Table 2.

Study Characteristics

Study	Study Design	Location and Population	Measures of IPV	IPV prevalence	Vulnerability factors for higher IPV during COVID-19	Quality Appraisal
Agüero, 2021	• External database • January 2007 until July 2020 • Data tracking of calls to the helpline Línea 100	• Peru • Data at the month and state level on the volume of calls per 100,000	• Helpline calls	• Increase in IPV • 48% increase in calls to the helpline since the pandemic • Prior to the pandemic, 60% of women were experiencing domestic abuse	• Not driven by any demographic background characteristics, even when considering the pre-determined prevalence of domestic violence	• 26/39 (66.67%) • Data collection method did not allow for information about the sample to be reported (i.e., caller demographic characteristics)
Blackhall et al., 2020	• Prospective data collection beginning on March 26th, 2020	• United Kingdom • People presenting with maxillofacial injuries across five hospitals	• Hospital records	• No reported data on changes in IPV • 529 patients over six weeks • 44 cases of facial fractures with 17 cases related to domestic abuse	• n/a	• 24/39 (61.54%) • Limited justification for the choice of statistical analysis
Bracewell et al., 2020	• Unspecified timeframe • Semi-structured interviews and discussions following the start of the pandemic	• United Kingdom • n = 15 survivors of IPV • n = 6 healthcare providers	• Thematic analysis of transcribed participant interviews using NVivo v12	• Increase in IPV • 50–70% increase in initial requests for support via email from both survivors and wider services in the 3 months from April 2020 compared to the previous 3 months	• Confinement to the home • Increased stalking and surveillance opportunities • Loss of employment or working from home • Delays in the criminal justice system	• 24/39 (61.54%) • No consideration of the sample size requirements • Limited consideration of the validity and reliability of qualitative methods
Davis, et al., 2020	• April 22nd to April 30th • Cross sectional study	• USA • N = 2045 non-institutionalized adults • Stratified quota sampling to ensure sample characteristics of sex, age, and race were representative • Respondents needed to be at least 18 years old	• Four-item tool to assess IPV perpetration and victimization since the outbreak of COVID-19 • Self-reported demographic data, recent health history, and degree of personal social distancing	• No reported data on changes in IPV	• Reporting COVID-19 positive (almost 3 times more likely to experience IPV) • Living in states with higher spread of the COVID-19 virus • Losing a job due to COVID-19 (2.5–3 times more likely to experience physical IPV)	• 31/39 (79.49%) • No consideration of the sample size requirements • Limited justification for the choice of statistical analysis
Gebrewahd et al., 2020	• April to May 2020 • Community based cross-sectional study consisting of interviews and a self-administered questionnaire	• Ethiopia • N = 682 reproductive age women	• Violence against women was measured using thirteen WHO questions regarding psychological, physical, and sexual violence	• No reported data on changes in IPV • Psychological violence was the predominant type of violence (13.3%) • 8.3% of participants suffered from physical violence such as slapping or throwing objects • No participants experienced physical violence by threatening to use or using a gun, knife, or other weapon • Sexual violence experienced by 5.3% participants	• After controlling for the confounding effects of age, IPV was associated with education level, occupation, type of marriage and husband’s age • 68.3% of participants were married with arranged marriage, and 31.7% of participants were living with their self-chosen husband • 52.9% of respondents had low family income • 34.6% of the women had no education • 5.3% of participants had no children • 62.8% of women reported that their husband had a drinking habit • 40.2% reported that their husband displayed aggression	• 29/39 (74.36%) • No justification for the choice of statistical analysis
Gosangi et al., 2021	• External database • March 11, 2020 to May 3, 2020 • Examined data from the same period for 2019, 2018, and 2017 • Examined referrals to the institutional domestic violence and intervention and prevention program since 1997	• USA • Survivors of domestic violence referred to the program (96% female in 2020)	• Frequency of referrals to the domestic violence intervention and prevention program	• Decrease in number of survivors seeking hospital care during the pandemic (62 in 2020; 104 in 2019; 106 in 2018; 146 in 2017, p < 0.001) • The incidence of physical IPV and the severity of injuries were greater during the pandemic • The number of survivors of physical abuse was 26/62 (42%) in 2020 compared to 42/342 (12%) in 2017–19 (p = 0.01), • Number of survivors of severe grade injury was (38%) in 2020 compared to 17% from 2017–2019 (p = 0.03)	• No significant differences in age, gender, or marital status between years • Race (significantly more white people presenting and less African American during COVID-19) • Higher incidence of survivors of high-risk abuse (strangulation, weapons, stab wounds, and burns)	• 30/39 (76.92%) • No consideration of sample size requirements • Limited justification for the choice of statistical analysis • Single location prevents the results from being generalized to other geographic regions
