Utilization of Norwegian crisis shelters before and during the COVID-19 pandemic

Abstract

Background:

The measures introduced to control the COVID-19 pandemic, including lockdowns and physical distancing, exerted considerable influence on society.

Objectives:

The aims of this study were to examine (1) the prevalence of people seeking Norwegian crisis shelters for domestic violence during the first period of the COVID-19 pandemic (2020) compared to the year before, (2) the demography and type of violence among first-time visitors and (3) to compare the utilization of the crisis shelters and characteristics of the users between shelters in the capital and the other shelters throughout Norway.

Design:

Observational study.

Method:

The study was based on data from each crisis shelter in 2019 and 2020. Comparison between the 2 years were based on corresponding periods (12 March until 31 December) and analysed with t-test and chi-square tests.

Results:

Total use of crises centres, residential stays and daytime visits were lower during the pandemic (n = 7102) compared to the pre-pandemic period (n = 11 814). There was a shift from daytime visits to phone contacts when the restrictions were established. There was a higher proportion of residential stays versus daytime visits during the pandemic (21.5%) compared to the pre-pandemic period (15.4%) (p ⩽ 0.001). The proportion of first-time users was higher during the pandemic compared to pre-pandemic period both for residents (52.4% vs 47.1%) and daytime visitors (10.9% vs 9.0%). Among first-time crisis shelter residents during the pandemic period, fewer reported having children at home compared to the pre-pandemic period. The background of the crisis-shelter users did not differ between the capital and rest of Norway, but the capital had relatively more residents with psychological violence and threats during the pandemic.

Conclusion:

The utilization of Norwegian crisis shelters, especially daytime visits was lower during the pandemic. There was a shift in daytime contacts from visits to phone contact at the pandemic outbreak. To ensure that information about available crisis shelters reaches the total population, these shelters should be prepared for a higher volume of phone contacts in a future pandemic situation.

Introduction

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was discovered in January 2020, and in March the same year, The World Health Organization (WHO) declared a pandemic. ¹ The COVID-19 pandemic influenced society in many fundamental ways. Lockdowns and physical distancing to control the pandemic are examples of measures that were implemented.^1,2 These types of measures influence people’s everyday life significantly. For vulnerable groups, changes in one’s social context may influence living conditions and health considerably, ³ as might be the case for men and women exposed to domestic violence during the COVID-19 pandemic.^4,5

Domestic violence has been clearly documented as a large global health problem among women for many years. ⁶ A recent Lancet study ⁷ points to the critically high lifetime prevalence of violence against women (27%), also before the COVID-19 pandemic. The authors based their study on data from the WHO Global Database on Prevalence of Violence Against Women, covering about 90% of the global population of women and girls. In addition to pointing at the general problem, the authors stress the urgent need to intervene in the post-COVID-19 period, to initiate multisectoral collaboration against domestic violence, and to strengthen the response from the public.

Public regulations due to COVID-19, led to more people spending working hours as well as leisure time behind closed doors. Hence, there was an unintended potential for an increase in domestic violence, including abuse of partners, older people and children. ⁸ A comprehensive review from 2021 reported an increase in domestic violence cases during the lockdown in North America, Europe, Asia-Pacific Area, and Africa. ⁹ Another review from 2021 concluded that the domestic violence increased in response to the COVID-19 measures of stay-at home/lockdown. This review included 18 studies from the United States, Mexico, Italy, Sweden, Australia, Argentina and India, but not Norway. ¹⁰ However, in a study from the Netherlands, no significant difference was found between families before and during the coronavirus crisis regarding incidents of ‘no violence’, ‘moderate’ or serious’ violence. ¹¹

In Norway, the prevalence of intimate partner violence (IPV) during an individual’s lifetime is about 27%, and the incidence of life-threatening IPV is 9% for women and 2% for men.^{12 –14} It seemed that in Norway, IPV increased during the COVID-19 lockdown. The number of police reports on IPV increased by 54% during the lockdown in Norway in 2020, including the severity of the reported incidents increased. ¹⁵ In the lockdown period, the increased risk of domestic violence was highlighted in the media, and the staff at the crisis shelters in Norway were classified as ‘key personnel’. However, this may also indicate a higher awareness for serious domestic violence in the police during COVID-19. ¹⁵

