The current practices of intervention with batterers

A review of papers about intervention with batterers published during this period (2000-2013) was conducted. Social sciences databases were checked. Papers about programs for a specific public or programs with a broad intervention focus, and with female and homosexual offenders were excluded.

Results

Thirty-six studies that described 37 intervention programs fulfilled the inclusion criteria. In general, the analysed programs adopted a group format (70.3%) and a cognitive-behavioural (56.8%) or psychoeducational (18.9%) intervention model (32.4% assumed to adopt a Duluth model). Concerning the effectiveness, results showed success rates of 39.4%-97%, dropout rates of 10%-58% and recidivism rates of 0%-65.9%.

Discussion

The effectiveness of intervention with batterers remains controversial, which seems to be due to the different methodologies used in the studies. Despite the inconsistencies, programs for perpetrators are an important way to reduce intimate partner violence recidivism.

Intimate partner violence; batterers; intervention programs; intervention efficacy

Passive-aggressive responses*** (-)Authors	Research design	Results

Bowen²⁷ ⁽¹⁾ (Domestic Violence Perpetrator Program – DVPP)	Quasi-experimental	Group cohesion (+); Leaders support (+)
	(pre-test, post-test and 5 months follow-up)	Expressiveness: G1** (+)
	G1 = 2 hours	Anger and aggression: G1** (+)
	G2 = 2.5 hours	Order and organization: G2* (+)
		Leader control: G2* (+); Innovation: G1 (+)

Bowen et al.^{20 (1)} (Domestic Violence Perpetrator Program – DVPP)	Quasi-experimental	Dropout = 33%
	(pre-test, post-test and 11 months follow-up)	Re-offending = 21%
	G1 = completers	Demographic characteristics: G1 x G2: ns
	G2 = dropout	Psychological characteristics: G1 x G2: ns
		Criminal history characteristics: G1 x G2** (+)

Bowen et al.^{28 (1)} (Domestic Violence Perpetrator Program – DVPP)	Quasi-experimental	Re-offending = 10.8%
	(pre-test, post-test and 5 month follow-up)	Cognitive domain, arousal domain, behavioural domain, sympathy for the battered women, chance locus of control: GN x GO: ns
	GN = non-offenders
	GO = offenders

Buttell²⁹	Quasi-experimental	Moral reasoning: ns
	(pre-test and post-test)

Buttell and Carney^{30 (2)}	Quasi-experimental	Spouse-specific behaviour, controlling behaviour and propensity for abuse: ns
	(pre-test and post-test)

Buttell and Carney^{31 (3)}	Quasi-experimental	Passive-aggressive responses*** (-)
	(pre-test and post-test)	Assertiveness, controlling behaviour, propensity for abuse: ns
	G1 = African American	G1 x G2: ns
	G2 = Caucasian

Buttell and Carney^{32 (2)}	Quasi-experimental	G1 = Self-deception*** (+)
	(pre-test and post-test)	G2 = Propensity for abusiveness*** (+)
	G1 = African American	Total sample
	G2 = Caucasian	Passive-aggressive responses* (-); controlling behaviour* (-); propensity for abusiveness** (-)

Buttell and Pike³³	Quasi-experimental	Success = 72.5%
	(pre-test, post-test and 12 months follow-up)	Alcohol (-); Drug (-); Violence*** (-)
		Control (-); Stress-coping (-)

Carney and Buttell^{34 (3)}	Quasi-experimental	Dropout = 58
	(pre-test and post-test)	Interpersonal dependency: ns

Chan³⁵	Quasi-experimental	Physical assault* (-)
	(pre-test and post-test)	Psychological aggression* (-)

Cissner and Puffett³⁶	1 year follow-up	Success = 68%
(The Domestic Violence Accountability Program)		Recidivism: in-program arrest = 27%; 1 year post sentence arrest = 27%; 1 year post-program arrest = 18%
		Criminal contempt arrest = 7%
		Violent arrest = 12%

