key: cord-0839972-fiyk0zbj
authors: Veneti, Lamprini; Salamanca, Beatriz Valcarcel; Seppälä, Elina; Starrfelt, Jostein; Storm, Margrethe Larsdatter; Bragstad, Karoline; Hungnes, Olav; Bøås, Håkon; Kvåle, Reidar; Vold, Line; Nygård, Karin; Buanes, Eirik Alnes; Whittaker, Robert
title: No difference in risk of hospitalisation between reported cases of the SARS-CoV-2 Delta variant and Alpha variant in Norway
date: 2021-12-11
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2021.12.321
sha: fac8ae760b7d2f21e12a9dabb2847654bd4177e6
doc_id: 839972
cord_uid: fiyk0zbj

Objectives To estimate the risk of hospitalisation among reported cases of the Delta variant of SARS-CoV-2 compared to the Alpha variant in Norway. We also estimated the risk of hospitalisation by vaccination status. Methods We conducted a cohort study on laboratory-confirmed cases of SARS-CoV-2 in Norway, diagnosed between 3 May and 15 August 2021. We calculated adjusted risk ratios (aRR) with 95% confidence intervals (CIs) using multivariable log-binomial regression, accounting for variant, vaccination status, demographic characteristics, week of sampling and underlying comorbidities. Results We included 7,977 cases of Delta and 12,078 cases of Alpha. Overall, 347 (1.7%) cases were hospitalised. The aRR of hospitalisation for Delta compared to Alpha was 0.97 (95%CI 0.76–1.23). Partially vaccinated cases had a 72% reduced risk of hospitalisation (95%CI 59%–82%), and fully vaccinated cases had a 76% reduced risk (95%CI 61%–85%), compared to unvaccinated cases. Conclusions We found no difference in the risk of hospitalisation for Delta cases compared to Alpha cases in Norway. Our results support the notion that partially and fully vaccinated cases are highly protected against hospitalisation with COVID-19.

Multiple variants of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID-19, have been observed worldwide. Some of these variants have been designated as variants of concern (VOC), defined by the World Health Organisation as variants associated with increased transmissibility, increased disease severity or change in clinical disease presentation, and/or decreased effectiveness of public health and social measures or available diagnostics, vaccines, and therapeutics (World Health Organisation, 2021a) . Variants of concern include the Alpha In Norway (population 5.4 million), testing activity for COVID-19 is high, with consistently 3-5% of the population tested weekly (defined as one or more tests per person within a seven-day period) from March through August 2021. Mathematical modelling estimated that consistently over 50% of all cases weekly were detected in this period (Norwegian Institute of Public Health, 2021a). Sequencing capacity in Norwegian laboratories was rapidly scaled up from early December 2020, and the capacity to screen for variants or perform whole genome sequencing (WGS) was further increased following reports of widespread transmission of the Alpha variant in the UK. From early April 2021 until mid-August over 70% of cases diagnosed had available data on the variant of SARS-CoV-2 that caused their infection.

Alpha has been shown to be more easily transmitted than non-VOC variants (Davies et al., 2021) , and was the dominant circulating SARS-CoV-2 variant in Norway during the third wave of infections in the winter and spring of 2021. It was also associated with a 1.9-fold increased risk of hospitalisation compared to non-VOC variants . Similar associations were observed in other European countries (Bager et al., 2021a; Funk et al., 2021; Grint et al., 2021) . Local and national nonpharmaceutical interventions and increasing vaccination coverage gradually decreased transmission, and Norway started its national reopening plan during the spring (Norwegian Institute of Public Health, 2021a).

The first case of the Delta variant was diagnosed in Norway in April 2021, and local transmission was first evident in the beginning of May. Delta superseded Alpha as the dominant circulating variant in early July, accounting for over 90% of new infections by the end of that month. This coincided with the start of the fourth wave of SARS-CoV-2 infections, and a subsequent increase in the number of new hospitalisations (Norwegian Institute of Public Health, 2021a). There is evidence of increased transmissibility (Campbell et al., 2021; Dhar et al., 2021; Public Health England, 2021a) and lower vaccine effectiveness against infection (Lopez Bernal et al., 2021; Seppälä et al., 2021; Sheikh et al., 2021) for Delta compared to Alpha. In addition, studies from Scotland (Sheikh et al., 2021) , England (Twohig et al., 2021) , Denmark (Bager et al., 2021b) and Ontario, Canada (Fisman and Tuite, 2021) have suggested that infection with the Delta variant increased the risk of hospitalisation by 1.5 to 2.8-fold in those settings.

