Evaluate and manage the subfertile couple

In one line

Subfertility is a problem of a couple, not a woman: conception requires an egg, a sperm, a patent path between them and a receptive uterus, so the work-up tests each of those in parallel, and the single most consequential management decision is matching the treatment to the cause — letrozole for the anovulatory woman with polycystic/metabolic ovarian syndrome, ICSI for severe male factor, IVF for tubal disease — while in South Africa the harder reality is that the evidence-based answer and the deliverable one frequently diverge.

Mechanism & pathophysiology

A spontaneous conception needs five things to happen in sequence within one cycle: an oocyte must mature and be released; sperm in adequate number and quality must reach the fallopian tube; the tube must be patent and functional enough to capture the egg, host fertilisation and transport the embryo; the endometrial cavity must be normal and the endometrium receptive; and implantation must succeed. Subfertility is the failure of that chain, and every cause maps onto a broken link. Reasoning from the chain rather than from a memorised list is what lets you order the work-up sensibly and interpret a normal result correctly — a normal mid-luteal progesterone tells you the ovulatory link is intact and shifts your suspicion downstream, it does not "rule out infertility".

Ovulatory disorders are conventionally grouped the way the WHO frames anovulation, because the group dictates the treatment. Group I is hypothalamic–pituitary failure — low gonadotrophins, low oestrogen, the hypogonadotrophic state of functional hypothalamic amenorrhoea (low body weight, excessive exercise, stress), Kallmann syndrome, or pituitary causes; these women have an unstimulated axis and respond to pulsatile GnRH or gonadotrophins, not to anti-oestrogens, because there is no oestrogen-negative-feedback loop for an anti-oestrogen to interrupt. Group II is hypothalamic–pituitary–ovarian dysfunction with normal gonadotrophins and oestrogen — overwhelmingly polycystic/metabolic ovarian syndrome (PMOS, previously PCOS), the single commonest cause of anovulatory subfertility, in which disordered gonadotrophin secretion, insulin resistance and intra-ovarian hyperandrogenism arrest follicles short of dominance. Group III is ovarian failure — high gonadotrophins, low oestrogen — premature ovarian insufficiency, where the follicular pool is exhausted and ovulation induction is futile; the route to a pregnancy is oocyte donation. Hyperprolactinaemia sits slightly outside the WHO scheme but acts by suppressing GnRH pulsatility; it is a discrete, treatable cause that a single prolactin level detects. The mechanistic payoff is direct: you cannot pick an ovulation-induction agent until you know which ovulatory link is broken, because an anti-oestrogen works in Group II and fails in Groups I and III. The metabolic and endocrine groundwork for PMOS is developed in polycystic-metabolic-ovarian-syndrome; here the focus is the fertility decision that follows from it.

Tubal and pelvic factor is mechanical obstruction or functional damage to the path. In South Africa this link carries disproportionate weight: a high background prevalence of Chlamydia trachomatis and gonococcal pelvic inflammatory disease, and of genital tuberculosis, produces tubal occlusion, hydrosalpinx and peritubal adhesions at a rate that shapes the entire epidemiology of subfertility on the continent. A hydrosalpinx is doubly harmful — it blocks the tube and the retrograde flow of inflammatory fluid into the cavity halves IVF success until the tube is dealt with. Endometriosis damages the same link through adhesions, distorted tubo-ovarian anatomy and an inflammatory peritoneal environment hostile to gametes.

Male factor contributes to roughly half of all couples (as a sole or contributing cause) and is a failure of the sperm link — a problem of production (primary testicular failure, varicocele, cryptorchidism, post-orchitis, genetic causes such as Y-microdeletions or Klinefelter syndrome), of transport (obstruction, congenital bilateral absence of the vas in CFTR carriers, post-vasectomy), or of delivery (ejaculatory or erectile dysfunction). The mechanistic point that governs treatment is that severe oligo- or azoospermia is bypassed, not cured: ICSI injects a single viable sperm and sidesteps every step of natural sperm function, which is why the discovery of profound male factor reroutes the whole couple straight to assisted conception.

