key: cord-0765489-zu07vbc4
authors: Juul, F. E.; Jodal, H. C.; Barua, I.; Refsum, E.; Olsvik, O.; Helsingen, L. M.; Loberg, M.; Bretthauer, M.; Kalager, M.; Emilsson, L.
title: Mortality in Norway and Sweden before and after the Covid-19 outbreak: a cohort study
date: 2020-11-13
journal: nan
DOI: 10.1101/2020.11.11.20229708
sha: 9450a58c674715d05c152a2d8f26eb1a0389516d
doc_id: 765489
cord_uid: zu07vbc4

Objectives: Norway and Sweden are similar countries regarding ethnicity, socioeconomics and health care. To combat Covid-19, Norway implemented extensive measures such as school closures and lock-downs, while Sweden has been criticised for relaxed measures against Covid-19. We compared the effect of the different national strategies on all-cause and Covid-19 associated mortality. Design: Retrospective cohort. Setting: The countries Norway and Sweden. Participants: All inhabitants. Main outcome measures: We calculated weekly mortality rates (MR) with 95% confidence intervals (CI) per 100,000 individuals as well as mortality rate ratios (MRR) comparing the epidemic year (29th July, 2019 to 26th July, 2020) to the four preceding years (July 2015 to July 2019). We also compared Covid-19 associated deaths and mortality rates for the weeks of the epidemic in Norway and Sweden (16th March to 26th July, 2020). Results: In Norway, mortality rates were stable during the first three 12-month periods of 2015/16; 2016/17 and 2017/18 (MR 14.8 to 15.1 per 100,000), and slightly lower in the two most recent periods including during epidemic period (2018/19 and 2019/20; 14.5 per 100,000). In Sweden, all-cause mortality was stable during the first three 12-month periods of 2015/16; 2016/17 and 2017/18 (MR 17.2 to 17.5 per 100,000), but lower in the year 2018/19 immediately preceding the epidemic (16.2 per 100,000). Covid-19 associated mortality rates were 0.2 per 100,000 (95%CI 0.1 to 0.4) in Norway and 2.9 (95%CI 1.9 to 3.9) in Sweden. The increase in mortality was confined to individuals in 70 years or older. Conclusions: All-cause mortality remained unaltered in Norway. In Sweden, the observed increase in all- cause mortality during Covid-19 was partly due to a lower than expected mortality preceding the epidemic and the observed excess mortality, was followed by a lower than expected mortality after the first Covid-19 wave. This may suggest mortality displacement.

Covid-19 was declared an international health crisis by the WHO on 30 th January, and a pandemic on 11 th March, 2020 (1,2). As of 17 th September, 945,962 deaths have been officially recorded as a consequence of the disease worldwide (3) , and some claim that the death toll is even higher (4, 5) .

Assessments of the burden of Covid-19 have been hampered by lack of comprehensive data on the disease and of the benefits and harms of the measures against it (6) . Cause-specific death rates are prone to bias, especially for a disease with a high asymptomatic burden and large differences in testing and reporting of causes of death between countries. Therefore, allcause mortality trends may provide a more reliable alternative to assess burden of an epidemic in different countries and regions (4, 5) .

Norway and Sweden are similar countries regarding ethnicity, governmental and administrative systems, socioeconomics and public health care systems, and both countries have reliable, timely and complete registration of deaths (7, 8) . To combat the epidemic, Norway implemented extensive measures such as school closures and lock-downs, and reported few Covid-19 associated deaths and little severe disease. Sweden, on the other hand, reported more Covid-19 associated deaths and disease, and has been criticised for their relaxed measures against Covid-19 (9). The substantially different national strategies to control the Covid-19 epidemic in the two countries is a natural experiment that enables difference-in-difference analyses (10) to evaluate the benefits and harms associated with the epidemic and its measures.

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint Starting 12 th March, both countries strongly emphasised general measures against Covid-19; social distancing with preferred distance of at least 1 meter; hand wash and disinfection, and self-quarantine for all individuals with symptoms suggestive of Covid-19. Contact tracing, isolation and follow-up of Covid-19 transmission was strictly enforced by public health authorities in both countries (26,27).

