In one line
Safe gynaecological surgery is a system, not a single skill: the operation itself is only the middle of a chain that runs from honest risk assessment and reasonable-patient (Castell) consent, through a checklist-disciplined theatre with timed antibiotic and risk-stratified VTE prophylaxis, into an enhanced-recovery pathway that gets the patient eating, mobile and home — and the consultant's job is to make every link in that chain deliberate rather than reflexive.
Performing the procedure is the lesser challenge; keeping the patient safe around it is the harder one. Most complications are prevented before the knife touches skin, by selecting and optimising the right patient, by disclosing risk the way the law now requires, and by hard-wiring the prophylaxis that the evidence has settled. A residual baseline risk remains even after fully appropriate care, so an event should trigger a no-blame review of the system rather than being treated as an automatic failure.
Mechanism & pathophysiology
Perioperative care rests on the physiology that makes surgery dangerous. Three overlapping mechanisms account for most of what is done around an operation.
The surgical stress response. Tissue injury, fasting, pain and anxiety trigger a stereotyped neuroendocrine and inflammatory cascade. Afferent nociceptive and inflammatory signals drive the hypothalamic–pituitary–adrenal axis (cortisol, ACTH) and the sympathetic nervous system (catecholamines), while damaged tissue releases cytokines (IL-6, TNF-α, IL-1). The downstream state is catabolic and pro-inflammatory: insulin resistance (functional diabetes even in non-diabetics), sodium and water retention, protein breakdown, immune suppression and a hypercoagulable shift. This matters because the response is modifiable. Almost every enhanced-recovery component — not starving the gut, carbohydrate loading, regional/opioid-sparing analgesia, minimal-access surgery, euvolaemia, early feeding — is an intervention that blunts one arm of this response. Reframed that way, enhanced recovery is not a hotel-comfort programme; it is the deliberate attenuation of a maladaptive stress physiology — which is why it works.
Virchow's triad and surgical VTE. Venous thromboembolism is the perioperative death you are most often blamed for missing, and its mechanism is Virchow's three pillars, every one of which surgery aggravates. Stasis: anaesthesia abolishes the calf-muscle pump, the patient is immobile on the table and in bed, and pelvic surgery and a gravid-sized uterus mechanically compress pelvic veins. Endothelial injury: pelvic dissection, retraction and energy devices directly damage vessel walls. Hypercoagulability: the stress response itself raises fibrinogen and clotting factors and suppresses fibrinolysis, and malignancy adds tissue-factor-driven, cytokine-mediated procoagulant activity — which is exactly why cancer surgery sits in a higher risk tier than benign surgery of identical magnitude. The triad is the through-line that explains both who needs thromboprophylaxis (the more boxes a patient ticks, the higher the tier) and which prophylaxis you reach for (mechanical methods attack stasis; pharmacological methods attack hypercoagulability; you frequently need both).
The infection mechanism. Surgical site infection follows when an inoculum of bacteria at the wound exceeds host defences during the decisive window — the few hours around incision and closure when the wound is open and the haematoma/clot is colonisable. The single most important deduction from that mechanism is timing: a tissue antibiotic concentration above the minimum inhibitory concentration must already be present at incision, not started afterwards, which is the entire rationale for giving prophylaxis in the hour before the knife rather than in recovery. Prolonging antibiotics for days afterwards does not extend that window of benefit; it only selects for resistance.
With the physiology in place, each downstream decision stops being a memorised rule and becomes a consequence.
Assessment
The pre-operative assessment exists to answer two questions: is the planned operation the right one for this patient now, and what must I fix or mitigate before I expose her to it. Work it as systematically as a booking clinic.
- Risk stratification and functional capacity. Anchor the global risk in the ASA physical-status class and, more usefully, in functional capacity — can she climb two flights of stairs or do equivalent ~4 MET activity without limiting symptoms? Poor or unassessable functional capacity in major surgery is the trigger for cardiopulmonary referral, not a reflex battery of tests. Apply NICE NG45 to preoperative testing: investigations are targeted to the surgery grade and comorbidity, not ordered routinely — a fit woman for minor day surgery needs essentially none.
- Cardiac risk. Identify active cardiac conditions (unstable angina, decompensated heart failure, significant arrhythmia, severe valve disease) that mandate optimisation or postponement. For stable disease, the combination of surgical magnitude and functional capacity drives the decision; a resting ECG and focused assessment usually suffice for intermediate-risk gynaecological surgery.
- Anaemia — correct it before, not transfuse around it. Preoperative anaemia independently predicts transfusion, infection, longer stay and death, and it is common in the SA gynaecology population (heavy menstrual bleeding, fibroids, malignancy, iron deficiency, HIV). The principle is to diagnose and treat the anaemia in clinic — iron (oral, or intravenous where time is short or oral is failing) for iron deficiency — rather than accept a low haemoglobin and plan to transfuse perioperatively. Patient blood management starts weeks before the list.
- Diabetes and glycaemia. Assess control (HbA1c), plan the perioperative regimen, schedule the patient early on the list, and target reasonable glycaemia — surgical hyperglycaemia worsens infection and the stress response amplifies it. Poor control is a reason to optimise, not a reason that fixes itself on the day.
- Smoking. Smoking impairs wound healing and oxygenation and raises pulmonary and wound complications; even short-interval cessation before surgery helps, and the pre-op clinic is a cessation opportunity, not just a risk-noting exercise.
- Frailty, not age. In the older woman, frailty (assessed with a tool such as the Clinical Frailty Scale) predicts outcome far better than chronological age. A frail patient warrants a frank conversation about goals, a tailored (often less radical) plan and proactive geriatric/anaesthetic input.
- The SA host factors. Test for and manage HIV (and optimise ART), screen for and treat anaemia and undernutrition, and remember TB and the realities of access — a patient who must travel far for follow-up needs a plan that does not assume easy return.
The assessment is not a form; it is the moment you decide what to fix. With the patient selected and optimised, the next obligation is consent — and that has changed.
Consent — the Castell standard and the SA frame
The law of consent has shifted from what a reasonable doctor would disclose to what a reasonable patient would want to know, and a consultant must be able to state the standard precisely.
In South Africa the governing authority is Castell v De Greef 1994 (4) SA 408 (C), which replaced the old Bolam/Sidaway "responsible body of doctors" test with the material-risk standard. A risk is material if a reasonable person in the patient's position, if warned of the risk, would be likely to attach significance to it, or the medical practitioner is or should reasonably be aware that this particular patient would be likely to attach significance to it. Castell was decided in 1994, twenty-one years before the United Kingdom reached the same test in Montgomery v Lanarkshire Health Board [2015] UKSC 11, so Montgomery is a useful UK parallel rather than the standard SA is held to. Two consequences follow. First, materiality is not a fixed percentage. A small numerical risk of a consequence the patient would dread (a young woman's risk to fertility, a singer's risk to her voice) can be material, while a larger risk of a trivial outcome may not be. Second, consent is a dialogue: the duty is to discuss the material risks of the proposed procedure and the reasonable alternatives (including conservative management and doing nothing), tailored to this woman's circumstances, values and questions, not to recite a generic list at her.