In order to understand the impact of the Delta variant on the burden of COVID-19 in Norway, and support preparedness planning in the hospital sector, we used linked individual-level data to estimate the risk of hospitalisation among reported cases of the Delta variant compared to reported cases of the Alpha variant, accounting for demographic characteristics, vaccination status and underlying comorbidities. We also estimated the risk of hospitalisation by vaccination status.

We obtained data from the Norwegian national preparedness registry for COVID-19 (Norwegian Institute of Public Health, 2021b). The preparedness registry contains individual-level data from central health registries, national clinical registries and other national administrative registries. It covers all residents in Norway, and includes data on all laboratory-confirmed cases of COVID-19 in Norway, all hospitalisations among cases and COVID-19 vaccinations (supplementary materials, part 1).

We conducted a cohort study, including cases who first tested positive for a SARS-CoV-2 infection between 3 May (week 18) and 15 August (week 32) 2021, who had a national identity number registered, and who had been infected with the Alpha or Delta variant. Reinfections (>6 months since previous positive test and/or determined to be a reinfection by the national reference laboratory based on sequence data) were included. During the study period, SARS-CoV-2 tests were available free of charge for everyone, including those with mild or no symptoms, close contacts, and individuals in quarantine. All positive and negative tests were registered in the national laboratory database. Variants were identified based on whole genome sequencing (WGS) using Illumina or Nanopore technology, partial sequencing by Sanger sequencing or PCR screening for selected targets (Norwegian Institute of Public Health, 2021c). We extracted data up to 30 August 2021, ensuring at least 15 days of follow-up since last date of sampling.

We defined hospitalisation as hospital admission following a positive SARS-CoV-2 test, where COVID-cause of admission were excluded from the study population in order to avoid bias. All admissions to hospital, regardless of the length of stay, were included.

In Norway, Comirnaty (BioNTech-Pfizer, Mainz, Germany/New York, United States) and Spikevax (mRNA-1273, Moderna, Cambridge, United States) were the two most frequently administered vaccines in the study period, using a two-dose schedule (Norwegian Institute of Public Health, 2021d). We defined SARS-CoV-2 cases according to their vaccination status:

1. Those unvaccinated with a COVID-19 vaccine before positive test.

2. Those vaccinated with 1 dose of a COVID-19 vaccine <21 days before positive test.

3. Partially vaccinated -those who tested positive ≥21 days after their first dose of a COVID-19 vaccine, and <7 days after the second dose.

Fully vaccinated -those who tested positive ≥7 days after their second dose Nasreen et al., 2021) with at least the recommended minimum interval between doses depending on the type of vaccine (Norwegian Institute of Public Health, 2021e), or 7 days after their first dose if they had previously been diagnosed with a SARS-CoV-2 infection ≥21 days before vaccination. Cases who received the Janssen vaccine were considered fully vaccinated 21 days after one dose.

Some people have underlying comorbidities that cause them to have a moderate or high risk of severe COVID-19, regardless of age. These individuals were prioritised for vaccination in Norway (Norwegian Institute of Public Health, 2021f). We categorised cases into three groups: i) no underlying comorbidities, ii) medium risk comorbidity and iii) high risk comorbidity, as detailed in supplementary materials, part 1.

We described cases in terms of variants, vaccination status, demographic characteristics, underlying comorbidities and hospitalisation. We also described cases in terms of admission to an intensive care unit (ICU) and COVID-19 related deaths (Norwegian Institute of Public Health, 2021g).