Halford et al., 2020	• External database • Crimes recorded between March 8, 2020 and April 2, 2020 are compared to expected rates based on data from 2016 to 2019 • Use of crime data in the UK Police Force over the last 5 years, covering 2000 square miles and around 1.5 million people	• United Kingdom	• Crime records	• 1 week after lockdown, domestic abuse decreased 45% compared to an average of the previous 4 years • Declined significantly by over 40 percent one week after lockdown (March 16, 2020), with weekend peaks being less prominent • More likely to reflect a reduction in reporting and recording resulting from social distancing measures	• Vulnerable children category was captured separately from domestic abuse, and found a decrease, but this was more likely a result of social distancing measures and school closures • Vulnerable child indications as a proportion of all crime also decreased	• 27/39 (69.23%) • Data collection method did not allow for information about the sample to be reported (i.e., demographic information) • Limited information about validity and reliability of statistical analysis
Hamadani et al., 2020	• May 19, 2020 to June 18, 2020 • Interrupted time-series (randomly selected a subset of participants from an RCT)	• Bangladesh • 3016 mothers, of whom N = 2424 agreed to participate	• IPV questions based on the validated WHO multi-country survey tool addressing emotional, physical, and sexual violence • Single item scale asking whether IPV had changed since the onset of stay-at-home orders	• Emotional, physical, and sexual violence increased among women experiencing IPV before COVID, with more than half reporting an increase • Of participants reporting emotional abuse, 66.8% reported an increase in insults, 66.0% humiliation, and 68.7% intimidation • 56.3% of participants reported increasing physical violence • 50.8% of participants reported increasing sexual violence	• n/a	• 31/39 (79.49%) • Limited consideration of sample size requirements • Limited rational for choice of data collection tools
Hassan et al., 2020	• External database • March 1st, 2020 through April 11 2020 (examined March 1–21 and then March 22 to April 11) which aligns with initiation of stay at home orders on March 21, 2020 • Chart review performed of all plastic surgery emergencies	• USA • Patients presenting in the emergency department for trauma consults	• Hospital records	• Increase in IPV • In the pre stay at home order period, 2 surgeries were the result of domestic violence (2.3%) compared to 5 surgeries in the period following the order (8.1%)	• n/a	• 24/39 (61.54%) • Limited consideration of sample size requirements • Limited rational for choice of data collection tools
Iob et al., 2020	• External database • March 21, 2020, to April 20, 2020 • Secondary use of data from University College London’s COVID-19 Longitudinal Social Study	• United Kingdom • N = 44,775 stratified sample of adults age 18 + , with 51% female and 49% male	• Psychological abuse: experiences of being bullied, controlled, intimidated, or hurt by someone over the past week • Physical abuse: experiences of being harmed physically or hurt over the last week • Both items rated on a 4-point scale from ‘not at all’ to ‘nearly every day	• No reported data on changes in IPV • Psychological abuse reported by 8.3% of participants (9.4% living alone, 11.7% living with adults other than their partner; 6.8% living with their partner or spouse) • Physical abuse reported by 2.9% of participants (3.7% living alone; 4.8% living with another adult other than their partner; and 2.0% living with their partner or spouse)	• Having children in the household (9.8% psychological and 3.3% physical) compared to households without (7.8% psychological and 2.8% physical) • Low income • Pre-existing chronic physical illnesses • Mental illness • Unemployment due to disability • Seeking work and unemployed • Overcrowding • A COVID-19 diagnosis	• 30/39 (76.92%) • Limited consideration of sample size requirements • Limited justification for the choice of statistical analysis
Jacob et al., 2020	• External database • March and April 2016, 2017, 2018, 2019, and 2020 • Cross sectional study using data from an external database	• Australia • Single center study at Westmead hospital, which services 1.5 million people • Trends in trauma admissions	• Hospital records	• No change in IPV • No significant differences in the number of trauma admissions due to domestic violence	• n/a	• 25/39 (64.10%) • Limited rational for choice of data collection tools • Limited justification for the choice of statistical analysis • Single center study
Jetelina et al., 2021	• 14-day period in April 2020 • Cross sectional online survey	• USA • N = 2411 participants completed the survey, of which N = 1759 reported having an intimate partner and were included in the analyses	• Extended- Hurt, Insulted, Threatened, Scream scale (E-HITS) is a 5-item scale to assess the frequency of IPV from 1 = never to 5 = frequent • Five-point Likert scale indicating whether IPV had gotten better or worse since COVID-19	• No change in IPV • IPV reported by 18% • Of respondents positive for IPV, 54% stated IPV stayed the same during COVID-19, 17% stated IPV worsened, and 30% stated it improved • The risk of IPV worsening was 4.38 times greater for physical compared to emotional abuse, and 2.31 times higher for sexual abuse compared to other forms of abuse	• Lower among women (OR = 0.63; 95% CI 0.50 to 0.82) compared to men • Job or income loss due to the pandemic (OR = 1.63; 95% CI 1.17 to 2.27) • Children being present had no significant effect on rates of IPV or severity change in IPV during the COVID-19 pandemic	• 25/39 (64.10%) • Non-stratified convenience sample with a low response rate of 5.7%
