Changes in ambulance departures for assault calls during COVID‐19 pandemic restrictions

Abstract

Restrictions related to COVID‐19 changed the daily behavior of people, including the expression of violence. Although an increased incidence of violent behavior, especially domestic violence, was expected during the pandemic, retrospective analyses have yielded mixed results. Records of ambulance departures to address injuries caused by assaults in the Pilsen region, Czech Republic, during the restrictive measures during the national state of emergency were compared to data from 3 previous years using general linear models. The number and severity of assaults were analyzed for the whole sample and separately for patients of either sex, for residential or nonresidential locations, and for domestic violence. Controlling for the seasonal effects, the number of assaults decreased by 39% during the pandemic restrictions compared to the 3 previous years. No difference was found between the effects of restrictions on assaults resulting in an injury of a male or female patient. The decrease was specifically pronounced in the sample of assaults in nonresidential locations, while no effect of restrictions was observed in assaults in residential locations and domestic assaults. Pandemic restrictions were associated with a decreased incidence of violent assaults that required ambulance services. Although the incidence decreased especially in those assaulted outside of their homes, we found no support for an increase in domestic violence or violence against women. Pandemic restrictions may have served as a protective rather than a risk factor for assaults severe enough to warrant a call for ambulance services.

1. INTRODUCTION

In 2020 orders to stay home (lockdowns) were issued by many governments to limit the transmission of SARS‐CoV‐2. The lockdowns, in combination with other efforts, were helpful in limiting the spread of the illness (Dzúrová & Květoň, 2021). However, it appears that the long‐term involuntary containment of people in their homes during the lockdown periods may have been associated with changes in their behavior, including the perpetration of violence. Some of these changes may be reflected in an altered pattern of use of health services including ambulance departures to address assault injuries.

1.1. Reports from trauma centers

Changes in service use patterns have been observed particularly in American Emergency Departments (EDs), surgical departments, and trauma centers. A retrospective study examining trauma volumes before, during, and after a 2‐month COVID‐19 lockdown period at a single center in San Francisco, California, found an overall reduction of trauma volume during lockdown periods (Matthay et al., 2021). However, the incidence of violence‐related injuries persisted. Indeed, the proportion of violence‐related injury mechanisms accounted for 46% of injuries during the lockdown. Trauma team activations were used to assess trauma volume with a total of 20,129 trauma team activations from January 2015 through June 2020 included in the study. A similar retrospective study examining the effects of a COVID‐19 lockdown on trauma volume using the registry of the Virginia Commonwealth University found that the proportion of traumas due to violence showed a nonsignificant decrease from 17.6% in the control period to 12.6% during the lockdown. Furthermore, the proportion of chronic alcohol abuse was significantly higher in the lockdown period (15.5% vs. 6.8%, p < .01; Leichtle et al., 2020). In this study, patient demographics and details of trauma sustained during a lockdown period of 61 days in March and April 2020 were compared with the trauma sustained during control periods spanning the same dates in 2018 and 2019. A total of 1317 trauma activations were included.

In another retrospective chart review from a trauma center in Las Vegas, Nevada, all orthopedic trauma consults placed during the 45 consecutive days of the state‐wide lockdown period in March and April 2020 were examined (Lubbe et al., 2020). Calendar‐matched periods in 2018 and 2019 were used as controls. A total of 1113 consult notes were reviewed. Overall, the results showed that orthopedic trauma consults (proxies of trauma volume) decreased during the lockdown in comparison with the control periods. However, there was an increase in gunshot wounds during the lockdown. This finding was consistent with other reports (Abdallah et al., 2021; Sutherland et al., 2021). A similar retrospective chart review during the same 2020 lockdown period was implemented at a rural trauma center in South Carolina (Rhodes et al., 2020). A total of 2900 patients presented to the ED during the lockdown period. The patient set (cases) was compared with 7008 ED patients (controls) admitted in 2019 during a period that was calendar‐matched with the 2020 lockdown. A statistically significant increase in assaults was detected during the 2020 lockdown period in the cases in comparison with controls.

