key: cord-0303078-uhnd4jjx
authors: Kraft, K. B.; Elgersma, I. H.; Lyngstad, T. M.; Elstrom, P.; Telle, K. E.
title: COVID-19 vaccination rates among health care workers by immigrant background. A nation-wide registry study from Norway.
date: 2021-09-21
journal: nan
DOI: 10.1101/2021.09.17.21263619
sha: 64bb7d134cdbff202a007e4d19ca049ef2dba258
doc_id: 303078
cord_uid: uhnd4jjx

Background: Studies have suggested that some minority groups tend to have lower vaccination rates than the overall population. This study aims to examine COVID-19 vaccination rates among health care workers (HCWs) in Norway, according to immigrant background. Methods: We used individual-level, nation-wide registry data from Norway to identify all HCWs employed full-time at 1 December 2020. We examined the relationship between country of birth and COVID-19 vaccination from December 2020 to August 2021, both crude and adjusted for e.g. age, sex, municipality of residence, and detailed occupation codes in logistic regression models. Results: Among all HCWs in Norway, immigrants had a 9 percentage point lower vaccination rate (85%) than HCWs without an immigrant background (94%) at 31 August 2021. The overall vaccination rate varied by country of birth, with immigrants born in Russia (71%), Serbia (72%), Lithuania (72%), Romania (75%), Poland (76%), Eritrea (77%), and Somalia (78%) having the lowest crude vaccination rates. When we adjusted for demographics and detailed occupational codes, immigrant groups that more often worked as health care assistants, such as immigrants from Eritrea and Somalia, increased their vaccination rates. Conclusion: Substantial differences in vaccination rates among immigrant groups employed in the health care sector in Norway indicate that measures to improve vaccine uptake should focus specific immigrant groups rather than all immigrants together. Lower vaccination rates in some immigrant groups appears to be largely driven by the occupational composition, suggesting that some of the differences in vaccine rates can be attributed to variation in vaccine access.

As of 5 September 2021, 91% of health care workers (HCWs) in Norway had received either one or two doses of a vaccine against SARS-CoV-2, the virus causing coronavirus disease 2019 (COVID-19). 1 High vaccine uptake in a population is crucial for vaccine effectiveness. Furthermore, a high vaccine uptake among HCWs is vital to secure the operation of the health care services and the health of vulnerable patients.

Studies have suggested that some minority groups tend to be more vaccine hesitant than majority groups. [2] [3] [4] A survey in Norway on attitudes toward COVID-19 vaccines indicated that individuals born in Eastern Europe, Western Asia and Africa tended to be more vaccine hesitant. 2 A study from the United Kingdom (UK), using representative survey data collected between 24

November and 1 December 2020 showed that vaccine hesitancy was more prevalent among Black, Pakistani/Bangladeshi, and mixed non-UK/Irish White ethnic groups. 3 A systematic review over observational studies about the pandemic influenza vaccine H1N1 in the US showed that African American and Latinos had disproportionally lower vaccination rates than other groups. 4 Together, these surveys indicate that there may be larger vaccine hesitancy in certain immigrant groups, possibly resulting in a lower vaccine uptake.

Reports from the COVID-19 vaccination program in European countries have indicated that there are differences in vaccination rates based on country of origin. [5] [6] [7] A Norwegian report from April 2021 showed that persons above 75 years born in Norway, Sweden and Denmark had the highest vaccination rates (≥90%), while persons born in Somalia had the lowest vaccination rates (34%). 5 Similar, the Public Health Agency of Sweden reported in April 2021 that for people aged 80 years and older, vaccination rates for non-immigrants were at 91%, versus 59% and 44% for people born in North Africa and other African countries. 6 A preprint study on HCWs in UK find that a larger proportion of white HCWs have received a vaccine against COVID-19 than HCW with another ethnicity. 7 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

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

The copyright holder for this this version posted September 21, 2021. ; https://doi.org/10.1101/2021.09.17.21263619 doi: medRxiv preprint

To improve knowledge on vaccination in minority groups, this study aims to examine registered COVID-19 vaccination among immigrant groups across HCWs in Norway from December 2020 to August 2021.

The study focuses on HCWs, since they have been prioritised in the COVID-19 vaccination programme and are vital in securing the operation of the health care services and the health of vulnerable patients. The Norwegian vaccination program has prioritised essential HCWs who are critically hard to replace, or HCW with direct patient contact. 8 However, a study finds that hesitancy may exists in a great extent among HCWs in Turkey. 9 At 31 August, all HCWs should have been offered the COVID-19 vaccine. Since there may be differences in demographic and socioeconomic composition for immigrant groups, which in turn may be related to differences in vaccine uptake, we also adjust for age, sex, and detailed occupational groups.

We have utilised individual-level data from BeredtC19, which is an emergency preparedness register aimed to provide rapid knowledge during the COVID-19 pandemic. 10 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The establishment of BeredtC19 forms part of the legally mandated responsibilities of NIPH during epidemics. Institutional board review was conducted, and The Ethics Committee of South-East Norway confirmed (June 4th 2020, #153204) that external ethical board review was not required

The study population included all HCWs in Norway with a contract of equal or more than hours per week and aged between 20 and 65 at 1 st of December 2020. Persons who were not permanent residents of Norway longer than six months (tourists, asylum applicants, etc.) could not be included.

