Address gender-based violence, social determinants and the public-health dimensions of women's health

In one line

Gender-based violence is a population-level public-health emergency that surfaces as ordinary gynaecological and obstetric presentations, and the consultant's task is to hold three frames at once — the social determinants that produce the disease, the time-critical clinical-forensic response to the woman in front of you (HIV PEP within 72 hours, emergency contraception within 120 hours, evidence to a court standard), and the prevention agenda that should make the next case unnecessary.

This chapter sits one level above the recognition-and-acute-drill groundwork covered for the Intermediate exam — revise the safe-enquiry technique, the J88, the acute sexual-assault sequence and the statutory architecture at the Intermediate gender-based violence chapter. The work here is the consultant layer: reasoning about GBV as a determinant of women's health at population scale, defending the evidence behind a screening or response decision, and arguing the medico-legal calls a specialist is expected to own.

Mechanism & pathophysiology

The mechanism of GBV is social, not cellular, and a consultant who reaches only for individual-level explanations will misread the epidemiology. Violence against women is produced by an ecological cascade: societal norms that sanction male control over women, community-level poverty and weak sanctions, relationship-level conflict and male dominance, and individual histories of childhood abuse and harmful alcohol use. South Africa concentrates almost every one of these drivers — entrenched gender inequity, extreme income inequality, a normalisation of violence inherited from a violent history, high firearm availability and pervasive alcohol misuse — which is why the country sits at the extreme tail of the global distribution rather than merely above average.

The "pathophysiology" that matters clinically is how this social exposure converts into measurable disease in the woman's body. Three pathways do most of the work:

• Direct injury. Blunt and penetrating trauma, strangulation, and the obstetric consequences of abdominal trauma — placental abruption, preterm birth, low birth weight, fetal loss. Strangulation deserves separate weight: it is a sentinel marker of near-lethal violence and a predictor of subsequent femicide, and it can cause delayed airway or neurological compromise with minimal external signs.
• Reproductive coercion and its sequelae. Forced or coerced sex, contraceptive sabotage and forced continuation or termination of pregnancy generate unintended pregnancy, repeat unsafe abortion, and recurrent sexually transmitted infection. The syndemic interaction with HIV is the SA-defining feature: violence increases HIV acquisition (through forced sex, inability to negotiate condom use, and the biological vulnerability of genital trauma) and HIV-positive women face elevated violence on disclosure — the two epidemics amplify each other.
• Chronic stress physiology and mental ill-health. Sustained exposure drives depression, anxiety, post-traumatic stress disorder, substance use and suicidality, which in turn worsen antenatal engagement, adherence and obstetric outcome. In pregnancy this is not abstract: violence may begin or escalate in pregnancy, and intimate-partner homicide is a recognised contributor to maternal mortality.

The unifying idea is that GBV behaves like any other major determinant of women's health — it has a dose-response relationship with poor outcomes, it clusters with poverty and HIV, and it is, in principle, preventable. Treating each presentation as an isolated clinical event misses that the woman with recurrent STI, the late-booker, the unexplained abruption and the woman with intractable "non-specific" pelvic pain may be expressions of the same upstream cause.

The social-determinants lens does the same explanatory work for women's health more broadly that the violence model does for GBV specifically. Health is patterned by the conditions in which women are born, grow, live and work — income, education, food security, water and sanitation, housing, access to services — and these gradients explain more of the variance in maternal and reproductive outcomes than any single clinical factor. In South Africa the gradient is steep and stratified along the historical fault lines of race and geography: a rural woman in a former homeland district faces a maternal-mortality risk, an HIV exposure and a violence exposure that a wealthy urban woman does not, and the difference is produced upstream of any clinic. The public-health frame insists that the determinant, not just the disease, is a legitimate target — that contraceptive stock-outs, the distance to the nearest functioning theatre, and the alcohol outlet density of a neighbourhood are as much "pathophysiology" of poor women's health as any biochemical pathway. A consultant who can name the gradient and its drivers reasons about the population the patient came from, not only the patient.

Assessment

Assessment operates at two altitudes — the individual encounter and the population the patient represents — and a consultant is expected to move between them.

