Violence against Women during the COVID-19 Pandemic: From Children to the Elderly

Round 1

Reviewer 1 Report

The authors analyze data of the Portuguese Association for Victim Support for the year 2020 to study whether COVID-19 lockdown measures have contributed to a change in violence against women.

The research question is interesting. However, I have a number of major concerns with the study.

Main concerns:

1) I am unclear about the theoretical foundation and conceptual framework of the paper. 

How do the analyses speak to the research questions?

Why were infants and young children included in the study? In what ways are they seeking help as we would expect help seeking behavior of adults?

Are the reported outcomes exclusive? IPV can include physical, mental and sexual violence. What does it tell us about the pandemic that IPV was reported most frequently. Presumably nothing, if there is no comparison to pre-Covid years.

2) Having read the paper, I did not understand what the statistical test or method was that was applied to receive the results in tables 4 and 5. Which months were compared to which other months and how? How were these expected values estimated? How can we think of these expected values? How were the standardized residuals estimated? 

3) The analysis only includes data for 2020. Any observed patterns in the data by months may simply reflect seasonality. The authors do not address this issue at all.

4) The paper is hard to follow. The conducted analyses are not motivated in a coherent way and the discussion does not follow from the results, nor does it include relevant limitations. The draft has language errors throughout. Tables are lacking important information and notes describing the table content.

Author Response

We would like to thank the reviewer for the suggestions throughout the manuscript. We will provide a response to them in the following paragraphs:

1) I am unclear about the theoretical foundation and conceptual framework of the paper.

How do the analyses speak to the research questions?”

Research questions were adequately changed in the revised manuscript after an overall update due to the different reviewers’ suggestions.

Why were infants and young children included in the study? In what ways are they seeking help as we would expect help seeking behavior of adults?

This manuscript results from an investigation project that analyzed the violence perpetrated on victims of all ages during the pandemic. Therefore, not studying young female victims would limit the understanding of the type of violence or crimes these young victims were subject to during the pandemic and lockdown. As referred to in the manuscript (lines 116 to 121 of the original manuscript, lines 117 to 125 of the revised one), the requests for help may come from the victims themselves or from other people (hence, adults), so anyone that contacts APAV may request help for a young person. It is not possible to evaluate the different behavior of adults or children considering the existing variables of the database.

Are the reported outcomes exclusive? IPV can include physical, mental and sexual violence. What does it tell us about the pandemic that IPV was reported most frequently. Presumably nothing, if there is no comparison to pre-Covid years.

This part was amended in the revised manuscript considering the suggestions.

2) Having read the paper, I did not understand what the statistical test or method was that was applied to receive the results in tables 4 and 5. Which months were compared to which other months and how? How were these expected values estimated? How can we think of these expected values? How were the standardized residuals estimated?

Even though the revised manuscript no longer have these tables, lines 203 to 207 of the original manuscript mention the statistical test that was used and how to interpret it, as well as a reference to the article by Agresti (2007) that explains how it is conducted.

3) The analysis only includes data for 2020. Any observed patterns in the data by months may simply reflect seasonality. The authors do not address this issue at all.

The revised manuscript includes data from 2019 and 2020 to allow a comparison between a pandemic and a non-pandemic year.

4) The paper is hard to follow. The conducted analyses are not motivated in a coherent way and the discussion does not follow from the results, nor does it include relevant limitations. The draft has language errors throughout. Tables are lacking important information and notes describing the table content.

All suggestions have been considered and changes have been made in the revised manuscript.

Reviewer 2 Report

This paper is very interesting and contributes to the recently available knowledge on the issue of violence against women during the pandemic. It is with great satisfaction and interest that I read data from Portugal, for the whole year 2020.

Please find below some remarks that aim to increase the scientific interest of this article. Some remarks are minor and related to the form, others are more important and would require a reworking of the "Results" part.

1- It is not clear to the reader what the statistical unit of the sample is: 6475 women or 6475 requests for help during the year (the same woman may have called several times). I think it would be useful to mention somewhere how many calls the APAV received for male victims, for the same period. Then it is clear that the sample only concerns female victims.

2- How are the types of violence defined? A large proportion is "violence by an intimate partner" without distinguishing between these situations. However, it can be psychological, physical or sexual violence.

3- Check whether 42 or 43 deaths. Why are they called "homicides" in Table 1 and "femicides" in Table 2? Among these deaths, do the authors know if any were caused by IP violence?

