A systematic review of infectious disease outbreaks and violence against women and girls: changes in magnitude, mechanisms and lessons for the future

Abstract

Infectious disease outbreaks (outbreaks) are increasing across the globe due to climate change, urbanisation and changes in land use, and many of their response measures impact risk factors for violence against women and girls (VAWG). We conducted a systematic review to consolidate existing evidence on the impact of any outbreaks, and their public health responses, on the change in magnitude of VAWG and mechanisms facilitating violence among women and girls in low-income and middle-income countries. Though our search strategy aimed to capture studies from any outbreak since 2014, all quantitative evidence on VAWG impacts, and all but one qualitative study, focused on the COVID-19 pandemic, and only three studies disaggregated outcomes for women versus girls. Overall, our synthesis of the evidence points to increased VAWG during the first year of the COVID-19 pandemic compared with pre-pandemic levels. We identified five broad mechanisms through which violence occurred against women and girls: (1) income loss due to economic shutdown, financial insecurity, and/or job loss, (2) movement restrictions, (3) changes in access to public services, (4) fear of exposure to infectious disease, and (5) a legacy of mistrust in health systems from previous outbreaks. Our study demonstrates the novelty of VAWG monitoring during outbreaks, the need for increased surveillance and the known mechanisms to date through which VAWG may be perpetrated during outbreaks. By implementing both short-term protective measures and long-term structural reforms, outbreak responses may not only break the cycle of VAWG exacerbated by public health emergencies but build resilient systems that protect women, girls and marginalised populations before, during and after crises.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• There is growing recognition that infectious disease outbreaks can exacerbate risk factors for violence against women and girls (VAWG), particularly through economic strain and social isolation. However, systematic evidence on the magnitude of this violence and its mechanisms remains limited, especially in low-income and middle-income countries (LMICs).

WHAT THIS STUDY ADDS

• This study provides the first comprehensive synthesis of the effects of the COVID-19 pandemic and Ebola outbreaks on VAWG. Our findings point to an overall increase in VAWG during the first year of the COVID-19 pandemic in LMICs, and five key mechanistic drivers across COVID-19 and Ebola: income loss, movement restrictions, reduced access to services, fear of exposure to infectious disease and a legacy of mistrust in health systems from previous outbreaks.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Findings underscore the need for systematic VAWG monitoring during all outbreaks and we call for integration of VAWG prevention and response in public health emergency planning. Policymakers and practitioners can use these insights to design outbreak responses that mitigate VAWG risks through both immediate protective measures and long-term structural reforms.

Introduction

Infectious disease outbreaks (‘outbreaks’) are increasing in frequency, magnitude and impact across the globe¹ due to factors including climate change, urbanisation and changes in land use, and are one of the leading causes of death worldwide.² Given their ability to spread quickly with profound safety and health impacts across a diverse range of communities, outbreaks place substantial burdens on healthcare systems, social systems and economies of affected communities. In low-income and middle-income countries (LMICs), where limited health system resourcing intersects with poor living conditions, vulnerabilities to outbreaks are exacerbated.2,4

LMICs are often at unique risk of experiencing outbreaks of preventable diseases, such as malaria, cholera, measles and Ebola, despite there being known measures to minimise many types of outbreaks. Several factors play a role in preventing these outbreaks, but many of these measures require available resources and finances. High-income countries have comparatively advanced healthcare and surveillance systems, established research funding for preventing and curing diseases, vaccine availability, and efficient health communication mechanisms.^{5 6} The impacts of outbreaks can be exacerbated by gaps that are pronounced in LMICs, such as limited healthcare access, overcrowding, poor sanitation and malnutrition.^{6 7} These limitations ultimately contribute to higher rates of disease transmission, illness prolongation and death.^{6 8} Multi-sectoral interventions in the community, including water and sanitation, are important to consider for both prevention, response, control and mitigation efforts.⁹

