From subtle symptoms to disclosure: Communication in violence detection in primary care Authors: Orlane Pivert, Clémence Loubrieu, Julie Magret, Nicolas Naiditch, Stéphanie Mignot Published: 2026 ABSTRACT AbstractBackgroundIdentifying victims of interpersonal violence remains a challenge for general practitioners, who must rely on subtle clinical cues and foster an environment conducive to disclosure.ObjectivesTo describe the symptoms, behaviours, and communication strategies that facilitate the recognition of interpersonal violence in primary care consultations.MethodsFor this qualitative study, general practice consultations by three physicians in the Poitou-Charentes department, France, were audio-recorded from June to September 2019. Adult patients (≥18 years, fluent in French) were included unless they rejected audio recording. Only the audio-recording of consultations during which interpersonal violence was disclosed were analysed following an inductive approach to identify categories and themes until saturation was reached.ResultsIn total 25/326 patients disclosed experiences of violence. Detection combined verbal cues (vague somatic complaints, ambiguous trauma references) by the patient and open-ended or direct questions by the general practitioners. Techniques included linking symptoms to violence, broad funnel questions, and most effectively, direct queries, such as ‘Did someone hurt you?. The three physicians adapted their approach, balancing respect for the patient readiness and clinical conviction, often revisiting the topic from different perspectives. Empathy and a trusted doctor-patient relationship proved crucial for violence disclosure.ConclusionOur findings highlight the value of the physicians’ tenacity and empathic attitude for identifying hidden situations of violence, and suggest that integrating these strategies in routine general practice can enhance detection. The findings underscore the need of training physicians in patient-centred communication and suggest future research avenues for refining detection protocols in primary care. Background Violence is defined as the intentional use of physical force or power, threatened or actual, against oneself, another person or a group or community that results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment or deprivation [1]. It is commonly categorised into self-directed, interpersonal, and collective violence. Interpersonal violence is divided into two categories: family and intimate partner violence (IPV), and community violence. Violent acts can be of a physical, sexual and/or psychological (involving deprivation or neglect) nature [1]. Interpersonal violence represents a major global public health issue. The World Health Organisation (WHO) estimates that nearly one in three women worldwide experiences physical or sexual violence by an intimate partner or non-partner in her lifetime [2]. According to UNICEF, one in four children under five lives with a mother who is a IPV victim and 15 million adolescent girls (15–19 years of age) have experienced forced sex [3]. Besides the immediate trauma, the impact of violence extends across the lifespan and affects both mental and physical health. Survivors present an increased risk of depression, anxiety, post-traumatic stress disorder, suicidal behaviours and substance use, as well as somatic conditions such as chronic pelvic pain, fibromyalgia, and gastrointestinal disorders [4–6]. Recognising violence victims in primary care is critical, but remains challenging. Validated screening tools exist, such as the Woman Abuse Screening Tool (WAST), the Hurts, Insult, Threaten and Scream (HITS) and WAST-short tests [7,8], yet their direct application in general practice is often unrealistic. In agreement, a 2021 systematic review on IPV screening programmes found high screening reach (median = 82%), but substantial variability in positivity and referral rates. This emphasises that screening efficacy depends also on institutional support, training and referral infrastructure and not on tools alone [9]. As patients rarely disclose violence explicitly, presenting instead with somatic or psychological complaints, it is important to explore how signs of interpersonal violence can emerge naturally during a routine medical consultation. Therefore, the aim of this study was to identify the symptoms, behaviours, and attitudes expressed by patients in general practice consultations that may suggest medium- or long-term consequences of interpersonal violence. Methods Study design and participants This qualitative study was conducted in general practice settings in the Poitou-Charentes department (Nouvelle-Aquitaine region), France. Three female physicians participated: one established GP in an urban practice in Poitiers, and two locum doctors, at the beginning of their career, in semi-rural or urban group practices. The aim was to identify patients’ symptoms, signs and behaviours during medical consultations that could suggest medium- or long-term consequences of past interpersonal violence. Participants were adult (≥18-year-old) patients who consulted their GP. Inclusion criteria were: sufficient proficiency in French to understand the study and to give their oral consent for audio recording their routine consultation. Consultation audio-recording was carried out for half or one day on several days. Data collection Routine general practice consultations were audio recorded using a digital recorder placed in the physician’s office. Consultations were conducted as usual, without any modification in the approach or management. Patients were invited to speak freely, and the detection of potential interpersonal violence was based on the physician’s usual clinical judgement, in line with their professional practice. At the end of each consultation, each physician completed a short questionnaire in which they reported the patient demographic information, socio-professional status, context of the consultation, and immediate impressions, including the patient’s attitudes and behaviours. Consultations without mention of interpersonal violence were deleted immediately after the post-consultation questionnaire and not transcribed. Sampling continued until thematic saturation was reached. This was defined as the absence of new violence-related symptoms, behavioural patterns, or general practitioners’ communication strategies across