Când lipsa serviciilor sociale devine urgență psihiatrică

Introduction

Domestic violence, including its extreme forms such as femicide (the gender-based killing of women), is a major public health and social problem, with devastating consequences for victims and communities. While traditional efforts have focused on protecting victims, growing research suggests that comprehensively addressing perpetrators, through social and psychiatric interventions, is essential to prevent recidivism and reduce violence in the long term⁽¹⁾. This article explores the links between mental disorders and recurrent violent behavior, the types of psychiatric and social interventions available to perpetrators, the ethical and legal challenges involved, and examples of international good practice, based on recent medical literature (2017-2025, with a focus on articles indexed in the PubMed database).

A general observation of the way in which femicide is represented in the public narrative in Romania, in newspaper articles, leads us to the idea of outlining a social problem or, more likely, to that of awareness at a social level; for the local space, cultural factors orient towards social dimensions that describe the family as a space of violence and institutional interaction⁽²⁾. The role of gender stereotypes, traditional mentalities, poverty and lack of institutional support as cultural factors perpetuate the phenomenon^(3,4). The general theme in the newspapers is centered on the dimension of the phenomenon and the gravity of the situation; articles frequently emphasize the number of cases reported in 2025 (e.g., “the 30^th case”, “the 32^nd case”, “the 33^rd case”), which creates a framework of ongoing crisis^(5-7). It is highlighted that femicide is a systemic phenomenon – not isolated cases – and it is correlated with socially tolerated domestic violence and the inefficiency of the authorities’ response. Repeated references to official statistics (Eurostat, NGOs reports) are to contextualize the dimension of the phenomenon⁽⁶⁾. Civic and political reactions are another theme, presenting public protests, petitions with tens of thousands of signatures, pressure from civil society for a distinct law on femicide⁽⁸⁾. NGOs (e.g., FILIA Center, “Schimbarea suntem NOI”) appear as central actors that formulate concrete measures: effectively monitored protection orders, gender education in schools, real support for victims. There are mentions of the involvement of politicians and intentions for legislative reform. The psychological and preventive dimension is reached in many cases; an interview in which a psychologist explains the stages of an abusive relationship, the alarm signals, the mechanisms of manipulation and emotional dependence is instructive. The cycle of violence and the “normalization” of aggressive behavior are discussed⁽⁶⁾. The manner of expression and tone of the discourse is alarming and urgent, using terms such as “earthquaking”, “with bestiality”, “silent crisis”, “complicit state” to convey the gravity of the phenomenon. There is an alternation between the factual account (data, legal measures, press releases from the prosecutor’s office) and the emotional description (content warnings, violent images, empathy for victims). The narrative structure is dual; in case reports we find the chronology of the facts and the legal consequences, and in analytical articles/interviews we find contextualization, recommendations and calls to action. The language is specialized in some sections (legal terms, psychological concepts such as “gaslighting”, “Stockholm syndrome”) and explained in a way that is understandable to the public⁽⁶⁾. The representation of the aggressor is done through the modes of action (extreme and visible violence – beatings, kidnappings, killing by firearms or by brutal methods – dragging with a car, repeated blows)⁽⁵⁾, through persistence and escalation (often the aggressor has a history of violence – e.g., child abuse, violation of protection orders) and through action in public space (some acts are committed in full view of the world – restaurants, streets, which shows the absence of fear of consequences). The treatment by the authorities seems to be quite consistent: preventive arrest in most cases, but previous preventive measures (protection orders) prove ineffective^(5,7). In the reports, the aggressor is sometimes presented with the presumption of innocence (binding legal formula), but in other articles the emphasis falls on his clear culpability and dangerousness. The lack of an effective mechanism for monitoring aggressors (e.g., electronic bracelets are mentioned, but not widely implemented) is another common aspect presented⁽⁸⁾. There are no positive or justifying representations of the aggressor; social support of the aggressor or concern for his public image remain unrepresented. In the journalistic “psychological” analysis, the aggressor is described as having toxic behavioral patterns, possible previous traumas, but without minimizing their responsibility. NGOs and activists emphasize that the state acts passively towards the aggressors, which leads to the repetition of the crimes⁽⁶⁾. Journalistic texts jointly construct a “national crisis” narrative; there is a tension between the official discourse (which respects the procedure and the presumption of innocence) and the activist/journalistic discourse (which demands immediate action). The representation of the aggressor is consistently negative, but the main emphasis falls on the victims and the institutional deficiencies, not on the detailed profiling of the perpetrators.

Local issues

In Romania, protecting victims of domestic violence has become a major priority in recent years. Since 2020, the intense media coverage of the cases has favored the amendment and republishing of Law no. 217/2003 on the prevention and combating of domestic violence⁽⁹⁾, which also regulates the procedure for issuing protection orders: provisional (issued by the police for a maximum of five days) and definitive (issued by the court, up to 12 months). Their number is increasing. According to the Romanian Police, “in the first four months of 2025, 3788 provisional protection orders were issued, of which 1464 were converted into definitive orders; police officers intervened in 40,030 cases of domestic violence, mostly in rural areas; 57% were hitting or other violence and 12% were threats”⁽¹⁰⁾. The phenomenon probably remains underreported, as data are dispersed between institutions (People’s Advocate; Public Ministry; Ministry of Internal Affairs).

The current regulation does not provide for residential services dedicated to evicted aggressors, who are directed to night shelters for homeless people, managed locally. Article 31 of the law establishes the procedure for orientation, escort and declaration of the accommodation address, under the coordination of the police and mobile teams. In practice, many shelters refuse to accept people with psychiatric disorders, citing a lack of resources, and the residential infrastructure remains deficient⁽¹¹⁾.

This lack of options frequently determines the referral of evacuated aggressors to psychiatric services, with a request for evaluation and possible hospitalization. Hospitalization is regulated by Law no. 487/2002⁽¹²⁾ and Order of the Ministry of Health no. 488/2016, which require the existence of a psychiatric emergency, the patient’s consent or the conditions for involuntary hospitalization. In the absence of these criteria, services cannot be reimbursed. The pressure on psychiatrists is increased, as they face insistent requests and ethical risk situations (“Where should we take him if we don’t hospitalize him?”).

