In one line
Pregnancy hides tuberculosis and pneumonia while making both more dangerous, and the highest-risk period is not the pregnancy but the first months after delivery. Screen at every visit, treat with the standard regimens because they are safe, and remember that a woman who is breathless on the postnatal ward has less reserve than she looks like she has.
Where this sits. The organisms and the host response underneath this chapter are taught in the Primary chapter on microorganisms in pregnancy and the puerperium. The HIV layer that changes the presentation and the urgency is the Finals chapter on HIV in pregnancy, and the critically ill mother is obstetric critical care.
The infection that kills mothers without being written down
Non-pregnancy-related infection sits among the leading causes of maternal death in South Africa, and the largest part of it is respiratory. Globally, tuberculosis contributes to 6 to 15% of all maternal mortality. About 200 000 women develop tuberculosis in pregnancy or the puerperium each year, split as roughly 151 000 during pregnancy and 49 000 afterwards.
Look at how that 200 000 divides, because the split is stranger than it first appears. Pregnancy lasts nine months and the postpartum period, as counted here, lasts a few. Yet nearly a quarter of the cases fall in the shorter window. Expressed as risk rather than counts, the incidence risk ratio is 1.4 in pregnancy (95% CI 1.1 to 1.7) and 1.9 postpartum (95% CI 1.5 to 2.5).
So the dangerous period is after she has been discharged, when nobody is looking. That single observation reorganises how you think about the topic, and the reason for it explains almost everything else in this chapter.
Why a pregnancy tolerates a fetus, and what that costs her lungs
A fetus is, immunologically, half foreign tissue that the mother does not reject. Achieving that requires a real shift, not a vague dampening: maternal immunity moves away from cell-mediated, Th1-dominant responses and toward humoral, Th2-dominant ones, with expanded regulatory T-cell populations at the maternal-fetal interface.
That trade is excellent for the pregnancy. It is poor for anything the body controls with cell-mediated immunity.
Work out which infections that puts at risk and the list writes itself: the intracellular organisms. Tuberculosis is the obvious one, and the same logic covers listeria, and the fungal and viral infections that trouble immunosuppressed patients generally. This is the identical mechanism that makes pregnancy hard for a woman living with HIV, because her cell-mediated arm is already depleted, and the two deficits add (HIV in pregnancy).
Now run the mechanism forward past delivery. The tolerance was required for as long as there was a fetus. Once the placenta is delivered, that suppression lifts, and it lifts quickly.
A restored Th1 response meeting an infection that was quietly established during the tolerant months produces exactly what an immunologist would predict: inflammation appearing or worsening at the moment immunity returns. That is why the postpartum risk ratio is higher than the antenatal one, and it is the same phenomenon as immune reconstitution after starting antiretroviral therapy.
The clinical consequence is uncomfortable, because it inverts normal obstetric practice. The period when she most needs someone looking for tuberculosis is the period after we have discharged her.
Why it gets missed, which is not carelessness
Set the symptoms of early tuberculosis beside the symptoms of a normal third trimester.
| Tuberculosis | Also a normal pregnancy |
|---|---|
| Fatigue | Yes |
| Breathlessness | Yes, from splinting and raised oxygen demand |
| Poor appetite | Yes |
| Failure to gain weight | Masked, because she gains weight regardless |
| Night sweats | Common in normal pregnancy |
| Cough | The only one that is not |
Every feature except the cough is expected in a well pregnant woman, and weight loss, the sign that would ordinarily alarm you, is concealed by the fetus and the uterus growing whether or not she is wasting. Pregnancy does not merely fail to reveal tuberculosis. It actively camouflages it.
Which is precisely why the recommendation is not to screen the symptomatic, but to screen everybody. WHO recommends that all pregnant women in high tuberculosis-burden countries and all pregnant women living with HIV are screened at every antenatal visit.
South Africa goes one step further, and the step is the important one. If every symptom of tuberculosis is camouflaged by pregnancy, a symptom screen alone will keep missing the disease however often you repeat it. So the National Consolidated Guidelines separate what happens at booking from what happens afterwards:
| Who, and when | What the guideline requires |
|---|---|
| Living with HIV, first or booking antenatal visit — newly diagnosed, known and on treatment, or known and not yet on treatment | A TB-NAAT for all of them, regardless of symptoms, plus any other investigation indicated, such as urine LAM. Start or continue TLD |
| Living with HIV, any follow-up antenatal or postnatal visit | Symptom screen; send a TB-NAAT if one or more symptoms are present |
| Not living with HIV, any visit | Symptom screen; send a TB-NAAT if one or more symptoms are present |
The booking test is a systematic screen rather than a symptomatic one, and that is the whole point: it is the only step in the pathway that does not depend on symptoms the pregnancy is hiding. Every pregnant and breastfeeding woman is then evaluated for tuberculosis at every visit thereafter, through the antenatal period, at delivery, and for as long as she breastfeeds.
Two rules attach to the result. A NAAT that is negative while the symptoms persist is not a rule-out: investigate further, and add urine LAM if her CD4 is below 200 with symptoms, or if she has been admitted for any medical condition, whatever her CD4. A NAAT that is negative in a woman with no symptoms returns her to symptom screening at every subsequent visit, at delivery, and through breastfeeding.
Alongside this sits a national danger-sign list, which is worth memorising because it is the point at which she stops being a clinic problem: a respiratory rate of 30 or more, a heart rate of 120 or more, unable to walk unaided, breathless at rest or while talking, coughing up a tablespoon or more of fresh blood, or drowsy, confused, agitated or losing consciousness. Any of these means discuss with a doctor or refer.
Making the diagnosis when you cannot rely on the symptoms
Sputum is still the test that matters. Send it for Xpert, which returns both the organism and rifampicin resistance, and add culture, because a woman starting a six-month regimen deserves a susceptibility result. Pregnancy changes none of this.
What pregnancy does change is the willingness to order a chest radiograph, and the hesitation causes real harm. A chest film with abdominal shielding delivers a fetal dose that is negligible: far below any threshold associated with fetal effects, and vanishingly small beside the consequences of undiagnosed tuberculosis in a mother and her newborn. Radiation is the wrong thing to be frightened of here. Where the clinical suspicion exists, take the film.
Two further habits pay off in this population. First, remember that HIV alters the picture: lower-zone and disseminated disease, less cavitation, and a higher yield from extrapulmonary sampling. Second, remember that a substantial minority of tuberculosis in pregnancy is extrapulmonary, so lymph node aspiration, pleural fluid, urine lipoarabinomannan testing in advanced HIV, and an endometrial biopsy where genital tuberculosis is suspected all have a place. Establish her HIV status and, if she is positive, her CD4 count and viral load, because that changes both the likely site and the urgency.