Norway had a complete lockdown from 12 March 2020. This included closing schools, kindergartens, universities and most businesses. In addition, there were restrictions relating to how many people you were allowed to visit and where you could travel. ¹⁶ The restrictions were gradually removed from May 2020. The national borders were opened again 15 July 2020. However, due to a major increase in COVID-19 cases some weeks after, most restrictions were re-introduced in the autumn 2020, until the first vaccinations started in December 2020. There were differences between the municipalities in Norway. The region including the capital Oslo had more restrictions and for longer periods than most other parts of Norway. This was due to higher numbers of COVID-19 cases in the Oslo area. ¹⁷

In Norway, the crisis shelters spread throughout the country are open 24 hours a day. ¹⁸ These are accessible for women, men, and children together with parents, who need a safe place to stay, live and hide from a violent partner, or other person in a close relationship for a shorter or longer period. The threshold for access to the crisis shelters is low. Staying at a shelter is free of charge and does not require referrals. The shelters receive public funding and support from the authorities. All municipalities in Norway must offer their inhabitants such services if they are subjected to domestic violence and need help, according to a Shelter Act that was enacted in 2010. During the pandemic, the Norwegian Domestic Violence Shelters adapted their activity to the prevailing infection control measures. Most of the crisis shelters reported being able to help victims of violence and collaborate with public services as before COVID-19 despite of the reduced and modified services in line with the infection control measures. ¹⁹

To the best of our knowledge, there is a sparsity of studies examining the impact of COVID-19 on utilization of domestic violence shelters. A Canadian study of 15 shelters found that infection control measures reduced the space available for people to stay and live at the shelters, thus restricting service availability and provision. ²⁰ Statistics Canada released a report on Canadian residential facilities for victims of abuse during the years 2020/2021 of the pandemic. Workers at the crisis shelters reported that accommodation capacity was the greatest challenge during the COVID-19 pandemic. In total, the proportion of people seeking crisis shelter was 31% lower than reported in 2017/2018, when data had last been collected. ²¹

However, in Norway, it is not known to what extent the crisis shelters were used and by whom. The first aim of this study was therefore to examine changes in the prevalence of people seeking refuge at Norwegian crisis shelters for domestic violence during the first period of the COVID-19 pandemic (2020) compared to the year before, 2019. Second, we aimed to study the characteristics of first-time visitors during the pandemic, regarding demography and type of violence reported, and to compare this with a similar pre-pandemic period. Finally, we aimed to compare the utilization of the crisis shelters and characteristics of the users (demography and type of violence) between crisis shelters in the capital of Norway, that is, Oslo, and the remaining shelters in Norway.

Methods

Setting

In Norway, the local authorities are required to offer daytime services and crisis shelters (open 24 hours a day) for victims of domestic violence. There are 44 crises shelters distributed throughout the country. ¹⁸ This study was based on data yearly reported from each crisis shelter to The Norwegian Directorate for Children, Youth and Family Affairs (Bufdir). ¹⁸ The collected information is used for annual statistics reports from the centres. The data are reported anonymously, and the participants gave informed agreement to be registered (92.4% accepted to give information exceeding use). All visits are reported, first-time visit for a person at a crisis shelter is noted as such, but possible repetitive visits from the same person cannot be identified. The two reporting forms are used primarily for administration overview and are not validated. All methods were carried out in accordance with relevant guidelines and regulations as given by the Declaration of Helsinki.

Data

Data for the year 2019 and 2020 were provided from Bufdir for this study. Due to COVID-19 lockdown 12 March 2020, only data from the period 12 March until 31 December for both 2019 and 2020 were included in the analyses and are referred to as study period (Figure 1). The year 2019 is referred to as ‘pre-pandemic’ and 2020 as ‘during the pandemic’. To explore changes due to time, sub-analyses for utilization of the crisis shelters in 2019 and 2020 were compared for the period 1 January until 11 March.