Evacuated aggressors, often without material resources or documents, become additionally vulnerable, especially those with preexisting mental disorders. The lack of a functional network of centers and poor coordination between institutions lead to a “dilution of responsibility”, in which coercive solutions prevail to the detriment of therapeutic and social ones.

Practical situation – case presentation

A patient with minimal psychiatric history (one hospitalization and one presentation to psychiatry over a year ago; diagnosis of simple schizophrenia, with treatment followed for about a month, then interrupted) is brought to the hospital by ambulance, accompanied by the police, at the request of the father who claimed an aggression. A provisional protection order was issued⁽⁹⁾. The patient has no other relatives or social support, and he was taken in without money, documents, medication, clothes or shoes.

Psychiatric examination: oriented, cooperative, coherent patient, without hallucinations, delusions or behavioral disorders; critical judgment preserved. The patient denies the assault, stating that the father requested police intervention due to a conflict over money; there were no witnesses. During the transport and assessment, he did not exhibit hetero- or self-aggressive behavior. He refuses hospitalization and does not accept the previous diagnosis, considering it erroneous.

In this situation, a psychiatric emergency is not found, and the legal criteria for involuntary hospitalization are not met (Law no. 487/2002; Ministry of Health Order no. 488/2016). However, the emergency crew insists on hospitalization, citing the absence of alternative accommodation and the risk of further aggression. At the time of the assessment, the existence of an “imminent” risk, as defined in the legal norms, cannot be assessed.

Analysis: the case highlights the risk of systemic abuse of psychiatric hospitalization as a substitute solution to the lack of social services. The rhetoric of “stay a few days, then leave” often translates into long-term hospitalizations, sometimes until the expiration of the protection order. The consequences include stigmatization, loss of skills, difficulties in social reintegration, and the patient’s continued vulnerability. In parallel, the doctor is placed in a moral vulnerability, oscillating between respecting the patient’s autonomy and institutional pressure to prevent a possible future risk.

Methodology

A methodology combining narrative review and a quantitative, bibliometric method (using the VOSviewer program) was used. The narrative review used the data extracted through the quantitative method and was compared with the national situation.

The PubMed platform was queried using the following formula: (“domestic violence” [MeSH] OR “intimate partner violence” [MeSH]) AND (“aggressor” OR “perpetrator”) AND (“mental disorder” OR “mental illness” OR “psychiatric”) AND (“community treatment order” OR “restraint” OR “seclusion” OR “de-escalation” OR “social support” OR “protective services”). A total of 62 results were obtained, in the period 1982-2025; up to the year 2000 (three results), only a limited number of sources (five articles) were identified. Most papers are from the period 2017-2025 (with a maximum of six papers per year in 2017, 2019 and 2022).

Correlation between mental disorders and recurrent aggressive behaviors

Recent studies confirm a significant association between certain mental disorders and the tendency towards repeated domestic violence. Men who commit violence against their partners frequently present high rates of mental health problems – from personality disorders (e.g., antisocial or borderline traits) to affective disorders (depression, anxiety) or posttraumatic stress disorder (PTSD) – compared to the general population^(13,14).

For example, a meta-analysis indicates that effective treatment of mental health problems is associated with a reduction in intimate violence: reductions in psychiatric symptoms in perpetrators are accompanied by reductions in violent behavior⁽¹⁴⁾. Alcohol and substance use disorders are also common among perpetrators, and they play a major role in precipitating violent episodes. A classic study found that 53% of men arrested for domestic violence met the criteria for alcohol use disorder and 31% for drug use disorder⁽¹³⁾. Acute substance use (alcohol, cocaine, marijuana) was significantly associated with increased aggression⁽¹⁵⁾. Mood disorders also contribute to risk: depressive symptoms and PTSD (especially hyperarousal and anger) can increase the likelihood of violent outbursts^(16,17).

People with severe mental disorders (e.g., schizophrenia, bipolar disorder) show a disproportionately high involvement in domestic violence, both as perpetrators and victims. A 2024 study of 942 patients with severe mental disorders reported that 22% of them had committed physical aggression against a partner within a year (and 27% were victims of partner violence)⁽¹⁸⁾. The relationship between victim and aggressor can often be bidirectional in the context of serious mental pathology, with anger playing a complex mediating role in this cycle of violence⁽¹⁸⁾. These data do not imply that mental disorders inevitably cause violence – most people with mental illness are not violent. However, the presence of diagnoses such as antisocial or borderline personality disorder, addiction or PTSD may increase the risk of recurrent aggressive behavior and recidivism after the first incident^(13,14). In addition, it has been observed that perpetrators with higher levels of psychopathology are more likely to drop out of rehabilitation programs and reoffend, compared to those without such problems⁽¹⁴⁾. This suggests the need for interventions to also target the mental health component to break the cycle of violence.

In Romania, the link between mental health problems, alcohol consumption and intimate partner violence overlaps with several relevant local realities: episodic excessive alcohol consumption is among the highest in the EU (approximately 35% of adults monthly; for men, approximately 53%), which amplifies the risk of escalating domestic conflicts and potentiates the effects of affective, personality or PTSD disorders on repeated violent behavior⁽¹⁹⁾. In parallel, public policies have introduced tools aimed at reducing recidivism: the Electronic Monitoring Information System (SIME) for aggressors – extended nationwide from 1 October 2024 – strengthens the enforcement of protection orders and creates an operational window for referral to psychiatric and addiction services when risks are identified⁽²⁰⁾. At the service level, ANES reports mapping and standardizing the support network for victims and aggressors (including multidisciplinary training), which supports the integration of mental health and addiction interventions into perpetrator programs – a direction aligned with evidence that reducing psychiatric symptoms and problematic consumption decreases the incidence of violence⁽¹¹⁾.

Psychiatric interventions applied to aggressors

Psychiatric interventions aim both at treating comorbid mental disorders in aggressors and at managing impulsivity and the risk of violence. Drug treatment is personalized according to the specific diagnosis: for example, antidepressants (SSRIs) can help reduce irritability and impulsivity in depression or PTSD, mood stabilizers and antipsychotics may be indicated in people with bipolar, schizoaffective or other psychoses who manifest aggression, and anxiolytic medication or treatments for ADHD may be useful in controlling anger and self-control deficits in some cases. The goal of pharmacotherapy is to reduce the psychological symptoms that fuel violent behavior (e.g. paranoia, intense anger, uncontrolled impulsivity).