Many outbreak response measures impact risk factors for violence against women and girls (VAWG), or “any act of gender-based violence that results in, or is likely to result in, physical, sexual, or mental harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life”,¹⁰ putting women and girls at an increased risk.¹¹ For example, control measures implemented to prevent outbreak spread, such as lockdowns during COVID-19, may lead to increased close proximity with family members and intimate partners, and heightened tensions within households. With outbreaks straining healthcare systems, weakening economies and disrupting typical functioning of societies, individuals are exposed to increased levels of stress related to job loss, illness, finances and lifestyle changes that are prone to turn into acts of domestic violence (DV).^{11 12} They can also exacerbate existing problems, by making individuals more violent or by taking away other social interactions, meaning women and girls are trapped with their abusers. Outbreaks in LMICs push those who are economically vulnerable into poverty, which may force families to consider child marriage as a means of alleviating financial stress.¹² As a result of school closures and mandated lockdowns, the likelihood of school drop-out, child labour, child marriage and adolescent pregnancy may increase, as well as result in reduced access to age-appropriate health education and information.

Several reviews synthesise qualitative or quantitative evidence limited to COVID-19 responses in early 2020, certain subcategories of VAWG such as domestic violence, or in high-income countries only. This review is unique in that it seeks to (1) synthesise evidence across disease outbreaks in a 10-year period, (2) integrate qualitative and quantitative data, (3) use comparable quantitative data for comparability, (4) include all types of VAWG, and (5) is specific to LMICs, to pull out learnings specific to those countries that are most vulnerable to infectious disease outbreaks.13,19 Ultimately, the purpose of this systematic review is to quantify and consolidate evidence on the impact of outbreaks and resulting public health responses on VAWG in LMICs. The results of this review are intended to assist public health practitioners, healthcare and other social service workers, researchers, and policymakers to prepare for and respond to VAWG in the context of future outbreaks in LMICs, in order to reduce proven risks, and protect women’s and girls’ rights around the world in situations where they are most vulnerable.

Methods

We conducted this systematic review informed by Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and include the completed checklist as a supplementary file (see online supplemental file S1). The protocol is not registered because it was initiated rapidly in response to emerging evidence about public health emergencies’ impacts on VAWG in LMICs, but all relevant information needed to replicate the search is in the methods below.

Search strategy

We conducted an original search (13 May 2024), and an updated search (21 January 2026) of GenderWatch through Proquest, MEDLINE through EBSCO, and PsychINFO using search terms and Boolean operators related to VAWG, outbreaks and LMICs. We selected outbreaks that are likely to result in acute emergency responses, using the WHO’s Disease Outbreak News (DONS) as reference for our time period of interest.²⁰ The full list of search terms is available in online supplemental table 6. We searched for all studies published on or after 01 January 2015 through 21 January 2026. We selected this start date for several reasons: (1) to capture the earliest published research examining VAWG during the 2014–2015 West African Ebola outbreak, which represented a watershed moment in recognising public health emergencies’ (PHE) impacts on VAWG, (2) the 2015 publication of the Inter-Agency Standing Committee (IASC) Guidelines for Integrating Gender-Based Violence Interventions in Humanitarian Action together with the launch of the Sustainable Development Goals (SDGs) marked a shift toward more systematic VAWG documentation, and (3) a preliminary scan of pre-2015 literature revealed that research on VAWG during infectious disease outbreaks was sparse and lacked the methodological rigour necessary for systematic synthesis.

Study criteria

We considered studies to be eligible for inclusion in the systematic review if they explicitly described the effect of an outbreak or its control measures on the incidence of VAWG and/or women and girls’ experiences of VAWG and the mechanisms through which the VAWG was perpetrated. We included quantitative, qualitative and mixed methods studies that were published in Arabic, English, French or Spanish to capture the broadest scope of evidence within the technical capacities of the review team. We excluded studies that did not take place in at least one LMIC, due to our specific interest in these geographies. Importantly, no quantitative studies focusing on non-COVID-19 outbreaks met inclusion criteria for the final review stage. Therefore, to strengthen our ability to quantify the impact of infectious disease outbreaks on changes in the magnitude of VAWG, we only included quantitative articles that compared pre-COVID-19 estimates of VAWG to a post-COVID-19 period in 2020. Studies were also excluded if findings were unable to be disaggregated by gender and country.

