Infertility causes much distress, particularly in those districts of Africa where as many as 30 per cent of families are childless. Primary Mother Care describes what health centres can do for it (2.3), and if they have done their job properly, there is little more that you can do in a district hospital. You may decide that you have other priorities, and that infertility is so unrewarding that you are not going to try to treat it. If you do decide to do so, make it part of your family planning activities and promote an integrated ''fertility service', which is concerned with both too much and too little fertility.

Typically, about 60 per cent of infertility is caused by the adhesions that follow PID (pelvic inflammatory disease). Repairing tubes that PID has blocked is an expert's task, and even then the success rate is low, so you may decide that there is little point in investigating or referring these patients. If you decide to do so, you can: (1) Do a hysterosalpingogram which will tell you where a block is. Many district hospitals find these too expensive and time-consuming. (2) Insufflate the tubes, which is cheaper, but gives less reliable information. Also, the instrument often leaks, and you can make mistakes. (3) Do a laparoscopy which again is expensive and time-consuming.

Some couples will be childless because the wife is not ovulating. You can find this out by: (1) Taking her history. If she has regular cycles, she is almost certainly ovulating. Failure to ovulate is typically associated with irregular cycles or amenorrhoea. (2) Dilating and curetting her during the second half of the cycle, and sending the scrapings for histology. (3) Asking her to keep a temperature chart, as described in Primary Mother Care. She may be sufficiently intelligent and motivated to do this, particularly if she is a member of the hospital staff or a teacher. A regular 0.5[de]C temperature rise 14 days before the start of menstruation is good evidence that she is ovulating. Lack of this rise, especially if her periods are irregular or scanty, is strong evidence that she is not doing so. (4) If you are fortunate enough to have a laparoscope or laprocator, you can examine her ovaries (15.4), to see if they are scarred, showing that she has ovulated. At the same time you can test the patency of her tubes, by injecting a blue dye though her cervix, and seeing if it appears in her peritoneal cavity.

If she is not ovulating, refer her. If you cannot refer her, you may be justified in inducing ovulation with clomiphene, which is comparatively safe, if she can afford it. One contributor considers it has no place in this manual. Only an expert should give her bromocriptine or the gonadotrophins. Use a temperature chart to monitor your success.

A very occasional patient is sterile as the result of tuberculous endometritis; sterility is its most common presentation. Treating tuberculosis is not difficult, but it is unlikely to make her fertile[md]she has about an 8 per cent chance of conception, but only a 2 per cent chance of a live child. She also runs an increased chance of an ectopic pregnancy.

INFERTILITY The health centre staff should have taken a history, examined both partners, and sent the husband's semen for examination. If their workup is incomplete, complete it (M 2.3).

HUSBAND. His seminal fluid must be examined within 2 hours. It is normal if: it has a volume of 2 ml to 6 ml, it is liquid after 30 minutes, it has 60% of motile sperms, and if it has 20 million sperms or more per ml, less than 15% of them being abnormal.

If he has a low sperm-count, suggest they abstain from sex until the 12th to 14th day of the cycle, to increase his sperm- count at the time of ovulation. If he has pus cells in his ejaculate, treat his infection.

WIFE. Curette her late in the second half of her cycle. Either: (1) Use a microcurette of the Novak or similar type, as an outpatient. Or, (2) dilate and curette her under anaesthesia. Put half the curettings into formol saline for histology, and the other half into a sterile bottle for culture and, if possible, for guinea pig inoculation for tuberculosis. Indicate on the request form for histology that you want to know if she is ovulating. Remember not to overload a pathology service which is overloaded already.

HYSTEROSALPINGOGRAM [em]CAUTION ! (1) Before you start, do a pelvic examination to exclude pregnancy and active pelvic infection. (2) Do a hysterosalpingogram within 10 days of the patient's last period, and not in the premenstrual or active menstrual phases of the cycle. (2) Wear a lead apron.

EQUIPMENT. An intracervical cannula, preferably of the Leech Wilkinson screw-in type. A Miller cannula causes less trauma to the cervix, but does not make such a good seal with it, unless she is under general anaesthesia, which allows you to use more force. A 20 ml syringe filled with a water-soluble contrast medium such as ''Urografin'. Avoid oily contrast media.