Norway and Sweden have similar, single-payer, public health care systems with universal coverage. In each country, all residents are assigned an individually unique national registration number, which provides information on sex and date of birth and allows linkage to national registers with data on socioeconomics characteristics, health and disease, hospitalisation and death. Because reporting of death and other health-and socioeconomic variables is mandatory for all residents national registries are close to 100% complete (28) .

Since its emergence in February 2020, Covid-19 is categorised as a communicable disease with mandatory, immediate reporting of all positive cases to the Norwegian Institute of Public Health in Norway and the Public Health Agency in Sweden.

We retrieved weekly numbers of all deaths (regardless of cause) in Norway and Sweden for the last five years until the 26 th July, 2020 from Statistics Norway (SSB) (29) , the National Board of Health and Welfare (Sweden) (30) and Statistics Sweden (SCB) (31) . These registries are complete with three (Norway) and two (Sweden) weeks delay in registration of deaths. 26 th July, 2020, was the last observation in the present study. All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint

We calculated mortality for the following five 12-month periods before the endo of the observation period on July 26 th , 2020. Mondays were defined as the start of each week: The 2019/20 period includes the Covid-19 epidemic and is compared to each preceding 12month period, and to the mean of the four preceding 12-month periods, for the whole populations of Norway and Sweden. We separately analysed predefined age groups 0-69 years, 70-79 years, and ≥80 years (24) . In sensitivity analysis, we explored different weekly cut offs, but the results did not change.

All Covid-19 associated deaths, i.e. deaths among individuals with a positive Covid-19 test up to 30 days before death, until 3 rd August, 2020 in both Norway and Sweden, were retrieved from the Institute of Public Health in Norway (32) and the Public Health Agency of Sweden (33), stratified by age. We calculated proportions of Covid-19 associated deaths in each predefined age group.

We calculated weekly MR per 100,000 inhabitants for Norway and Sweden based on the total number of individuals living in Norway on 1 st January of the current year (34) and in Sweden on 31 st December the previous year (35); e.g. MR for all weeks of 2020 were based on the All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint number of individuals registered on 31 st December, 2019 in Sweden, and on 1 st January, 2020 in Norway.

We calculated the mean number of deaths and mortality rates per week, and present the lowest and the highest number of deaths throughout the 12-month periods. We compared mortality rates in the 2019/20 period to the four preceding 12-month periods (2015-2019) and calculated mortality rate ratios (MRR). In addition, we calculated all-cause mortality rates and Covid-19 associated mortality rates for the weeks of the epidemic (16 th March to 26 th July, 2020), using the total number of deaths and all Covid-19 associated deaths, respectively, and the corresponding populations of 2020. We estimated 95% confidence intervals (CI) for MRRs assuming the number of deaths to follow a Poisson distribution.

All data analysed in this report are publicly available, and no ethical approval was needed. 

No patients nor the public was involved in the design, conduct nor interpretation of this study.

The population in Norway and Sweden increased slightly from 2015 to 2020, from 5,165,802 in Norway and 9,747,355 in Sweden in 2015, to 5,367,580 in Norway and 10,327,589 in Sweden in 2020. The age distribution of the two populations was similar: 88% of the All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint inhabitants of Norway and 85% of the inhabitants of Sweden were 69 years or younger; 8% and 10% of the inhabitants were aged 70-79 years in Norway and Sweden, respectively, and 4% and 5% of the inhabitants were aged 80 years or older, respectively. The mean number of deaths per week varied between 766 (2015/16) and 790 (2016/17) in Norway, and between 1,651 (2018/19) and 1,777 (2019/20) in Sweden ( Table 2) .

In Norway, mortality rates were stable during the first three was similar to the first three 12-month periods (MRR 0.99 to 1.00) but was 7% higher than the immediately preceding period (2018/19) (MRR 1.07; 96% CI 1.06 to 1.08).

In all age groups, the mortality rates during the epidemic period (2019/20) were lower or similar to the mean of the four preceding periods in both countries (Table 3, Figure 2A -C).