Study screening and data analysis

We imported each of the records from the database searches into Covidence.²¹ IS, MM and three research assistants held three team meetings prior to screening to review the inclusion and exclusion criteria and practice study screening procedures. After excluding duplicates, two independent team members screened each publication’s title, abstract and full text; IS and MM resolved conflicts in weekly meetings. The data extracted for each article included study location, type of outbreak and response, form(s) of VAWG investigated (online supplemental table 7), study population characteristics, study methodology and major findings. Definitions used to categorise forms of VAWG are specified in online supplemental table 7. Forms of VAWG were designated into two categories: (1) domestic violence, relating to all forms of violence perpetrated within the domestic and private sphere, and (2) non-domestic violence and unspecified VAWG, relating to all forms of violence perpetrated in the public sphere or by someone other than an intimate partner or family member. We imported the data into a Google Spreadsheet for analysis. The team analysed data according to study type, population, type of VAWG and location of VAWG perpetration. We classified the location of VAWG perpetration to inform policymakers and public health practitioners about where interventions would be most effective.

Outcomes and reporting

We report on two outcomes of interest: changes in the magnitude of VAWG—measured quantitatively through changes in incidence, prevalence, frequency or severity—and mechanisms through which VAWG was perpetrated—measured quantitatively and qualitatively. For changes in the magnitude of VAWG, we first discuss population-based studies—studies where participants or records prospectively and/or retrospectively report perceived changes in VAWG over at least two reference time periods (‘longitudinal studies’) and studies where participants or records report perceived changes in VAWG during one reference period (‘single time point studies’). We then present studies examining changes in service utilisation based on hotline, social service or health facility-based service provision (‘service utilization studies’). Due to our analysis of these data by VAWG type, totals reported may be greater than the number of studies (eg, one study may report both decreased physical violence outside of the home and increased child marriage). For quantitative studies, major findings were statistical data related to our outcomes of interest, that is, changes in VAWG across time. We present overall descriptions, then findings according to where VAWG perpetration took place. Our second outcome of interest is the mechanisms through which VAWG was perpetrated according to where the VAWG took place. Qualitative findings were identified by applying inductive and thematic coding to extracted text around VAWG. Each theme represented a mechanism through which VAWG was perpetrated.

Quality appraisal

We used an adapted version of the Effective Public Health Practice Project (EPHPP) Quality Assessment Tool for Quantitative Studies²² to appraise the quality of each quantitative study assessing changes in the magnitude of VAWG included in the review. We simplified the EPHPP to account for the fact that studies investigating outbreaks are unlikely to have a counterfactual. We used eight questions—judged as strong, moderate or weak—to guide our quality appraisal. The questions spanned selection bias, study design, data collection methods, withdrawals and dropouts, and analyses. Two team members collaboratively assigned each publication a total score based on the majority of the responses to the eight questions; if there were equal numbers of two scores, the publication received the lower score. A third team member reviewed one third of the quality scores for consistency. An online supplemental file S2 includes quality questions, possible responses and ratings.

Results

Study characteristics

Across searches, we identified 2869 records. We removed 645 duplicates, and an additional 1979 publications based on title and abstract screening. After checking publication availability and screening for eligibility, we reviewed the full text of 245 publications and extracted data from 111 publications. In this review, we synthesise findings from a final sample of 53 quantitative studies assessing changes in magnitude of VAWG and 80 quantitative, qualitative, and mixed methods studies exploring VAWG mechanisms during the COVID-19 pandemic and Sierra Leone’s Ebola outbreak (see figure 1).23,133

Figure 1. Consort diagram of publications included in this systematic review. IDO, infectious disease outbreak; LMIC, low-income and middle-income country; VAWG, violence against women and girls.

Despite including search terms for several outbreaks that have occurred since 2014, including Ebola, cholera and Zika, only quantitative and qualitative studies exploring the impact of COVID-19 and one qualitative study on Ebola met inclusion criteria. The research team eliminated the few peer-reviewed studies that discussed the relationship between VAWG and another infectious disease outbreak during the abstract review stage, largely due to lack of direct data collection or analysis; in other words, few academic research studies have been conducted on the impact of other infectious diseases on VAWG and of those few, we did not identify any eligible studies for the criteria in this review.