The clinical clues to GBV are a pattern, not a complaint. The presentations that should prompt enquiry include injuries inconsistent with the given history, delay in seeking care, repeat attendance with vague or somatic complaints, recurrent or treatment-resistant STI, unintended or concealed pregnancy, late or absent antenatal booking, a partner who attends every visit and answers for her, chronic pelvic pain without pathology, and depression, substance use or self-harm. None is diagnostic alone; the skill is in reading the constellation.

Routine universal screening versus clinical enquiry. This is a genuine point of guideline divergence. WHO does not recommend universal screening (asking every woman in every health-care encounter), because trials have not shown that screening alone improves outcomes and it carries opportunity and safety costs; WHO instead recommends clinical enquiry — asking when assessing conditions that may be caused or complicated by IPV. Several national bodies (and many antenatal programmes) favour routine antenatal enquiry on the grounds that pregnancy is a window of both heightened risk and reliable contact with the health system. The defensible consultant position is to state the distinction explicitly: case-finding driven by clinical suspicion is universally supported; whole-population screening is not endorsed by WHO on current evidence, and any enquiry must meet minimum safety conditions (privacy, no third party, a trained responder and a referral pathway in place) before it is offered.

The enquiry itself must be safe before it is thorough. See her alone — the controlling companion may be the perpetrator — in a private, uninterrupted space with a chaperone for examination. Ask once, gently, without expressions of doubt. Establish stability and immediate safety first; resuscitation of major trauma, haemorrhage and strangulation overrides both history-taking and any forensic process.

Quantify lethality risk, because it changes the urgency of the protective response: a history of strangulation, escalating violence, weapons in the home, explicit threats to kill, recent separation, and pregnancy itself are recognised markers that the next assault may be fatal. The femicide surveillance data give this a hard edge — intimate-partner femicide is the leading category of female homicide in South Africa, so a woman volunteering these markers is describing a measurable mortality risk, not a relationship difficulty, and the assessment must treat it as such.

Document to a forensic standard from the first contact. A consultant supervising a service is responsible for the quality of its medico-legal records, which may be the only objective account read in court years later. The principles: record observations in objective descriptive terms with measurements and a body map; quote the survivor's account in her own words and attribute it as history; never write conclusory opinions that cannot be defended (the absence of genital injury is not evidence of consent, and writing it as such is a serious error); complete the J88 contemporaneously and legibly; and maintain an unbroken chain of custody for every specimen — unlabelled, unsealed or improperly handled evidence is forensically worthless and can sink a prosecution.

Investigations are driven by the clinical and forensic picture: urine βhCG in all of reproductive age; a baseline HIV test (a positive baseline means she was already living with HIV, which converts PEP to treatment); STI screening and hepatitis B status; and trauma- or toxicology-directed bloods where the history demands them (including where drug-facilitated assault is suspected). The interpretation matters as much as the test: a negative βhCG does not exclude an at-risk conception window and does not remove the indication for emergency contraception; a reactive baseline HIV test is a clinical finding requiring linkage to treatment, not a contraindication to the rest of the pathway.

At population level, the assessment a consultant must be able to make is of the burden and its drivers in their own service: the antenatal IPV-disclosure rate, the proportion of unintended pregnancies, the HIV co-burden, and whether a functioning sexual-assault pathway actually exists on site. The South African numbers anchor this. Globally, around one in three women experience lifetime physical or sexual violence; roughly 30% of ever-partnered women report lifetime physical and/or sexual intimate-partner violence (WHO 2021 estimates). South Africa exceeds this: the 2016 Demographic and Health Survey put lifetime IPV among ever-partnered women at approximately 22%, and national femicide surveillance recorded an overall femicide rate of 11.1 per 100 000 women and an intimate-partner femicide rate of 4.9 per 100 000 in 2017 — more than four times the estimated global rate and higher than any other country with comparable data (Abrahams et al., 2024).

Management

Management spans the same two altitudes. For the individual it runs immediate → ongoing → long-term; for the population it is a prevention agenda. The non-negotiable principle threaded through all of it: care is never made conditional on the woman reporting to police, and time-critical prophylaxis never waits for the forensic process.