An important measure in serious cases is the non-voluntary hospitalization (involuntary psychiatric hospitalization) of the aggressor, applied according to the mental health law, when he poses an imminent danger to himself or others. In cases of domestic violence associated with acute psychosis, mania or other severe disorders, involuntary hospitalization may be necessary to prevent immediate harm to the victim and to allow the initiation of treatment. However, this intervention is a last resort, since it deprives the person of their liberty; from an ethical point of view, it must be strictly justified by the need to protect the life and integrity of those involved. During hospitalization, in addition to pharmacological stabilization, psychotherapeutic interventions for anger management and counseling can be initiated, in parallel with planning for the victim’s safety.

After stabilization, in order to avoid relapse into violent behavior, some countries resort to Community Treatment Orders (CTOs). Through a CTO, the aggressor with a mental disorder is legally obliged to follow outpatient treatment (medication, attendance at psychiatric checkups, therapy), as a condition for remaining in the community. The aim is to prevent psychiatric decompensation that could lead to a relapse into violence. However, the effectiveness of CTOs in reducing violent behavior is questionable. A 2025 meta-analysis found no significant differences in rates of violence or criminal offenses between offenders under CTO and those in voluntary treatment – with the expected benefits of CTO decreasing as the quality of studies increased (the results of RCTs being neutral)⁽²¹⁾. This is likely because the risk of violence is increased by comorbid or nonclinical variables that are beyond the scope of CTOs. The authors concluded that CTOs, as a single measure, fail to reduce aggression because they do not address concomitant risk factors (such as substance abuse, prior trauma, or poor socioeconomic background) that go beyond the strictly medical realm⁽²¹⁾. In other words, the legislative control of treatment has limited impact in the absence of comprehensive psychosocial support.

Another possible psychiatric intervention, especially in the judicial context, is the referral of aggressors with severe mental health problems to forensic psychiatry services or departments. In these settings, aggressors who have committed serious crimes can receive psychiatric treatment in a secure setting, either as a court-ordered safety measure (e.g., mandatory medical hospitalization instead of detention, if criminal irresponsibility for medical reasons is found), or as part of the criminal sentence (e.g., therapeutic programs in prison). Such interventions aim at both clinical stabilization and behavioral responsibility.

In addition to medication and hospitalization, specialized psychotherapies play an essential role. Aggressors can benefit from individual psychotherapy (e.g., cognitive-behavioral therapy focused on anger management, conflict resolution, and restructuring dysfunctional beliefs about relationships), as well as group therapy for domestic violence (aggressor groups, facilitated by therapists, where empathy, responsibility-taking and nonviolent communication techniques are worked on). In practice, it has been found that many aggressors reach health services complaining of psychological symptoms (anxiety, depression, insomnia) or substance abuse problems, without explicitly mentioning violent behaviors⁽²²⁾. If the therapeutic intervention is limited to prescribing medications (e.g., antidepressants) for these symptoms, without directly addressing the abusive behavior, the results may be insufficient – some aggressors have perceived this as unhelpful⁽²²⁾. In contrast, when given access to psychotherapeutic interventions focused on abusive behavior, many perpetrators showed greater interest and commitment to change. For example, one participant interviewed in a qualitative study attributed the improvement in his violent behavior to the combination of psychiatric treatment (for his mental health issues) and participation in a perpetrator program while incarcerated. This case illustrates the importance of an integrated approach: medication treatment can control the individual’s emotional or impulsive vulnerabilities, while psychotherapeutic programs develop healthy relationship skills and anger management mechanisms.

Psychiatric interventions for perpetrators include: pharmacotherapy appropriate to the diagnosis (to reduce impulsivity and symptoms that may fuel violence), involuntary hospitalization in cases of acute psychopathological risk, community treatment orders to ensure adherence to treatment after hospitalization, individual and group psychotherapy focused on domestic violence and, when necessary, referral to forensic care systems. The literature emphasizes that such interventions have maximum effect when integrated into a broader, multidisciplinary framework.

In Romania, involuntary hospitalization of a person with a mental disorder is regulated by the Mental Health Law no. 487/2002 (republished in 2012), and it is applied when there is imminent danger to oneself or others, with procedures and guarantees provided for in the regulatory framework. In parallel, the Criminal Code provides for medical safety measures: “compulsion to medical treatment” (outpatient, ordered by the court) and “medical hospitalization” (in specialized units), with the possibility of transforming outpatient treatment into hospitalization if it is not respected – a mechanism functionally close to the so-called CTO, but regulated criminally, not by a special law dedicated to CTO. For serious cases or crimes, referral to forensic psychiatry/hospitals “for safety measures” offers treatment in a secure setting, under judicial control. On the victim protection component, Law 217/2003 provides for the protection order and, since 2018, the “provisional protection order” issued by the police for five days; since 2021, electronic monitoring of aggressors through the Electronic Monitoring Information System (SIME) has been introduced and expanded nationally in stages, reaching large-scale implementation from 1 October 2024. At the same time, the CPT (European Committee for the Prevention of Torture and Inhuman or Degrading Treatment or Punishment – Council of Europe) evaluations from 2023 indicate pressure on the infrastructure and human resources in psychiatry (over 8800 beds in 36 psychiatric hospitals and another 5400 in county departments), highlighting the need for adequate conditions and care – an essential aspect when discussing involuntary hospitalization and safety⁽²³⁾. In the post-discharge area, the Government approved in 2024 a 2024-2029 Action Plan that includes integrated case management and strengthening community services – aligned with your conclusion regarding the need for a multidisciplinary framework and psychosocial support⁽²⁴⁾. Recent local literature reports the phenomenon of the “revolving door” in services (frequent readmissions), which reinforces the idea that pharmacotherapy alone is not enough and that psychosocial and adherence interventions are critical⁽²⁵⁾.

Social intervention models for rehabilitation and relapse prevention

Social interventions play a crucial role in the rehabilitation of aggressors and the prevention of relapse, complementing psychiatric treatment. These interventions aim to educate, empower and socially reintegrate people who have committed violence, while also addressing the contextual factors (family, community, cultural) that favor aggressive behavior.