Additionally, we excluded persons without known municipality of residence (n=8). We identified Table   2a and b for the detailed codes). 11 The study population consists therefore of workers with patientcontact, such as physicians, nurses and health care assistants, as well as cleaning staff, but not administrative or kitchen staff. In total, the study population numbers 356 053 persons.

The residents of Norway were divided into three groups based on definitions by Statistics Norway: is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted September 21, 2021. ; https://doi.org/10.1101/2021.09.17.21263619 doi: medRxiv preprint United States of America (USA), and Serbia. When the results are disaggregated by country of birth, Norwegian-born to immigrant parents are not included in their parents' country of birth, but with the rest of the population. When reporting results for immigrants as one group, we include all immigrants (i.e. not only those in the 26 countries).

The outcome is a binary variable capturing if the person had received at least one dose of a vaccine against SARS-CoV-2 in Norway from 28 December 2020 to 31 August 2021.

Among HCWs, timing of access to COVID-19 vaccines differed across occupations, and HCWs are therefore divided into six occupational groups (physicians, specialist nurses, nurses, nursing associates, health care assistants and others). In the regression analyses we used more detailed categories (physicians, specialist nurses, nurses, nursing associates, health care assistants, ambulance personnel, cleaners, radiographers (and associated staff), bioengineers, social educators, dentists, occupational therapists, physiotherapists, psychologists, and other health care workers). The other included categorical variables were operationalized as follows: sex (male/female), age (<29, 30-39, 40-49, 50-59, >60), registered PCR-positive test (yes/no), COVID-19 infection (yes/no), medical risk group for severe COVID-19 (yes/no), institution of employment (hospital, other specialist health service, primary health care service (excl. nursing homes), home care service, nursing home, or other health service), and municipality of residence (356 municipalities of Norway).

Since the outcome variable is binary, we employed logistic regression models with robust standard is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted September 21, 2021. ; https://doi.org/10.1101/2021.09.17.21263619 doi: medRxiv preprint adjusted for age, sex, municipality of residence, COVID-19 infection, medical risk group, occupation, and health care service of employment. For three of the municipalities there was no variation in the outcome variable (100% vaccination rate), resulting in automatically being dropped from the logistic regression model. Hence, the three municipalities were excluded from the whole sample (160 observations in total).

We included 356 053 HCWs aged between 20 and 65 years, mean (SD) age 41 (13). 81% were women and 11% were living in the capital Oslo. Immigrant HCWs numbered 62 339 and had a similar mean age at 41 (11), a lower percent of females (76%) and a higher share living in Oslo (19%). There were 5 477 Norwegian-born to immigrant parents with a considerably lower mean age at 29 (8) and higher percentage living in Oslo (41%). Descriptive statistics on the sample including immigrants by country of birth are presented in the supplementary table 1. Table 1 presents the crude vaccination rates by immigrant background and occupation. Overall, 92% of all HCWs had been vaccinated by the end of our observation period on 31 August 2021 (Table 1) .

Norwegian-born to immigrant parents had an overall vaccination rate of 85% and 88%. This pattern holds across occupational groups, although the sizes of the differences vary with occupation.

The lowest vaccination rates were among HCWs from Russia (71%), Serbia (72%), Lithuania (72%), Romania (75%), Poland (76%), Eritrea (77%), and Somalia (78%). The immigrant groups with the highest vaccination rate were from Vietnam (97%), Thailand (96%), Sri Lanka (95%), Denmark (95%), UK (94%), and India (94%). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint In general, physicians (97%) and specialist nurses (97%) had the highest vaccination rate, while health care assistants (89%) had the lowest. The tendency for vaccination rates to be higher for higher-educated HCW, has been evident since the start of vaccination in Norway in January 2021, with the differences declining somewhat in the last month (see supplement figure 1 ). The pattern can also be observed for within several of the immigrant groups. Physicians and health care assistants from Lithuania, for example, have a vaccination rate of 92% and 66%, while physicians and health care assistants from Pakistan have a vaccination rate of 94% and 87%. How many HCWs who worked as health care assistants varied by country of birth ( Figure 1) . Several of the immigrant groups, especially those from Syria, Eritrea, Afghanistan and Somalia, had a disproportionally large share of the persons working as health care assistants. In the whole population 25% of HCWs were health care assistants, while the share of HCWs who work as health care assistants among immigrants from Syria, Eritrea, Afghanistan and Somalia were 65%, 49%, 46% and 42%. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted September 21, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted September 21, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

The copyright holder for this this version posted September 21, 2021. ; https://doi.org/10.1101/2021.09.17.21263619 doi: medRxiv preprint

Overall, both immigrants and Norwegian-born with immigrant parents have lower vaccination rates than non-immigrants in the crude and adjusted models. Vaccination rates vary between immigrant groups based on country of birth. Some of the differences are likely to related to different occupational composition by country of birth.