Leslie & Wilson, 2020	• External database • January 1, 2020, to May 27, 2020, compared with data of the same period in 2019 • Examine police calls for service	• USA • People calling police for concerns related to domestic violence	• Crime records	• Increase in IPV • 7.5% increase in IPV calls for service to the police during March through May of 2020, compared to the same period in 2019 • The increase began over a week before the first stay-at-home order, and the effects were largest during the first five weeks • IPV calls were up 9.7% in first 5 weeks	• Not driven by any specific demographic group • Households without a history of domestic violence drove the increase	• 28/39 (71.79) • Limited rational for choice of data collection tools • Limited justification for the choice of statistical analysis
Mohler et al., 2020	• External database • Baseline period from January 2, 2020, to March 16, 2020, compared to March 20, 2020, to April 18, 2020, in LA, and March 24, 2020, to April 21, 2020, in Indianapolis • Time series of verified crime reports per day and mean calls per day	• USA • Trends in volume of calls for service related to assaults in Los Angeles and Indianapolis	• Crime records	• Increase in IPV • Overall assault calls were unchanged in both locations, but regression of daily calls for service rate against stay-at-home order indicator, when controlling for the day of the week and week of month effects, shows a significant increase in domestic violence in both cities	• n/a	• 25/39 (64.10%) • Limited rational for choice of data collection tools • Limited justification for the choice of statistical analysis
Murphy et al., 2020	• External database • January 2017 to April 26, 2020 • Retrospective access of orthopaedic services use data	• United Kingdom • People referred to the on-call orthopaedic service at a UK District hospital which provides trauma care to a population of 625,000 people	• Hospital records	• No significant differences in referrals during the first 8 weeks of 2020 • During the remainder of COVID-19 there was a significant reduction in overall referrals • Cases of domestic abuse were comparable to previous years	• n/a	• 23/29 (58.97%) • Limited rational for choice of data collection tools • Limited justification for the choice of statistical analysis
Olding et al., 2021	• Existing database • March 23, 2021, to April 29, 2021 • Compared to data from the same period during the previous two years • Prospective data collection and review of past hospital data	• United Kingdom • Data collected on all patients presenting with traumatic penetrating injuries	• Hospital records	• Decrease in IPV • Trauma numbers fell by 35% overall • Two cases of injury due to IPV and four injuries related to IPV more broadly • IPV as a percentage of trauma’s decreased from 2018 (96%) to 2019 (89%) to 2020 (63%)	• n/a	• 14/29 (35.90%) • Small sample size, particularly when dividing the sample based on aetiology • Limited justification for the choice of statistical analysis
Øverlien, 2020	• April 8th, 2020, reminder email April 20th, survey closed April 23rd • Cross sectional online survey	• Norway • N = 46 Refuges providing services to survivors of domestic abuse • As of 2010, both men and women could seek support and shelter	• The survey contained 15 items with fixed answers, follow‐up questions and opportunity for the refuges to provide supplement qualitative descriptions and examples	• No reported data on changes of prevalence of IPV • 56% of clinician respondents reported a decrease in the number of requests from clients • Clinicians commonly reported that it is too quiet, and that survivors are not receiving adequate support	• 83% of refuge staff reported being most concerned about children (for reasons including IPV and inadequate support in households) • 43% reported that their clients believed the presence of children in the home during the day increased the risk of violence and abuse • 57% of staff said their clients reported the virus control measures and the COVID-19 crisis, were increasing stress for abusers	• 26/29 (66.67%) • Limited rational for choice of data collection tools • Limited justification for the choice of statistical analysis
Pakenham et al., 2020	• Unspecified, since COVID-19 • Online survey	• Italy • N = 1035 Italian citizens at least 18 years of age • 45.4% of the sample was either married or living with a partner, and 54.6% were single, widowed or divorced	• Twelve COVID-19 risk factors combined to form the COVID-19 lockdown Index (e.g., increase in IPV, COVID-19 diagnosis, duration of lockdown)	• No change in IPV • The mean domestic violence rating was 1.27, denoting that on average, participants did not report an increase in domestic violence • 5.3% of participants reported a moderate to high increase in domestic violence	• n/a	• 29/39 (74.36%) • Lack of clear description of research setting and target population