Figure 1.

Flow chart Users of Norwegian crises shelters 2019-2020. Flow chart presenting total number and number of first-time users during the study period, from 12 March (lockdown in 2020 due to pandemic) until 31 December, the years 2019 and 2020.

Variables

The Norwegian crisis shelters are used both by daytime visitors for help and support, and for residential stays that can last for several days. In the analyses, we have distinguished between the different use of the crisis shelters. This was possible as the reporting forms were different for the two kinds of use: daytime visit and residential stay. The two reporting forms included mostly similar questions. Both forms included which crisis shelter was used and whether it was the first daytime visit/resident’s first stay at a crisis shelter the current year. Furthermore, we included the following variables from both forms: Information on date on arrival, the user’s gender, age, employed (not (i.e. disability pension/student/work at home) versus full-time or part-time work), non-Norwegian origin (both parents born abroad vs others), children at home (yes/no), children brought along to the crises centre (yes/no), ‘What forms of violence and abuse has the user been exposed to’ with multiple choices including psychological violence, physical violence, threats and sexual assault (yes/no). The form also included if there were ‘One or more assaulters’ (one vs two or more).

For residential stays, three additional variables were included in the reporting form and used in this study. Date of departure (so length of stay was calculated), information whether the assaulter was of non-Norwegian origin (both parents born abroad vs others) and ‘Was the perpetrator(s) under the influence of drugs during the abuse/assault’. The last variable was categorized into intoxicated yes (always or sometimes) vs no (never or don’t know).

To illustrate the use of crisis shelters during the study period, from 12 March until 31 December, we divided the period to 10 time-span groups with 29 or 30 days in each. From January 2020, the reporting form used to register daytime visitors included information on phone contacts. There was no comparison information from 2019; however, we have included phone contacts for the 2020 study period in the graph. Even though the COVID-19 restrictions were national, the capital of Norway, had a stricter interpretation of the regulations through the first year of the pandemic. Therefore, data from the capital were compared with data from the rest of the country.

Statistical analysis

Power calculation was done for the primary aims, with power of 0.76 and higher (two-sided significance level α = 0.05). Data concerning the utilization of crisis shelters by distribution of residential stays and daytime visits within the two periods were presented with numbers, percentages, and tested with the chi-square test. To illustrate the utilization of crisis shelters by time, the occurrence by 2 weeks per 100,000 inhabitants was calculated and presented in a graph. The data for residents and daytime visitors was presented separately, with differences in proportion of first-time users between the two periods shown with numbers, percentages, and chi-square tests.

Descriptive statistics for the characteristics of first-time crisis shelter residents and first-time daytime visitor were provided by mean, standard deviation (SD), median, interquartile range (IQR), numbers, and percentages provided for the total study population.

Analyses were carried out for only first-time users, to ensure that we had unique users. Furthermore, men were excluded due to the differential gender distribution of the other characteristics. The distribution of user characteristics, forms of violence, and abuse the user has been exposed to, as well as information regarding the assaulter, were presented for each period. Possible differences between the two periods were analysed by t-test and chi-square tests. Tests for differences in occurrence between periods, across capital and the rest of Norway was performed with an interaction term in linear regression (age and length of stay) and for the categorical variables Mantel–Haenszel test for homogeneity.

Missing data

Some of the variables (age, employed, non-Norwegian origin, more than one assaulter, non-Norwegian assaulter, and intoxicated assaulter) had missing information. The percentages in the tables are based on completed answers for each variable, and the number that was missing for each variable were presented in the table’s footnotes.

All analyses were performed in STATA, and a significance level α = 0.05 was used.

We followed the STROBE Guidelines when preparing the manuscript.

Results

Total use of crises centres, residential stays and daytime visits was lower during the pandemic (n = 7102) compared to pre-pandemic period (n = 11 814) (Table 1). Especially daytime visits were reduced so the distribution of residential stays and daytime visits gave a higher proportion of residential stays during the pandemic (21.5%) compared to the pre-pandemic period (15.4%) (p ⩽ 0.001). The same pattern was found for both the capital and the rest of Norway. For the period 1 January to 11 March, the residential stays constituted 21% both years.