A first pillar is represented by the programs for reeducating aggressors, often known as Batterer Intervention Programs (BIPs). These programs, usually carried out in groups, combine psychoeducational techniques with cognitive-behavioral interventions. A classic example is the Duluth model, originally developed in the USA, which emphasizes the awareness by aggressors of their attitudes of control and dominance over their partners and the promotion of equal relationships. Modern programs, however, have evolved beyond Duluth, incorporating evidence-based strategies. For example, they include elements of motivational interviewing to increase perpetrators’ commitment to change and components inspired by dialectical behavior therapy (DBT) to improve emotion regulation and stress tolerance⁽¹⁴⁾. A systematic review published in 2024 found that interventions that added substance abuse or mental health treatment components to perpetrator programs had promising results, compared with standard programs that did not address these aspects⁽¹⁴⁾. In fact, reductions in mental health problems (e.g., reductions in depression, anxiety) in perpetrators were associated with decreases in intimate violence, suggesting that addressing these problems as part of rehabilitation helps prevent recidivism⁽²⁶⁾.

Social support is the second essential element. Many perpetrators face socioeconomic difficulties – unemployment, housing instability, lack of a support network – which can fuel stress and, implicitly, violent behavior. Community-based social support interventions attempt to reintegrate the perpetrator prosocially: they help them find a job, continue their education or training, and access financial or legal counseling. By stabilizing these aspects, external stressors that can precipitate a relapse into violence are reduced. At the same time, support groups (such as antiviolence men’s circles) or mentoring can provide positive role models and long-term accountability. Psychosocial education is integrated into many programs. This involves structured sessions in which perpetrators learn nonviolent communication skills, peaceful conflict resolution techniques, anger and jealousy management, and empathy for victims. The effects of their violence on their partners and children are explained to perpetrators, in order to help them take responsibility and stop minimizing or denying it. Research shows that, by participating in such programs, positive changes in attitudes can be observed: a decrease in sexist beliefs and the tendency to blame the victim, an increase in awareness of the seriousness of the acts committed, and the development of prosocial attitudes^(27,28). In addition, some participants also report improvements in their own symptoms of depression or anxiety, as well as an increase in perceived social support, which contributes to sustainable change⁽¹⁴⁾.

Modern social intervention models also emphasize community and institutional involvement. An internationally recognized example is the Coordinated Community Response (CCR) model, which brings together multiple agencies: probation, police, courts, social work, victim support organizations and mental health services. CCR recognizes domestic violence as a complex problem that affects the entire social system, and requires the synchronized intervention of all these actors⁽²⁹⁾. For example, in a city implementing the CCR model, when an abuser is released on parole, probation works with a local abuser program to provide counseling, while police monitor compliance with restraining orders, and victim organizations maintain contact with the partner to ensure her safety. This integrated approach maximizes the chances of successful rehabilitation and minimizes the risk that the abuser will “slip” through the cracks of the system and reoffend.

The effectiveness of interventions is still being studied, but meta-analyses suggest that well-designed programs can have a modest but significant effect on reducing recidivism. For example, a meta-analysis of 59 controlled studies found a significant (although small) decrease in the recurrence of physical abuse among perpetrators who completed a program compared to those who did not⁽³⁰⁾. A critical aspect, however, is program completion: perpetrators with alcohol or other disorders are at higher risk of dropping out, which reduces the effectiveness of the intervention. Fortunately, some studies show that, for those perpetrators who manage to complete the program, the recidivism rate does not differ significantly between those with and without alcohol problems – a sign that the intervention was beneficial regardless of the presence of alcoholism, as long as the participants remained involved^(14,31). To increase retention, more and more programs are adopting motivational components and addressing individual barriers (e.g., providing transportation assistance to sessions, flexible hours, additional individual counselling for those with special needs). Promising results of including motivational strategies indicate significant reductions in recidivism and improved adherence to treatment⁽³²⁾.

A notable example of good practice is the Drive project in the UK, which targets high-risk offenders with particularly dangerous repetitive behavior. The project combines intensive case management – each offender is monitored by a multidisciplinary team, including social workers, substance abuse counselors and psychologists – with multi-institutional cooperation (police, social services, victim NGOs). In the pilot phase, this program reported dramatic reductions in violent behavior: for example, the incidence of physical abuse fell by 82%, and the incidence of sexual abuse decreased by 88% among perpetrators included in the program compared to the pre-program situation (https://drivepartnership.org.uk). Although these results should be interpreted with caution (as they are not from a randomized trial but from a program evaluation), they highlight the potential of personalized interventions and close supervision in cases of dangerous perpetrators. Finally, the involvement of extended family and the community can support rehabilitation. Some programs invite (with the consent of the victims) family members or community mentors to attend certain sessions, in order to strengthen the perpetrator’s support network and help them stay on the path to change once the formal program has ended. Community antiviolence education (awareness campaigns, involvement of community leaders, clergy, etc.) creates a social environment less tolerant of violence, which puts positive pressure on aggressors to maintain their nonviolent behavior.

In Romania, social interventions for perpetrators are linked to the measures in Law no. 217/2003 (with amendments from 2018-2024) and to related instruments: the protection order (court) and the provisional protection order issued by the police for five days, complemented from 2024 by the national extension of electronic monitoring (SIME) – useful for supervising compliance with restrictions and contact with services. On the rehabilitation component, the counseling of perpetrators is methodologically regulated (including standardized progress reporting), and specialized services have been developed in recent months/years under the subordination of local authorities – e.g., DGASPC (General Directorate of Social Assistance and Child Protection) counseling centers, District 6 Bucharest; Dolj), which offer psychological/social counseling, rehabilitation and reintegration, voluntarily or at the disposal of the court⁽³³⁾. The “coordinated community response” (CCR) approach is promoted in interinstitutional guides and national working materials (police – probation –social assistance – health – NGOs), with an emphasis on the rapid circulation of information and the complementarity of programs for aggressors with victim support (www.transcena.ro); the practical efficiency of these guides remains to be assessed. In practice, probation and courts can attach obligations to participate in programs for aggressors; the expansion of SIME and legislative updates in 2024 increase the capacity to monitor compliance with measures and contact with community services (https://probatiune.just.ro/).