4- Be precise, lines 164 and 169, page 5/11, the age of 65 is included in both groups... 

5- In the tables, improve the readability = Table 2, put back the age groups corresponding to the 3 groups (<= 17; 18-64; 65 or more). In general, always include a unit and a decimal for percentage values = for example, 52.0 or 0.9, even if the value is zero. However, in the text, the decimal point is not necessary for percentages; this makes reading more difficult, so round up to the nearest even number.

6- Table 2: I don't understand why you include a 'Total' line with 100% if the same crime can be counted several times, it is not logical. Problems with the percentages, therefore… cases are not independent.

7- This is the Chi-squared test and not the Qui-squared test (multiple occurrences).

8- On page 5, lines 158-159, the authors say that they did not impute missing data because it is categorical data. This is a weak argument. I am not sure that missing data should be imputed but the reason is certainly not that. Many research works in social sciences or biomedical sciences are published with imputation of data for discrete variables. In general, imputation of missing data reduces estimation bias due to the selection effect of non-response. However, in the case of these data, the authors may not have enough data (number of variables/questions) to allow an imputation that could be considered valid.

9- In the recording of data, how are sexual violence distinguished from physical violence? In the case of rape, for example, is the same victim's complaint counted both in "physical violence" and in "sexual violence"?

10- I find that there is an indecision on the objectives of the paper: 1/ to better analyse the violence suffered by women in Portugal during the pandemic; 2/ to describe what response is given to the requests during 2020 (analysis of practices during the pandemic) or 3/ to show if the period of confinement (April) is different from the other months. The authors should choose more clearly which question(s) they want to answer. I find tables 3 and 5 weakly interesting. I would have found it interesting to know whether the answers given to women are the same or not depending on whether it is a violence by the partner or other contexts of violence (stratified analysis, you have enough observations to do this). On this subject of violence against women, there is now a real body of knowledge, and we know that situations of violence must be contextualised. The social responses (legal, medical or instrumental) cannot be the same depending on whether it is a case of violence by a partner, violence at work or in public spaces. This is why your work would be more informative if you analysed your data further. In order to know if April differs from other months (i.e. lockdown) you should compare the situation in April 2020 with that of April 2019, as there may be seasonal phenomena.

11- The difference between "emotional support" and "psychological support" is not clear (see table 3 and the commentary). Please specify further.

12- In discussion, it should also be mentioned that a drop in calls may also be partly due to a drop in violent altercations. If travel is limited, violence in the public space is reduced, as is violence by friends or relatives, kept at a distance. Within the couple, the permanent presence of the woman at home satisfies the control needs of a dominating partner, who no longer has recourse to other coercive measures to exert his influence. 

13- Works on the subject could be cited, in particular:

-       Romito P, et coll. Intimate partner violence against women during the COVID-19 lockdown in Italy: a multicenter survey involving anti-violence centers. Violence against Women, 2022. DOI: 10.1177/10778012221079374.

-       Romito P et coll. Partner’s violence during the COVID-19 lockdown and women’s fear: a study involving anti-violence centres in Italy. Journal of Gender-Based Violence 2022. DOI: 10.1332/239868021X16425822014024.

Author Response

We would like to thank the reviewer for the suggestions throughout the manuscript. We will provide a response to them in the following paragraphs:

1- It is not clear to the reader what the statistical unit of the sample is: 6475 women or 6475 requests for help during the year (the same woman may have called several times). I think it would be useful to mention somewhere how many calls the APAV received for male victims, for the same period. Then it is clear that the sample only concerns female victims.

We have clarified the statistical unit and we have added a phrase that includes all requests for help that were received during 2019 and 2020.

2- How are the types of violence defined? A large proportion is "violence by an intimate partner" without distinguishing between these situations. However, it can be psychological, physical or sexual violence.

In the revised manuscript we have added the distinction.

3- Check whether 42 or 43 deaths. Why are they called "homicides" in Table 1 and "femicides" in Table 2? Among these deaths, do the authors know if any were caused by IP violence?

Amended.

4- Be precise, lines 164 and 169, page 5/11, the age of 65 is included in both groups...

Amended.

5- In the tables, improve the readability = Table 2, put back the age groups corresponding to the 3 groups (<= 17; 18-64; 65 or more). In general, always include a unit and a decimal for percentage values = for example, 52.0 or 0.9, even if the value is zero. However, in the text, the decimal point is not necessary for percentages; this makes reading more difficult, so round up to the nearest even number.