ANAESTHESIA. No anaesthesia is usually needed, but if she is very anxious, premedicate her with diazepam 30 minutes beforehand and do a cervical block.

METHOD. You can do a salpingogram in the X-ray department. If possible, screen her during injection of the dye. If not, lie her on her back on the X-ray table with her hips and knees flexed, and the plate under her pelvis. A tube[nd]plate distance of a metre is satisfactory and no grid is needed.

Insert Cusco's speculum and clean her cervix with cetrimide. Hold her cervix gently with a single-toothed tenaculum, lightly closed to the first ratchet; she should feel little pain.

Expel all air from the syringe and cannula, inject 20 ml of contrast medium firmly into her cervix, and take a film. If she has a cornual block, 20 ml will not go in. If possible, take a second film some hours later. The dye should have spread into her peritoneal cavity. If it remains loculated, this suggests adhesions and impaired fertility.

TUBAL INSUFFLATION INDICATIONS. Although theoretically simple, false results are not uncommon. If insufflation is the only method of investigation you have, this suggests that expert tubal surgery is unlikely to be available, which should make you question the value of insufflation.

EQUIPMENT. An insufflator, a source of a carbon dioxide, and preferably a device for recording the pressure graphically. If necessary you can use air.

METHOD. Give her a general anaesthetic and put her into the lithotomy position. Insert a Sims' speculum. Insert the insufflator into her cervix and fill her vaginal canal with fluid, so that the cannula is submerged, and you can see if there is a leak. Discharge some carbon dioxide, and listen over her lower abdomen for the sound of it bubbling out of her tubes. Measure the rise in pressure of CO[,2] before free flow occurs. If her tubes are patent, pressure will peak, and flow occur below 40 mm Hg. If they are blocked it may rise as high as 160 mm.

If you are using air, use a maximum of 250 ml, and don't go above 100 mm Hg, because of the risk of air embolism.

LAPAROSCOPY [s7]AND DYE INJECTION Under general anaesthesia insert a Miller cannula into her cervix. Insert a laparoscope, as for tubal ligation (15.4), and tilt her head down until you see a good view of her pelvis. If you cannot see clearly, insert the Verres needle (with the valve closed) in the midline suprapubically, and use this to manipulate her tubes. Inject 10[nd]20 ml of methylene blue dye diluted 1:10 in sterile water, and look for dye spilling from the ends of her tubes.

Normal tubes Her fimbriae look healthy and the dye spills through easily. It may spill on one side only, but if both tubes look healthy, they are probably both patent.

Cornual block No dye enters her tubes. As your assistant injects the dye, the region of their insertion into her uterus blanches slightly.

Fimbrial block Her tubes are often distended; their fimbriae are clubbed and sealed over the ostia, and may be adherent to her ovaries. As you inject the dye, the thin walls of her tubes allow you to see it entering them. Usually, no dye spills out. Sometimes the fimbrial block is partial, so that only a little spills.

ANOVULATORY INFERTILITY CLOMIPHENE is only indicated for anovulatory infertility. If you cannot refer her and can afford it, consider giving her clomiphene. Warn her of the increased incidence of multiple pregnancies. Unless she is also receiving gonadotrophins, this risk is small. Give her 50 mg daily from the second to the sixth day of her menstrual cycle, or at any time if cycles have stopped, to a maximum of 6 courses. Monitor ovulation with a temperature chart. If she does not ovulate increase the dose by 50 mg amounts each month, to a maximum of 200 mg daily for 5 days.

CAUTION ! (1) Only give clomiphene to adequately investigated patients with patent tubes and fertile husbands. Don't use it randomly on all infertile patients. (2) It is contraindicated in hepatic disease, ovarian cysts, pregnancy, and abnormal uterine bleeding. (3) Side-effects include visual disturbances, ovarian hyperstimulation (very rare unless it is used with gonadotrophins; if it occurs stop treatment), hot flushes, nausea, vomiting, depression, insomnia, breast tenderness, weight gain, rashes, dizziness and hair loss. It may make her ovaries tender, and simulate an acute abdomen.

Mr Printer. Please take in the ''Not in series figure' ''Fertile cycles'.