On the effectiveness of programs, recent international literature indicates modest but significant effects of well-designed interventions and more consistent ones when mental health or substance abuse components are integrated (a meta-analysis by Babcock et al., 2024; a systematic review by Sousa et al., 2024)^(14,30). For high-risk offenders, Drive (UK) intensive case management models, anchored in a robust CCR, have reported substantial reductions in abuse in independent evaluations, although based predominantly on program evaluations, not RCTs. These findings support the “multi-component” direction that this paper proposes.

Ethical and legal challenges in protecting the aggressor without endangering the victim’s safety

Addressing aggressors raises numerous ethical and legal dilemmas, as it involves a delicate balance between the rights and well-being of the aggressor – as a citizen and often a patient with health problems – and the right to safety of the victim. Any intervention intended to “protect” or help the aggressor must be evaluated in light of the potential risks to the victim, so that the help offered to the aggressor does not result in exposing the victim to new dangers.

A first challenge is respecting confidentiality in the therapeutic relationship with the aggressor versus the obligation to protect potential victims. Mental health professionals may find themselves in the position of receiving disclosures about acts of violence from an aggressor patient. Ethically, confidentiality is a central value in medicine; however, if there is an imminent risk to someone’s life or integrity (for example, the patient declares concrete intentions to harm his or her partner), the obligation to warn and protect the person concerned arises (the Tarasoff principle, recognized in some jurisdictions). In practice, aggressors are often reluctant to disclose their violent behaviors to doctors or therapists, precisely for fear of legal consequences. A qualitative study highlighted that anxiety related to confidentiality is a major obstacle: some aggressors avoided discussing violence issues with doctors for fear that the information would be noted in the medical file and lead to the involvement of the police or child protection services⁽²²⁾. One participant even stated that he would rather suffer the consequences of a mutually abusive relationship than seek professional help, for fear that he could lose custody of his or her children if the family problems became known to the authorities. This situation highlights the dilemma: how can we encourage perpetrators to seek help (and be honest with the therapist) while ensuring a certain degree of confidentiality, without compromising the need to intervene when the safety of the victim (or children) is at stake?

Another ethical challenge is related to coercive interventions (such as involuntary hospitalization or forced treatment). On the one hand, forcing a perpetrator to undergo treatment (psychiatric or otherwise) may be justified by the risk they pose, thereby preventing harm to others. On the other hand, from a human rights perspective, any deprivation of liberty or treatment administered without consent is an infringement on the autonomy of the individual. CTOs, for example, have been criticized by civil rights organizations, who argue that forcing someone to take medication in the community violates personal freedom, especially if that person has not (yet) committed a serious crime⁽²¹⁾. Ethically, the question arises: to what extent can we justify violating the will of an aggressor patient for the sake of preventing potential violence? The consensus in good practice guidelines is that such measures should only be used when the risk of violence is clear and present, and that they must be accompanied by procedural guarantees (periodic reviews, right to appeal, legal support for the patient).

From a legal perspective, there are instruments designed to protect the victim that can sometimes conflict with the aggressor’s need for recovery. A court-issued protection order, for example, prohibits the aggressor from having any contact with the victim for a specified period. This is essential for the victim’s safety, but it raises the issue of the aggressor’s continuity of treatment: if the aggressor and the victim lived together, once the aggressor is evicted by order, he may be left homeless – a situation that aggravates his social and psychological vulnerability. Where does the aggressor go after the restraining order? Some countries have started to develop shelters or counseling centers for aggressors (similar to shelters for victims), precisely to offer them a safe place to stay and follow the intervention program, instead of ending up on the streets or in a chaotic environment that could increase the risk of recidivism. In Romania, for example, the updated Law 217/2003 provides for the possibility of counseling aggressors and even forcing them to participate in social reintegration programs, but the necessary infrastructure (specialized centers for aggressors) is still insufficiently developed. Legally, forcing the aggressor to undergo counseling or treatment (as part of the punishment or as a complementary measure to a protection order) raises questions about how this is implemented and monitored. If the aggressor does not show up for the imposed counseling sessions, what happens next? The answer varies: failure to comply with the plan can be penalized (e.g., by transforming the punishment into a custodial sentence) or administratively, but the follow-up mechanisms must be clearly established. Another ethical aspect is stigmatization. Labelling aggressors exclusively as “criminals” can stand in the way of their rehabilitation. Professionals must find a balance between clearly condemning the act (to avoid any justification of violence) and treating the aggressor with a certain empathy as a patient, when he has real psychological problems. Some voices criticize the focus on aggressors, considering that resources should be invested only in victims; but others argue that the rehabilitation of aggressors is in itself a strategy to protect future victims. The ethical challenge here is to offer the aggressor a chance for recovery (the right to treatment, to human dignity) without minimizing the victim’s suffering or creating the impression of excusing the behavior. The concept of responsibility is fundamental: the aggressor must be involved in interventions not as a simple “sick” lacking control, but as an individual responsible for his actions, but who also has the capacity to change with the appropriate support.

Regarding the international legal framework, there is a growing consensus that interventions for aggressors must be an integral part of the strategy to combat domestic violence. Article 16 of the Istanbul Convention⁽³⁴⁾ (Convention on Preventing and Combating Violence Against Women, 2011) obliges signatory states to establish programs to treat perpetrators, both in the context of domestic violence and sexual offences. The Convention emphasizes that these programs should focus on the safety of victims and on changing the violent attitudes and behavior of perpetrators. Therefore, from a legal point of view, protecting the perpetrator (in the sense of providing them with the opportunity for rehabilitation and preventing recurrence) is not in contradiction with, but closely linked to protecting the victim – both require complementary interventions. The challenge lies in implementing these requirements: for example, ensuring that there is communication between programs for abusers and services for victims (so that if an abuser becomes dangerous or abandons treatment, the victim is alerted immediately), or developing protocols to guide professionals in the aforementioned confidentiality dilemmas.

The ethical and legal challenges in working with domestic abusers arise from the need to reconcile the aggressor’s interest in recovery with the imperative of victim safety. The solution lies in a clear framework of legal accountability for aggressors (protection orders, judicial monitoring, mandatory treatment) combined with ethically provided therapeutic and social support (confidentiality limited by safety considerations, coherent and evaluated interventions). Transparency towards victims regarding the steps taken with the aggressor, as well as their involvement in safety plans, is essential. Professional ethics require that professionals do not “abandon” the aggressor (as this would implicitly mean abandoning possible future victims), but also that they do not privilege him to the detriment of victims. Any intervention strategy must be victim-centered, even when focusing on the aggressor – a guiding principle that helps maintain the right balance.