For individuals 80 and older, mortality rates were highest in 2016/17 in both Norway and in Sweden, while for individuals in Norway younger than 80 years mortality rates were declining from 2015 to 2020 (Table S1 , Figures S4, S5) . In Sweden, a similar pattern is seen among younger than 70 from 2015-2019.

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. (Table S3 ). This corresponds to 18 fewer deaths in Norway and 232 more deaths in Sweden per week over an 18-week period.

Our study shows that although Covid-19 associated mortality rate was almost 15-fold higher in Sweden than in Norway during the epidemic, all-cause mortality was not higher in Sweden compared with three of the four preceding years. An increase in all-cause mortality was only observed in comparison to the immediately preceding period (2018/19), because mortality was lower than in the previous years. The excess mortality was confined to individuals older than 70 years. In contrast, mortality rates were lower than expected for all ages in Norway and individuals younger than 70 years in Sweden.

All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint At the beginning of the Covid-19 epidemic, extensive social measures were introduced in Norway in the form of restrictions and prohibitions to limit the spread of Covid-19 (13) (14) (15) (16) (17) (18) (19) , while the public health authorities in Sweden chose a much less intrusive strategy (20) (21) (22) (23) (24) (25) .

The Swedish strategy against Covid-19 has therefore received intense international attention and criticism (9), notably because reported mortality rates in Sweden have been higher than in comparable countries such as Norway. In Sweden, however, mortality was lower than expected in the months preceding the epidemic. This finding may suggest mortality displacement.

Mortality displacement (37) entails temporarily increased mortality (called excess mortality) in a population as a result of external events, such as heat waves (38) , or epidemics like influenza (39) or Covid-19. The observed temporary excess mortality likely arises because people in vulnerable groups die weeks or months earlier than they would otherwise, due to the timing and severity of the unusual external event. The excess mortality is therefore preceded or followed by periods of lower than expected mortality. The period preceding the excess mortality in Sweden during the Covid-19 epidemic, characterised by lower mortality than usual, might be due, at least partly, to a mild influenza season during the winter of 2019-20 (40) . Further, after the Covid-19 epidemic, we might see a decline in morbidity and mortality below normal levels in Sweden, as the oldest and frailest have already died. Consistent with the theory of mortality displacement and catch-up of the Covid-19 effects in Sweden. Figure 1 indeed indicates a lower than expected mortality rate in Sweden in the last weeks of follow-up in our study. It is too early, however, to reliably estimate this effect.

Mortality rates were lowest in 2019/20 for all ages in Norway. In Sweden 2019/20 was similar to previous year (2018/19) and lower than preceding periods for individuals younger All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint than 70 years. This may be due to successful low-intensity and low-cost measures against Covid-19 such as social distancing, personal hygiene, self-quarantine for individuals with symptoms, and contact tracing in both countries. The rate of registered infectious diseases other than Covid-19 (such as E. coli gastroenteritis and N. gonorrhoea infections) has indeed decreased (41) .

Societal burdens related to the measures against Covid-19 were substantially higher in Norway than in Sweden (Table 1) The fight against Covid-19 required extensive health care resources (42, 43) , and studies from other countries indicate that morbidity and mortality from causes other than Covid-19 have changed after the outbreak of the epidemic (44) . A limitation of this study is that we only examined all-cause and Covid-19 associated mortality, and no other specific causes of death, as this information is not yet available for the study period. A Swedish study reviewing the medical records of 122 people (a total of 51% of all deaths in the study-region) found that 70% of the deaths were Covid-19 associated, however only 15% of the cases were indeed Covid-19 specific (45) . In addition, the observational study design does not provide an All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint opportunity to draw conclusions about the causal relationship between the social measures in Norway and Sweden and the two countries' differences in mortality rates.