We calculated adjusted risk ratios (aRR) with 95% confidence intervals (CIs) using multivariable logbinomial regression. Our outcome of interest was hospitalisation. Variables considered as possible confounders in our analysis were variant (Alpha or Delta), vaccination status (4 levels), age (4 age groups), sex, country of birth (3 levels), period of sampling (biweekly as categorical variable, and

week as continuous variable), county of residence (12 levels), and underlying comorbidities (3 levels).

Model selection for the multivariable binomial regression was conducted using the likelihood ratio test and the Akaike Information Criterion. We kept the variables variant (due to the main aim of the study) and sex (demographic characteristic associated with risk of hospitalisation in previous analyses in Norway ) in our multivariable analysis, even if they were not significant. We also checked for interactions between all our co-variates by including interaction terms in our models. We conducted the main analysis separately for some of the groups of variables (subgroup analysis) to ensure that the associations remained robust for variant and vaccination status.

In addition to our main analysis, we conducted a number of sensitivity analyses by extending or restricting our study population (for example, including only cases who had WGS results), by adjusting our outcome definitions (for example, including all cases who were hospitalised regardless of main cause of admission) and by changing our analysis method (for example, using Cox regression) to further explore if our main results were robust (supplementary materials, part 2.1).

We also assessed the power of our study to detect a range of potential effect sizes for the risk of hospitalisation with the Delta variant, compared to Alpha (supplementary materials, part 2.2).

Statistical analysis was performed in Stata version 16 (Stata Corporation, College Station, Texas, US), and R version 4.1.0.

Ethical approval for this study was granted by Regional Committees for Medical Research Ethics -South East Norway, reference number 249509. The need for informed consent was waived by the ethics committee. 

During the study period, 347 (1.7%) cases were hospitalised with COVID-19 as main cause of hospitalisation. Among Delta cases 107 (1.3%) were hospitalised, compared to 240 (2.0%) among Alpha cases (Table 1) . Time from positive test to hospitalisation was ≤15 days for 344/347 hospitalised cases. Three cases were hospitalised 17-27 days after positive test. No additional hospitalisations >15 days following positive test were observed in subsequent data extractions with updated data. The median time from testing to hospitalisation was slightly shorter for Delta cases (5 days, IQR: 1-7) than Alpha (6 days, IQR: 3-8.5; Wilcoxon rank-sum p value = 0.016).

In the univariate analysis, the crude RR of hospitalisation among those infected with Delta compared to Alpha was 0.68 (95%CI 0.54-0.85) suggesting a lower risk of hospitalisation among Delta cases. In our multivariable model, after adjusting for sex, age group, country of birth, vaccination status and underlying comorbidities, no difference was found in the risk of hospitalisation between Delta and Alpha, with an aRR of hospitalisation of 0.97 (95%CI 0.76-1.23) ( Table 2 ). Week of sampling and county of residence were not significant predictors in the multivariable model and were excluded from the final model presented here (a sensitivity analysis where we included these two variables is presented in supplementary materials, part 2.1). When we checked for interactions, only an interaction between age group and vaccination status was detected. In order to simplify our main results presented here, we decided to not include the interaction term in our main model. The association between vaccination status and hospitalisation in the subgroup analysis by age group and other variables is presented in supplementary materials, part 2.5. No interaction was found in our main multivariable analysis between variant and vaccination status. In Table 3 , we present the subgroup aRR estimates for Delta compared to Alpha which confirmed the findings in the main analysis. The aRR of hospitalisation among unvaccinated cases for Delta compared to Alpha was 1.10 (95%CI 0.84-1.45). Our results were robust in all our sensitivity analyses (supplementary materials, part 2.1).

The crude RR for hospitalisation among fully vaccinated compared to unvaccinated cases changed when we adjusted for other factors due to confounding. In our multivariable model, after adjusting for sex, age group, country of birth, variant and underlying comorbidities, partially vaccinated cases had a 72% reduced risk of hospitalisation (95%CI 59%-82%) and fully vaccinated had a 76% reduced risk of hospitalisation (95%CI 61%-85%), compared to unvaccinated cases (Table 2 ). In supplementary materials, part 2.5 we present our subgroup analysis for the risk of hospitalisation by vaccination status that seemed to be robust with our main findings. In our subgroup analysis by variant type, we found that partially vaccinated cases had a 77% (95%CI 58%-87%) and 72% (95%CI 50%-85%) reduced risk of hospitalisation and fully vaccinated cases a 79% (95%CI 59%-89%) and 70% (95%CI 39%-85%) reduced risk of hospitalisation compared to unvaccinated cases infected with the Alpha and Delta variant respectively.