Cissner and Puffett³⁶	1 year follow-up	Success = 77%
(Interborough Developmental Consultation Centre – IDCC)		Recidivism: in program arrest = 29%; 1 year post sentence arrest = 36%;
		1 year post program arrest = 14%
		Criminal contempt arrest = 14%
		Violent arrest = 16%

Connors et al.³⁷ (The Moderate Intensity Family Violence Prevention Program)	Quasi-experimental	Interpersonal relationship (jealousy)* (+); Aggression* (+)
	(pre-test and post-test)	Violence* (-); Empathy for the victims* (+)
	G1 = Ready to change	Acceptance of responsibility*** (+)
	G2 = Not ready	Acknowledges the use of power and control*** (+)
		Recognizes cognitive distortions*** (+)
		Understands the abusive patterns***(+)
		Motivation to change* (+); Participation in program and performance* (+);
		Responses to potential risk situations* (+); Readiness to change* (+)
		G1*** x G2

Connors et al.³⁸ (The High Intensity Family Violence Prevention Program)	Quasi-experimental (pre-test and post-test)	Interpersonal relationship (jealousy)* (-); Aggression* (-); Violence*** (-)
		Acceptance of responsibility*** (+)
		Acknowledges the use of power and control*** (+)
		Empathy to the victims* (+); Recognizes cognitive distortions* (+); Understands the abusive patterns* (+); Motivation to change* (+); Responses to potential risk situations (interpretation) (+); Readiness to change* (+)

Echeburúa and Fernández-Montalvo^{39 (4)}	Quasi-experimental	Success: 68.2%
	Pre-test	Cognitive bias about women (-); Cognitive bias about use of violence (-);
	Post-test	Empathy: ns; Trait-anger: ns; Anger expression (-); Impulsivity (-); Self-esteem (+); Symptomatology* (-)

Echeburúa et al.^{40 (4)}	Quasi-experimental	Success = 92%
	(pre-test and post-test)	Cognitive bias about women (-); Cognitive bias about use of violence (-);
	G1 = Completers	Empathy: ns; Symptomatology* (-); Hostility* (-); State-anger* (-); Impulsivity: ns;
	G2 = Dropout	Self-esteem: ns; Inadaptation: ns; Motivation to change (G1* (+) x G2)

Echeburúa et al.⁴¹	Experimental (pre-test, post-test and 1, 3, 6 and 12 months follow-up)	Success: 87.96% Cognitive bias about women* (-); Cognitive bias about the use of violence* (-); Empathy* (+); Trait-anxiety* (-); Trait-anger* (-); Depression* (-); Self-esteem* (+); Inadaptation* (-)

Authors	Research design	Results

Feder and Dugan⁴² (The Broward experiment)	Experimental	Success = 97%
	(pre-test and post-test)	Self-reports of abuse: ns; Victim reports of abuse: ns
	G1 = Experimental group	Recidivism (official reports): ns
	G2 = Control	Beliefs about wife beating, attitudes toward women, treats domestic violence as a crime, responsibility, likelihood for repeated abuse: ns

Lawson et al.⁴³	Quasi-experimental	Secure attachment* (+); Anxiety and avoidance attachment: ns;
	(pre-test and post-test)	Comfort with closeness and depending of others = G1** (+) x G2 x G3
	G1 = Secure-changed	Avoidance and closeness with others = G1 x G2* (+) x G3* (+); Depression and anxiety* (-) G1* (-) x G2 x G3
	G2 = Secure-unchanged	Functioning*** (+) G1 x G2 x G3: ns
	G3 = Insecure	Partner violence* (-) G1 x G2 x G3: ns

Lila et al.⁴⁴	Pre-test and post-test	Perceived severity* (+)
		Responsibility assumption* (+)
		Risk of recidivism* (-)