Table 1.

Distribution of utilization of crises shelters for residential stays and daytime visits in Norway by period. ^a

Crisis shelter	Period				p value
	Pre-pandemic		During pandemic
	N	%	N	%
Residential stays	1818	15.4	1528	21.5
Daytime visits	9996	84.6	5574	78.5	⩽0.001
Total	11,814		7102

^aPre-pandemic period (12 March until 31 December 2019), pandemic period (12 March until 31 December 2020).

The utilization of crises centres per 100,000 inhabitants for pre-pandemic and during pandemic by 2 weeks is illustrated in Figure 2. For residential stays, there was a lower utilization during the pandemic, especially in the first and last period in the capital Oslo (Figure 2(a)). For the rest of Norway, the same pattern was found for the first period (12 March until middle of July) but it was not as pronounced. The utilization of daytime visits showed a marked reduction during pandemic compared with the year before except for August (Figure 2(b)). For the capital there were more phone contacts (Figure 2(c)) than daytime visits during the pandemic. Phone contacts were not registered in 2019.

Figure 2.

Utilization of crisis shelters per 10,000 inhabitants by 2 weeks from 01 January 2019 to 31 December 2020 by capital of Norway and rest of Norway: (a) residential stays, (b) daytime visits, (c) phone contacts, registration started 01 January 2020.

There was a higher proportion of first-time users during the pandemic compared to pre-pandemic period both for residents (52.4% vs 47.1%) and daytime visitors (10.9% vs 9.0%) in Norway (Table 2). In the capital, this trend was not found among residents where the proportion of first-time residents was 46% in both periods, while for the first-time daytime visitors, there was an increase (20.0% during the pandemic and 13.9% pre-pandemic).

Table 2.

Proportion of first-time residents and daytime visitors at Norwegian crises shelters by period. ^a

Crisis shelter	Period						p value
	Pre-pandemic			During pandemic
	Total	First-time		Total	First-time
	N	N	%	N	N	%
Residents	1670	786	47.1	1394	730	52.4	0.003
Daytime visitors	9166	821	9.0	5248	570	10.9	⩽0.001

^aPre-pandemic period (12 March until 31 December 2019), pandemic period (12 March until 31 December 2020).

Table 3 shows the characteristics for first-time residents and first daytime visitors in Norway for the study period from March 2019 to December 2020. The median number of 24 hour stays for first-time residents at the crisis shelter was 25.9 days. Most of the first-time shelter residents were not employed and had non-Norwegian origin, while the opposite distribution was found among daytime first-time visitors, where the fewest were employed, and had non-Norwegian origin. There was a high occurrence of reported violence and abuse both among residents and daytime visitors. The majority of users were women (crisis shelter residents 90.4%, daytime visitors 88.0%) and our further analyses include only women.

Table 3.

Characteristics of first-time residents and among daytime visitors at a Norwegian crises shelter. Data from 12 March (lockdown in 2020 due to pandemic) until 31 December (2019 and 2020).

Characteristic		Shelter residents		Daytime visitors
Age	Mean	35.7	SD: 12.5	36.8	SD: 12.1
24 hours stay	Median	25.9	IQR: (3–37)
		n	%	n	%
Gender, women		1368	90.4	1222	88.0
Not employed		1015	69.0	624	47.7
Non-Norwegian origin a		964	64.5	536	41.9
Children brought along		672	44.3	116	8.3
Children at home		169	11.2	648	46.6
Psychological violence		1310	86.4	1204	86.6
Physical violence		956	63.1	682	49.0
Threats		853	56.3	736	52.9
Sexual assault		241	15.9	221	15.9
More than one assaulter		220	15.0	166	12.5
Assaulter
Non-Norwegian origin a		713	53.3
Intoxicated b		541	53.4
Total number		1516		1391

IQR: interquartile range.

^aBoth parents born abroad.

^bThe assaulter was under the influence of drugs, Always or sometimes.