Amended.

6- Table 2: I don't understand why you include a 'Total' line with 100% if the same crime can be counted several times, it is not logical. Problems with the percentages, therefore… cases are not independent.

Amended.

7- This is the Chi-squared test and not the Qui-squared test (multiple occurrences).

Amended.

8- On page 5, lines 158-159, the authors say that they did not impute missing data because it is categorical data. This is a weak argument. I am not sure that missing data should be imputed but the reason is certainly not that. Many research works in social sciences or biomedical sciences are published with imputation of data for discrete variables. In general, imputation of missing data reduces estimation bias due to the selection effect of non-response. However, in the case of these data, the authors may not have enough data (number of variables/questions) to allow an imputation that could be considered valid.

We have reformulated the reason for not inputting data.

9- In the recording of data, how are sexual violence distinguished from physical violence? In the case of rape, for example, is the same victim's complaint counted both in "physical violence" and in "sexual violence "?

We have added a more detailed section in the “procedure” section, concerning how the data is retrieved by the APAV professionals to the victims. When someone requests help, for instance for a case of rape, APAV professionals follow an internal protocol with specific questions about the type of crime reported. So, in this case, they would ask if there was physical violence beyond the necessary for the rape to happen (e.g., hitting the victim's head repeatedly until unconscious, which would be classified as both sexual and physical violence) or if the perpetrator resorted to other means to (e.g., drugs to make the victim unconscious, which would be classified as sexual violence).

10- I find that there is an indecision on the objectives of the paper: 1/ to better analyse the violence suffered by women in Portugal during the pandemic; 2/ to describe what response is given to the requests during 2020 (analysis of practices during the pandemic) or 3/ to show if the period of confinement (April) is different from the other months. The authors should choose more clearly which question(s) they want to answer. I find tables 3 and 5 weakly interesting. I would have found it interesting to know whether the answers given to women are the same or not depending on whether it is a violence by the partner or other contexts of violence (stratified analysis, you have enough observations to do this). On this subject of violence against women, there is now a real body of knowledge, and we know that situations of violence must be contextualised. The social responses (legal, medical or instrumental) cannot be the same depending on whether it is a case of violence by a partner, violence at work or in public spaces. This is why your work would be more informative if you analysed your data further. In order to know if April differs from other months (i.e. lockdown) you should compare the situation in April 2020 with that of April 2019, as there may be seasonal phenomena.

We thank the reviewer for these suggestions. We have specified the objectives to focus more on the VAW and less on the type of response given by the victim’s association or the seasonality. As also suggested by other reviewers, we have added the data from 2019 to compare to 2020 and allow an understanding of how the VAW has changed during a non-pandemic and pandemic year.

11- The difference between "emotional support" and "psychological support" is not clear (see table 3 and the commentary). Please specify further.

With the changes in the manuscript, this part was removed.

12- In discussion, it should also be mentioned that a drop in calls may also be partly due to a drop in violent altercations. If travel is limited, violence in the public space is reduced, as is violence by friends or relatives, kept at a distance. Within the couple, the permanent presence of the woman at home satisfies the control needs of a dominating partner, who no longer has recourse to other coercive measures to exert his influence.

We have added that hypothesis to the discussion.

All suggestions have been considered and changes have been made in the revised manuscript.

Reviewer 3 Report

Overall, the study has an interesting goal, but it is not fulfilled through the process and presents major fragilities regarding:

i) literature review

- It is suggested the definition of violence against women from Action against violence against women and domestic violence (Istanbul Convention) (Council of Europe, 2011)

- The data presented is massively from 2020, which is a small part from all the data available regarding the pandemic context. Therefore, there are much more updated data that would be interesting to underline the project, including portuguese official data. Also, the United Nations study on Violence Against Women that is cited (2013) is not the latest version (2021) and together with other two studies, the authors slide to generalization using broad terms like “general” and “universal”.

- It should be clearer the relation between the increase of VWA and decrease of official reports – it should be unequivocal what is perceived as “requests” - towards a domestic violence structure, towards authorities, towards familiar/social help? Again, the data that underline this point is not the most recent.

ii) The Method

- Regarding the method, the criteria for crimes categorization is doubtful – Reported crimes were more often for intimate partner violence, followed by psychological violence related crimes. When there is psychological violence within a intimate relationship (as it occurs commonly), how is categorizes? Does it mean that there are situations that are not within the spectrum of domestic violence? Is it gender violence? If it is not, it should be evident from start (violence against women or violence that takes place also against women?)