Lin et al.⁴⁵	Quasi-experimental	Overall violence: GE* (58.6%) x GC (41.4%) (3 months follow-up)
	(3, 6 and 9 months follow-up)	Physical violence: GE (20%) x GC (13.4%): ns
	GC = Control	Verbal violence: GE* (58.6%) x GC (43.7%) (3 months follow-up)

Novo et al.²⁶ (Galicia Reeducation Program for Gender Violence Offenders)	Quasi-experimental	Psychological symptoms*** (-)
	(pre-test and post-test)	Obsessive-compulsion** (-)
		Interpersonal sensitivity*** (-)
		Depression* (-); Anxiety (-)
		Hostility*** (-)
		Phobic anxiety** (-)
		Paranoid ideation* (-); Psychoticism* (-)


Pascual-Leone et al.²⁵ (Relating Without Violence)	Quasi-experimental	Dropout = 10%
	(7, 8 and 9 month, 1, 2 and 3 years follow-up)	Recidivism (3 years follow-up) = 15.6%
	GE = Intervention	GE* (-) x GC (7 and 8 months follow-up)
	GC = Control	GE x GC: ns (9 months, 1, 2 and 3 years follow-up)

Ramírez et al.⁴⁶	Quasi-experimental	Aggression: ns (hostility** (-) (GE x GC: ns)
	(pre-test and post-test)	Benevolent sexism*** (-) (GE* x GC); Hostile sexism* (-) (GE x GC)
	GE = Treatment	Jealousy* (-) (GE x GC: ns); Marital satisfaction* (+) (GE* x GC)
	GC = Control	Conflict tactics** (-) (GE x GC: ns)
		Responsibility* (+) (GE* x GC); Responsibility minimization*** (-) (GE x GC: ns)
		Emotional abuse** (-) (GE x GC: ns)
		Personality (NEO-PI): ns (openness to experience (+); kindness(+) (GE x GC: ns)
		Personality (I-7):Impulsiveness**(-) (GE x GC); empathy(+) (GE x GC: ns); venturesomeness(-) (GE x GC: ns)
		Anger (anger trait* (-); external anger expression*** (-); internal anger expression* (-); internal anger control* (+); anger expression and control* (-) (GE x GC: ns)

Rosenbaum et al.⁴⁷ (Mens’ Educational Workshop – MEW)	Quasi-experimental	Success
	(post-test and 20 months follow-up)	G1 (56.5%) x G2*** (79.5%)
	G1 = Self-referred	G1 (39.4%) x G2* (75.2%) (20 sessions)
	G2 = Court-referred	Recidivism
	G3 = Completers	G1 (13.4%) x G2** (4.9%)
	G4 = Non-completers	G3* (-) x G4
	7, 10 e 20 sessions	10 and 20 sessions = > recidivism

Sarto and Esteban⁴⁸	Post-facto	Dropout = 55.1%
	(3, 6 and 12 months follow-up)

Schmidt et al.⁴⁹	Quasi-experimental	Success = 52%
	(pre-test and post-test)	Statements that support abusive behaviour (-); Statements that support a nonviolent relationship (+); Motivating factors for behaviour change** (+)

Stith et al.⁵⁰ (Domestic Violence Focused Couples Treatment)	Quasi-experimental	Success
	(pre-test and 6 month and 2 years follow-up)	T1 (70%) x T2 (73%): ns
	T1 = Multi-couple group	Beliefs about wife beating: T1(-) x T2 x T3 (M(-) x W)
	T2 = Individual couple	Marital satisfaction: T1** (+) x T2 x T3
	T3 = Comparison	Acceptance of marital aggression: T1*** (-) x T2 x T3
		Recidivism
		6 months: T1(25%)* x T2(45%) x T3(67%)
		2 years: T2*(0%) x T3 (50%); T1⁺(13%) x T3; T1 x T2⁺