Resident, comparing the capital with rest of Norway

Among first-time female residents, a lower proportion reported having children at home (8.7% vs 13.1%) and that the assaulter had a non-Norwegian background (49.3% vs 56.4%) in the pandemic period compared to pre-pandemic period (Table 4). These findings were not different between the capital and the rest of the country. However, there were differences between the regions regarding reported psychological violence and threats. There was increased reporting of these types of violence during the pandemic in the capital, while for rest of the country lower numbers of threats were reported during the pandemic.

Table 4.

Female first-time residents at a Norwegian crisis shelter, comparison of period ^a , total and for capital compared with rest of Norway.

						Norway
	Total					Capital				Rest of Norway
	Pre-pandemic		During pandemic		p value b	Pre-pandemic		During pandemic		Pre-pandemic		During pandemic		p value c
	Mean	SD	Mean	SD		Mean	SD	Mean	SD	Mean	SD	Mean	SD
Age	35.4	11.9	35.7	12.5	.669	33.6	10.4	32.5	11.8	35.7	12.1	36.0	12.6	.540
Mean of 24 hours stay	26.9	35.8	24.8	30.7	.250	26.2	39.1	22.5	20.3	27.0	35.2	25.1	31.7	.750
	n	%	n	%		n	%	n	%	n	%	n	%
Not employed	472	69.5	457	70.0	.852	67	70.5	49	74.2	405	69.4	408	69.5	.641
Non-Norwegian origin d	444	64.6	423	63.9	.779	81	84.4	53	82.8	363	61.4	370	61.9	.768
Children brought along	340	48.6	306	45.8	.306	36	35.6	25	37.3	304	50.8	281	46.8	.504
Children at home	92	13.1	58	8.7	.008	8	7.9	3	4.5	84	14.2	55	9.15	.863
Psychological violence	602	86.0	589	88.2	.231	83	82.2	63	94.0	519	86.4	526	87.5	.050
Physical violence	449	64.1	426	63.8	.887	67	66.3	52	77.6	382	63.8	374	62.2	.096
Threats	408	58.3	363	54.3	.141	55	54.5	44	65.7	353	58.9	319	53.1	.040
Sexual assault	118	16.9	114	17.1	.918	22	21.8	9	13.4	96	16.0	105	17.5	.131
More than one assaulter	98	14.5	87	13.3	.529	26	28.6	13	21.0	72	12.3	74	12.5	.317
Assaulter
Non-Norwegian origin d	339	56.4	301	49.3	.014	58	90.6	43	86.0	281	52.3	258	46.1	.738
Intoxicated e	231	54.7	273	54.4	.914	21	53.9	20	47.6	210	54.8	253	55.0	.585
Total number	700		668			101		67		599		601

The number that was missing for the variables: age n = 22, connected to work-life n = 36, non-Norwegian origin n = 19, more than one assaulter n = 40, intoxicated n = 444, assaulter origin n = 157.

^aPre-pandemic period (12 March until 31 of December 2019), pandemic period (12 March until 31 December 2020).

^bTest for different occurrence of current variable between pre-pandemic and pandemic period.

^cTest for interaction between period and capital versus rest of Norway.

^dBoth parents born abroad.

^eThe assaulter was under the influence of drugs, Always or sometimes.

Daytime visitors, comparing the capital with the rest of Norway

A higher proportion of first-time female daytime visitors was not employed during the pandemic period (53.1%) compared to pre-pandemic period (47.2%), and fewer reported experiences of threats during pandemic compared to pre-pandemic (47.8 vs 56.2%) (Table 5). However, these findings were not significantly different between the two regions. There were differences between the regions for bringing children to the centre: For the capital, this was reduced during pandemic but increased for rest of Norway; however, only three women in the capital reported that they brought along children during pandemic. Also, we found less reported psychological violence during pandemic in the capital compared with rest of Norway.

Table 5.

Female first time daytime visitors at a Norwegian crisis shelter, comparison of period ^a , total and for capital compared with rest of Norway.