Records of all trauma patients requiring consultation between March 1 and April 11, 2020, at the department of plastic surgery at the University of Chicago Medical Center were retrospectively reviewed (Hassan et al., 2020). March 1–21, 2020 was the period preceding the start of the lockdown (control period) and the lockdown period was March 22 to April 11, 2020. There were 88 consultations in the control period and 62 during the lockdown. The main findings were “an increase in the percentage of assault‐related injuries including those associated with domestic violence” and a decrease in traffic accidents. The number of assaults increased from 18 to 21 as did the number of domestic violence cases (from 2 to 5 cases).

Data from Europe suggest mixed findings regarding violence exposure during periods of lockdowns during the pandemic. For example, a report from a large ED in Berlin, Germany compared the ED cases (n = 107) observed during a 35‐day COVID‐19 lockdown period with controls observed during a calendar‐matched control period (n = 168; Maleitzke et al., 2021) and found a significant decrease of total cases during the lockdown period, but significant increases of incidence proportions for injuries due to domestic violence (RR = 2.41), as well as for substance abuse. In a French study, patients requiring urgent care for upper limb injury during a 55‐day COVID‐19 lockdown period (cases) (n = 279) were compared with control patients (n = 784) seen during the equivalent time period in 2019 (Pichard et al., 2020). The results showed a decrease in the rates of road, work, and leisure accidents and an increase in domestic accidents and undefined “aggressions.” The increase in “aggressions” was not statistically significant. Data from a single hospital in Denmark show no difference in risk ratio for violence comparing the number of patients admitted in 2020 with the average of the previous 2 years (Trier et al., 2022). An increase in the incidence of minor injuries was found but there was no change in the severe injuries when not distinguishing the cause and reduction in traffic‐related injuries. A decline of 24.2% in total traumatic injuries was observed in a Swedish register‐based retrospective study comparing the patients admitted during the first wave of COVID‐19 with the respective period of the previous year (Bäckström & Wladis, 2022). An insignificant increase in the proportion of assaults (from 7.2% to 8.4%) was also noted. Records of patients presenting in the King's College Hospital in London show a decrease in the number of patients presenting due to gunshot injuries or injuries inflicted with a sharp object during the first two lockdowns (spring and autumn 2020) compared to the respective periods a year before (Hickland et al., 2022). A minor increase in the third lockdown (winter 2020–2021) was found but no tests of statistical significance were provided. Further, no difference was found in intimate partner violence (IPV), and “[a]necdotal evidence from the hospital's emergency department (ED) doctors” is reported, indicating that IPV “decreased in the first national lockdown of 2020 but remained at normal levels during subsequent lockdowns” (p. e127).

Finally, studies from Asian and Pacific countries found decreases in the total trauma volume and no change in assault‐related injuries. A retrospective cohort study of 263 patients admitted to an Indian trauma center for injuries during a 68‐day lockdown period in 2020 reported that a total of 9 injuries were due to assault (Dhillon et al., 2020). In a calendar‐matched control group of 2019 admissions (n = 611), there were eight injuries due to assault. During the lockdown, there was a decrease in the total trauma volume, but no change in the small absolute numbers of injuries due to assault. In a study from China, a total of 757 trauma center patients from two hospitals were observed during a 32‐day COVID‐19 lockdown period (“community quarantine”; Zhu et al., 2020). During the lockdown, the number of traffic and outdoor injuries first decreased, but then increased. The number of indoor injuries remained stable. This study had no control condition (no observations outside the lockdown period) and it is not clear how the injuries were sustained. In a single trauma center in India, 5485 patients were admitted during a lockdown, and the month after the lockdown was released. Although the total number of patients increased post‐lockdown, no change was observed in the subcategories of assaults or domestic violence (Vatsya et al., 2022). Data from a single trauma center in Saudi Arabia reveal a decrease in trauma patients seeking help during lockdown compared to respective periods of the 2 previous years together with an increase in the proportion of those with injuries secondary to an assault (Hakeem et al., 2021). A retrospective, the register‐based study observed a decrease in major trauma admissions in New Zealand, with no difference in the proportion of self‐inflicted, unintentional, and “other” injuries, comparing the periods of restrictions in 2020 with the respective period in 2019 (McGuinness & Harmston, 2021).