Only 21 publications explicitly included girls in their study criteria, meaning that the preponderance (n=91) of studies included women only. Among the studies that did include girls, only three examined outcomes or mechanisms among girls, separately from outcomes or mechanisms among women: one study explored violence against girls, aged 13–18 years, during the COVID-19 pandemic,¹²¹ a second study on child marriage disaggregated findings for adolescents ages 14–17 years and adult women,¹⁰⁸ and a third disaggregated between rape cases (women) and sexual defilement cases (adolescent girls).⁴⁸

57 unique countries are represented in the final sample, with 31 publications from Eastern and Southern Africa, 23 from South Asia, 17 from the Middle East and North Africa, 13 from Latin America and Caribbean, nine from West and Central Africa, five from Europe and Central Asia, and four from the East Asia and Pacific region. 11 articles spanned multiple regions. An online supplemental file S3 includes characteristics for all publications included in this systematic review.

Changes in magnitude of VAWG during the COVID-19 pandemic

Out of the 53 quantitative publications assessing changes in magnitude of VAWG during the first year of the COVID-19 pandemic (2020), compared with the pre-pandemic period, there were 24 longitudinal studies, 12 single point studies, and 17 service utilisation studies. We judged 14 publications to provide strong evidence and 39 to provide weak evidence (see table 1). Across longitudinal studies and VAWG types, the preponderance of comparisons found that VAWG increased (n=33 with significance testing and one without) or stayed the same (n=23 with significance testing) during the COVID-19 pandemic; only seven comparisons with significance testing found a decrease in a type of VAWG. Longitudinal studies with significance testing are included in table 2. Table 3 contains results for service utilisation studies with significance testing. Of the comparisons in service utilisation studies with significance testing, six found evidence of increased reports of VAWG, four of similar VAWG reporting, and three of decreased VAWG reporting. Of service utilisation comparisons made in studies without significance testing, one found evidence of increased reports of VAWG, two of similar VAWG reporting and four of decreased VAWG reporting. Among single point studies, seven found increased reports of VAWG, two found similar VAWG and two reported decreased VAWG before and during the first year of the COVID-19 pandemic. See online supplemental file S4 for more information about each study and individual study citations.

Table 1. Quality assessment of quantitative studies included in this review.

Author(s)	Selection bias		Study design	Data collection method		Withdrawals and dropouts		Analyses	Overall
Abu-Elenin et al26 (2022)
Abujilban et al28 (2023)
Abujilban et al29 (2022)
Agüero32 (2021)
Aolymat35 (2021)
Asik & Nas Ozen37 (2021)
Atilla et al38 (2023)
Bardales Mendoza et al43 (2022)
Bhattaram et al45 (2022)
Chhabra et al47 (2022)
Chikhungu & Nkwunonwo48 (2026)
Decker et al51 (2022)
Decker et al52 (2022)
Dehingia et al53 (2023)
El-Nimr et al56 (2021)
Farmani et al58 (2021)
Fereidooni et al60 (2023)
Forry et al61 (2021)
Gaga et al62 (2022)
Gambir et al63 (2024)
Hamadani et al64 (2020)
Hernández-Vásquez et al67 (2023)
Hoehn-Velasco et al68 (2021)
Kaswa72 (2021)
Kiarie et al74 (2022)
Kitutu et al75 (2025)
Lamichhane et al78 (2021)
Joly et al79 (2025)
Mahapatro et al82 (2023)
Maharlouei et al83 (2023)
Mahmood et al85 (2022)
Mahmood et al86 (2022)
Mahmud & Riley87 (2021)
Maji et al88 (2022)
Miller et al93 (2022)
Diaz et al95 (2025)
Ozgurluk et al101 (2023)
Pande et al102 (2022)
Porter et al104 (2021)
Pourmehdi105 (2024)
Puri et al106 (2023)
Ravindran & Shah109 (2023)
Roman et al111 (2023)
Roy et al112 (2022)
Sharma & Khokhar115 (2021)
Shitu et al117 (2021)
Sönmez Güngör et al119 (2023)
Steinert et al121 (2023)
Tripathi et al122 (2022)
Wood et al127 (2022)
Woofter et al128 (2024)
Zagloul et al132 (2022)
Zsilavecz et al133 (2020)

, strong; , moderate; , weak.

Table 2. Harvest plot of direction and significance of results from longitudinal studies exploring the relationship between COVID-19 and violence against women and girls (VAWG), by study directionality and VAWG type^*.