Immediate — the acute sexual-assault response (the time-critical core). The clinical-forensic priorities run largely in parallel, but the prophylaxis clock dominates the sequence:

Two figures must be exact because candidates lose marks on them and patients lose protection: HIV PEP within 72 hours (the benefit falls steeply with delay; give the full 28-day course, not a starter pack, with the SA first-line TLD regimen) and emergency contraception within 120 hours. Both lose efficacy with every hour, and neither may be delayed for the evidence kit — the kit can follow, the prophylaxis cannot wait. Confirm the precise regimen and window against the current NDoH PEP and contraception guidance at the point of care, as these are among the most frequently updated numbers in practice.

Immediate — the IPV (chronic) response uses LIVES, not rescue. Most GBV the gynaecologist meets is chronic intimate-partner violence surfacing through routine care, where the registrar's job is to recognise, validate, assess danger, document and connect — not to direct her decisions. The WHO first-line support framework is LIVES: Listen closely, with empathy and without judgment; Inquire about her needs and concerns; Validate her experience by showing you believe her; Enhance her safety with a concrete protective plan; Support her by connecting her to information, services and social support. Validation ("this is not your fault; help is available") and an explicit safety plan are themselves clinical interventions, not pleasantries.

Ongoing. Arrange and protect follow-up: repeat HIV testing within the PEP window, adherence support, completion of the hepatitis B and STI courses, and mental-health review — watchful waiting over the first one to three months with definitive trauma-focused therapy (CBT or EMDR) for established PTSD, escalated sooner if she is incapacitated or acutely suicidal. Address the safety plan at every contact; risk is dynamic and separation is itself a high-risk period.

Long-term. The woman cured of the acute episode still carries chronic risk and chronic sequelae — recurrent presentation, mental ill-health, reproductive harm. Long-term management is continuity of care embedded at district level so the tertiary centre is not the only safety net, linkage to protection orders (Domestic Violence Act 116 of 1998), social work and shelters, and recognition that her children may be the next index case.

Population — prevention is management at scale. A specialist working in a high-burden setting cannot treat their way out of GBV; the upstream agenda is part of the job. The WHO/UN Women RESPECT framework names seven evidence-informed prevention strategies — Relationship-skills strengthening, Empowerment of women, Services ensured, Poverty reduction, Environments made safe, Child and adolescent abuse prevention, and Transformed attitudes, beliefs and norms. The health-sector contributions a consultant actually controls are the "S" (a functioning, properly resourced sexual-assault and IPV response) and the secondary-prevention reach of routine reproductive care — contraceptive access, safe abortion, adolescent sexual and reproductive health — each of which reduces the unintended pregnancies, unsafe abortions and repeat exposures through which GBV does much of its damage.

The prevention frame maps onto the classic three levels and a consultant should be able to place an intervention at the right one. Primary prevention stops violence before it starts — the RESPECT strategies, school-based programmes that shift gender norms, alcohol-availability regulation, and economic-empowerment interventions; these sit largely outside the clinic but a specialist with a public voice can advocate for them. Secondary prevention is early detection and the interruption of harm — clinical enquiry, the antenatal window, contraceptive access that prevents the coerced pregnancy, and PEP that prevents HIV seroconversion after assault. Tertiary prevention limits the consequences of established harm — trauma-focused mental-health care, management of chronic pelvic pain and reproductive sequelae, and the protective and social linkage that prevents re-victimisation. The South African operationalisation of "Services" is the Thuthuzela Care Centre model — a one-stop, 24-hour, multidisciplinary site with co-located medical, forensic, psychosocial and criminal-justice services and an on-site police presence, designed to reduce the secondary trauma of fragmented care and to raise prosecution rates by getting the forensic chain right from the start. Knowing that the country has only a modest number of these centres against an enormous burden is part of an honest answer: the model is sound, the coverage is the limiting factor, and naming that gap is more useful than reciting the ideal.

Reproductive justice as public health

Access to contraception and to safe abortion is not a separate ethical topic bolted onto GBV; it is the same public-health argument. Coerced and unintended pregnancy is both a consequence of violence and a vector for further harm, and the historical SA burden of unsafe abortion — the toll that the Choice on Termination of Pregnancy Act 92 of 1996 was written to end — is the clearest demonstration that restricting access does not reduce abortion but only makes it lethal. A consultant defending reproductive-health access on public-health grounds argues from outcomes: where contraception and safe termination are accessible, maternal mortality from unsafe abortion falls and the cascade of repeat unintended pregnancy is interrupted. Adolescent sexual and reproductive health is the highest-yield extension of this logic, because adolescence concentrates both the risk of coercion and the lifetime dividend of getting access right. The reproductive-justice frame — that the conditions for safe, autonomous reproductive decisions are unequally distributed along exactly the same social lines as violence — is the language in which this is now argued, and it is defensible from the data, not merely from principle.