The reports from trauma centers from around the world suggesting discordant results have used convenience patient samples, with no claim to represent any general population. The patients selected for these studies were those whose injuries were serious enough to warrant at least surgical evaluation; this means that most victims of violence, whose injuries, if any, were less severe, were not represented in these studies. These studies also failed to provide standardization in the definition, duration, and enforcement of the lockdowns. Thus, there are many reasons for the variation between the results regarding the presence or absence (or increase vs. decrease) of violence during the lockdown. However, there is one area of agreement: there was a general decrease in trauma volume during the lockdown. This decrease is due in part to the reduction of traffic accidents and outdoor sports injuries, but maybe also due to the fear of possible COVID‐19 infection should one go to the ED for the treatment of a minor injury that might be managed at home (Moynihan et al., 2021).

1.2. Evidence from reviews

Waseem et al. (2021) analyzed 57 studies of the effects of the covid pandemic on trauma from multiple European countries, along with the United States, New Zealand, Australia, China Hong Kong, India, Iran, Pakistan, and South Africa, reporting a reduction in trauma rates together with a higher proportion of traumatic injuries secondary to interpersonal violence or deliberate self‐harm and increasing proportion of injuries sustained at home. Beiter et al. (2021) analyzed 28 reports comparing community violence rates or proportions during COVID against the previous baseline published until December 2020 from multiple European countries, United States, Australia, South Africa, India, Pakistan, Brazil, and Turkey. Of these, 14 studies found no difference, while 8 studies observed an increase and 6 studies a decrease in violent trauma.

Across multiple studies from developed countries, the number of patients admitted with orthopedic trauma decreased during the pandemic by 52% on average, with a relative increase in the proportion of traumas secondary to assaults and falls occurring at home, and a relative decrease in other aetiologies, such as sport and traffic injuries (Waseem et al., 2021). Change in the rates of injuries related to assaults, however, has not been sufficiently examined, as most studies reported the proportions of trauma secondary to assaults rather than assault rates (Beiter et al., 2021).

1.3. Reports using other sources of data

To address the possible effects of lockdown on aggression, the Buss–Perry Aggression questionnaire (BPAQ; Buss & Perry, 1992) was administered to 5928 US adults during the first 6 months (April–September 2020) of the COVID‐19 pandemic (Killgore et al., 2021). Data across the 6‐month period were compared between those currently under lockdown and those not under lockdown. Results indicated that during the period of June through August 2020, total aggression scores and physical aggression scores were significantly higher for persons who were currently under lockdown relative to those who were not. In a Mexican study examining the impact of COVID‐19 lockdown on crime against women (Hoehn‐Velasco et al., 2021) using the national crime register as the source of data which includes the entire population of women, results indicated that sexual crimes and domestic violence declined during the lockdown and then increased again to the pre‐lockdown level. In Mexico, the official national stay‐at‐home order went into effect on March 23, 2020, and it continued until May 30. The reasons for the decline are complex. For example, some, but not all, municipalities banned the sale of alcohol. The ban on alcohol was associated with statistically significant reductions in domestic violence and crimes against women.

Although reductions were noted in Mexico, in the United States, many police departments reported increases in domestic violence and other violent behavior during the lockdown period in March 2020 (Boserup et al., 2020). These departments were in the states of Oregon, Texas, Alabama, and New York. Police reports of violence are usually generated as a response to citizens’ complaints. The decision to complain depends on the seriousness of the resulting injury (if any), the victim's level of fear (which may partly depend on the community's tolerance of violence), and the opportunity to make the complaint. In some cases of domestic violence, the perpetrator controls the victim's access to telephones (Chatterjee et al., 2018). If that happens in many homes, group results may show a paradoxical (and spurious) decrease in domestic violence. Other studies supported the increase in emergency calls related to domestic violence increased in police and nongovernmental organizations' statistics, specifically in the United States (Leslie & Wilson, 2020; Mohler et al., 2020), although some studies also found mixed results (Kofman & Garfin, 2020), as well as no change (Sorenson et al., 2021). A slight decrease in domestic violence was observed in the criminal records of Chicago (McLay, 2022).