Uterine factor breaks the receptivity link — submucosal fibroids distorting the cavity, intrauterine adhesions (Asherman syndrome, classically after sepsis or aggressive curettage), a uterine septum, or endometrial pathology. Age and ovarian-reserve decline is the link that degrades silently in every woman: oocyte number and, more importantly, oocyte quality fall with age, with a clear inflection in the mid-to-late thirties and a steep decline after 40, driven by rising meiotic aneuploidy. This is the one variable no treatment reverses, which is why a woman's age is the most powerful single prognostic factor in the entire field and why delay is itself a harm. When every link tests normal the couple has unexplained subfertility — a real diagnosis of exclusion in perhaps a quarter of couples, not a synonym for an incomplete work-up.

Assessment

Subfertility is defined as failure to conceive after 12 months of regular unprotected intercourse; investigate earlier — at 6 months — when the woman is ≥35, or when the history flags a cause: oligo/amenorrhoea, known or suspected tubal disease (prior PID, ectopic, pelvic surgery, genital TB), suspected male factor, or where age and reserve leave no time to waste. Investigate immediately, not after a waiting period, for an obvious cause such as amenorrhoea or azoospermia.

The cardinal principle is that the work-up is of the couple, done in parallel, and history is taken from both partners.

• History (her): menstrual pattern (regular cycles ~every 21–35 days predict ovulation; oligo/amenorrhoea predict an ovulatory disorder), age, duration of trying, previous pregnancies and their outcomes, prior pelvic infection/PID/STIs, TB exposure or treatment, pelvic or abdominal surgery, dysmenorrhoea or dyspareunia (endometriosis), galactorrhoea or visual symptoms (prolactinoma), and hirsutism/acacanthosis (PMOS). Coital frequency and timing, smoking, alcohol, weight, and chronic disease.
• History (him): previous paternity, childhood undescended testes or mumps orchitis, genital trauma or surgery, infections, systemic illness, medication and anabolic-steroid or testosterone use (suppresses spermatogenesis — a common, reversible, easily missed cause), occupational heat/toxin exposure, smoking and alcohol, and sexual function.
• Examination: her BMI and body-fat distribution, signs of hyperandrogenism, thyroid, galactorrhoea, and a pelvic examination for masses, fixed retroversion or nodularity (endometriosis), and cervical/vaginal abnormality. His testicular volume and consistency, the presence of both vasa, and a varicocele.

The investigations test the five links directly:

• Confirm ovulation — mid-luteal serum progesterone, drawn 7 days before the expected next period (day 21 of an idealised 28-day cycle, but timed to the individual cycle — a day-21 level in a 35-day cycle is meaningless). A value comfortably above the laboratory's ovulatory threshold confirms a corpus luteum formed. In oligomenorrhoea the test is repeated or timed by tracking. A regular cycle is itself good evidence of ovulation, so in a woman with reliably regular menses the progesterone is confirmatory rather than diagnostic.
• Ovarian reserve — antral follicle count on transvaginal ultrasound and anti-Müllerian hormone (AMH). Both estimate the quantity of the remaining follicular pool and predict the ovarian response to stimulation; neither reliably predicts natural fertility or oocyte quality, and a low result must never be used to deny a woman a chance or to tell her she cannot conceive — it informs the choice and dose of treatment and the urgency, not the verdict. Day-3 FSH/LH/oestradiol adds information where AMH is unavailable. A high FSH with low oestradiol points to diminished reserve or POI.
• Tubal patency. In a woman with no history suggesting pelvic pathology, first-line is an outpatient test — hysterosalpingography (HSG) or, where available, hysterosalpingo-contrast-sonography (HyCoSy), which avoids ionising radiation and assesses the cavity simultaneously. Where the history predicts pelvic disease (PID, endometriosis, prior ectopic or surgery, suspected TB), laparoscopy and dye is the more informative first test because it both diagnoses and allows treatment of adhesions, endometriosis and tubal disease at one sitting. The choice of tubal test is therefore driven by the pre-test probability of pelvic pathology, not by reflex.
• Semen analysis is the cornerstone of the male work-up and the cheapest, highest-information investigation in the whole pathway — never defer it. Produced after 2–7 days' abstinence and assessed against the WHO 2021 (sixth-edition) lower reference limits: semen volume ≥1.4 mL, sperm concentration ≥16 million/mL, total sperm number ≥39 million per ejaculate, total motility (progressive + non-progressive) ≥42%, progressive motility ≥30%, vitality ≥54%, and normal morphology ≥4%. An abnormal result is repeated after ~3 months (a full spermatogenic cycle, and to exclude a transient insult such as a febrile illness) before it is acted on, unless azoospermic. Azoospermia mandates referral for endocrine (FSH, LH, testosterone), genetic (karyotype, Y-microdeletion, CFTR) and, where relevant, imaging assessment to separate obstructive from non-obstructive causes — the distinction determines whether surgical sperm retrieval can succeed.
• Uterine cavity is assessed on the transvaginal scan done for the antral count, and where the history or scan raises suspicion (recurrent loss, prior cavity surgery or sepsis, an abnormal HSG) by saline-infusion sonography (SIS) or hysteroscopy, the latter being the reference standard and therapeutic in the same procedure.