Among the 107 patients hospitalised with Delta, 16 (15%) were admitted to ICU compared to 40 (17%) of 240 patients hospitalised with Alpha. Among the 56 cases admitted to ICU, 40 were unvaccinated and 14 had been vaccinated <21 days before positive test. There were 24 deaths total among the study cohort, of which 20 were reported as COVID-19 related (Norwegian Institute of Public Health, 2021g). Of these 20, five were Delta cases, and 15 Alpha cases. Additional analyses were not done on these outcomes due to small numbers.

In this study, we have analysed individual-level data on laboratory-confirmed cases of COVID-19 in Norway and hospitalisations among cases within the study period, as well as demographic characteristics, vaccination status and underlying comorbidities. Earlier analyses of other VOC in Norway showed an increased risk of hospitalisation for Alpha and Beta compared to non-VOC , in line with others (Bager et al., 2021a; Funk et al., 2021; Grint et al., 2021; Fisman and Tuite, 2021) . Here our findings indicate no difference in the risk of hospitalisation for SARS-CoV-2 cases infected with the Delta variant compared to the Alpha variant in Norway, in contrast to published estimates from other countries. An analysis from Scotland suggested an adjusted hazard ratio for hospitalisation of 1.85 (95%CI 1.39-2.47) for Delta compared to Alpha (Sheikh et al., 2021) .

In England, a similar association was observed (hazard ratio 2.26, 95%CI 1.32-3.89) (Twohig et al., 2021) . The study from England (hazard ratio 2.32, 95%CI 1.29-4.16) and Denmark (aRR 3.01, 95%CI 2.02-4.50) (Bager et al., 2021b) suggested increased risk of hospitalisation in an unvaccinated cohort, which our study also does not support. In addition, the study from Ontario, Canada estimated an adjusted odds ratio for hospitalisation of 1.45 (95%CI 1.27-1.64) for Delta compared to N501Ypositive VOC (Alpha, Beta or Gamma) as well as a 2.01-times increased risk of ICU admission (95%CI 1.60-2.47) and 1.69-times increased risk of death (95%CI 1.16-2.35) (Fisman and Tuite, 2021) .

In comparing estimates, the study settings need to be considered, with each conducted in a different population, time period and healthcare system. For example, any differences in SARS-CoV-2 testing criteria and activity, and capacity to screen for variants may lead to differences in the subset of Alpha and Delta cases diagnosed between the studies. Outcome definitions and analysis methods could also have played a role, however it is unlikely that these factors have impacted the results presented in the different studies to an extent that would explain the different associations observed. Our statistical methodology would have been sufficient to detect a comparable increase in risk as the other studies if it existed, and our sensitivity analyses gave robust results using different outcome definitions.

Our results highlight the importance of taking local epidemiological characteristics into account, when endeavouring to understand the effect that different variants have on the COVID-19 epidemic in different settings. Our results are representative of a young cohort of SARS-CoV-2 cases in a country with broad testing criteria, and high testing activity and capacity to screen for variants. In the study period, the health system operated well within capacity, criteria for hospital admission were consistent and hospital treatment was available to all those who would benefit. Aside from local restrictions in the event of outbreaks, there were no notable lockdowns. There was high vaccination coverage among populations at greater risk of severe COVID-19, and vaccination coverage was steadily increasing as Delta superseded Alfa as the dominant variant (Norwegian Institute of Public Health, 2021a).