	Prospective†						Retrospective†
	Increase			Decrease	No change		Increase			Decrease		No change
Domestic violence (DV)
Unspecified DV	60	106	127		51	104	117					38	121
Emotional or psychological intimate partner violence (IPV)	106						26	28	38	28		56	119
							58	62	64
							83	86
Financial IPV							26	38	86			56	119
Non-partner DV					37	104
Physical IPV	87	106	127	37			26	56	62	28	119	26	28	38
							78	86				64	86
Sexual IPV	106				127		56	62	83	29	119	37	38	64
							86
Verbal IPV							38
Non-DV and unspecified VAWG
Online violence		53								132		119	132
Physical or sexual violence		37										78
Unspecified VAWG							61

^*Only studies with significance testing are included in this table.

^†Direction in the table reflects the change in VAWG comparing values before and during the COVID-19 pandemic.

IPV, intimate partner violence; VAWG, violence against women and girls.

Table 3. Harvest plot of direction and significance of results from service utilisation studies exploring the relationship between COVID-19 and violence against women and girls (VAWG), by VAWG type^*†.

	Increase			Decrease			No change
Domestic violence (DV)
Unspecified DV	32	67	111	68	85	111	45
Physical IPV	101						133
Other and unspecified forms of VAWG
Physical or sexual violence	74						102
Unspecified VAWG	109						58

^*Only studies with significance testing are included in this table.

^†Direction in the table reflects the change in VAWG comparing values before and during the COVID-19 pandemic.

IPV, intimate partner violence; VAWG, violence against women and girls.

Domestic violence (DV)

The majority of publications exploring changes in magnitude of VAWG investigated domestic violence (DV). Among longitudinal and single point studies, 34 comparisons reported increased DV, 19 reported no change, and only six reported decreases. More longitudinal studies reported significant increases than decreases or similar VAWG for emotional or psychological intimate partner violence (IPV) (nine out of 12 comparisons found increases; one without significance testing), financial IPV (three out of five comparisons found increases; one without significance testing), and verbal IPV (one out of one comparison found an increase; a second study without significance testing confirmed this finding). For physical IPV, eight of 16 comparisons found increases, with two additional studies without significance testing also reporting increased IPV; the remaining comparisons found no change or a decrease. Similarly, for unspecified DV, four of eight comparisons found increases, while two studies without significance testing reported both increased and decreased violence. Two studies with significance testing and one without reported no change in non-partner DV. Five comparisons with significance testing found increased sexual IPV, four with significance testing found similar sexual IPV, and two comparisons with significance testing found decreased sexual IPV—another two studies without significance testing also reported increased sexual IPV. One study without significance testing reported both increased and unchanged child marriage.

Non-DV and unspecified VAWG

13 studies assessed changes in non-DV and unspecified VAWG. Nine studies were population-based. Two studies, one with and one without significance testing, reported increased violence against women during the COVID-19 pandemic. One study with and one study without significance testing reported increased femicide during the COVID-19 pandemic. Regarding physical or sexual violence, two studies with significance testing found increased and similar incidence. A third study without significance testing confirmed this. While two comparisons reported no change in online violence, two other comparisons reported increases and decreases—all four studies used significance testing.

Changes in service utilisation

More studies with significance testing found decreases or no change in service utilisation for VAWG (n=3 and n=4, respectively), compared with increases (n=3). Three comparisons with significance testing found increases in DV reporting, three found decreases (plus three comparisons without significance testing), and one comparison with and one without significance testing reported no change in DV. Two service utilisation studies of physical IPV reported increased and similar VAWG. One study with significance testing reported increased physical or sexual violence, one with and one without significance testing reported no change, and one study without significance testing found decreased examinations for physical violence during the pandemic compared with pre-pandemic. Two studies reported a significant increase and no change in unspecified VAWG during the pandemic. There is a small body of mixed evidence on increases in non-DV service utilisation during the COVID-19 pandemic.