The crucial interpretive discipline is to read each result as evidence about one link, and to keep the work-up proportionate. In a SA district or regional setting the resource-appropriate sequence is deliberate and cheap-first: take both histories and examine both partners; send a semen analysis and a correctly timed mid-luteal progesterone; check prolactin and TSH if cycles are irregular; obtain a baseline pelvic ultrasound; and only then commit to a tubal test, choosing HSG/HyCoSy or laparoscopy by the pelvic-disease probability. AMH and formal reserve testing belong with the patient who is heading towards ovarian stimulation or ART. Add an HIV test for both partners and screen for the infections that matter to a pregnancy — this is not optional in the SA context, both because it changes pre-conception care (optimise ART, achieve viral suppression, the SA framework being PVT-era HIV care) and because HIV-serodiscordant couples have a specific, manageable conception pathway that depends on knowing both statuses.

Management

Order the plan immediate → ongoing → definitive, and let the cause dictate the route.

Immediate / preconception and lifestyle — for every couple, started at the first visit and never deferred while investigations run. Folic acid 5 mg daily where there is any risk factor (and 0.4 mg otherwise), rubella immunity check and vaccination if needed, optimisation of chronic disease, cervical screening if due, and frank advice on the modifiable factors that genuinely move the odds: weight (a raised BMI reduces fecundity, worsens PMOS anovulation and lowers ART success; even modest loss can restore ovulation, and a very low BMI is equally a cause), smoking (lowers fertility in both partners and accelerates ovarian ageing — cessation is one of the highest-value interventions available), alcohol moderation, and correctly timed, regular intercourse (every 2–3 days throughout the cycle, which is more reliable than ovulation-test-driven timing). These cost nothing, and in PMOS-related anovulation a structured weight and lifestyle programme is genuinely first-line, with pharmacological ovulation induction layered on top — not instead of it.

Definitive — matched to the cause.

Anovulatory PMOS (WHO Group II). Letrozole is first-line ovulation induction. The aromatase inhibitor lowers oestrogen, releasing the hypothalamus from negative feedback to drive FSH and a single dominant follicle; given as 2.5 mg daily for 5 days from early in the cycle (escalating to 5 and 7.5 mg in non-responders), its advantage over clomiphene is a higher live-birth and ovulation rate and a more physiological, predominantly mono-follicular response. Clomiphene citrate (50 mg for 5 days, titrated to 150 mg) remains an effective and very cheap alternative where letrozole is unavailable, at the cost of its anti-oestrogenic effect on endometrium and cervical mucus and a higher multiple-pregnancy rate. Metformin is not an effective stand-alone ovulation-induction agent — its place is adjunctive (it improves the metabolic milieu and may add to clomiphene in the insulin-resistant or clomiphene-resistant woman, and has a role in reducing OHSS risk in gonadotrophin/IVF cycles), and it should not be offered alone as the route to a baby. Gonadotrophins (low-dose FSH with ultrasound and oestradiol monitoring) and laparoscopic ovarian surgery ("drilling") are second-line for the clomiphene/letrozole-resistant woman. The non-negotiable safety rule across all ovulation induction is monitoring: ovulation induction without cycle monitoring risks both high-order multiple pregnancy and, with gonadotrophins, ovarian hyperstimulation syndrome — the agent that makes a clinic look successful by producing twins and triplets is the agent that fills the neonatal unit.

Hypothalamic anovulation (Group I). Correct the underlying cause first — restore weight and reduce excessive exercise. Where induction is needed, the unstimulated axis demands gonadotrophins (or pulsatile GnRH), not an anti-oestrogen; correct hypothyroidism and treat hyperprolactinaemia with a dopamine agonist (cabergoline or bromocriptine), which alone restores ovulation in the great majority of hyperprolactinaemic women.