Our study also underlines the need for more research to further understand the association between the Delta variant and severe disease. In this study we did not have access to data on clinical disease severity among cases, and the number of ICU admissions and COVID-19 related deaths were low in both groups. Our results therefore cannot directly conclude as to whether there is a difference in virulence for Delta compared to Alpha. What our results do suggest is that other factors, such as age, country of birth, underlying comorbidities and vaccination status are associated with the risk of hospitalisation among SARS-CoV-2 cases in Norway. However, even if the risk of hospitalisation among Delta and Alpha cases in Norway is similar, risk of infection with Delta is higher given evidence of increased transmissibility (Campbell et al., 2021; Dhar et al., 2021; Public Health England, 2021a) and lower vaccine effectiveness against infection (Lopez Bernal et al., 2021; Seppälä et al., 2021; Sheikh et al., 2021) . This must be considered as prevention and control measures are weighed up in view of the burden of disease in society, capacity in the healthcare system and progress of vaccination programmes. In Norway, vaccine effectiveness against laboratory-confirmed infection with the Delta variant has been estimated to be 22% among partially vaccinated and 65% among fully vaccinated persons . Our results suggest that partially and fully vaccinated cases infected with the Delta variant are highly protected against hospitalisation, in line with published estimates from elsewhere (Fisman and Tuite, 2021; Public Health England, 2021b; Sheikh et al., 2021; Statens Serum Institut, 2021) . This highlights the importance of ensuring high vaccination uptake.

The vast majority of vaccinated cases received the mRNA vaccine Comirnaty, which did not allow us to investigate whether vaccine type had an additional impact on the risk of hospitalisation (supplementary materials, part 2.4).

Sampling effects can bias the estimate of risk when using surveillance data. For example, if a larger proportion of milder cases were diagnosed in the Delta cohort, this could underestimate the risk of hospitalisation for Delta compared to Alpha. As we did not have data on relevant parameters that would have helped us to explore this further, such as clinical disease severity or viral loads, we cannot rule out this bias. However, such bias should be considered in light of the consistent COVID-19 testing strategy in Norway during the study period. In addition, it has been suggested that, when comparing two variants for a post-infection outcome at a time when one variant is in the process of supplanting the other, that in fact a relatively larger proportion of severe cases of the new variant (in this case Delta) could be diagnosed (Seaman et al., 2021) . This underlines one of several challenges with using surveillance data to determine the relative disease severity of new variants in an evolving epidemic setting (Bager et al., 2021b) .

There are some limitations with our analysis. While our sample size was marginally larger than the study from Scotland (Sheikh et al., 2021) , both in terms of number of cases overall and number of Delta cases, our power calculations indicated that our study may be underpowered if Delta was associated with a small increased risk of hospitalisation compared to Alpha (supplementary materials, part 2.2). In addition, the method used to determine underlying comorbidities will likely underestimate the true prevalence, as only individuals that have been in contact with health services are identified. Data on medications used and procedure codes are currently not taken into account, which would improve the definitions and detect more individuals with underlying comorbidities.

Our findings indicate no difference in the risk of hospitalisation for cases infected with the Delta variant of SARS-CoV-2 compared to the Alpha variant in Norway. This is a more encouraging finding than previous studies for the ongoing response to the COVID-19 pandemic in settings where the Delta variant is circulating, although evidence of increased transmissibility and lower vaccine effectiveness against infection for Delta must also be considered. Our results highlight the importance of taking local epidemiological characteristics into account, when endeavouring to understand the effect that different variants have on the COVID-19 epidemic in different settings.

Data on protection against severe disease are crucial to guide future vaccination strategy, and the results from this study support the notion that partially and fully vaccinated cases are highly protected against hospitalisation with COVID-19.

All co-authors were involved in the conceptualisation of the study. RW drafted the study protocol and coordinated the study. MLS, KB, OH, RK and EAB contributed directly to the acquisition of data. 

The authors declare that they have no competing interests.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

The dataset analysed in the study contains individual-level linked data from various central health registries, national clinical registries and other national administrative registries in Norway. The researchers had access to the data through the national emergency preparedness registry for COVID-19 (Beredt C19), housed at the Norwegian Institute of Public Health (NIPH). In Beredt C19, only fully anonymised data (i.e. data that are neither directly nor potentially indirectly identifiable) are permitted to be shared publicly. Legal restrictions therefore prevent the researchers from publicly sharing the dataset used in the study that would enable others to replicate the study findings.

