Skip to main content


 Last update: January 11, 2021

The Covid-19 pandemic has impacted societies in ways we are only beginning to understand. Like many other researchers, I tried to help bring understanding to the crisis while it unfolded. My work has focused on how to best make sense of the overwhelming amounts of data that are continuously being publicized about the state of pandemic.

In April 2020, we published a short research note titled “Evaluating Interventions on the Spread of Corona” arguing that comparative Google trends results could be used to quickly get a sense of the causal effect of new restrictions. Since speed was of the essence, we argued that search queries would allow faster evaluation than e.g. PCR tests. The stricter restrictions imposed in Norway had an immediate effect on the number of symtom queries.

Around the same time, I started publishing a graphical dashboard of the Swedish death count. The Public Health Agency of Sweded publishes statistics by actual date of death. Most other countries only report newly added deaths, without information about when each person died. Since it often takes some time for a death to be identified, this can be misleading. However, including death dates creates new problems. Mainly, the reporting delay means that the last few days are still not completed, which creates a false sense of an always decreasing trend. To account for this I added a graphical illustration of expected additional deaths. This later grew into a full-fledged nowcasting model documented in the paper below.

Covid-19 deaths in Sweden

The Covid-19 dashboard is available at

Work in Progress

  • Nowcasting Covid-19 statistics reported with delay: a case-study of Sweden

    Under review at Scientific Reports

    with Joacim Rocklöv and Jonas Wallin

    The new corona virus disease - COVID-2019 - is rapidly spreading through the world. The availability of unbiased timely statistics of trends in disease events are a key to effective responses. But due to reporting delays, the most recently reported numbers are frequently underestimating of the total number of infections, hospitalizations and deaths creating an illusion of a downward trend. Here we describe a statistical methodology for predicting true daily quantities and their uncertainty, estimated using historical reporting delays. The methodology takes into account the observed distribution pattern of the lag. It is derived from the removal method, a well-established estimation framework in the field of ecology.