Tubal and pelvic factor. The decision is tubal surgery versus IVF, and it turns on the severity of the damage and the woman's age. Selective tubal surgery can be appropriate for mild distal disease or proximal occlusion in a young woman with otherwise good prognosis, but for bilateral or severe tubal disease IVF is the more effective route, bypassing the tube entirely. A hydrosalpinx must be dealt with before IVF — salpingectomy (or proximal tubal occlusion) of the affected tube roughly restores IVF success that the hydrosalpinx fluid otherwise halves, and is one of the clearest cause-and-effect interventions in the field. Genital TB deserves specific mention in SA: it damages the endometrium as well as the tubes, carries a poor reproductive prognosis even after anti-TB treatment, and frequently leads directly to IVF or, with severe Asherman-type endometrial damage, to surrogacy.

Male factor. The pathway escalates with severity: mild abnormalities may still allow natural or IUI conception and warrant attention to reversible causes (stop exogenous androgens, treat infection, lifestyle, consider varicocele repair in selected men); moderate-to-severe abnormalities route to IVF; and severe oligo/asthenospermia or surgically retrievable azoospermia route to ICSI, which needs only a handful of viable sperm. Obstructive azoospermia is amenable to surgical sperm retrieval with ICSI; non-obstructive azoospermia with high FSH and small testes carries a poor retrieval prognosis and may end in donor sperm. A genetic cause (Klinefelter, Y-microdeletion, CFTR) should be sought and counselled before ICSI, because ICSI transmits the heritable defect that natural selection would otherwise have filtered.

Unexplained subfertility. With a normal couple work-up and a younger woman, expectant management with optimised timing for a defined period is legitimate, because a meaningful proportion conceive naturally and the active alternatives carry cost and risk. The escalation is towards IVF as the most effective intervention; the older the woman, the shorter the justifiable period of expectant or low-intensity management, because reserve and oocyte quality are falling while she waits. Empirical clomiphene and unstimulated IUI have repeatedly failed to beat expectant management in this group and are not routine.

When to refer to ART, and the SA access reality. Refer for assisted reproduction when the cause mandates it (severe tubal disease, severe male factor, failed ovulation induction, advancing age with low reserve) or when first-line treatment has not succeeded. The honesty required of a SA consultant is that the referral and the access are different things. South Africa has three public-sector ART units — Steve Biko Academic Hospital in Pretoria, and Groote Schuur and Tygerberg in Cape Town — meeting an estimated 10–13% of the need, so that patients in seven of nine provinces must travel hundreds of kilometres to reach a state IVF cycle. Even where a subsidised cycle is available, ovarian-stimulation medications are not state-funded and must be paid for out of pocket (commonly R8,000–R15,000 per fresh cycle), which constitutes a catastrophic cost for a large fraction of patients and produces dropout rates as high as 70% between diagnosis and treatment among the poorest. The defensible plan therefore states the evidence-based route and names the deliverable one: maximise the cheap, high-impact interventions that need no IVF unit (weight and lifestyle, smoking cessation, correctly directed letrozole with monitoring, treating a hydrosalpinx, stopping a man's testosterone), counsel realistically about access and cost, and — crucially — do not let a woman age out of her own fertility on a public waiting list without telling her plainly that time is the variable no treatment recovers.

Guidelines compared

The substantive divergences worth holding are: the letrozole-first shift (the 2023 international PCOS guideline made explicit and the WHO endorsed by adding letrozole to its Essential Medicines List for ovulation induction in 2023 — a change directly relevant to SA formulary advocacy); the consistent NICE position against empirical oral agents and unstimulated IUI in unexplained subfertility, which often surprises candidates who expect "try clomiphene first"; and the recognition that NICE CG156 has been replaced by NG257 (2026) — citing the withdrawn number is a dated error.