An increase in domestic and IPV against women during the pandemic was expected even before the widely cited report on calls to the police (Boserup et al., 2020), based on several factors: a decreased ability of victims to leave households shared with the perpetrator, an assumed decrease in readiness of public institutions to act against domestic violence during lockdowns, and the experience from other pandemics such as the Ebola pandemic in Western Africa (Bradbury‐Jones & Isham, 2020; John et al., 2020; Kofman & Garfin, 2020; Peterman et al., 2020; Sacco et al., 2020). Some medical reports, however, revealed a decrease in patient admissions related to domestic or sexual violence. Gosangi et al. (2021) compared patients “screening positive for or reporting IPV” and the victims of physical intimate partner violence (IPV) between March 11 and May 3, 2020, with the same period of the 3 previous years treated at “a large urban academic medical center located in the north‐eastern United States.” The number of IPV victims decreased from an average of 114 between 2017 and 2019 to 62 in 2020, while the number of physical IPV victims increased from an average of 14 to 26. Muldoon et al. (2021) compared the number of patients of the Sexual Assault and Domestic Violence Program at The Ottawa Hospital between March 4 and May 5, 2020, to the same period in 2018, recording a decrease of 55.84%, equaling a weekly decrease of 4.66 patients. In the Czech Republic, violence against women was examined in a qualitative study showing an increased demand for services from nongovernmental organizations supporting victims of domestic violence (Niklova & Moree, 2020), while records of governmental intervention centers indicated no change in domestic violence (Vojtiskova et al., 2020).

The existing literature provides an inconsistent picture of the development of violence‐related injuries during the pandemic and related restrictions. Although the overall number of traumatic injuries decreased, especially for injuries sustained outdoors, according to most reports, the heterogenous results addressing injuries secondary to an assault warrant further investigation.

1.4. Present study

In the current study, we examined the rate of violent assault in the Czech Republic during the pandemic lockdowns. Ambulance departures were used as a measure of assault rates consistent with other studies (Backe & Andersson, 2008; Cusimano et al., 2010; Fullerton et al., 1998). Specifically, we examined the records of the emergency centers of the Pilsen Region in Czechia, servicing its 590,461 inhabitants, thus increasing the size of the data set and improving generalizability compared to data from single hospitals. Ambulance records are ideal to examine assault rates because they contain quantification of the severity of the injuries sustained, allowing researchers to rule out possible differences in the willingness to call an ambulance due to the pandemic. Reduction in care utilization due to the pandemic was especially pronounced in less severe cases (Moynihan et al., 2021), and as such, would lead to an increase in the average severity of the injuries, as the hesitations would be more likely to affect requesting ambulance departures for minor rather than life‐threatening injuries (as reported by Gosangi et al., 2021). We analyzed the records of the ambulance departures in the Pilsen region, Czech Republic, comparing the periods of COVID‐related restrictions (March 12 to May 17, 2020, and October 5, 2020, to April 11, 2021; see Appendix A) with records from the 3 previous years. Particularly, we examined the effects of COVID‐related restrictions on the incidence of physical assaults leading to ambulance departures. Based on previous warnings against gender‐based and domestic violence (Bradbury‐Jones & Isham, 2020; John et al., 2020; Kofman & Garfin, 2020; Mittal & Singh, 2022; Peterman et al., 2020; Sacco et al., 2020) during the pandemic restrictions and research differentiating between injuries secondary to assaults sustained at home or elsewhere (Shepherd et al., 2021), we included the sex of the victim, the residential versus nonresidential location of an assault, and if the injuries were sustained because of domestic assaults. To detect possible changes in willingness to request an ambulance, records of the severity of the injuries sustained were analyzed along with the frequency of assaults.

2. METHODS

2.1. Setting

In 2020, the Pilsen Region had 590,461 inhabitants served by 25 Emergency Medical Centres. About 1200 daily calls to the medical emergency line are answered at the regional headquarters located in the city of Pilsen. The ambulance operation system is fully covered by compulsory health insurance for all citizens of the Czech Republic. Emergency phone operators at the Emergency Medical Services are paramedical workers with professional qualifications, including a university degree in the field of paramedical science. Paramedical workers complete 3 months of training targeted at operation procedures, legislation, receiving an emergency call, the evaluation and assignment of urgency, and providing first aid by phone. The training process is monitored by monthly tests and a final examination. The medical team at the place of ambulance departure further checks, and if necessary, corrects data obtained by the phone operators, including medical status and reasons for departure. These corrected data were used in our analyses.