Mechanisms of VAWG during the COVID-19 pandemic and Ebola outbreaks

We identified five mechanisms through which VAWG was perpetrated during the COVID-19 pandemic and Ebola outbreaks. The mechanisms included (1) income loss due to economic shutdown, financial insecurity, and/or job loss, (2) movement restrictions, (3) changes in access to public services, (4) fear of exposure to infectious disease, and (5) a legacy of mistrust in health systems from previous outbreaks. In the publications we reviewed, there was evidence to suggest that each of the factors contributed to both DV and other forms of VAWG. Figure 2 displays each mechanism, the number of studies providing evidence for the mechanism, and whether each of those studies identifies factors as contributing to DV or non-DV and unspecified VAWG. An online supplemental file S5 identifies studies that provide evidence for each mechanism. Overall, the evidence indicates these factors are all closely related, often synergistic and overlapping.

Figure 2. Mechanisms of violence against women and girls (VAWG) during the COVID-19 pandemic and Sierra Leone’s Ebola outbreak. DV, domestic violence.

The most commonly discussed mechanism in the literature, which was identified in 66 publications, was income loss due to economic shutdown, financial insecurity and/or job loss. While several articles reported an effect of income loss on non-domestic VAWG, the majority of studies focused on DV. For instance, a qualitative study among healthcare providers providing services to refugees in South Asia identified male partner’s job loss, and the accompanying anger and fear, as a motivator for unprovoked violence against their female partners.⁴⁹ Another study in South Asia similarly identified male job loss as a motivator of anger and insecurity, which led to general frustration and anxiety reflected through violence against female partners.⁷⁶ A multi-country study in Africa found that the stress of male partner job loss, and their inability to contribute to meeting basic needs in their household, led to higher rates of violence against female partners.⁵⁵ Another study in one country in Africa found that individuals with the least wealth lost jobs, faced reduced salaries or livelihoods that could not be continued during COVID-19, leading to domestic VAWG.⁵⁴ A similar study in rural West Africa described COVID-19 as an ‘enforcer’: when women could not provide economically for their husbands due to job loss, they needed to ‘compensate’ with sex, even when they did not want to.³⁴ Financial insecurity, expressed as food insecurity by participants in a mixed methods study in Indonesia, led to DV, and ultimately to child marriage.¹⁰⁸

This mechanism was also enacted through changed family structure, temporary, precarious labour, and abusers’ changed alcohol and drug use, all as results of job loss and economic insecurity. 18 studies explored changed family structure, interactions and responsibilities, including the sudden inability to avoid abusers. Studies in across geographical regions reported mothers and children being exposed to family members, including husbands, more often at home, leading to increased risks for sexual violence against women and their children.^{83 98} Even when women remained employed during the pandemic, one study in the Middle East identified male partner distress, including increases in household chores, as a motivator for abuse against women while they worked from home.⁶⁵ Unlike many other articles exploring DV, these researchers also identified the negative impact of mothers-in-law, who perpetrated abuse against women themselves, while also encouraging their sons to do so. Another study reported that the pandemic forced many families to relocate and stay with relatives, which exposed them to sexual violence, which was most commonly perpetrated by male relatives, including brothers, fathers and uncles.¹¹⁰ One publication specifically addressed increased vulnerability of female sex workers during the lockdown as they navigated low negotiating power for sex with condoms and other client sexual abuse.⁴¹ Increased drug and alcohol abuse, coupled with stay-at-home orders, was identified as enabling DV in the Asia-Pacific region.⁹¹ Economic shutdown, financial insecurity and job loss drove VAWG both within and outside of domiciles.

Movement restrictions, which we define as government orders limiting travel to prevent the spread of COVID-19, were found to be contributory to DV in 57 publications and related violence was often enacted through social isolation and police officer conduct during enforcement of restrictions. A media-based content analysis in South Asia revealed that women who were forced to stay home with their abusers during lockdown faced dependency, exploitation and violence. The authors suggested that even if mechanisms existed for the women to report violence, they would not have, due to being isolated with their abusers.³³ A study in the Middle East found that forced cohabitation during the lockdown exacerbated tensions within relationships that were present before the lockdown, which further increased women’s risk of VAWG.⁶⁵ COVID-19-induced changes in social interactions negatively impacted women and girls’ experiences of VAWG during the pandemic. Social isolation often increased VAWG by confining women and girls in close proximity to perpetrators of violence and reducing typical social support from relatives and friends due to movement restrictions. Three studies in Nigeria explored the extent to which police perpetrated sexual misconduct during lockdowns and found that women had to trade sex or tolerate sexual harassment from officers for free movement or protection when breaking lockdown rules for various reasons, including their livelihood or to access health services.23,25