Examples of international good practices

Several countries and international organizations have developed innovative models of intervention for aggressors which can serve as good practices. A common element of these models is the integrated approach that combines coercive measures with therapeutic support, and which involves cross-sectoral cooperation.

In the Nordic countries, for example, there is a long tradition of voluntary programs for aggressors. Norway was among the first countries to establish counseling centers for violent men, the Alternative to Violence program – ATV (https://dvota.at/), back in the 1980s. ATV offers long-term therapy to aggressors who voluntarily enroll or are referred by the courts, with promising results in reducing recidivism, being later replicated in Sweden and Denmark. The practice of these centers emphasizes understanding the aggressor’s personal dynamics (history of childhood abuse, attachment problems, etc.) combined with making them responsible for their current behavior. Evaluations have shown high levels of satisfaction among the partners of these men, who have reported a reduction or cessation of violence after their involvement in the program. In the UK, in addition to the aforementioned Drive project, there are national standards (the Respect accreditation) for programs for batterers. One element of good practice here is cooperation with victim services: batterer programs include a victim safety specialist in the team, who maintains contact with their partners throughout the program, offering them support and periodically checking on their safety. This ensures that intervention on the batterer does not take place in a vacuum, but simultaneously with support and protection for those directly affected by the violence. British evaluations, such as Project Mirabal (https://projectmirabal.co.uk/), have identified six areas of positive change in participating batterers, from stopping physical violence to improving communication and taking responsibility, suggesting that well-implemented programs can produce tangible changes in family life.

In Spain, since 2004, legislation has required all perpetrators convicted of gender-based violence to participate in rehabilitation programs (sometimes called “re-educational psychotherapy”) as part of their sentence. The Spanish prison system and probation services have developed such standardized programs, for example, the PRIA program – Programa de Intervención con Agresores⁽³⁵⁾. Published results indicate a low recidivism rate among those who complete the programme, compared to those who either did not follow it or dropped out. The key here seems to be the integrated approach: many Spanish perpetrators were also included in treatment for alcohol addiction or received family therapy where appropriate, thus showing respect for the complexity of the factors leading to violence. At the European level, the Council of Europe, through its monitoring body GREVIO, has collected and published examples of promising practices⁽³⁴⁾. The recommendations emphasize that programs for perpetrators can be effective if they are culturally adapted (for example, in migrant communities or ethnic minorities, where the causes of violence may be influenced by war trauma or cultural norms, programs need to integrate this perspective). The importance of training professionals was also highlighted: psychologists, social workers, mediators who work with perpetrators need specialized training in the field of gender-based violence, to avoid re-victimization or alliance with the perpetrator against the victim (a subtle risk that can arise in therapy). A good practice is to include supervision of teams working with perpetrators, to continuously reflect on the ethics of the intervention and the safety of all parties.

Last but not least, technology and innovation are making their way: in Canada and Australia, for example, mobile apps and telephone counseling lines have been piloted for men who feel they are losing control – basically, hotlines for aggressors, where they can call anonymously before committing an act of violence, to receive guidance from counselors in managing the emotions of the moment. While it is too early to rigorously assess the effectiveness of such initiatives, they represent an effort to prevent the acute episode of violence by intervening at the aggressor’s very moment of crisis.

International good practices show that interventions for perpetrators work best when: (1) they are part of an integrated approach (involving justice, health, social services and the community); (2) they prioritize the safety of victims and maintain a connection with them throughout the program; (3) they address the perpetrator’s related problems (mental health, addictions, unemployment, etc.); (4) they are adapted to the cultural and individual context of each case; (5) they are subject to continuous evaluation and improvement based on the results obtained. Integrating these principles into public policies and everyday practice is the direction recommended by experts to reduce the incidence of domestic violence and femicide in the long term⁽¹⁾.

In Romania, policies on intervention against aggressors have gradually aligned with European standards, especially after the ratification of the Istanbul Convention (Law no. 30/2016)⁽³⁶⁾. In recent years, methodological guides and interinstitutional mechanisms have been developed that reflect the principles of the integrated approach found in international models, such as the collaboration of justice – social assistance – health – NGOs. GREVIO (Group of Experts on Action against Violence against Women and Domestic Violence), in 2022, highlighted Romania’s progress in the field of victim protection and encouraged the expansion and standardization of programs for aggressors, in order to avoid their uneven application across counties⁽³⁴⁾. The recommendations coincide with international principles: the safety of the victim must remain a priority, and the intervention against the aggressor must be correlated with effective monitoring and multi-sectoral cooperation. On the service component, counseling centers for aggressors have been gradually developed under the subordination of local authorities (e.g., DGASPC), some in partnership with NGOs, offering psychological and social counselling, both voluntarily and as an obligation imposed by a protection order or sentence. This model partly reflects Nordic practices (Norway, Denmark), where voluntary participation is encouraged, but it can also be correlated with judicial referrals. Romania has also started to adopt the principles of CCR (Coordinated Community Response), through interinstitutional protocols and pilot projects at local level, especially in large cities, aiming at rapid coordination between probation, police, courts, social services and victim support organizations, a mechanism similar to those implemented in the USA, UK or Scandinavia. European evaluations (including GREVIO and CPT 2023) have shown that progress in Romania is promising, but important challenges remain: the lack of a uniform network of standardized programs for aggressors, insufficient training of specialized staff, and the lack of a system for evaluating the effectiveness of programs, as exists in the UK (e.g., Respect Accreditation, Project Mirabal).

Bibliometric image

General description of the map

The map displays the co-occurrence network of terms in PubMed articles (Figure 1). The nodes represent MeSH terms or keywords, and the links represent the frequency with which they appear together in the same article. The color of each cluster is the result of the VOSviewer clustering algorithm, and the size of the node indicates the frequency of occurrence.

The red cluster (left, top and center) is focused on terms related to young ages, family relationships and domestic violence. The green cluster (right, top) is focused on older ages, risk factors and epidemiological studies. The blue cluster (bottom right) is focused on interventions, social support, mental disorders and sexual abuse.

