal epidemics? http://www.eurosurveillance.org/ViewArticle.aspx?ArticleId=20246 Dionne & Dostie (2007) New Evidence on the Determinants of Absenteeism Using Linked Employer-Employee Data https://www.jstor.org/stable/25249126 Groenewold et al (2015) Exploring National Surveillance for Health-Related Workplace Absenteeism: Lessons Learned From the 2009 Influenza A Pandemic https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4552194/ Hafner et al. (2015) Health, wellbeing and productivity in the workplace (RAND) https://www.rand.org/content/dam/rand/pubs/research_reports/RR1000/RR1084/RAND_RR1084.pdf Henderson et al (2009) Public Health and Medical Responses to the 1957-58 Influenza Pandemic http://online.liebertpub.com/doi/pdfplus/10.1089/bsp.2009.0729?cook=& Thanner et al. (2011) Understanding estimated worker absenteeism rates during an influenza pandemic.

Figure 2. Global Distribution of Epidemic Figure 3. Variation of GDP by EPI Cluster

Ben Oppenheim1, Mark Gallivan1, Nita Madhav1, Naor Brown1, Volodymyr Serhiyenko1, Nathan Wolfe1, Patrick Ayscue1

1Metabiota, Inc., 425 California Street, 2nd Floor, San Francisco, CA 94104, United States

Global Preparedness for the Next Pandemic

Current methods for assessing country preparedness typically rely on self-assessment and focus almost exclusively on public health metrics. However, the ability to detect, communicate, and respond during epidemics have complex enabling requirements beyond the health sector. We developed an epidemic preparedness index (EPI) that rates and ranks 188 countries by ability to detect, communicate, and respond to epidemics. We incorporated data produced by international organizations, including WHO and the World Bank. We performed validation based on WHO outbreak reporting, vaccination during the 2009 influenza pandemic, and MERS case-fatality ratios.

Figure 1. Epidemic Preparedness Framework

Katrin Binder2, Julia Weinelt2, Gunther Kraut2

2Munich Re, Königinstraße 107, 80802 München, Germany

Ben Oppenheim, PhDMetabiota, Inc.Email: boppenheim@metabiota.comWebsite: www.metabiota.com

We strive to make the world more resilient to epidemics.

Epidemic Risk Insurance: Use Case for the Preparedness

Introduction

Methods

We created a weighted combination of five sub-indices (Figure 1), constructed from multiple indicators. Public health infrastructure (e.g., medical personnel) Physical and communications infrastructure (e.g., logistical capacity) Economic resources (e.g., health sector resources) Institutional capacity (e.g., political stability) Public health communication (e.g., education) Weights were derived from a multi-round, anonymous, expert Delphi process with eleven participants.

Results

EPI consists of a global relative ranking of countries on a 0-100 unit-less scale, which for further analyses were categorized into five clusters using the k-means technique Highest EPIs were identified in Western Europe, North America, Australia. Lowest EPIs were identified in Western and Central Africa, Western Asia, and Southeast Asia (Figure 2). A high variation of GDP was observed within each EPI cluster (Figure 3). Better EPI score showed statistically significant correlation with faster outbreak reporting times, greater vaccine coverage during the 2009 influenza pandemic, and lower case-fatality ratios for MERS.

Discussion

EPI considers a range of capacities beyond the health sector and removes self-reporting biases Variation in the EPI indicates opportunity to improve global preparedness The EPI can be used in infectious disease modeling studies to account for differences in country-level response. Future enhancements include adding intra-country variation and disease specificity A comprehensive understanding of epidemic risk will enable governments and organizations to make informed decisions about preparedness planning and mitigation strategies. Such strategies could include epidemic risk insurance (see box below).