- It should be clear why are adult women considered starting form 17 years from age (legal age) and children considered before 11 years.

- The lockdown is briefly mentioned once and not specified that is decreed by portuguese government and it should be traced comparisons between pre and post lockdown period and the same period before and after 2021 (e.g., same months in 2019 and 2021).

- When there is mentioned a higher frequency of sexual crime (172), the affirmation lacks comparison and data presented in table 2, specifically the total n and percentage of IPV and Offenses (cf. 127) and would be interesting to understand why the emotional support is considered in all types of support.

- According to table 2, the types of violence more prevalent were psychological and sexual, but in the discussion, it is contradicted, underlining psychological and physical violence.

- Also, it is referred that the “requests for help” decreased due to restriction policies, but later is mentioned that remote support had increased. Therefore, the causality should be developed.

- It are 42 femicides mentioned, 4 of them children or adolescents and this massive data is invisible during the study.

Author Response

We would like to thank the reviewer for the suggestions throughout the manuscript. We will provide a response to them in the following paragraphs:

1. i) literature review

- It is suggested the definition of violence against women from Action against violence against women and domestic violence (Istanbul Convention) (Council of Europe, 2011)

Amended.

- The data presented is massively from 2020, which is a small part from all the data available regarding the pandemic context. Therefore, there are much more updated data that would be interesting to underline the project, including portuguese official data. Also, the United Nations study on Violence Against Women that is cited (2013) is not the latest version (2021) and together with other two studies, the authors slide to generalization using broad terms like “general” and “universal”.

We have included more updated studies, including the ones referred by the reviewer.

- It should be clearer the relation between the increase of VWA and decrease of official reports

Amended.

– it should be unequivocal what is perceived as “requests” - towards a domestic violence structure, towards authorities, towards familiar/social help? Again, the data that underline this point is not the most recent.

Amended.

1. ii) The Method

- Regarding the method, the criteria for crimes categorization is doubtful – Reported crimes were more often for intimate partner violence, followed by psychological violence related crimes. When there is psychological violence within a intimate relationship (as it occurs commonly), how is categorizes? Does it mean that there are situations that are not within the spectrum of domestic violence? Is it gender violence? If it is not, it should be evident from start (violence against women or violence that takes place also against women?)

 This was amended in the revised manuscript.

- It should be clear why are adult women considered starting form 17 years from age (legal age) and children considered before 11 years.

Amended.

- The lockdown is briefly mentioned once and not specified that is decreed by portuguese government and it should be traced comparisons between pre and post lockdown period and the same period before and after 2021 (e.g., same months in 2019 and 2021).

Amended. As the manuscript is originated from an investigation project that occurred during 2020, it is not possible to retrieve data from 2021.

- When there is mentioned a higher frequency of sexual crime (172), the affirmation lacks comparison and data presented in table 2, specifically the total n and percentage of IPV and Offenses (cf. 127) and would be interesting to understand why the emotional support is considered in all types of support.

Due the changes suggested by other reviewers, in the revised manuscript this section no longer exist.

- According to table 2, the types of violence more prevalent were psychological and sexual, but in the discussion, it is contradicted, underlining psychological and physical violence.

Amended.

- Also, it is referred that the “requests for help” decreased due to restriction policies, but later is mentioned that remote support had increased. Therefore, the causality should be developed.

Amended.

- It are 42 femicides mentioned, 4 of them children or adolescents and this massive data is invisible during the study.

Amended.

All suggestions have been considered and changes have been made in the revised manuscript.

Round 2

Reviewer 1 Report

I appreciate that the authors included data from 2019. However, even after the revision my concerns of fundamental nature with regards to points raised in my previous report remain. I am concerned about the theoretical foundation of the paper and its empirical implementation.

How is violence against women (VAW) defined with respect to the empirical analysis undertaken? As far as I understand it includes women who identify as victims of crime during the call as well as third party callers who report about a crime committed to a woman. Is that correct? If yes, please state that clearly in the manuscript. The description as it is currently, is not stringent at all and scattered over different paragraphs.

Why are thefts and other types of crime included? Why is this considered VAW in general? What is the theoretical consideration of including it with respect to COVID and confinement measures?