Cluster analysis and connection to theoretical directions

The red cluster – Domestic violence and family relationships includes the following major terms: “humans”, “male”, “female”, “child”, “infant”, “adolescent”, “parents”, “child abuse”, “mother-child relations”, “parenting”, “domestic violence”, “intimate partner violence”, “substance-related disorders” and “longitudinal studies”. Interpretation: it is the thematic center of the map, dominated by the term “humans” which acts as a pivot between all clusters. It strongly correlates with the theoretical direction that focuses on the relationship between mental disorders and recurrent aggressive behaviors, as it includes age-related terms (“infant”, “adolescent”, “child”), which reflects studies on aggressors in family contexts and on the impact on children, terms related to substance use (“substance-related disorders”), that indicate a clear link to the psychiatric comorbidities discussed (alcohol, drugs), and terms related to family dynamics (“parenting”, “mother-child relations”), that suggest the inclusion of family and transgenerational dynamics in the analysis.

The close links between male, female and domestic violence confirm the focus of studies on gender violence. This cluster covers exactly the epidemiological background discussed in the narrative review: distribution by gender, age, the role of childhood abuse in the perpetuation of violent behaviors and the interaction with substance abuse.

The green cluster – Epidemiology and risk factor in older age includes the following major terms: “aged”, “middle aged”, “80 and over”, “cross-sectional studies”, “prevalence”, “risk factors”, “crime victims”, “socioeconomic factors”, “sex factors”, “elder abuse”. Interpretation: it highlights the segment of domestic violence among older adults, including “elder abuse”. Keywords such as “prevalence” and “cross-sectional studies” indicate a significant body of epidemiological research. Those such as “risk factors” and “socioeconomic factors” show a connection to the social determinants discussed in the “Models of social intervention” section of the review. “Crime victims” connects to both aggressors and the bidirectional victim-aggressor analysis in severe psychiatric pathology. This cluster is more related to the part of the review on contextual factors and socioeconomic vulnerabilities, but also to the ethical challenges in the elderly (where dependency and social isolation complicate intervention).

The blue cluster – Interventions, social support and sexual abuse includes the following major terms: “adult”, “social support”, “child abuse”, “sexual”, “stress disorders”, “post-traumatic”, “spouse abuse”, “adaptation”, “psychological”, “research design”, “united states” and “sex offenses”. Interpretation: it is closest to the theoretical direction “psychiatric and social interventions” in the review. The term “social support” correlates with community interventions and the support network for aggressors. Other keywords, such as “stress disorders”, “post-traumatic adaptation” and “psychological”, are central elements in psychotherapy and recovery of aggressors with PTSD. The appearance of the keywords “spouse abuse” and “sex offenses” refers to the legal-forensic area of interventions. The presence of “research design” and “united states” suggests a large base of empirical studies (mostly from the USA) that test the effectiveness of programs. The cluster almost directly includes topics from the sections on forensic psychiatry, community treatment orders, group programs for aggressors, but also connections to evaluating the effectiveness of interventions.

Pivot nodes and inter-cluster links: “humans” (central pivot, connects all clusters), “male” and “female” (connect epidemiology – green, and psychosocial studies – blue), “social support” (blue) and “risk factors” (green) (bridge social interventions and prevalence studies), “substance-related disorders” (red) (connect family epidemiology with psychiatric interventions).

The correlation mental disorders-violence is clearly reflected in the red cluster through “substance-related disorders”, “child abuse” and “domestic violence”. Psychiatric interventions are visible in the blue cluster through “stress disorders”, “post-traumatic”, “adaptation”, “psychological” and “spouse abuse”. Social interventions are key bridges: “social support” (blue) and “socioeconomic factors” (green). Ethical and legal challenges are indirectly suggested by the presence of “crime victims”, “sex offenses”, “research design” (legal approaches and intervention evaluation). International good practices are present especially through the connections between the USA (as the main study area) and social factors; these suggest that many conclusions come from Western contexts, which aligns with the observations in the report on cultural adaptation.

Discussion

The role of social support and integrated interventions

From the analysis of the bibliometric map and the correlation with the literature, several clear observations emerge about social support and interventions for aggressors. Social support is a pivot in rehabilitation and prevention of recidivism. The term “social support” appears centrally in the blue cluster, linked to both psychiatric terms (“stress disorders”, “post-traumatic”, “adaptation”, “psychological”) and forensic terms (“spouse abuse”, “sex offenses”). In the literature, social support has a dual role: reducing external stressors (unemployment, homelessness, social isolation) that can precipitate recidivism and reinforcing behavioral change through support networks (men’s anti-violence groups, mentoring, community circles). The studies incorporated in the review show that perpetrators included in programs that combine mental health treatment with social support (housing, vocational reintegration) have a lower risk of recidivism than those who receive psychiatric intervention alone. Coordinated Community Response (CCR) models and the British Drive project confirm that the simultaneous involvement of multiple agencies (police, probation, social services, NGOs) increases victim safety and supports the process of change of perpetrators. Interventions for perpetrators should be multidimensional and integrated.

The map shows that the terms in the intervention area (“social support”, “PTSD”, “adaptation”, “psychological”, “research design”) connect with both epidemiological clusters (red and green), confirming that the interventions are grounded in clinical and contextual data.

Psychiatric interventions focus on pharmacotherapy adapted to the diagnosis (SSRIs, mood stabilizers, antipsychotics), individual and group psychotherapy (CBT, DBT, motivational interviewing), focused on reducing impulsivity, anger management and restructuring violent beliefs, and on involuntary hospitalization and community treatment orders (CTO) in high-risk cases, but the effectiveness of CTO is questionable in the absence of social support. Social interventions involve programs (Batterer Intervention Programs), modernized with mental health and addiction treatment modules, and the integration of socioeconomic support (housing, employment) to eliminate contextual risk factors. The link between social support and interventions is evident in the fact that programs with an integrated approach (medical plus social) are the ones that yield better results. The lack of social support can undermine even effective psychiatric treatment – hence the need for individualized reintegration plans.

As a general observation of the map-literature analysis, we can say that social support is not a secondary element, but a major predictor of the success of interventions on aggressors. Effective interventions are multidisciplinary, integrated and personalized, acting simultaneously on mental health, violent behavior and on socioeconomic and environmental factors.