Changed access to public services, including school closures, also increased exposure to violence and was explored in 35 publications. One study in Africa reported that women were locked out of their homes in the middle of the night by their abusers, and were unable to access social services, leaving them exposed both to COVID-19 and VAWG outside of their domicile.³⁰ In another study in a different African country, researchers identified school closure as a reason for child marriage: without school to occupy them, and with the pandemic’s associated financial stresses, girls participated in relationships with older men, leading to early pregnancies and child marriage.⁵⁷ While women wanted to access family planning services, they were unable to access them during the COVID-19 pandemic lockdown in rural West Africa, resulting in sexual IPV, as they disagreed with their husbands about how many children they should have.⁷³ Respondents in another study believed girls were becoming pregnant due to lack of supervision when schools closed in an East African refugee camp, leading to unintended pregnancy and child marriage.¹¹⁶ Decreased access to services, closures and disruptions made women and girls vulnerable to violence.

Eight publications identified fear of exposure to infectious disease as associated with non-domestic VAWG and DV. For example, a publication from South Africa discusses women’s choice to go outside in this way: male partners used psychological abuse to make women fearful of getting COVID-19, so the women had to decide whether to possibly endure death inside their homes or to leave, contract COVID-19, and possibly die outside of their home.⁵⁴

Lastly, we identified one additional pathway that was unique to the study exploring health-seeking behaviours during the Ebola and COVID-19 outbreaks in Sierra Leone: a legacy of mistrust in health systems from previous outbreaks.⁴⁶ This study noted that there were cases of healthcare workers perpetrating violence against individuals during the Ebola outbreak and that, if a woman’s experience of IPV resulted in injury (especially if the injury included bleeding), the woman also faced a risk of being quarantined and treated poorly when seeking help for the injury in the formal healthcare system. This fear and mistrust of the healthcare system affected health-seeking behaviours during the Ebola outbreak. Participants reported still holding these same fears during the COVID-19 outbreak 7 years later; although the response measures for COVID-19 differed from those for Ebola, respondents nonetheless remained less inclined to seek care for experiences of violence.

Discussion

This systematic review examines the extent to which infection control measures during infectious disease outbreaks exacerbated existing vulnerabilities to VAWG in LMICs, including DV and violence committed by others. We found evidence for increased exposure to VAWG during the COVID-19 pandemic and Sierra Leone Ebola outbreak through five mechanisms: (1) income loss due to economic shutdown, financial insecurity, and/or job loss, (2) movement restrictions, (3) changes in access to public services, (4) fear of exposure to infectious disease, and (5) a legacy of mistrust in health systems from previous outbreaks. In addition to identifying these factors contributing to VAWG, our quantitative results describe changes in the magnitude of VAWG during the COVID-19 pandemic. While the evidence was nuanced, overall, more studies found significant increases in VAWG, as compared with significant decreases or similar rates of VAWG.

Throughout this review, we identified gaps in the published literature, the foremost of which is the total absence of quantitative and near total absence of qualitative research exploring VAWG in the context of any outbreak other than COVID-19, despite the many localised and pandemic-level outbreaks experienced around the globe in the past 10 years.¹ This gap indicates that VAWG tracking in relation to infectious disease outbreaks is new, and the evidence in this systematic review points to the need for more systematic VAWG monitoring moving forward to understand different VAWG outcomes and mechanisms across various types of outbreaks.

Additionally, there is a lack of quantitative evidence on specific forms of VAWG within the DV category, such as child marriage, financial IPV, non-partner DV and verbal IPV, as well as the non-DV category, such as online violence, physical or sexual violence. All but three studies in our sample combined data on women and girls or only investigated violence against women. Future research must investigate whether and how violence perpetrated against girls is different from violence against women to ensure that policies and programmatic interventions reach those who are vulnerable, regardless of their age. There is also a lack of geographical representation which could be supplemented with additional research: we only identified nine studies from West and Central Africa, five from Europe and Central Asia, and four from East Asia and Pacific region. We also believe that there are opportunities to design studies that use standardised outcome measures, with significance testing, to provide evidence that can be aggregated across studies. Future research should include longitudinal data from panel studies, allowing for comparisons within large populations before and after outbreaks.