2.2. Measures

Records of ambulance departures in the Pilsen region, Czech Republic, to an event evaluated as an assault by the emergency phone operators between January 1, 2017, and April 30, 2021, were received from the Emergency Medical Service of the Pilsen Region. The categorization of an event as an assault is based on the emergency phoneline operators' evaluation, based on the description of those calling the line claiming that the suspected injury was caused intentionally by an act of aggression (the assault also includes attacks by animals which were omitted from the data set used). The data received included the date, brief verbal description of the emergency call and local situation by the phone operators and/or paramedics, sex of the patient, and NACA classification of injury severity (Dami et al., 2015) recorded by the paramedics examining the patient. Based on the verbal descriptions, the location of the assault was coded, distinguishing between assaults at residential locations (including accommodations with presumed long‐term residence, such as dormitories, workers’ hostels, retirement, or children's homes) and other locations, such as outdoor areas, bars, working premises, and so forth. Domestic violence was coded based on the verbal descriptions by the phone operators, with domestic violence including cases when the phone operator labeled a family member or life partner as the alleged perpetrator; all other assaults were coded as nondomestic. Records of 3747 departures were received, resulting in 2866 cases after removing duplicates, assaults by animals, and assaults with unknown sex of the patient or unknown location.

2.3. Data analysis

To examine the effect of covid restrictions on the frequency of assaults, generalized linear regression based on quasi‐Poisson distribution was chosen as the most appropriate measure combining feasibility for count data with overdispersion with relative simplicity (Osgood, 2000; Walby et al., 2016). The effect of COVID‐related restrictions on weekly assault rates was analyzed, with the first week starting on Monday, January 2, 2017, and the last on April 19, 2021. Restrictions were defined as the period of national emergency state, including March 12 to May 17 and October 10, 2020, to April 11, 2021 (see Appendix A for details). To analyze the possible effects of sex, the location of the assault, and possible specifics of domestic violence, the model included interactions of these variables with restrictions. To control for seasonal fluctuations in rates of violence observed in Europe (Morken & Linaker, 2000; Rock et al., 2008; Tiihonen et al., 2017; Uittenbogaard & Ceccato, 2012), each week was further labeled with a number approximating seasonal effect, calculated as the difference between the given month and June in months (i.e. 0 for June, 1 for July, 6 for December, 5 for January, etc.). Possible shifts in the severity of injuries were tested, after assaults with missing NACA values were dropped, by the Mann–Whitney test on all assaults. Analysis was conducted using R version 3.6.3 and packages tidyverse (Wickham et al., 2019) and sjPlot (Lüdecke, 2021).

3. RESULTS

As displayed in Table 1, there was a reduction in ambulance departures to assault‐related injuries during restrictions (IRR = 0.50; p < .01), while seasonality had no effect (p = .52). In terms of interactions, there was an additional increase in ambulance departures to injuries secondary to an assault in residential locations similar in magnitude to the reduction of all assaults (IRR = 1.92; p < .01). There were no difference in departures for male versus female patients with assault‐related injuries (p = .36) or to departures to assaults related versus not related to domestic violence (p = .34).

Table 1.

Incidence rate ratios of weekly ambulance departures to injuries secondary to an assault depend on seasonality, time, restrictions, and interactions of restrictions with the sex of the patients, location of the assaults, and domestic versus nondomestic assaults

Predictors	IRR	CI	p
(Intercept)	2.42	2.07–2.83	<.001
Seasonality	0.97	0.94–1.00	.052
Restrictions	0.50	0.32–0.75	.001
Sex [male]	3.04	2.65–3.50	<.001
Residential	0.40	0.35–0.46	<.001
Domestic	0.09	0.07–0.11	<.001
Restrictions * Sex [male]	0.82	0.53–1.28	.364
Restrictions * Residential	1.92	1.28–2.86	.002
Restrictions * Domestic	1.36	0.69–2.49	.340

Note: Statistically significant results are in bold.

Abbreviations: CI, 95% confidence interval; IRR, incidence rate ratio.

Severity of the injuries secondary to an assault leading to ambulance request was lower during the time of restrictions (W = 491,761; p < .01; median 2.00 vs. 2.00; mean 2.20 vs. 2.31 ¹ ) in the whole sample, as well as in the subsample of males (W = 256,416; p < .05; median 2.00 vs. 2.00; mean 2.21 vs. 2.34) and in the nonresidential locations subsample (W = 136,598; p < .05; median 2.00 vs. 2.00; mean 2.31 vs. 2.18), while it did not differ in the subsample of ambulance departures to residential locations (p = .46), females (p = .46) or in domestic assaults (p = .57).