						Norway
	Total					Capital				Rest of Norway
	Pre-pandemic		During pandemic		p value b	Pre-pandemic		During pandemic		Pre-pandemic		During pandemic		p value c
	Mean	SD	Mean	SD		Mean	SD	Mean	SD	Mean	SD	Mean	SD
Age	36.4	12.0	36.8	11.9	.722	35.6	9.9	36.7	11.3	36.6	12.6	36.8	12.0	.611
	n	%	n	%		n	%	n	%	n	%	n	%
Not employed	322	47.2	247	53.1	.050	73	43.5	53	58.9	249	48.4	194	51.7	.101
Non-Norwegian origin d	273	42.1	203	43.0	.769	85	56.7	54	58.1	188	37.8	149	39.3	.977
Children brought along	63	8.9	44	8.6	.893	16	8.8	3	2.5	47	8.9	41	10.6	.016
Children at home	320	44.9	255	50.0	.081	81	44.8	50	41.0	239	45.0	205	52.8	.086
Psychological violence	623	87.5	435	85.3	.265	159	87.9	95	77.9	464	87.4	340	87.6	.047
Physical violence	356	50.0	254	49.8	.946	92	50.8	61	50.0	264	49.7	193	49.7	.899
Threats	400	56.2	244	47.8	.004	88	48.6	47	38.5	312	58.8	197	50.8	.746
Sexual assault	115	16.2	92	18.0	.386	20	11.1	19	15.6	95	17.9	73	18.8	.386
More than one assaulter	87	12.7	52	10.9	.341	22	12.9	12	12.1	65	127	40	10.6	.768
Total number	712		510			181		122		531		388

The number that was missing for the variables: age n = 178, connected to work-life n = 75 non-Norwegian origin n = 102, more than one assaulter n = 60.

^aPre-pandemic period (12 March until 31 of December 2019), pandemic period (12 March until 31 December 2020).

^bTest for different occurrence of current variable in pre vs pandemic period.

^cTest for interaction between period and capital versus rest of Norway.

^dBoth parents born abroad.

Discussion

In our cross-sectional study of utilization of Norwegian crisis shelters before and during the COVID-19 pandemic, we found that the use of the crisis shelters was lower, especially regarding daytime visits, during the pandemic. However, there was a higher proportion of first-time users of the shelters during the first nine and a half months after the COVID-19 outbreak compared to the same period the year before. Phone communication was the dominating contact form for daytime users of the shelters during the pandemic, especially in the capital Oslo. There was no change in user characteristics, types of violence, or assaulters in Norway, but subfigures from the capital show a higher percentage of residents reporting psychological violence and threats during the pandemic compared to the rest of the country.

Knowing that the prevalence of domestic violence increased during the pandemic in many countries,^2,9,10 it is difficult to understand the reduced use of the crisis shelters without additional information from the shelters. Victims of domestic violence are likely to have spent more time with their abusers than normal, which may have increased tensions and created more opportunities for violence. Moreover, the fact that people spent less time spent in the public sphere reduced the possibility and the opportunity to disclose signs of violence for others. Thus, COVID-19 and the resulting regulations for infection prevention, may have silenced the victims of domestic violence. ⁹

The reduced use of crisis shelters during the pandemic lockdown was unexpected as the effects of the lockdown, including loss of work and income, children staying at home, problems with getting hold of drugs, may have increased stressful domestic situations and put victims at higher risk of violence. ²² It is also possible that the stay-at-home policy provided the perpetrators with increased control of their partners’ movements and contacts, could have reduced the perceived need for serious threats and acts of violence.^{23 –26} Less stress in daily life due to the lockdown, could have eased tensions in the relations between the perpetrator and the victim. Thus, the use of crisis shelters may not necessarily reflect the level of violence in the community.