Guidelines compared

The bodies a candidate is expected to reconcile agree on the time-critical clinical core and diverge mainly on screening and on the structure of prevention.

The recent direction of travel worth flagging: the WHO 2021 prevalence estimates reset the global denominator (one in three women), the 2024 SA femicide analysis confirmed that the country remains the global outlier, and SA PEP guidance has standardised on dolutegravir-based TLD as a single daily tablet — a meaningful adherence gain over older multi-tablet PEP. Confirm any section-level statutory ambit and the exact current PEP/EC regimen against the governing Acts and the live NDoH guidance before acting, as these are periodically revised.

The evidence & the controversy

Three contested questions separate a consultant answer from a textbook one.

Does screening for IPV actually help? The honest reading of the trial evidence is that screening tools improve detection but have not been shown, on their own, to improve health outcomes — which is precisely why WHO stops short of recommending universal screening and recommends clinical enquiry instead. The controversy is that this can be misread as "don't ask". The defensible synthesis is that detection without a functioning response is futile or even harmful, so the question is not "screen or not" but "is there a safe, resourced pathway to refer into" — where there is, enquiry is justified; where there is not, building the pathway comes first. In the SA antenatal setting, the high background prevalence and the captive, repeated contact of antenatal care strengthen the case for routine enquiry relative to a low-prevalence high-income setting — context changes the calculus.

Mandatory reporting of adult IPV — protection or paternalism? Some jurisdictions mandate clinician reporting of IPV to police. WHO explicitly recommends against this for competent adults, on the evidence that it deters women from disclosing and seeking care and removes their control at the moment control matters most. The SA position aligns: the competent adult woman decides whether to open a case, and treatment is unconditional — while child abuse remains a firm mandatory-report exception. The live ethical tension is the duty to warn an identifiable third party at imminent risk versus the woman's autonomy and the harm of breaching her trust; the defensible approach is a structured one — assess lethality, maximise her safety with her, reserve breach of confidentiality for a genuine, specific, imminent threat to a third party, and document the reasoning — rather than a blanket rule in either direction.

Are current trends and exposures part of the GBV picture? Two threads currently shape the discourse and reward a candidate who can weigh them rather than recite them. First, technology-facilitated abuse — image-based sexual abuse, GPS-enabled stalking, coercive control through a partner's surveillance of a phone — is increasingly the medium of IPV, and it interacts with the safety plan (a safety plan that ignores the tracked phone is incomplete). Second, the periodic claims that environmental exposures — endocrine-disrupting chemicals, contaminants in sanitary products — are a hidden driver of women's reproductive harm: these belong in the social-determinants conversation but should be presented as contested and not yet established as causal at population scale, and explicitly distinguished from the well-evidenced determinants (poverty, inequity, HIV, alcohol) that actually drive the SA burden. Naming the speculative as speculative, while holding the line on what the data do support, is the consultant move.

Does the response actually work, and how would a service prove it? The harder appraisal question is not whether to respond but whether a given response improves outcomes — and the GBV literature is humbling here. Much of the early enthusiasm for screening, advocacy and brief counselling rests on trials that improved process measures (disclosure, referral uptake) without convincingly moving health outcomes, and several well-designed interventions have been null. This does not licence therapeutic nihilism — first-line support, time-critical prophylaxis and a functioning forensic pathway are justified on first principles and on the prevention of discrete harms (an HIV seroconversion averted, an unintended pregnancy prevented) that do not need a trial to be worth doing. But a consultant asked to commission or defend a service should reason like an evaluator: define the outcome that matters (completion of the 28-day PEP course, prosecution rate, repeat-presentation rate), measure it, and resist the temptation to declare success on the basis of activity alone. The SA evidence base is itself a model of this discipline — the femicide surveillance programme has run repeated national mortuary-based surveys for over two decades precisely so that the country can measure whether anything is changing, which is why it can state, with confidence and with confidence intervals, that South Africa remains the global outlier. Surveillance that measures the real endpoint is itself a public-health intervention.