4. DISCUSSION

The number of ambulance departures for assault calls decreased during pandemic restrictions and the severity of the injuries was also reduced during the long period of restrictions, albeit to a lesser extent. With the severity of assaults not increasing, we interpret the observed decrease in ambulance departures as decreased incidence of assaults rather than a decrease in willingness to call the ambulance. Our results are consistent with findings of reduced frequency of traumatic injuries observed in the United Kingdom (Olding et al., 2021), the United States (Leichtle et al., 2020), and decreases in the number of assault‐related injuries observed in Berlin, Germany (Maleitzke et al., 2021), London, UK (Olding et al., 2021), Richmond, Virginia, US, (Leichtle et al., 2020), and Pietermaritzburg, South Africa (Zsilavecz et al., 2020), as well as our observation of a decrease in assault‐related hospitalizations in the Czech Republic using nationwide data (Pisl et al., 2022). A recent review of the literature found no difference in the incidence of trauma during the pandemic (Beiter et al., 2021); however, this may be due to the inclusion of studies reporting increased proportion rather than the rate of violent trauma. Some differences may also be explained by low crime rates and firearm possession rates in the Czech Republic (10 times lower compared to the United States; Alpers & Picard, 2021a, 2021b), as disproportions in effects of the pandemic on violent trauma may be related to firearm accessibility (Beiter et al., 2021) and increase in gun violence was observed in the United States (Abdallah et al., 2021; Kim & Phillips, 2021).

No difference was found between the effect of restrictions on male and female patients and assault‐related injuries decreased in nonresidential locations while stagnating in the residential ones. With the severity of the injuries remaining stable and decreasing, there was no evidence of an increase in hesitation with respect to ambulance calls. Therefore, our results suggest a decrease in the number of assaults rather than in willingness to seek medical help. With respect to location, our results are in line with results from Cardiff, UK, indicating a decrease in the frequency of assaults in patients injured “outside the home” during lockdown (Shepherd et al., 2021) and a relative increase in injuries sustained indoors reported by multiple studies worldwide (Waseem et al., 2021). With respect to violence against women, our results are consistent with the decreases observed in Mexico (Hoehn‐Velasco et al., 2021), although this effect was associated with the ban of alcohol sales in several Mexican states. In Czech Republic, the consumption of alcohol was reported to decrease only moderately during the pandemic (Kilian et al., 2021), remain stable (Csémy et al., 2021), or even increase (Winkler et al., 2020). Considering that Czech nationwide register data revealed a larger decrease in assault‐related hospitalizations in males than in females during lockdowns (Pisl et al., 2022), no difference between males and females may be also the result of a lack of statistical power.

No differences in either rate of ambulance departures to domestic assaults or the severity of the injuries sustained were observed (taking into consideration the confidence intervals of the effects of restrictions per se as well as their interaction with domestic violence), suggesting that pandemic restrictions had no effect on the incidence of domestic violence or that this effect was too subtle to be confirmed. These results are consistent with the Czech study of governmental center's records finding no difference comparing the first wave of pandemic restrictions till June 2020 with the previous year (Vojtiskova et al., 2020). These results are also within the range of other European studies: an increase from 14 to 20 cases of domestic violence was observed in Berlin, Germany (Maleitzke et al., 2021), a decrease from 41 to 19 cases of IPV and an increase in domestic violence (excluding IPV) from 0 to 2 in South London, UK (Olding et al., 2021), and change from 10 in the control period (of 98 days) to 5 in the lockdown period in Catania, Italy (54 days; Di Franco et al., 2021). The stable incidence rate and statistically insignificant increase in the proportion of domestic violence observed in our study are also compatible with a recent meta‐analysis (Piquero et al., 2021) that reported an increase in domestic violence, predominantly in the United States, while not distinguishing between studies reporting rates and proportions. Our results do not show a pattern of decrease in the number of intimate partner violence victims accompanied by increased severity suggesting postponing of help‐seeking observed in the north‐eastern US (Gosangi et al., 2021). The contradiction between our data and some police and helpline reports of increased domestic violence (Boserup et al., 2020; Leslie & Wilson, 2020; Mohler et al., 2020), including those offering help to domestic violence victims in the Czech Republic (Niklova & Moree, 2020), may have several explanations. First, the potential increase in domestic assaults may be masked by the victims being less likely to admit the true cause of their injuries when requesting an ambulance during the pandemic, due to the inability to leave premises shared with the perpetrator (Niklova & Moree, 2020). In this case, we would, however, be likely to observe an increase in the severity of the injuries, although more complex interactions cannot be ruled out based on a single measure. Second, domestic violence may have other forms than physical assaults examined in our study: for instance, sexual, emotional, psychological, or economic violence, and stalking (Sacco et al., 2020) that may have followed different patterns than physical violence during the restrictions. Third, the increased demand for services of respective nongovernmental organizations (NGOs) may be to some extent explained by their increased promotion or medialization of domestic violence during the pandemic. The dynamics of the incidence of different forms of domestic violence under restrictions should be addressed by further research, including the interaction of readiness to admit domestic violence with the medialization of the topic and societal changes.