Lack of information about open crises shelters during the lockdown may have further limited the use of the shelters. Pandemic restrictions may also have influenced the possibility of reaching the crisis shelters by physical visits. Restrictions in transport possibilities, use of home offices, and a generally lockdown of society might explain why there were fewer users, as pointed out by a Norwegian survey among leaders at the crisis shelters. ¹⁹ Most of the crisis shelter leaders saw a reduction in numbers of requests during the lockdown, while the rates returned to normal when the strictest control measures were removed. Similar results are described in a qualitative study among crisis shelter staff in Canada. ⁵

Our study shows a shift from daytime visits to phone contacts beginning on 12 March 2020 when the restrictions were established. For the capital, phone communication dominated daytime contacts for the subsequent 5 months, even though the shelters were open for domestic violence victims. The capital had relatively more COVID-19 cases than the rest of the country. The population density is higher in the capital than in the rest of Norway, and the restriction of the population during the pandemic was stricter in the capital, with lockdown for longer periods, and longer periods of restrictions in, for example, visiting other people, presence at work and had reduced accessibility of public transport. When physical attendance is difficult, a change to phone contact is logical, and this has also been found in studies from other parts of the health care sector. ²⁷ Therefore, it might be useful, looking forward, to have a plan for increasing the number of phone lines to the crisis shelters, and to have more staff available to operate the phones. A qualitative study from Sweden mentions these kinds of challenges for social work professionals at women’s crisis shelters. ²⁸

This study found a higher proportion of first-time contacts during the pandemic than pre-pandemic, both for Norway as a whole and the capital in particular. The stay-at-home policy might have mobilized a new group of users that managed the violent situation before the pandemic, for example, by having the possibility to leave home for work or other activities. Furthermore, recurrent users might have used the crisis shelter less during the pandemic period due to their confinement in their homes. During the pandemic, there were also fewer women with residential stays that reported to have children at home, compared to pre-pandemic. It is likely that victims of domestic violence are protective of their children, and reluctant to seek help that might put the welfare of their children at risk.

Visitors to the crisis shelters in the capital showed an increase in psychological violence and threats during the pandemic year that were not seen in the rest of the country. It is difficult to explain this finding and the differences were small. However, this could be related to the stricter regulations that were introduced in the capital during the first year of the pandemic, such as closed shops, staying at home, and travel restrictions. Moreover, partial and permanent loss of work due to massive lockdown in the capital, led to economic problems and increased stress for many people in Oslo.

Strength and limitations

A strength of this study is that it included the total registered use of all the crisis shelters in Norway for the 2 years 2019 and 2020 and 92.4% consented to participate in detailed registration, making the data representative for the whole country.

A limitation of the study is that phone contacts were not registered in 2019. It was therefore not possible to compare phone contacts before and during the pandemic. The registrations were anonymous so recurring stays or visits could not be revealed. However, first time users were recorded and constituted the data used in analyses regarding changes in user characteristics. Another limitation is the data quality of the registration forms used by the crisis shelters. The forms are, as far as we know, not tested for validity and reliability. Future research on similar topic should include qualitative information, like interviews with the shelter leadership. This would add important information not covered by the present study. Generally missing was lower in the pandemic period and lower for the rest of the country than in Oslo. The number of missing was below 3% except for information from shelter residents regarding information on assaulter (intoxicated 33% missing and origin 12% missing). Results from those two variables cannot be generalized. Since only univariate analyses was performed missing data did not impacted the other results.

Conclusion

The utilization of Norwegian crisis shelters, especially daytime visits, was lower during the COVID-19 pandemic than the year before. There was a shift in daytime contacts from visits to phone contact at the pandemic outbreak. The proportion of first-time users was higher during the pandemic compared to pre-pandemic period both for residents and daytime visitors, and fewer of the first-time residents reported to have children at home. The background of the crisis-shelter users did not differ between the capital and rest of Norway, but the capital had relatively more residents who experienced psychological violence and threats during the pandemic compared to the rest of the country.

Knowledge from this study is important for policy makers when planning for future crises or pandemics that would involve similar public restrictions and movement regulations. Efforts must be undertaken to ensure that information about available crisis shelters reaches the total population. The crisis shelters as well as other important help services should be prepared for higher volumes of phone contacts. Future planning should include consideration of ways to ensure the best possible care for people exposed to IPV when an unexpected event such as a pandemic occurs.