consecutive consultations. Two investigators (initials) independently reviewed the collected transcripts weekly using the RQDA software application (add a link or reference). Data collection took place between June and September 2019 and analyses were completed in November 2019. Data analysis The audio-recordings of violence-disclosing consultations were anonymised and transcribed verbatim at the end of each week during the data collection period. Their analysis was carried out following an inductive approach and in two steps:Thematic coding of the transcripts (coding and segmentation) using the RQDA software. Data were independently coded by two investigators (initials). Then, codes were discussed with a third investigator (initial) to ensure internal triangulation. Regular meetings were held to discuss the analytical process, compare findings, and reach a consensus on recurring themes.Open coding of the transcripts revealed the key concepts emerging from the data that were then grouped into subcategories through axial coding. Lastly, selective codes were formed by prioritising the axial codes into overarching categories. Thematic coding of the transcripts (coding and segmentation) using the RQDA software. Data were independently coded by two investigators (initials). Then, codes were discussed with a third investigator (initial) to ensure internal triangulation. Regular meetings were held to discuss the analytical process, compare findings, and reach a consensus on recurring themes. Thematic coding of the transcripts (coding and segmentation) using the RQDA software. Data were independently coded by two investigators (initials). Then, codes were discussed with a third investigator (initial) to ensure internal triangulation. Regular meetings were held to discuss the analytical process, compare findings, and reach a consensus on recurring themes. Open coding of the transcripts revealed the key concepts emerging from the data that were then grouped into subcategories through axial coding. Lastly, selective codes were formed by prioritising the axial codes into overarching categories. Open coding of the transcripts revealed the key concepts emerging from the data that were then grouped into subcategories through axial coding. Lastly, selective codes were formed by prioritising the axial codes into overarching categories. The objective was to detect recurrent signs, symptoms and consultation patterns associated with past experiences of interpersonal violence. This real-time analytical technique allowed testing and confirming the elicitation method effectiveness, assessing its reproducibility, and enriching the overall findings [10]. Ethics According to the French Jardé law on research involving human beings of 16 November 2016, a notification was first sent the Committee for the Protection of Persons on non-interventional type 3 research. Ethics approval was obtained on 7 June, 2019 (number 19.02.04.56426). Results Sample A total of 326 patients accepted the audio recording of their consultation and seven refused: two felt that the topic was not relevant to them, one refused to be questioned on the subject, and four did not provide any reason. During the consultation, violence was mentioned by 25/326 patients (19 women and 6 men; ∼7.7% of consultations). The age of these 25 patients ranged from 18 to 80 years and the mean consultation time was 33.23 min. Almost all socio-professional categories were represented, except skilled trades workers, small business owners and farmers (Table 1). All violence types were identified (Table 2). Women reported experiencing violence more frequently than men, particularly IPV, and were more likely to have experienced multiple forms of violence throughout their lives. Table 1.Characteristics of the patients and consultations included in the qualitative analysis (n = 25).Physician (M) patient (P)GenderPatient age category (years)Patient statusWork typeConsultation duration (minutes)M1P1Woman25–45KDOther13M1P2Woman18–25KPWhite collar worker42M1P3Man25–45KDBlue collar worker46M1P4Woman45–65NPIntermediate profession35M1P5Woman18–25KPStudent34M1P6Woman45–65KDIntermediate profession32M1P7Woman25–45KPHigher profession21M1P8Woman45–65KDOther42M1P9Man25–45KPStudent40M1P10Woman25–45KPIntermediate profession35M1P11Woman45–65KDWhite collar worker32M1P12Woman18–25KDWhite collar worker35M2P1Woman45–65KPOther22M2P2Woman25–45KPWhite collar worker31M2P3Man18–25KPOther35M2P4Woman45–65KPBlue collar worker21M2P5Man45–65KPBlue collar worker22M2P6Man45–65KPOther35M2P7Woman65–80KPOther22M2P8Woman45–65KPWhite collar worker27M2P9Woman45–65KPWhite collar worker33M2P10Man45–65KPHigher profession39M2P11Woman45–65KPIntermediate profession24M3P1Woman25–45NPHigher profession63M3P2Woman18–25NPOther49KD: patient known by the doctor; NP: new patient; KP: patient known by the practice but not by the doctor. Characteristics of the patients and consultations included in the qualitative analysis (n = 25). KD: patient known by the doctor; NP: new patient; KP: patient known by the practice but not by the doctor. Table 2.Types of interpersonal violence disclosed by patients during the consultations (n = 25).Violence typeSexual (adult)Sexual (childhood)Physical (partner)Sexual (partner)Psychological (partner)Violence in schoolWork place harassmentChild abuseFamily psychological (family)TotalM1P1 1 1M1P21 1M1P3 1 1M1P4 1111 4M1P5 1 1M1P6 1 1M1P7 1 1M1P8 1 1 2M1P9 1 1M1P10 1 1 2M1P11 1 1 2M1P12 11 2M2P1 1 1 2M2P2 1 1 2M2P3 1 1 2M2P4 1 1M2P5 1 1M2P6 1 113M2P7 1 1M2P8 1 12M2P9 1 1 2M2P10 1 1M2P111 1M3P1 1 1M3P2 1 113Total285184283 M = physician; P = patient. Types of interpersonal violence disclosed by patients during the consultations (n = 25). M = physician; P = patient. Symptoms Somatic manifestations Such symptoms were typically vague, chronic, treatment-resistant, prompting repeated consultations. Pain, chronic fatigue, cardiorespiratory discomfort, headache and digestive problems were commonly reported (Table 3). Such complaints were rarely explained by a clear biomedical cause and were perceived by physicians as possible bodily expressions of psychological distress. Table 3.Illustrative quotations on symptoms organised by themes and subthemes.ThemesSubthemesIllustrative quotations (M = physician, P = patient)Somatic manifestationsGeneral symptoms‘I’m very, very tired.’ P2 to M1‘And the hot flushes, they’re awful!’ P11 to M1‘And then I fainted. Here is the emergency room report.’ P2 to M2Chronic pain‘I have fibromyalgia.’ P7 to M1‘Well, yes, I also have pain here… when I move, um, no, I noticed that when I moved, my fingers moved slowly. You know, like I had a cramp, you know, it feels like a cramp. Anyway, my hand doesn’t bother me. What bothers me the most is when my shoulders and neck hurt. They hurt so much that I had to hold my neck to get up from the sofa.’ P8 to M2Sleep problems‘Already there were nightmares every night…’ P11 to M2‘I don’t sleep at night. Even with ZOPICLONE. It’s been going on for years.’ P4 to M2Eating and weight problems‘I eat compulsively and… I’ve gained 5 kilos in a month… I don’t know how to lose them, and it really bothers me…’ P4 to M2‘Because you’ve gained weight?’ ‘No, I’ve lost weight.’ ‘Well then!’ ‘But I don’t want to, I don’t like it. I weigh myself every day.’ ‘You weigh yourself every day!’ ‘Yes, I have a disorder, I have a disorder. Because of that. Well, I used to weigh 140 kilos, so…’ ‘Oh, right! So, you’ve lost weight then.’ ‘I lost 70 kilos and then I gained back, um…’ M1P9Specific organ problems‘And then there are the headaches, the migraines, it’s true that I often get them.’ P11 to M1‘It’s still warm. My heart feels like…’ ‘Does your heart feel like it’s beating fast?’ ‘(Silence). I can feel it.’ M1P4‘No, I can’t swallow anything because I feel nauseous.’ P7 to M1‘Well, I really have to say this… I’m constipated for life, eh!’ P9 to M1‘To urinate. It’s… 6 to 7 times a day, sometimes much more.’ P2 to M1‘And just before that, well, my ex, I had an operation… I mean, I had an abortion.’ P12 to M1‘Do you want to have more periods or fewer periods?’ ‘Less pain! (Laughter)’ M3P2‘And I’ve had a lump in my throat too. For too long. It’s been almost a year since I’ve had this lump in my throat. (…) I feel like I have tonsils although I had them removed (…) I’m afraid I have throat cancer.’ P5M1‘I had lots of little spots.’ P11 to M1Body or medical beliefs‘I was eating, and I felt myself drifting away. I felt a contraction in my brain.’ P2 to M2‘There are also my tubal clips. I have one that moves around and hurts me, but the last time I went to see a gynaecologist, he told me that operations are expensive… Sometimes I come across strange doctors…’ P8 to M2Psychological and emotional symptomsSelf-depreciation‘There’s that, and then there’s the fact that I don’t have confidence in myself. That’s something I’ve known for a long time.’ P2 to M3‘And then I’m not brave enough… brave enough for that…’ P10 to M2Dissociative symptoms‘Yes, but my brain is selective. When someone hurts me, after a while, I don’t remember it anymore…well, I realised that I don’t remember anything from early childhood up to the age of 12. It’s selective. My sister said to me, ‘Hey, we did this and that,’ but I don’t remember it… (…) Apparently, my brain just deleted it too. I realised that everything that was negative for me, well, it just fades away, I don’t remember it. If someone reminds me, yes, maybe, but otherwise… no.’ P8 to M2‘Whenever I had an image, something that happened to me, I would do everything I could to block it out… so that’s how I started to protect myself.’ P11 to M2‘My image of myself, I don’t know how to explain it. I mean, deep down, psychologically, I feel more like myself now.’ P9 to M1Emotional dysregulation‘I’m in a very bad mood. I’m going to have a lot of mood swings… (…) I will cry over nothing and, um… be very tense…’ P2 to M1‘Why are you crying?’ ‘Because I think too much.’ M3P2Anxiety-depressive symptoms‘After that, I made an appointment at the psychiatric hospital for severe depression and everything.’ P3 to M2‘I attempted suicide…’ P6 to M2Behavioural and coping responsesAddictions‘Actually, I’ve started again… I’m smoking again, but less. I smoke four a day, and they’re not regular cigarettes. They’re cigarettes with a cigarillo wrapper. They’re somewhere in between, really.’ P9 to M1‘I’m currently weaning myself off medications, so, um…’ P7 to M1Risky behaviours‘Well, except there. I’m expecting them, which is why I took a pregnancy test this morning, and I’m not pregnant, but I did have unprotected sex, without, um… Well, it’s not protected, but it’s not super dangerous either, but it’s not protected, so it’s still a little dangerous.’ P5 to M1‘I was shoplifting (…) Otherwise, I want to go to prison, but the solicitor doesn’t agree, so it’s been postponed’ P4 to M1‘Why haven’t you seen your GP since 2018? You were still on MIRTAZAPINE, LYSANXIA, STRESAM, BISOPROLOL, XARELTO, RAMIPRIL…’ The patient clears his throat. ‘How did you manage to renew your treatment?’ ‘Um… oh… it’s been a while…’ ‘You’re not taking it anymore?’ ‘No, no, no, I…’ M2P5Saviour role‘In any case, there’s this protective instinct towards others that makes it… Out of the question for that to happen. And then, likewise, disrespect makes me feel that way (…) But on the other hand, I have a terrible intolerance for injustice and aggression. If I see it in a car park, it has happened to me before, seeing a father slap a child, much too violently. Well, I intervened.’ P1 to M3‘The solicitor told me stories, there are hundreds of them, it’s outrageous. She said to me: Yeah, you don’t have a superhero cape, it’s up to us to do that.’ P3 to M1Relational and social difficultiesLoneliness‘It’s just that I’ve always taken care of everything on my own. I’ve always taken care of everything on my own and there are times when it’s difficult. (Sobs)’ P6 to M1‘And feeling lonely above all else’ P2 to M1Distrust‘I saw specialists (…) who found nothing.’ P10 to M2‘But I’d rather not take too many strange pills for my health.’ P1 to M2‘And I know that behind the scenes there are, um… there are little comments, little criticisms and, um… to hurt the person and…’ P2 to M1Need of recognition, resilience‘I went to physiotherapy this morning and, well, she found osteoarthritis. When she touched my fingers, they cracked. She said, ‘Oh dear,’ and I said, ‘Yes.’ Then I went to see the rheumatologist and just below my foot I have symptoms of rheumatism. But is that what’s causing it to go that far? My ankle hurts, my calf hurts, it hurts there… so is it muscular?’ P1 to M2Intrafamilial suffering‘Because I’m in the middle of a divorce’ P10 to M2‘No one in my family believes me except my daughter.’ P2 to M2Workplace difficulties‘I’m not thriving at work…’ P10 to M2‘So, I, um, I couldn’t go to work, so I’m also here because of a sick leave.’ P2 to M1‘My boss wants me to see a psychologist because I put a lot of effort into my work, and she thinks it’s too much. She told me… (emotions).’ P2 to M2Disability, unfitness‘I receive disability allowance.’ P4 to M1 Illustrative quotations on symptoms organised by themes and subthemes. Gynaecological and urological symptoms also emerged across narratives (Table 3). Sleep disturbances, appetite changes, and unexplained weight variations were described as part of an overall malaise rather than a single condition. Psychological and emotional symptoms Patients frequently reported emotional lability, anxiety, and inner tension, sometimes with episodes of irritability or anger (Table 3). The three physicians identified several consultations where depressive symptoms or anxiety were evident, but remained undeclared by the patient. Sometimes, patients revealed intrusive memories, flashbacks, or avoidance behaviours consistent with post-traumatic stress disorder. Behavioural and coping responses Some patients reported resorting to drugs, alcohol, tobacco, or other substances as a means of coping with stress (Table 3). The three physicians also identified risk-taking behaviours, including treatment non-adherence and impulsive actions, interpreted as indirect expressions of suffering. Other patients described a strong sense of responsibility towards others, positioning themselves as caregivers or protectors, sometimes to the detriment of their own health (Table 3). Relational and social difficulties Patients described difficulties in trusting others, feeling misunderstood, or avoiding discussions about personal issues. Experiences of divorce, family conflicts, or professional burnout were common. Some linked their current suffering to unresolved tensions at work or in intimate relationships (Table 3). Physicians’ communication approach The three physicians employed a three-step communication approach: (1) creating a safe space through empathic posture, (2) progressive questioning, and (3) persevering despite the patient’s initial hesitations. Tables 4 and 5 present illustrative quotations and Table 6 summarises these techniques. Table 4.Illustrative quotations for the empathic posture, organised by themes and subthemes.ThemesSubthemesIllustrative quotations (M = physician, P = patient)Giving timeLetting talk, listening‘Tell me…’ ‘I’ve come to see you, I’m starting a training course um… on the 3rd, but I got stuck, it’s starting to get better but at the same time, as a result, um… I’m going to ask you something else, so I got stuck all the way to the middle of my back, so I massaged, massaged, massaged, I put um…, as I have Widal’s syndrome, I didn’t take anything, they think it’s…. I have intolerances to things, so I avoid anything… chemical, unless I’m really in too much pain… but I’m just starting a hydrotherapy course… and I wanted to see if everything was working well… Ah if it’s a bit sticky here (pointing to the cervical spinal region), it’s a bit sticky here… there…’‘Ah yes.’ ‘I think it’s a bad position, it happened to me a long time ago, the doctor you’re replacing did an osteopathic manipulation on me, he put back a cervical vertebra, one but not the second I think… er I found out from the physiotherapist and then my mother that there was cervical osteoarthritis in the family…’‘OK’ ‘I had an accident at work, in fact we weren’t sitting well, so I managed to relax as much as possible, so I do stretches at home like that… so it goes back to 2015… but my right knee hit the floor because the little girl came into my legs, I tried to avoid her and poof I fell on my right knee… it hurts today I don’t know if it’s the weather I’m not sure… I’m not sure if it’s the weather… I’m not sure…’ M2P9Respect the silence‘Did someone hurt you?’ ‘Yes, there was violence, a lot of violence, I was hurt a lot… my ex-partner… me…(Silence) He used to hit me… (Silence) M2P7‘OK. Have you had any difficult situations before?’ ‘Hmm….’ ‘You say he was nice, as if you’d been in situations where there were bad guys…’ ‘Um… (Silence). Nothing serious’ M1P5‘OK… And what was it? Well, unless you don’t want to talk about it, but…’ ‘Yeah, I don’t want to talk about it, there was the court case two years ago, it was tough.’ M1P3Respect the timing, not force‘(Silence. Crying.) It wasn’t planned to talk about all this… (Laughs)’ ‘We don’t have to talk about it if you don’t want to.’ ‘(Laughs) It’s all very strange to me.’ ‘We don’t have to go on if you don’t want to’ ‘…well, I had a rather complicated childhood, we’re four children… my mum left when I was eighteen months old… it was very complicated, my dad drank a lot… when I was 13, I was put in a convent in Angouleme because he hit me, so there you go…’ M2P4Creating a space to speak outManaging third parties: getting them out of the room‘You were prescribed TANGANIL. (Silence) Is it possible to continue the consultation alone with you and for your daughter to go out?’ ‘No problem.’ (GIRL EXITS. Proceed to clinical examination.) ‘I’ll let you get settled in. I think we’ll be more comfortable. How long has this been going on?’ M2P2Active listening and empathyDemonstrate that you are listening‘Are you worried about anything?’ ‘Er… Well, I can see that you’re looking at the time, so I’ll go to my…’ ‘(looks towards the computer): Oh no, I didn’t look at the time, I was looking here!’ ‘Oh, I thought maybe… well, I don’t want to hold you up…’ ‘I’m listening, tell me.’ M1P5‘Tell me…’ M2P9‘Excuse me, I’m going to make you late!’ ‘Please do.’ M1P6Facilitating phrases or words‘Because?’ ‘It was his behaviour that meant that sometimes it wasn’t great?’ ‘Really? Is he hard?’ ‘He’s unstable in his head.’ ‘Hmm.’ M1P11‘OK, so?’ M1‘Yes?’ M2‘Go on then.’ M1‘Yeah?’ M3Reusing the same words, repeating sentences‘That’s all. But my body needed it. Otherwise, I don’t really like alcohol and its effects. It doesn’t make any sense at all.’ ‘You don’t think it makes sense?’ M2P3‘It’s the pressure.’ ‘Did he put pressure on you?’ M2P5‘Well, there were little things… serious things, you might say. Then, four years ago, my mother divorced So, I applied to …. It’s not a dismissal… it’s…’ ‘To get her name back…’ ‘And that’s it… revocation of the adoption!’ ‘What happened that was so serious?’ ‘Erm… there was abuse for two years when I was a child.’ M1P3Recognising, renaming, reformulating‘[…] You get an emotion here, it’s clear that something is affecting you…’ ‘Well, there’s… a man who’s no longer part of my life, since my aunt died, but who I’m very, very afraid of.’ M1P5‘What you went through was a form of violence.’ ‘Yes.’ M2P10‘Well, he managed to put me down and then, managed to… to get the upper hand.’ ‘Hmm. He dominated.’ ‘He dominated. It was very hard to get rid of him.’ M1P11‘Well, yes. But that’s no reason to beat you.’ M1Encouraging, reassuring‘Drug addict, alcoholic, er…, the whole shebang. And I stopped everything.’ ‘You stopped everything? Well done. That’s great.’ M2P1‘Yeah. It’s normal, it takes time, so it doesn’t just happen. It’s usually three or six months, so it takes a bit of time to get back on track. You’re getting back on track, but it takes time.’ M1 to P10 Illustrative quotations for the empathic posture, organised by themes and subthemes. Table 5.Illustrative quotations on how to talk about violence, organised by themes and subthemes.ThemesSubthemesIllustrative quotations (M = physician, P = patient)Taking the opportunityCall phrase‘Where? Nothing is noted in your file. What are your health problems?’ ‘We don’t know, it’s all in your head.’ ‘What do you mean?’ ‘I don’t know… since we can’t find it, it must be in your head…’ ‘And why do you say it’s in your head?’ M2P10‘We’ll say that with my friend it’s a bit complicated but…’ ‘Why is it complicated?’ M1P11‘And as I’m easy to manipulate, well, that puts me straight into addiction.’ ‘Is it because you’ve already been manipulated?’ M2P3Evocative context‘He did everything to me, so I don’t think so. I told him, I told him I don’t know how I stand with you. I know what he’s like!’ ‘Because he’s difficult?’ M1P11‘I’ve got my mother who’s being treated in a psychiatric unit, she collapsed, she’d consumed alcohol and she’s on a treatment that they make her take every day… there are times when she’s a bit absent… she took Lexomil at a rate of more than 3 a day for more than 15 years, I think that’s it.’ ‘Well, yes, that may have something to do with it…’ ‘Yes.’ ‘After your mum, if she has problems with alcohol and all that, I imagine she’s had some difficult life events?’ ‘Yes… and traumas too, which meant that…’ ‘OK’ ‘Assaults and all that, which weren’t her fault, but then I was brought up by my grandmother. We discussed it and… (hesitation)’ ‘And you too experienced aggression from your mother?’ ‘Er… My father.’ M1P1‘That’s what I was coming to talk about because it’s not going well at work.’ ‘Yeah’ ‘Both physically and mentally’ ‘OK, what’s going on?’ M1P2Linking symptoms to violenceLinking symptoms with a mental disorder‘That’s it! So, anxiety is a reaction to something around you. Is that OK?’ ‘It’s linked to a situation, isn’t it?’ ‘Exactly.’ M3P2‘Yeah, you probably have digestive sensitivity. Is it linked to… is it linked, for example, to moments of stress, things like that?’ ‘Ah, possibly.’ M1P12Find out when the symptom started‘And have you had this anxiety for a long time?’ P2: ‘Erm… Since the beginning of secondary school.’ ‘Since the beginning of secondary school? What happened?’ ‘Erm… (long hesitation), I had a lot of problems with my parents. After that, I think something happened that affected me enormously. It turned me against myself.’ M1P2‘How long have you been suffering from vertigo?’ ‘It’s been a while; I started when I was 46 and now I’m 67…’ ‘Did anything happen at the time?’ ‘Yes.’ ‘Did someone hurt you?’ ‘Yes, there was violence, a lot of violence, I was hurt a lot… my ex-partner…’ M2P7Ask whether the current situation is reminiscent of another‘Did it remind you about other things, something you’d experienced before as a child or things like that?’ ‘Yes.’ M1P8Ask whether the patient has already received any psychological help‘Who’s following you?’ ‘Well, I only had one appointment last Friday, with Mr P. (…)’ ‘And um… why are you going?’ ‘Um… because I need to talk to him because he’s a soldier (refers to her partner), and because things are complicated with my family.’ ‘Yeah?’ ‘I don’t know how to tell you; I’ve been through a lot and I need to talk about it too.’ M3P2‘Were you able to work on it, did you get help?’ ‘No.’ ‘Not really…’ ‘No.’ ‘Do you ever think about it?’ ‘But it’s… not anymore, I had a very serious sexual trauma when I was a young woman, at 20, but I knew, uh, I found the resources…’ M2P11Talking about a general situation, or about another patient to bring up the subject ‘And one of the things that’s trendy right now is violence against women. And uh… I think it would be better to also tell women that when you have stomach pain and pain during your period, the reason may be that you’re experiencing violence.’ ‘Oh yeah…’ ‘Yeah.’ M3P1‘I have patients who have that too, maybe that’s your case, following difficult life events. Have you ever been hurt?’ M1 to P12AskingFunnel questions‘Why do you have such low self-esteem?’ ‘No, I don’t think so…’ ‘You’d say you have confidence in yourself.’ ‘Yes… it mainly came about because of work… that’s when I started… to go off the rails, you know… because of the work schedule we had…’ ‘Oh yeah… how is your life at home? Do you live alone?’ ‘Yes, well, I’ve always been with my parents since I separated from my wife (…)’ M2P5‘Was it someone who put you down?’ ‘No, not at all.’ ‘Have you ever met people like that?’ ‘(Silence) Yes, someone I worked for.’ M2P5Direct questionsClinical examination […] ‘Have people ever hurt you in your life?’ ‘Yes’ M2P11Silence during the examination. ‘And then these hot flashes, they’re unbearable, they’re eating away at me.’ ‘Apart from your father, are there people who have hurt you in your life?’ M2P8More targeted questionsOn the aggressor’s jealousy, on the control exercised by and the ambivalence of the aggressor, on the victim’s fear of the aggressor, on the aggressor’s angry behaviour‘Oh, yes? He is hard?’ M1‘Does your partner get angry sometimes?’ M2‘He… can… bother you a bit? Look for conflict sometimes?’ M1‘Was he putting pressure on you’ M2‘It harras me?’ M1Using affirmative or even interpretative sentencesRather than interrogative sentences‘The only factor, one of the factors, is being overweight; it promotes early periods and having a life… that’s why I asked you if you were the oldest, if you had to endure more than others.’ M3‘Consulting an addiction specialist would also be beneficial, you’re right, because I think you’ve been harmed…’ M2 Illustrative quotations on how to talk about violence, organised by themes and subthemes. Table 6.Summary of the communication techniques used by physicians to facilitate disclosure of interpersonal violence.Consultation stepsCommunication techniquesPractical descriptionBuilding trust and creating a safe spaceEmpathic listening, giving time, respecting silencesLet the patient talk without interruption, use short prompts (‘I’m listening’), accept emotions and pausesCreating conditions for disclosureManaging third parties, ensuring confidentialityAsk the accompanying person(s) to leave the room when appropriate, explicitly reassure about confidentialityLinking symptoms to possible violenceSymptom–violence linking questionsExplore the symptom onset in relation to stressful or traumatic events, ask whether current difficulties echo past experiencesIntroducing the topic of violenceGeneral statements and funnel questionsStart with general comments about how common violence is, then move from open questions about home life to more focused questionsAsking directlySimple, non‑stigmatising direct questionsUse clear questions, such as ‘Has anyone ever hurt you?,’ ‘Is he/she complicated or harsh with you?’Deepening the assessmentExploring other violence typesAfter one violence type is disclosed, systematically ask about other violence types (psychological, physical, sexual, and others)Adapting to the patient readinessRespecting timing, not forcingStop or postpone questioning if the patient is not ready, propose to talk again later Summary of the communication techniques used by physicians to facilitate disclosure of interpersonal violence. Having an empathic posture Physicians began by listening in an empathic, non-directive manner, often inviting free speech that yielded extended narratives, reflecting deliberate space-making (Table 4). Such listening combined verbal cues and attentive non-verbal behaviour (eye contact, open posture, nodding). Respect for silence played a pivotal role. These pauses were recognised by patients as meaningful for enabling them to formulate their thoughts and feel truly heard (Table 4). The three physicians reinforced this supportive environment with short phrases, ‘yes’ ‘I’m listening’ ‘I understand,’ and when appropriate, mirrored the patient’s words to convey understanding. Concomitantly, they engaged in active observation: body posture, gaze, emotional expression, speech, and even changes in dress or personal hygiene. Patient autonomy remained paramount. The three physicians reminded patients that they could choose what to share, ‘We don’t have to talk about it if you don’t want to…’ (M2P4) and refrained from probing when disclosure was difficult (Table 4). Questioning strategies A ‘call’ sentence was used to address the patients’ comments or to pursue a lead that warranted further exploration. For example, when patients were unclear or only hinted at difficult experiences, physicians prompted them with interpretive extensions: ‘Am I to understand that he was very demanding of sexual relations?’ (M1P10), rather than letting implications go unspoken (Table 5). General, non-stigmatising questions were frequently employed (e.g. ‘Is he a difficult person?,’ ‘Have you been hurt?’), allowing the patient to interpret freely and avoiding direct focus on the aggressor. If clarification was needed, the three physicians provided concrete examples: ‘He’s hot tempered and distant angry… he sulks, it always has to be his way, it never goes well whatever you do, he always has something to say.’ (M2P1), which sometimes enabled patients to recognise that their situation was not normal and to make connections between their symptoms and home life (Table 5). The three physicians also used questions focused on the perpetrator’s behaviour (jealousy, control, ambivalence, aggression, insults, or isolation) and sometimes introduced the theme by explaining the cycle of violence, or by naming specific manipulative tactics, such as threats and guilt induction. Physical violence was addressed with simple, direct vocabulary. Questions became more targeted as the conversation developed, sometimes culminating in exploratory or interpretative statements: ‘Have people ever hurt you in your life?’ (M2P11). Funnel questions were used to gently explore well-being, relational history, and home life (Table 5). Persevere on questioning When they identified a violence type, the three physicians routinely screened for other types of violence that may have occurred, consistent with the complex, cumulative nature of the phenomenon. If a patient responded negatively but warning signs persisted, physicians reformulated their questions, sometimes several times if they suspected a history of violence: ‘And was there other violence? Verbal, I mean, and psychological, from what you’re telling me. And also physical or sexual?’ (M1P11). Some patients described vague or ‘fuzzy’ memories. Therefore, physicians had to patiently explore associations until gradually, significant disclosures emerged. Sometimes, patients completed the narrative, eventually linking their symptoms or current fears to childhood or adult trauma. Discussion Main findings This qualitative study found that violence detection in general practice involves three steps: empathic and safe attitude, facilitative communication strategies, and persistent adaptive questioning. In our study, disclosures often emerged after GPs created space for free speech, responded attentively to indirect references or ‘test disclosures,’ and noticed links between patients’ symptoms and their relational or domestic context. Direct, simple questions, such as ‘Has anyone hurt you?’ or ‘Is he a difficult person?’ proved both accessible and effective. Linking vague, chronic, or treatment-resistant symptoms to possible histories of violence appeared to be a particularly pragmatic strategy, as it often helped patients recognise connections they had not initially made. When suspicion persisted, GPs reformulated their questions, screened for other forms of violence, and continued the dialogue until patients were ready to disclose. Comparison with existing literature Similarly, a clinical review on IPV screening and physician interventions also recommends talking first about the subject, using a general statement about the prevalence of violence against women, followed by an open question ‘Because violence against women is so common in our society, I ask all my women patients about violence in their lives. Do you feel safe in your current relationship?’ [11]. In our study, this type of question was often used, allowing a very broad discussion. The other questions were ‘Is he harsh?,’ ‘Is he complicated?.’ The progressive, funnel-shaped diagnostic approach is recommended by the Belgian Scientific Society of General Medicine [12]. It starts with open questions, with various entry points, adapted to the context (e.g. home life), up to questions about true acts of violence (all types). Even if violence is denied, progressive questioning may bring it out during the consultation or later and opens a door to possible help. The WHO suggests using general remarks [1] and proposes five simple tasks for frontline support for victims of IPV and sexual assault [13]. It also recommends to use straightforward questions to identify threats, injuries, insults, brutality, control over money or dating, forced sex, as found in the present study and in a previous clinical review [11], for instance ‘Does your partner hit you, threaten you, try to control you?.’ Many standardised tools, such as the WAST, have been developed to support the identification of intimate partner violence in primary care. Although the French version has excellent sensitivity (97.7%) and specificity (97.1%) in primary care [7], our study was specifically designed to explore how violence emerges during routine consultations in French general practice, rather than to evaluate the implementation of a specific questionnaire. Typically, consultations with a GP are short, patients usually consult for other reasons, and the therapeutic relationship relies heavily on continuity and a face‑to‑face dialogue centred on the patient’s immediate concerns. Therefore, a formal screening instrument at each visit may be difficult to integrate into the daily workflow and could be perceived as distancing or intrusive by both patients and GPs, especially when the reason for the encounter is unrelated to violence. Our findings highlight how empathic listening, attention to somatic and psychological cues, and progressive questioning can be used during a consultation to facilitate disclosure. Recent studies conducted after our fieldwork provide additional context for interpreting our findings. A trial in the United States of America (USA) assessed both the implementation and effectiveness of an IPV screening programme. It found that implementation facilitation strategies can triple screening rates and double population-level IPV detection in primary care, although disclosure among screened patients remains stable at 7-10% [14]. Another USA trial showed that electronic health record-based confidential screening significantly outperforms routine nurse-led oral screening (past-year IPV prevalence: 1.5% versus 0.1%), highlighting the critical role of privacy and non-interruptive clinician alerts [15]. Recent European primary care surveys confirmed the suboptimal gender-based violence detection despite protocols. In addition, a survey of healthcare workers in the United Kingdom found that despite the high rates of domestic abuse also among primary care staff, workplace support is often lacking [16]. Similarly, a study in Spain stressed that gender-based violence detection in primary care needs to improve [17]. Strengths and limitations A strength of this study is the analysis of consultation audio-recordings, which is rarely used in qualitative research in medicine. This provided rich material for studying the behaviour and reactions of physicians and patients, as well as their interactions. The originality of this study lies in the fact that it focused on the detection of current and past violence, all violence types, not just domestic or sexual violence. This study has some limitations. First, the small number of positive cases limits the scope and findings remain exploratory. Additional studies with larger samples are required to consolidate these observations. Second, the lack of transcription and analysis of non-disclosure consultations precluded a comparative analysis; however, resource constraints justified prioritising the thorough analysis only of disclosure cases. Third, this study was limited to three physicians and one region. The three physicians were purposely selected to reflect the diversity in experience, age, and practice settings typical of French general practice. Their medical training background is in line with the French national standards. The study was carried out in France’s largest region, Nouvelle-Aquitaine. This region’s extensive mixed urban-rural demographic (84,000 km2, ∼6 M inhabitants) and socioeconomic diversity is representative of the French national primary care contexts. Moreover, thematic saturation, coder triangulation, and patient/practice diversity support transferability beyond the specific context. Fourth, all participating physicians were women, and this could have influenced the patients’ responses or the sensitivity of disclosure. Fifth, the detection rates might have been influenced by a social desirability bias. Future research should include GPs of different genders, backgrounds, and clinical approaches to better evaluate the universality and applicability of these findings. Sixth, data were collected between June and September 2019, which represents a six-year interval between fieldwork and publication. Nevertheless, several factors support the ongoing validity and relevance of our results. Interpersonal violence remains a persistent structural phenomenon and global prevalence estimates have remained largely stable over the past decade [18,19]. In France, IPV continues to concern a significant proportion of women consulting in primary care, without evidence of substantial declines in incidence or health service utilisation patterns since 2019. Furthermore, the core features of French general practice (i.e. short multi-purpose consultations, emphasis on therapeutic continuity, and reliance on the GP-patient trust) have not undergone major organisational changes that would fundamentally alter the GP-patient interaction dynamics. Our focus on enduring relational and communicative processes (empathic listening, symptom interpretation, progressive questioning) rather than time-sensitive epidemiological indicators further enhances the transferability