Strengths of our study included the sample size which allowed for meaningful analyses of male and female subgroups and accounted for the effects of assault severity, location, and seasonality. Compared to records of NGOs, intervention centers, police statistics, or Google searches, our data are less likely to be affected by changes in the medialization of the problem of violence during the pandemic or the hesitation of law enforcement services to act during the pandemic. The Czech Republic has a very low rate of violent crime (e.g., a murder rate of 0.61 per 100,000). Thus, there is a relatively limited “background noise” of general violence that might affect research on assaults.

Despite these strengths, there are limitations to our study that require comment. Our data only include assaults resulting in a request for an ambulance. Thus, we cannot distinguish between a change in the frequency of assaults and a change in the chance of an ambulance being called. With changes such as the closure of facilities serving alcohol, more assaults not resulting in ambulance requests may have happened on private premises. While our data do not show an increase in the severity of reported injuries that would be expected if this was the case, such an explanation cannot be fully ruled out. With respect to domestic violence, our results are based on a small sample and limited to cases labeled as such by emergency line operators. We note that only a small fraction of domestic violence is revealed even during personal emergency visits (Kothari & Rhodes, 2006).

The frequency of ambulance requests to assaults decreased substantially during the pandemic restrictions, by 39% during the pandemic restrictions. The decrease was primarily associated with assaults that occurred outside of private residences, while no effect of pandemic restrictions was found on assaults occurring at home or in relation to domestic assaults. No difference was observed between male and female patients. The lack of increase in the severity of the injuries secondary to assaults during the pandemic suggests that the reduction of calls for an ambulance was due to a decrease in the number of assaults rather than to a reduced willingness to call emergency services. Therefore, the records of ambulance departures in the Pilsen region, Czech Republic, suggest that pandemic restrictions may be a protective factor against physical violence that is severe enough to warrant a call for ambulance services.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

APPENDIX A. OVERVIEW OF RELEVANT COVID‐RELATED RESTRICTIONS IN CZECHIA

• 1.National state of emergency was defined by the laws and regulations 2020/69, 2020/156, and 2020/219 in the period between March 12 and May 17, 2020, and by 2020/391, 2020/521, 2020/593, 2021/21, 2021/59, and 2021/69 in the period between October 5, 2020, and April 11, 2021. This period almost fully overlaps with closure of bars and restaurants in the periods March 13, 2020, to May 10, 2020, October 22, 2020, to December 2, 2020, and December 18, 2020, to April 11, 2021, besides other measures being also in effect (restriction of public and private gatherings, closure of schools, etc).

All regulations available online at zakonyprolidi.cz/cs/yyyy‐nn with yyyy‐nn denoting year and number of the regulation (for instance, https://www.zakonyprolidi.cz/cs/2020-69).

Pisl, V. , Vevera, J. , Štěpánek, L. , & Volavka, J. (2023). Changes in ambulance departures for assault calls during COVID‐19 pandemic restrictions. Aggressive Behavior, 49, 76–84. 10.1002/ab.22055

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon request, depending on the consent of the Emergency Medical Service of the Pilsen Region.