However, external researchers are freely able to request access to linked data from the same registries from outside the structure of Beredt C19, as per normal procedure for conducting health research on registry data in Norway. Further information on Beredt C19, including contact information for the Beredt C19 project manager, and information on access to data from each individual data source, is available at https://www.fhi.no/en/id/infectiousdiseases/coronavirus/emergency-preparedness-register-for-covid-19/.

First and foremost, we wish to thank all those who have helped report data to the national emergency preparedness registry at the Norwegian Institute of Public Health (NIPH) throughout the pandemic. We also highly acknowledge the efforts that regional laboratories have put into establishing a routine variant screening procedure or whole genome sequencing at short notice and registration of all analysis in national registries for surveillance. Thanks also to the staff at the Virology and Bacteriology departments at NIPH involved in national variant identification and whole genome analysis of SARS-CoV-2 viruses. We also highly acknowledge the efforts of staff at hospitals around Norway to ensure the reporting of timely and complete data to the Norwegian Intensive Care and Pandemic Registry, as well as colleagues at the register itself. We would also like to thank Anja Elsrud Schou Lindman, project director for the national preparedness registry, and all those who have enabled data transfer to this registry, especially Gutorm Høgåsen at the NIPH, who has been in charge of the establishment and administration of the registry. We would also like to 

Vaccine effectiveness and duration of protection of Comirnaty, Vaxzevria and Spikevax against mild and severe COVID-19 in the UK. MedRxiv (Preprint). 2021 (cited

Risk of hospitalisation associated with infection with SARS-CoV-2 lineage B.1.1.7 in Denmark: an observational cohort study

Hospitalisation associated with SARS-CoV-2 delta variant in Denmark

Increased transmissibility and global spread of SARS-CoV-2 variants of concern as at

Estimated transmissibility and impact of SARS-CoV-2 lineage B.1.1.7 in England

Genomic characterization and Epidemiology of an emerging SARS-CoV-2 variant in Delhi

Evaluation of the relative virulence of novel SARS-CoV-2 variants: a retrospective cohort study in Ontario

Characteristics of SARS-CoV-2 variants of concern B.1.1.7, B.1.351 or P.1: data from seven EU/EEA countries, weeks 38/2020 to 10/2021

Severity of SARS-CoV-2 alpha variant (B.1.1.7) in England

Effectiveness of Covid-19 Vaccines against the B.1.617.2 (Delta) Variant

Effectiveness of mRNA and ChAdOx1 COVID-19 vaccines against symptomatic SARS-CoV-2 infection and severe outcomes with variants of concern in Ontario

Ukerapporter om koronavirus og covid-19. Oslo: Norwegian Institute of Public Health

Norwegian Institute of Public Health. Emergency preparedness register for COVID-19

Påvisning og overvåkning av SARS-CoV 2-virusvarianter

Coronavirus vaccine -information for the public

Norwegian Institute of Public Health. Vanlige problemstillinger om koronasertifikat

Who will get the coronavirus vaccine?

Spørsmål og svar om koronastatistikken og de interaktive diagrammene

SARS-CoV-2 variants of concern and variants under investigation in England

COVID-19 vaccine surveillance report Week 34. London: Public Health England

Adjusting for time of infection or positive test when estimating the risk of a post-infection outcome in an epidemic

2021 (cited

Vaccine effectiveness against infection with the Delta (B.1.617.2) variant, Norway

SARS-CoV-2 Delta VOC in Scotland: demographics, risk of hospital admission, and vaccine effectiveness

Spikevax og Vaxzevria i perioden, hvor alfa og delta varianterne har cirkuleret i Danmark

Hospital admission and emergency care attendance risk for SARS-CoV-2 delta (B.1.617.2) compared with alpha (B.1.1.7) variants of concern: a cohort study

Increased risk of hospitalisation and intensive care admission associated with infection with SARS-CoV-2 variants B.1.1.7 and B.1.351 in Norway

World Health Organisation. Weekly epidemiological update on COVID-19 -17