of these qualitative insights to current practice. Recommendations for general practice Recognising and facilitating disclosure of interpersonal violence Empathy is central to violence recognition and management in medical consultations. It allows GPs to identify subtle verbal and non-verbal cues and to establish the therapeutic alliance necessary to explore sensitive issues [20–22]. In our study, disclosures often followed attentive responses to indirect references, ‘test disclosures,’ symptom–violence associations, or non-verbal expressions of distress. From the patient’s perspective, empathy is associated with being heard, understood, and treated with humanity [22]. An empathic, non-judgmental posture fosters trust and creates a safe environment for disclosure. Emotional empathy may further support this relational dynamic [23,24]. Psychological violence is particularly difficult to detect, as patients may minimise, normalise, or fail to recognise abusive dynamics [25]. Awareness of behavioural patterns such as control, jealousy, ambivalence, and isolation is therefore essential. Supporting patients’ awareness and insight Many patients do not immediately recognise their experiences as violence and may attribute their symptoms to external circumstances or personal vulnerability [25,26]. Trauma-related mechanisms, including impaired insight, dissociation, or normalisation, may further hinder this recognition [27]. Helping patients to establish links between their symptoms and possible experiences of violence is a key component of care. This process should be gradual and patient-centred, avoiding premature labelling and allowing patients to develop their own understanding over time [28]. Communicating about violence: strategies and ethical considerations Effective communication relies on simple, non-judgmental, and behaviour-focused questions. Concrete experiences such as jealousy, control, isolation, or humiliation can help patients reflect on their situation without feeling judged or pressured. These conversations often need to be revisited across multiple consultations. Introducing the concept of violence may raise ethical considerations, as reframing experiences can imply a professional interpretation of the patient’s lived experience [29]. GPs should therefore adopt a cautious and collaborative approach, supporting insight while respecting the patient’s perspective, language, and readiness. Once violence is identified, appropriate referral and follow-up should be considered as part of ongoing care. In practice, the key points for GPs are set out in Box 1.Adopt an empathic, non‑judgmental attitude to foster trust, safety, and disclosure.Pay attention to indirect cues, test disclosures, and non‑verbal signs suggesting distress or control.Link vague, chronic, or treatment‑resistant somatic and psychological symptoms to possible histories of violence when appropriate.Explore concrete behaviours such as control, jealousy, isolation, humiliation, and intimidation rather than focusing only on physical acts.Use simple, open-ended, and non-threatening questions first, then progressively more direct questions about violence if suspicion persists.Avoid imposing labels too early; allow patients to describe their experiences in their own words and at their own pace.Revisit the discussion over several consultations when necessary, especially if warning signs remain despite initial denial.When violence is identified or strongly suspected, arrange appropriate referral, document concerns, and organise follow‑up as part of ongoing care. Adopt an empathic, non‑judgmental attitude to foster trust, safety, and disclosure. Adopt an empathic, non‑judgmental attitude to foster trust, safety, and disclosure. Pay attention to indirect cues, test disclosures, and non‑verbal signs suggesting distress or control. Pay attention to indirect cues, test disclosures, and non‑verbal signs suggesting distress or control. Link vague, chronic, or treatment‑resistant somatic and psychological symptoms to possible histories of violence when appropriate. Link vague, chronic, or treatment‑resistant somatic and psychological symptoms to possible histories of violence when appropriate. Explore concrete behaviours such as control, jealousy, isolation, humiliation, and intimidation rather than focusing only on physical acts. Explore concrete behaviours such as control, jealousy, isolation, humiliation, and intimidation rather than focusing only on physical acts. Use simple, open-ended, and non-threatening questions first, then progressively more direct questions about violence if suspicion persists. Use simple, open-ended, and non-threatening questions first, then progressively more direct questions about violence if suspicion persists. Avoid imposing labels too early; allow patients to describe their experiences in their own words and at their own pace. Avoid imposing labels too early; allow patients to describe their experiences in their own words and at their own pace. Revisit the discussion over several consultations when necessary, especially if warning signs remain despite initial denial. Revisit the discussion over several consultations when necessary, especially if warning signs remain despite initial denial. When violence is identified or strongly suspected, arrange appropriate referral, document concerns, and organise follow‑up as part of ongoing care. When violence is identified or strongly suspected, arrange appropriate referral, document concerns, and organise follow‑up as part of ongoing care. Conclusion These qualitative findings suggest that the identification of interpersonal violence in primary care depends on specific communicative qualities rather than on screening alone. Important qualities include non-judgmental listening, attention to indirect cues, progressive questioning, and repeated opportunities for disclosure over time. Our findings suggest that this communication-based approach could complement structured screening strategies, particularly when used selectively in patients presenting with vague, chronic, or relationally suggestive symptoms. Future studies should assess which communicative behaviours are most effective, whether brief screening prompts can be integrated into the clinical interview, and how these approaches compare with standardised protocols in different primary care settings. Supplementary Material Supplemental Material