Please be more clear and stringent with respect to the description of the study's sample selection from the APAV database.

One would expect that forms of violence that happen outside of the home decrease whereas crimes inside of the home, such as intimate partner violence, increase. The paper needs a stronger argument and some empirical or anecdotal support why IPV decreases whereas other forms of crime, those outside of the home increase. This is inconsistent with the literature. Why would this be different in Portugal?

For female children (younger than 17), intimate partner violence was reported to increase. However, I am puzzled that children have intimate partners at all. It might suggest that something went wrong in the data work.

Although, the study now includes 2019 data, the analysis relies on a simple comparison of means, disregarding any seasonality. This may be fine (journal's decision). I was unable to read the tables as its columns were displaced. In the text no standard errors, p-values or confidence intervals were reported. Those should be added. I also couldn't see any in the tables, but this might have bee due to their erreroneous representation.

Some of the literature is not clearly presented. For example for Ebert and Steinert 2021 and Gama et al. 2020 prevalence rates were reported, but without context or comparison rates. Ebert and Steinert 2021 could be better cited in the section on risk factors.

The authors should state clearly what this study contributes to the existing studies on VAW and IPV during COVID-19 in Portugal. 

The paper includes language errors throughout. The paper would benefit immensely from a more stringent writing. I am not sure if "senescence" is the adequate choice of words in the title.

Author Response

I appreciate that the authors included data from 2019. However, even after the revision my concerns of fundamental nature with regards to points raised in my previous report remain. I am concerned about the theoretical foundation of the paper and its empirical implementation.

How is violence against women (VAW) defined with respect to the empirical analysis undertaken? As far as I understand it includes women who identify as victims of crime during the call as well as third party callers who report about a crime committed to a woman. Is that correct? If yes, please state that clearly in the manuscript. The description as it is currently, is not stringent at all and scattered over different paragraphs.

We appreciate the feedback. We have clarified it more specifically in the Method section (118-120). We have considered any request for help that indicated that a female victim (who identifies as a woman) suffered a crime. We have also clarified how many of these victims were the ones that sought help in the victim's association and how many requests were from third parties.

Why are thefts and other types of crime included? Why is this considered VAW in general? What is the theoretical consideration of including it with respect to COVID and confinement measures?

We have considered all crimes that female victims have reported to the victim's support association, most of which are IPV (thus VAW), but we also wanted to describe the remaining types of crimes reported. We believe it would be an incomplete analysis of the data if we excluded those requests for help, especially as not many studies consider other crimes female victims suffer besides IPV. We have clarified it in this reviewed manuscript.

Please be more clear and stringent with respect to the description of the study's sample selection from the APAV database.

Amended.

One would expect that forms of violence that happen outside of the home decrease whereas crimes inside of the home, such as intimate partner violence, increase. The paper needs a stronger argument and some empirical or anecdotal support why IPV decreases whereas other forms of crime, those outside of the home increase. This is inconsistent with the literature. Why would this be different in Portugal?

We appreciate the feedback. APAV is one of the existing victim support associations. Despite being the main one in Portugal, one should not forget that some criminal cases may be reported directly to other victim support associations, to the police or to legal institutions, and not to APAV (not to mention the ones that go unreported). Our results include a significant number of data, however, it could be possible that more requesters for help would reach APAV in 2020 regarding non-IPV situations than in IPV ones. This could explain the general decrease in requests to APAV regarding IPV, but could be an increase in other institutions - nevertheless, it is not possible to confirm this hypothesis whatsoever.   

For female children (younger than 17), intimate partner violence was reported to increase. However, I am puzzled that children have intimate partners at all. It might suggest that something went wrong in the data work.

We understand the confusion. We have added a note mentioning that the youngest IPV victim was 15, which is plausible (e.g., Stöckl et al., 2014).

Although, the study now includes 2019 data, the analysis relies on a simple comparison of means, disregarding any seasonality. This may be fine (journal's decision). I was unable to read the tables as its columns were displaced. In the text no standard errors, p-values or confidence intervals were reported. Those should be added. I also couldn't see any in the tables, but this might have bee due to their erreroneous representation.

We have adjusted the tables. Due to the categorical nature of the data and the reduced number of variables with a low number of missing cases, we have decided to represent the results only descriptively, with the addiction of the growth rates.