Landmark trials & key evidence

The arithmetic the figures support: an intimate-partner femicide rate of 4.9 per 100 000 women, more than four times the global estimate, is the population-level expression of the same coercive control that presents in clinic as the late-booker and the woman with recurrent STI — the surveillance number and the clinic encounter are two ends of one disease.

Exam traps & red flags

• Making care conditional on reporting. HIV PEP, emergency contraception and injury care are provided regardless of whether she opens a criminal case. Conditioning treatment on a police report is both wrong and harmful.
• Letting forensics delay prophylaxis. PEP (72 h) and emergency contraception (120 h) lose efficacy by the hour; the evidence kit can wait, the drugs cannot.
• Quoting the windows wrong. PEP within 72 hours, EC within 120 hours — transposing them, or shortening EC to 72 h, is a recurrent error that costs the patient cover.
• Starting PEP without a baseline HIV test, or dispensing a starter pack instead of the full 28-day TLD course with adherence support.
• Over-claiming universal screening. WHO recommends clinical enquiry, not universal screening; presenting routine population screening as evidence-based for outcomes overstates the data. Equally wrong is reading "no universal screening" as "do not ask".
• Breaching a competent adult's autonomy. Adult IPV/sexual assault is not mandatorily reportable against her will; reporting child abuse (Children's Act) is the firm exception. Confusing the two — either way — is a serious medico-legal error.
• Reading a normal genital examination as evidence nothing happened. Most survivors of rape have no anogenital injury, particularly parous women; recording or implying otherwise can be used to discredit a true account.
• Missing strangulation. Minimal external signs can accompany life-threatening or delayed injury, and strangulation is a sentinel marker of future femicide — never dismiss it.
• Treating GBV as a social rather than clinical problem. Recurrent STI, unintended pregnancy, late booking, unexplained abruption and chronic pelvic pain are the disease presenting; missing the pattern is the commonest failure of all.
• Forgetting the population frame. A consultant answer locates the individual case in its social determinants and the prevention agenda; treating each case as an isolated event misses the point of the objective.

Evidence anchors

• WHO — Responding to intimate partner violence and sexual violence against women: clinical and policy guidelines (2013) — against universal screening / for clinical enquiry; EC within 5 days; PEP within 72 h; against mandatory IPV reporting.
• WHO — Health care for women subjected to intimate partner violence or sexual violence: a clinical handbook (WHO/RHR/14.26, 2014) — the LIVES first-line support model.
• WHO/UN Women — RESPECT women: preventing violence against women (2019) — the seven prevention strategies.
• WHO — Violence against women prevalence estimates, 2018 (released 2021) — ~1 in 3 women; ~30% ever-partnered lifetime IPV.
• Abrahams N, Mhlongo S, Chirwa E, et al. Femicide, intimate partner femicide and non-intimate partner femicide in South Africa, 1999–2017. PLoS Med. 2024;21(1):e1004330 — SA femicide and intimate-partner femicide rates; the global outlier.
• Horak J, Venter WDF, Wattrus C, et al. SAHCS 2023 Guideline for post-exposure prophylaxis. South Afr J HIV Med. 2023;24(1):1522 — PEP within 72 h; first-line TLD for 28 days; baseline HIV testing.
• Pronyk PM, Hargreaves JR, Kim JC, et al. Effect of a structural intervention for the prevention of intimate-partner violence and HIV in rural South Africa: a cluster randomised trial (IMAGE). Lancet. 2006;368(9551):1973–83 — Limpopo microfinance + gender/HIV training cut past-year IPV by 55% (aRR 0.45, 95% CI 0.23–0.91).
• South Africa NDoH National Sexual Assault clinical-forensic management guidance and the Thuthuzela Care Centre model — the SA one-stop, survivor-centred pathway (designated facilities; J88; sexual-assault evidence-collection kit).
• Criminal Law (Sexual Offences and Related Matters) Amendment Act 32 of 2007; Domestic Violence Act 116 of 1998; Children's Act 38 of 2005 (mandatory reporting of child abuse) — the SA statutory framework.
• South African Demographic and Health Survey 2016 — lifetime IPV among ever-partnered women approximately 22%.