Our study shows that all-cause mortality was largely unchanged during the epidemic as compared to the previous four years in Norway and Sweden, two countries which employed very different strategies against the epidemic. Excess mortality from Covid-19 may be less pronounced than previously perceived in Sweden, and mortality displacement might explain part of the observed findings. We hope that these findings can pave the way for a less Transparency declaration: Dr Frederik E Juul affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

License for publication: Dr Frederik E Juul has the right to grant on behalf of all authors and does grant on behalf of all authors, a worldwide licence to the Publishers and its licensees in perpetuity, in all forms, formats and media (whether known now or created in the future), to i) All rights reserved. No reuse allowed without permission. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint publish, reproduce, distribute, display and store the Contribution, ii) translate the Contribution into other languages, create adaptations, reprints, include within collections and create summaries, extracts and/or, abstracts of the Contribution, iii) create any other derivative work(s) based on the Contribution, iv) to exploit all subsidiary rights in the Contribution, v) the inclusion of electronic links from the Contribution to third party material where-ever it may be located; and, vi) licence any third party to do any or all of the above. perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint perpetuity.

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in

The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in The copyright holder for this this version posted November 13, 2020. ; https://doi.org/10.1101/2020.11.11.20229708 doi: medRxiv preprint

Statement on the second meeting of the International Health Regulations

Emergency Committee regarding the outbreak of novel coronavirus

detail/30-01-2020-statement-on-the-second-meeting-of-the-international-healthregulations-(2005)-emergency-committee-regarding-the-outbreak-of

WHO. WHO Director-General's opening remarks at the media briefing on COVID-19 -11

Coronavirus Cases

000 Missing Deaths: Tracking the True Toll of the Coronavirus Outbreak -The New York Times [Internet

The Economist. Covid-19 data -Tracking covid-19 excess deaths across countries | Graphic detail | The Economist

Challenges Estimating Total Lives Lost in COVID-19 Decisions

En person smittad med det nya coronaviruset har avlidit -Region Stockholm

Første koronadødsfall i Norge -NRK Norge -Oversikt over nyheter fra ulike deler av landet

Helsedirektoratet har vedtatt omfattende tiltak for å hindre spredning av Covid-19 -Helsedirektoratet

Omfattende tiltak for å bekjempe koronaviruset

Fastleger og legevakt -smittevern mot covid-19 (koronasykdom)

Råd til helsepersonell i spesialisthelsetjenesten om covid-19

Förordning om förbud mot att hålla allmänna sammankomster och offentliga tillställningar

Ytterligare begränsade möjligheter vid sammankomster och tillställningar | Polismyndigheten

Lärosäten och gymnasieskolor uppmanas nu att bedriva distansundervisning -Folkhälsomyndigheten

Tänk över om resan verkligen är nödvändig -Folkhälsomyndigheten

Ny fas kräver nya insatser mot covid-19 -Folkhälsomyndigheten

Nära hälften av alla planerade operationer inställda | SVT Nyheter

Data quality at the Cancer Registry of Norway: An overview of comparability, completeness, validity and timeliness

Norge under Korona/Covid-19 Statistikk om koronakrisen -SSB

Statistik om covid-19 -Socialstyrelsen

Ukerapporter om koronavirus og covid-19 -FHI

Veckorapporter om covid-19 -Folkhälsomyndigheten

Population -annually, per 1 january -SSB

SCB. Population in the country, counties and municipalities on 31 December

Direct impacts in cities

Mortality due to influenza in the United States -An annualized regression approach using multiple-cause mortality data

Influensa -uke 10

Reporting of notifiable infectiousdiseases during the COVID-19 response

Nasjonale tiltak

Covid-19-samfunnsøkonomisk vurdering av smitteverntiltak-andre rapport Rapport fra ekspertgruppe på oppdrag for Helsedirektoratet

COVID-19 pandemic and admission rates for and management of acute coronary syndromes in

Covid-19 oftast inte ensam dödsorsak bland äldre

Nå åpner universitetene, høyskolene og fagskolene igjen

Spørsmål og svar om koronavirus og studenter i Norge

Trust, threats, and consequences of the COVID-19 pandemic in Norway and Sweden: A comparative survey

Information till förskola, grundskola och gymnasier om covid-19

Covid-19 har påverkat vårdkontakter, operationer och väntetider. www.socialstyrelsen.se (accessed Sept 9

Kraftig nedgång andra kvartalet 2020

*Based on exchange rates as of 17 th September, 2020. 12-month periods are defined as: 2015/16: 27 th July