Some of the literature is not clearly presented. For example for Ebert and Steinert 2021 and Gama et al. 2020 prevalence rates were reported, but without context or comparison rates. Ebert and Steinert 2021 could be better cited in the section on risk factors.

Amended.

The authors should state clearly what this study contributes to the existing studies on VAW and IPV during COVID-19 in Portugal. 

Amended.

The paper includes language errors throughout. The paper would benefit immensely from a more stringent writing. I am not sure if "senescence" is the adequate choice of words in the title.

Amended.

Reviewer 2 Report

Highly improved, very interesting paper from Portugal data. I had some difficulties to read the new version of tables due to technical problems (columns too small), I will read  the tables with attention of the published paper.

Author Response

We appreciate the suggestions. We have adjusted the tables.

Reviewer 3 Report

Regarding this paper, it is suggested to review percentages from line 195 to 198 (e.g., percentage of attempted femicide/femicide increased or decreased?), as well as not mixing pre and post pandemic values as in line 261 (i.e., WHO 2013 and 2021) – It would be important to verify the percentage of attempted femicide/femicide requests for help pre pandemic to muscle the comparison.

The right side of Table 2 is not understandable and the text above it does not match e.g., n of IVP; in line 194 refers to n= 48 attempted femicide/femicide, 13 of them attempted).

Regarding the percentage presents of young victims, the numbers should be clearer. For instances, in line 216/217 it is said However, when comparing the growth of this type of violence, a decrease was noted in young victims (17 years old or younger; 68%), however above (line 207/208) is mentioned Concerning victims 17 years old or younger, it is registered an increase of 116% in psychological violence related crimes and an increase of 52% of IPV – How can you affirm an increase of the violence with an decrease of the decrease on the hel requests?  (as in line 178/179)

 It’s not clear, and would be very important to understand, (is posible) if this increase of VAW, specially IPV, if started in the pandemic context or if it escalated during 2020, specifically during lockdown and therefore it increased the requests for help (nonetheless the non-reported cases which will remain invisible), some of them addressing life risk.

Author Response

Regarding this paper, it is suggested to review percentages from line 195 to 198 (e.g., percentage of attempted femicide/femicide increased or decreased?), as well as not mixing pre and post pandemic values as in line 261 (i.e., WHO 2013 and 2021) – It would be important to verify the percentage of attempted femicide/femicide requests for help pre pandemic to muscle the comparison.

Amended.

The right side of Table 2 is not understandable and the text above it does not match e.g., n of IVP; in line 194 refers to n= 48 attempted femicide/femicide, 13 of them attempted).

We have fixed the table and we have clarified the attempted femicide/femicide confusion. 

Regarding the percentage presents of young victims, the numbers should be clearer. For instances, in line 216/217 it is said However, when comparing the growth of this type of violence, a decrease was noted in young victims (17 years old or younger; 68%), however above (line 207/208) is mentioned Concerning victims 17 years old or younger, it is registered an increase of 116% in psychological violence related crimes and an increase of 52% of IPV – How can you affirm an increase of the violence with an decrease of the decrease on the hel requests?  (as in line 178/179)

We understand the confusion. We have these three categories: 1) only psychological violence was reported; 2) only physical violence was reported; 3) both psychological and physical violence were reported.

Therefore, there was a decrease in reporting both psychological and physical violence (3.), but an increase in only psychological violence (1.). This may suggest that victims who suffered both physical and psychological violence in 2019 may have suffered less physical and more psychological violence in 2020 (however, there is no possibility to support this hypothesis as it is not possible to confirm if the victim has requested help multiple times because of the limitations inherent to the database).

 It’s not clear, and would be very important to understand, (is posible) if this increase of VAW, specially IPV, if started in the pandemic context or if it escalated during 2020, specifically during lockdown and therefore it increased the requests for help (nonetheless the non-reported cases which will remain invisible), some of them addressing life risk.

We understand the suggestion and agree on its importance, however, due to the fact we only analyzed a dataset and we did not have direct contact with the victims themself, it is difficult to say if it started or escalated during lockdown/2020. The increase in requests for help in 2020 may indicate an aggravation of it when compared to 2019, nonetheless, it would have been necessary to maybe compare maybe to earlier years (e.g., 2017 or 2018) to understand the effect and, unfortunately, the funded project was only for 2019 and 2020 data.

We appreciate, once again, the suggestions proposed by the reviewer.