The evidence & the controversy

The defining shift of the last decade is the displacement of clomiphene by letrozole as first-line ovulation induction in PMOS, and it rests on a clean head-to-head. Before it, clomiphene was unquestioned first-line — itself established by the PPCOS-I trial, which had shown clomiphene clearly superior to metformin for live birth and demoted metformin to an adjunct. Letrozole then beat clomiphene on its own terms in a large double-blind trial, with a higher live-birth and ovulation rate and a more mono-follicular (and so safer) response. The old objection — a theoretical teratogenicity signal raised by an early case series — has not been borne out by the trial and subsequent data, and the international guideline and WHO have moved letrozole into first place. The practical SA controversy is therefore one of supply and habit rather than evidence: clomiphene is everywhere and free, letrozole's availability in the public sector is uneven, and changing entrenched prescribing requires formulary and supply work, not more trials.

A second live argument is how aggressively to treat unexplained subfertility, and with what. The instinct to "do something" — empirical clomiphene, or stimulated intrauterine insemination — runs into the repeated finding that oral agents and unstimulated IUI do not beat simply timing intercourse, and that the gain from intervention comes mainly from IVF. The unexplained-infertility ovarian-stimulation evidence sharpens the cost/safety trade-off rather than resolving it: gonadotrophin stimulation produces the highest pregnancy rate but an unacceptable multiple-pregnancy burden, including triplets, while milder stimulation is safer but less effective. The defensible position is restraint then escalation — expectant management in the younger woman, IVF as the effective intervention, and a clear-eyed refusal to expose couples to the multiple-pregnancy risk of unmonitored or gonadotrophin-driven stimulation for marginal benefit.

A third thread is the medicalisation of declining fertility and the limits of testing. AMH has become a consumer product — marketed direct to women as a "fertility test" — and the consultant correction is firm: AMH and antral follicle count measure ovarian reserve (the response to stimulation), not the chance of natural conception and not oocyte quality, and a low value must never be weaponised to frighten a woman or to refuse her treatment. Set against this is the genuine and under-communicated biology of age-related decline, where the discourse runs the other way — social and elective oocyte cryopreservation is now openly debated, and the honest message a consultant owes a 38-year-old is that her age, not her AMH, is the dominant prognostic variable and that delay is itself a harm. Holding both truths at once — testing over-promises, but biology is real — is the mature position.

Finally, subfertility in South Africa carries an equity and rights dimension that the international literature underweights. Infertility is recognised by the WHO as a disease and access to care as a reproductive right, yet SA public ART meets barely a tenth of need, medications are unfunded, and the burden falls hardest on the poor and on women, who in many communities bear the social stigma of a couple's childlessness regardless of the actual cause — which is itself an argument for diagnosing the male factor early and visibly, rather than subjecting the woman to a cascade of tests for a problem that is half the time her partner's. Naming that inequity, and the cheap interventions that partly mitigate it, is part of a complete answer.

Landmark trials & key evidence

A short piece of arithmetic makes the letrozole case concrete. Moving cumulative live birth from 19.1% to 27.5% is an absolute increase of 8.4 percentage points (ARR ≈ 0.084), so the number needed to treat is ≈ 1/0.084 ≈ 12 — about one extra live birth for every twelve women given letrozole instead of clomiphene, from a drug of similar cost and a more favourable single-follicle safety profile. That yield, from a cheap and deliverable intervention, is precisely why the switch matters more in a resource-limited service than in a well-funded IVF programme: it buys live births without an ART unit.

Worked viva — structuring the subfertile couple

A representative stem: "a 33-year-old woman and her 36-year-old partner present to a regional clinic after 18 months trying to conceive; her cycles are 40–60 days apart, her BMI is 34, and she has acne and hirsutism." A high-scoring answer runs:

1. Frame it as a couple and a syndrome. "This is subfertility with a clear pointer to an ovulatory cause — the oligomenorrhoea, raised BMI and hyperandrogenism describe polycystic/metabolic ovarian syndrome — but I will not anchor on her. I assess both partners in parallel, and a semen analysis on her partner is one of my first investigations."
2. Investigate by link. "I confirm her ovulatory status with a mid-luteal progesterone timed to her own long cycle, check prolactin and TSH because her cycles are irregular, and arrange a pelvic ultrasound. I assess tubal patency — given no history of PID or pelvic surgery, an outpatient HSG or HyCoSy is appropriate first. I send HIV serology on both partners as part of preconception care."
3. Start the free, high-value interventions immediately. "Before any drug: 5 mg folic acid, rubella check, smoking cessation, and a structured weight-loss and lifestyle programme — in PMOS this is genuinely first-line and even modest weight loss can restore ovulation."
4. Choose ovulation induction by cause and by evidence. "If she remains anovulatory, letrozole is my first-line agent — it gives higher live-birth and ovulation rates than clomiphene and a safer, predominantly single-follicle response, with a number needed to treat of about twelve versus clomiphene for an extra live birth. I would add metformin only as an adjunct for its metabolic effect, never as the stand-alone route to a pregnancy, and I monitor cycles to avoid multiple pregnancy."
5. Build in the SA reality and the time axis. "I counsel that if first-line induction fails the next step is assisted reproduction, that public ART access is severely constrained and stimulation drugs are not state-funded, and that her age gives us a window we should not waste — so I escalate rather than drift."
6. Close the loop on her partner and on equity. "If the semen analysis is abnormal I repeat it after three months and act on the cause; severe male factor reroutes the couple to ICSI. Naming and treating the male factor early also protects her from carrying the social and diagnostic burden of a problem that is, half the time, not hers."

Exam traps & red flags

• Treating subfertility as the woman's problem. Male factor is a sole or contributing cause in roughly half of couples; sending a woman for a tubal work-up before her partner has produced a semen analysis is the commonest and least defensible error. The semen analysis is cheap, early and informative — never defer it.
• A mis-timed mid-luteal progesterone. It must be drawn ~7 days before the next period, timed to the individual cycle. A reflex "day 21" in a 35-day cycle is taken too early and reads falsely anovulatory.
• Metformin as a stand-alone fertility treatment. PPCOS-I showed metformin alone gives a live-birth rate far below clomiphene; offering it as the route to a baby in anovulatory PMOS is undertreatment. Letrozole is first-line; metformin is adjunctive.
• An anti-oestrogen for hypothalamic (Group I) anovulation. Clomiphene and letrozole act by interrupting oestrogen negative feedback; in a low-oestrogen hypogonadotrophic woman there is no feedback to interrupt and they fail. Group I needs gonadotrophins/pulsatile GnRH (and weight restoration first).
• Ovulation induction without monitoring. Gonadotrophins without ultrasound/oestradiol monitoring risk OHSS and high-order multiple pregnancy; unmonitored stimulation is how a clinic produces triplets.
• Not treating a hydrosalpinx before IVF. Hydrosalpinx fluid roughly halves IVF success; salpingectomy or tubal occlusion first is mandatory, not optional.
• Missing exogenous androgen use in the male. Testosterone or anabolic steroids suppress spermatogenesis and can cause azoospermia; this is reversible and easily missed if not specifically asked about.
• Weaponising a low AMH. AMH predicts response to stimulation, not natural conception or oocyte quality. Using it to tell a woman she cannot conceive, or to deny treatment, is wrong — it informs urgency and dose, not the verdict.
• Letting a woman age out on a waiting list. Age is the dominant prognostic variable; the SA access bottleneck means a candidate must counsel honestly about time, expedite the cheap effective steps, and not allow delay to become the untreated cause.
• Citing the withdrawn NICE number. NICE CG156 has been replaced by NG257 (2026); quoting CG156 as current is a dated error.

Evidence anchors

• Legro et al. — Clomiphene, metformin, or both for infertility in PCOS (PPCOS-I), N Engl J Med 2007
• Legro et al. — Letrozole versus clomiphene for infertility in PCOS, N Engl J Med 2014
• Diamond et al. — Letrozole, gonadotropin, or clomiphene for unexplained infertility (AMIGOS), N Engl J Med 2015
• Teede et al. / International PCOS Network — 2023 International Evidence-based Guideline for the Assessment and Management of PCOS, Hum Reprod 2023
• NICE NG257 — Fertility problems: assessment and treatment (2026; replaces CG156)
• WHO Laboratory Manual for the Examination and Processing of Human Semen, 6th edition (2021) — lower reference limits: volume ≥1.4 mL, concentration ≥16 million/mL, total sperm number ≥39 million, total motility ≥42%, progressive motility ≥30%, vitality ≥54%, normal morphology ≥4%.
• South Africa — public-sector assisted reproduction is provided by three academic units (Steve Biko Academic Hospital, Groote Schuur, Tygerberg), meeting an estimated 10–13% of need, with ovarian-stimulation medications not state-funded.