Taylor et al.⁵¹ (Victim Services’ Alternatives to Violence) (The Brooklyn Experiment)	Experimental	Recidivism
	(6 and 12 months follow-up)	Officially recorded incidents: GE** (-) x GC
	GE = Intervention	Victim-reported incidents: GE (-) x GC: ns
	GC = Control	Treatment = < recidivism
Taylor and Maxwell⁵² (Sacramento Jail Batterer Intervention Program)	Randomized experimental	Batterers
	(6 and 12 months follow-up)	Psychological abuse: GE x GC: ns;
	GE = Treatment	Threats of physical abuse: GE x GC: ns;
	GC = Control	Physical abuse: GE x GC: ns;
		Controlling behaviours: GE** (-) x GC
		Victims: GE x GC: ns
		Oficial data: GE (65.9%) x GC (65.7%): ns

Authors	Research design	Results
Alexander⁵³ (Stages of change – SOC; Cognitive-behavioural gender-re-education format – CBTGR)	Quasi-experimental	Physical and psychological aggression (post-test): ns
	(pre-test, post-test and 6 and 12 months follow-up)	Physical aggression (follow-up) (-) (G1* x G2)
	G1 = SOC	Psychological aggression (follow-up) (-) (G1 x G2: ns)
	G2 = CBTGR	Assumption of responsibility, attendance, treatment completion, working alliance, group cohesion: ns

Dunford⁵⁴	Randomized experimental (post-test and 1 year follow-up)	Success = 71%
(San Diego Navy Experiment)	T1 = Men’s group	Frequency of victimization: T1 x T2 x T3 x T4: ns
	T2 – Conjoint group	Official data: T1 x T2 x T3 x T4: ns
	T3 = Rigorous monitoring group	Prevalence and frequency of violence* (-)
	T4 = Control group

Hendricks et al.²⁴ (Stopping Abuse for Everyone – SAFE; Reasoning and Rehabilitation – R&R)	Quasi-experimental (6, 12 and 18 months follow-up)	Recidivism = 17.5%
	G1 = SAFE	(6-12 months follow-up = 23%; 12-18 months follow-up = 17%); G1* (14.4%) x G2 (32.4%); G3*** (10.6%) x G4 (38.8%)
	G2 = R&R + SAFE	Risk: G3 * (-) x G4; G1 x G2*** (+)
	G3 = Completed SAFE	Discharged score: G1 x G2** (+); G3 x G4* (+)
	G4 = Incomplete SAFE


Morrell et al.⁵⁵	Quasi-experimental (pre-test, post-test and 6 month follow-up)	Success = 86%
	T1 = Cognitive-behavioural group therapy	Recidivism: ns
	T2 = Supportive group therapy	Physical assault*(-); psychological aggression(-), injuries**(-), sexual coercion(-): T1 x T2: ns
		Readiness to change: ns (T1 x T2: ns)
		Self-esteem** (+) (T1 x T2: ns)
		Self-efficacy*** (+) (T1 x T2*)

Murphy et al.⁵⁶	Quasi-experimental (pre-test and post-test)	Self-esteem: T1* (+) x T2*** (+)
	T1 = Cognitive-behavioural	Physical aggression: T1* (-) x T2** (-)
	T2 = Compassion workshop

Scott et al.⁵⁷	Quasi-experimental (pre-test and post-test)	Dropout
	T1 = Standard program	T1 = 53.5% (46.5% completed the treatment); T2 = 15.8% (84.2% completed the treatment)***
	T2 = MET (motivational enhancing treatment) + standard program	Accountability: T1 = 2.34; T2 = 2.54
		Active engagement: T1 = 2.78; T2 = 2.72

Faculdade de Medicina da Universidade de São Paulo Rua Ovídio Pires de Campos, 785 , 05403-010 São Paulo SP Brasil, Tel./Fax: +55 11 2661-8011 - São Paulo - SP - Brazil
E-mail: archives@usp.br

Acompanhe os números deste periódico no seu leitor de RSS

[1] Address correspondence to: Olga Cunha. School of Psychology, University of Minho, Campus de Gualtar – 4710-057 – Braga, Portugal. E-mail: olgacunha27@hotmail.com