The aggressor’s perspective – thematic gaps

If we look at the bibliometric map from the aggressor’s perspective and think about intervention, support and social impact, we notice some clear gaps – i.e., areas insufficiently covered or not at all represented in the literature indexed in PubMed based on the formula used.

The motivational and commitment dimension

Observation: the map contains terms related to diagnosis (“PTSD”, “substance-related disorders”) and intervention (“social support”, “adaptation”, “psychological”), but no concepts about the aggressor’s motivation appear, such as “readiness to change”, “treatment engagement” or “dropout prevention”. This is important because studies show that, without active commitment from the aggressor, the effectiveness of programs decreases, regardless of the quality of the intervention (mentioned in the narrative review). The social consequence is logical; the lack of focus in research on motivation leaves a key element for reducing recidivism uncovered, which means less safety for victims and higher social costs.

Addressing the aggressor’s own trauma

Observation: the map contains words like “PTSD” and “child abuse” as terms, but these are more often used in the context of victims, not as factors that shape the profile of the aggressor. There seems to be a lack of in-depth analysis of the trauma suffered by the aggressor (complex childhood trauma, war experiences, sexual abuse suffered) and how the treatment of this trauma influences violent behavior. The social consequence derives from ignoring the aggressor’s trauma in interventions that can maintain the mechanisms of transgenerational violence (the cycle of violence).

Cultural and contextual perspectives

Observation: “united states” appears, but no terms related to cultural diversity, adaptation to minorities, migrants or indigenous communities. Keywords such as “cultural adaptation”, “minority groups”, “immigrants” or “refugees” are missing. The social consequence may be related to the fact that programs developed in Western contexts may be less effective or even inadequate in other cultures, increasing the risk of relapse if interventions are not adapted.

The impact of interventions on the aggressor’s social network

Observation: “social support” is present, but rather as a general protective factor. In the analyzed context, there seems to be a lack of research on how intervention on the aggressor affects the extended family, friends, colleagues – his direct and indirect social network. The social consequence may be based on the lack of this analysis, which can hide unintended negative effects (for example, the complete isolation of the aggressor without healthy relationship alternatives).

The accountability and restorative justice component

Observation: presence of “sex offenses” and “spouse abuse” indicate the connection with criminal justice, but there are no terms about “restorative justice”, “offender accountability” or “rehabilitation outcomes”. There is a lack of clear elements that would lead to measures and indicators of behavior change from the aggressor’s perspective (not just reducing incidents). If research focuses only on recidivism as an “outcome”, the opportunity to document profound changes in attitude and real social reintegration is missed.

Technological innovation in interventions

Observation: no terms related to technology, tele­medicine, mobile applications for self-management of violent impulses appear. Thus, keywords such as “telepsychiatry”, “mobile intervention”, “digital CBT” or “online perpetrator programs” are missing. Another social consequence may be emerging; the lack of this field limits the access of aggressors in isolated areas to support and treatment, which can maintain the cycle of violence.

The map shows that the current literature documents diagnosis, epidemiology and some of the interventions well, but research on the motivation and involvement of the aggressor, the aggressor’s own trauma, cultural adaptation, effects on the social network, empowerment and restorative justice, and on technological innovations in treatment is poorly represented.

Conclusions

Femicide and domestic violence are systemic phenomena, fueled by multiple factors (gender norms, poverty, problematic alcohol/substance use, institutional gaps). The public narrative in Romania correctly reflects the scale and urgency, but the effective solutions require integrated and consistent interventions, not just one-off reactions. Addressing aggressors is part of protecting victims. Psychiatric interventions (diagnosis, pharmacotherapy, psychotherapy, forensics when appropriate) reduce individual vulnerabilities (impulsivity, anger, psychotic/affective symptoms), but they are not sufficient without social support (housing, income, professional reintegration, support networks).

Coercive measures must be used proportionately and with guarantees. Involuntary hospitalization and legal obligations (protection orders, electronic monitoring, safety measures) are justified only in the event of imminent risk, and must be complemented by accessible therapeutic and social services. Control without support increases abandonment and does not sustainably reduce recidivism. The effectiveness of interventions on perpetrators is modest but real when programs are multicomponent: they combine psychotherapy focused on abusive behavior, treatment of comorbidities (mental health, addictions) and case management. Retention (abandonment prevention) is critical; motivational components and practical adaptations (flexible hours, transportation support) increase the adherence.

The coordinated community response (CCR) is the optimal framework: probation – police – courts – social assistance – health – NGOs work in sync, with the explicit priority of victim safety (including a dedicated victim specialist in perpetrator program teams).

The Romanian context is advancing, through the expansion of SIME, the consolidation of protection orders and the development of counseling services for perpetrators; however, infrastructure gaps persist (residential centers for evicted perpetrators, uniform network of standardized programs), territorial variations and training deficits for professionals. The ethics of intervention require a clear balance: confidentiality “with safety margins”, interinstitutional communication at high risk, accountability without stigmatization. The aggressor remains responsible for the acts, but has the right to adequate treatment; investing in his rehabilitation means prevents future victimization. Rigorous monitoring and evaluation require quality standards and accreditation of programs, clear indicators (recidivism, OP violations, retention, attitudinal changes), independent audit and systematic feedback from victims. Without evaluation, the expansion of programs risks reproducing inefficiencies.

Development directions include: (a) residential/temporary housing services for evicted aggressors; (b) integration of addiction treatment into all programs; (c) motivation/commitment modules and dropout prevention; (d) cultural adaptation and interventions for underrepresented groups; (e) use of technology (telepsychiatry, self-regulation applications, crisis lines for aggressors) with safety protocols. Immediate public policy priorities are dedicated funding for CCR and risk-intensive case management, creating national standards for perpetrator programs (with a victim safety component), promoting continuous multidisciplinary training, building operational bridges between SIME – police – probation – clinical services, and creating clear referral/counter-referral mechanisms.

The sustainable reduction of violence and femicide depends on ecosystems of intervention: strong legal measures, targeted psychiatric treatment and robust social programs, all delivered in a coordinated manner and continuously evaluated, with the victim at the center of decisions and with real accountability of the perpetrator.    

Corresponding author: Radu-Mihai Dumitrescu E-mail: dum_mihu@yahoo.com

Conflict of interest: none declared.

Financial support: none declared.

This work is permanently accessible online free of charge and published under the CC-BY licence.