The seven immigrant groups with the lowest vaccination rates displayed two distinctive patterns when all covariates were introduced. For five of the immigrant groups from Eastern Europe (Russia, Serbia, Lithuania, Romania and Poland), adjusting for compositional covariates did not explain much of the difference to rest of the population. Contrarily, for the two immigrant groups from East Africa with low vaccination rate (Eritrea and Somalia), we observe a considerable increase in the predicted probability of vaccination in the full model, suggesting that the lower vaccination rates in these groups can to a great degree be related to the occupational composition. The vaccination rate for health care assistants (89%) was lower than the other health care occupations (e.g. physicians (97%), nurses (93%)). In general, 25% of HCWs were health care assistants, while the share of HCWs from Eritrea and Somalia that were health care assistants were 50% and 43%. Other immigrant groups with a high percentage working as health care assistants, such as immigrants from Syria (65%) and Afghanistan (48%), showed a similar, although smaller, increase in the predicted probability when controlling for detailed occupation codes in the full model. For some of the Eastern European countries with lower vaccine rates occupational composition may be loosely related, but it does not appear to be as important as for Eritrea and Somalia. Therefore, there are likely to be other factors than the occupational composition related to the lower vaccine rates between immigrant groups.

A lower vaccination rate among health care assistants may be associated with both a lower education level and access. Supplementary figure 1 shows that health care assistants have had a lower vaccination rate than the other health care occupations during the entire period of vaccine roll-out, which is likely related to which occupations that have been prioritized first. Studies have also shown more vaccine hesitancy among groups with lower education level. 3 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint There are a few limitations that may bring some uncertainty to the study. First, some immigrant HCWs may have been vaccinated in their country of birth but not registered as such in Norway. We do not know the full extent of this, and it may be a reason for a lower rate, especially for the Eastern European countries. Additionally, there are HCW without a permanent identity number, which cannot be included since they are not registered with country of birth. It means that the sample are not capturing all immigrant HCW in Norway, and it may create some systematic differences. However, it is not likely that the magnitude of these limitations may alter the main results.

Which of the factors or barriers that have contributed the most to a lower vaccine rate for some groups has not been examined in this study. If vaccine hesitancy is a barrier, we do not know how deeply ingrained it is, or how easily it can change over time with more information about the vaccines. Since the population sample of this study is HCWs, so we could assume that HCWs have better, or at least equal, willingness to take a COVID-19 vaccine than the overall population. To ensure the highest possible vaccination rates in all groups we need more knowledge about this topic -we therefore call for more research to understand barriers to vaccination, including vaccine hesitancy.

Among HCWs, both immigrants (85%) and Norwegian-born to immigrant parents (88%) had lower vaccination rates than rest of the population (94%). There is a large variation between immigrant groups, with immigrants from Russia (71%), Serbia (72%), Lithuania (72%), Romania (75%), Poland (76%), Eritrea (77%), and Somalia (78%) having the lowest vaccination rate. Immigrants from Eritrea and Somalia more often worked as health care assistants, which in general had a lower vaccine rate than other health care occupations. When we therefore adjusted for demographics and detailed occupational codes the predicted probability of vaccination increased for Eritrea (0.849) and Somalia (0.854). The large variation between immigrant groups suggest that measures to improve vaccine uptake should focus on specific immigrant groups rather than all immigrants together. More research . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity.

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

The copyright holder for this this version posted September 21, 2021. ;

Ukerapport -uke 35. Oslo: Folkehelseinstituttet

Holdninger til vaksine, og etterlevelse av råd om sosial distansering og hygiene blant innvandrere i forbindelse med koronapandemien. Oslo: Folkehelseinstituttet

Predictors of COVID-19 hesitancy in the UK household longitudinal study

Disparities in H1N1 Vaccination Rates: a Systematic Review and Evidence Synthesis to Inform COVID-19 Vaccination Efforts

Oslo: Folkehelseinstituttet

Association of demographic and occupational factors with SARS-CoV-2 vaccine uptake in a multi-ethnic UK healthcare workforce: a rapid realworld analysis

Folkehelseinstituttets foreløpige anbefalinger om vaksinasjon mot covid-19 og om prioritering av covid-19-vaksiner, versjon 2. Oslo: Folkehelseinstituttet

The Knowledge and Behaviours of the Healthcare Staff Regarding Vaccine Rejection

Emergency preparedness register for COVID-19 (BEREDT C19)

SARS-CoV-2 in health and care staff in Norway

is required to investigate underlying reasons for differences in vaccination rates among HCWs with different country backgrounds.

The authors received no external funding

• Immigrant HCW had a lower vaccination rate than HCWs without an immigrant background • Vaccination rate varied country of birth, with immigrants from Eastern Europe and East Africa having the lowest vaccination rate • The vaccination rate immigrants from Eritrea and Somalia increased when we adjusted for demographics and occupation, suggesting that occupational composition is related to the vaccination rate. • The large variation between immigrant groups suggest that measures to improve vaccine uptake should focus on specific immigrant groups rather than all immigrants together.