Payne et al., 2020	• External database • February 2014 until March 2020 • Autoregressive integrated moving average (ARIMA) modeling was used to compute 6-month ahead forecasts of rates for criminal offenses	• Australia • Monthly offense rates per 100 000 people in Queensland	• Crime records	• Increase in IPV • Domestic violence order breach rate in March 2020 was 57.5 per 100 000 which was higher than the prior month (52.0 per 100 000) • Long term upward trend is likely a reflection of increased orders and not necessarily the number of breaches • Insufficient evidence to suggest breaches increased with COVID-19	• Job loss due to COVID-19 • Financial and emotional stress and strain that may not become evident until the future months • Disrupted routine activities	• 28/39 (71.79) • Limited rational for choice of data collection tools • Limited justification for the choice of statistical analysis
Piquero et al., 2020	• External database • January 1, 2020 to April 27, 2020 • Examining daily counts of all domestic violence incident reports	• USA • People filing domestic violence incident reports	• Crime records	• No change in IPV • Implementation of the stay-at-home order is not associated with a significant increase in domestic violence incidents • Not enough evidence to suggest an increase in IPV throughout the month after the stay-at-home order	• n/a	• 30/39 (76.92%) • Limited information reported within/from the incident reports. (i.e., no demographic information about the abusers or survivors)
Rhodes et al., 2020	• External database • March 16, 2019 to April 30, 2019 compared with the same time periods in 2020 • Use of data on past emergency department patients presenting at the American College of Surgeons	• USA • Patients arriving in the emergency department presenting with signs of assault • 2900 patients presented to the ED and 50 patients cause code for assault • Previously 7008 patients with 78 survivors of assault	• Hospital records	• Increase in IPV • Significant increase (p = 0.01) in assaults during COVID-19 particularly during the period after school closures • Overall trauma volume reduction • Significant increase in proportion of IPV related admissions • Mechanism of assault was largely penetrating: pierce (2.6% vs 18%) and knife (0% vs 12%)	• n/a	• 24/39 (61.54%) • Limited theoretical background provided • Limited justification for the choice of statistical analysis
Sabri et al., 2020)	• Unspecified, prospective data since COVID-19 • Qualitative in-depth interviews conducted via Zoom or phone. Purposive and snowball sampling methods were used to recruit participants	• USA • 45 in depth interviews conducted with survivors of IPV who spoke English, were over the age of 18, were born outside of USA, and were now living in USA permanently • 17 key-informant healthcare provider interviews were conducted as well	• Qualitative, in-depth interviews	• Increase in IPV • Reinforcing relationship between increased stressors and IPV due to COVID-19 • Increased severity and frequency of IPV for immigrant women in abusive relationships • Decrease in clients taking legal steps • Increased stalking and monitoring behaviour and financial abuse • Threatening to infect survivors with COVID-19	• Being undocumented (fear makes leaving abuse more difficult) • Increased proximity to abusers • Decreased privacy • Decreased ability of professionals to observe children • Decrease in protective mechanisms (community babysitting) • Fear of job loss and future financial strain • Unable to work or job loss due to additional childcare responsibilities • Closure of typical services and a shift to virtual services with lower access	• 31/39 (79.49%) • Did not consider research stakeholders in the study design
Sediri et al., 2020	• April 25, 2020 to May 6, 2020 • Snowball sampling • Cross sectional online survey	• Tunsia • N = 751 Females at least 18 years of age	• The history of domestic violence and its types before and during the lockdown were assessed by yes and no questions	• Increase in IPV • Significant increase in IPV during the lockdown (from 4.4 to 14.8%; p < 0.001) • Psychological abuse was the most frequent type of violence (96%)	• Violence during lockdown was associated with higher scores in depression, anxiety, and stress • Women who had experienced abuse before the lockdown were at an increased risk of violence during lockdown (p < 0.001; OR = 19.34 [8.71–43.00] • 95% of participants were at no particular risk for COVID-19, 4.2% were under quarantine, and 0.8% tested positive for COVID-19	• 24/39 (61.54%) • Non-stratified convenience sample with no data on the total number of people invited to participate
Weller et al., 2021	• External database • January 2020 to May 17, 2020 • Data from evaluation of services in London and Surrey were analysed anonymously and in aggregate in Excel and STATA	• United Kingdom • People seeking consultations for domestic violence	• Aggregate data using Excel and STRATA	• Increase in IPV • Initial reduction in the percentage of consultations (decreasing from 66% between March 9th and 22nd 2020 to 52% between March 23rd, 2020, and April 4th, 2020) • Increase was recorded in the percentage of IPV disclosures during lockdown (April 6th, 2020, to April 17th, 2020), compared with the 6 weeks prior to lockdown (February 10th 2020 to March 22nd 2020) with an increase from 0.22% (CI 0.14% to 0.32%) to 0.57% (CI 0.33% to 0.91%) (p = 0.002)	• n/a	• 15/39 (38.46%) • Limited theoretical background provided • Limited justification for the choice of data collection tools and statistical analysis