For many respiratory-related airborne diseases (eg, COVID-19, Respiratory syncytial virus (RSV), tuberculosis), the most effective control measures typically involve isolation, mask mandates and other practices that limit airborne transmission.^{13 134} During the height of the COVID-19 pandemic, some countries enacted various control strategies, such as transitioning to virtual school and work, enforcing lockdowns, social distancing, and other initiatives were implemented with the goal of controlling the spread of the virus. These control measures had significant implications for VAWG and must be taken into account in future outbreak response. Similarly, where relevant, actors introducing control measures must be aware of mistrust of health systems, and work with local communities to ensure that measures are adapted to and acceptable for their needs.

Finally, the identified mechanisms through which VAWG increases during infectious disease outbreaks highlight the need for policymakers to proactively integrate VAWG prevention and response into public health emergency preparedness plans. The majority of the reviewed mechanisms leading to VAWG were the direct or indirect consequences of measures taken to curb the spread of COVID-19 (and in one case, Ebola), such as restrictions on movement, limiting access to health and other social services, and full lockdowns. Although our search strategy allowed for the inclusion of studies that explicitly assessed the impact of policies or programmes to prevent VAWG during outbreaks, no such articles were found. However, our synthesis of the mechanistic evidence provides an initial foundation for future policy and research considerations. The overwhelming evidence linking containment policies to VAWG highlights the importance for policymakers to conduct rigorous risk-benefit analyses when developing and implementing outbreak response measures. While these measures may be necessary to contain disease spread, our findings demonstrate they can also exacerbate VAWG by increasing social isolation, exacerbating household distress, and leaving women and girls vulnerable to precarious labour or child marriage. Future research should also examine different policy levers that may be able to effectively intervene on the mechanisms identified in this review (eg, policies aimed at minimising income loss, limiting the purchase of alcohol, promoting the continuity of education, and ensuring access to safe and accessible health and other social services for women and girls).

Limitations

This study is not without its limitations, the foremost of which is the minimal data related to any outbreak other than COVID-19 (aside from the one qualitative study on the Ebola outbreak), despite the occurrence of many outbreaks (eg, cholera, Zika, etc) since 2015. While this limits our ability to account for mechanisms of VAWG during other outbreaks, it points to the need for more research in other contexts to build on our initial investigation. Due to the non-standardised reporting measures across studies (ie, incidence rates, risk ratios, ORs, statistical tests), we are unable to quantify the change in VAWG during the first year of the COVID-19 pandemic; however, we are able to substantiate the direction of change in VAWG by looking across study types and measures. Due to the relative scarcity of quantitative evidence on the topic, we decided to include population-based and service utilisation studies in our review. While these cannot be directly compared, service utilisation studies offer insight into the change in self-reports of VAWG in different geographies. Additionally, though we included publications in four languages, we may have missed relevant studies in local or regional languages, which limited the scope of the review. These limitations underscore the likelihood that our findings underestimate the complexity of VAWG during infectious disease outbreaks, reinforcing the urgency of improved surveillance, standardised measurement and inclusive research designs.

Conclusions

This systematic review reveals the extent to which COVID-19 and Ebola, and their associated public health response, resulted in increased acts of VAWG. We identify the mechanisms through which this VAWG occurred, providing evidence for the most efficient avenues to prevent VAWG in these types of outbreaks. While our research almost exclusively identifies the quantitative and qualitative impacts of the COVID-19 pandemic, it also reveals the importance of VAWG tracking during outbreaks. Systematic monitoring is necessary for future outbreaks, not just COVID-19, to understand how they exacerbate VAWG and to measure the progress of programmes to reduce and eliminate VAWG in these contexts. We hope that the results of this study will inform new research, policies, guidelines, and actions from government officials and public health practitioners to safeguard women and girls’ health and well-being in outbreaks to come.

The statements in this publication are the views of the author(s) and do not necessarily reflect the policies or the views of UNICEF.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.