Sepsis and septic shock in pregnancy

The pregnant patient is particularly susceptible to sepsis, owing to their borderline immune function. The typical pathogens of "puerpureal" or "childbed fever" were group A streptococci such as S.pyogenes. Toxic shock syndrome is not an uncommon feature. The college has examined this issue in Question 3.1 from the first paper of 2014 and the identical Question 14 from the second paper of 2009. Each time, the focus was on the difficulties in actually making a diagnosis of sepsis, the major causes of sepsis in pregnant patients, common bugs and antibiotics which are contrainidicated.

Probably the best article on this topic was by Barton et al (2012), but to be honest most of the college answers to the abovementioned SAQs was sourced directly from Oh's Manual, page 688 of the 7th edition. Beyond these brief notes any further reading will be at the cost of revising more important material, and should therefore be viewed as indulgently recreational.

Difficulties in assessing sepsis in the pregnant patient

Question 3.1 from the first paper of 2014 and Question 14 from the second paper of 2009 asked about "difficulties associated with the diagnosis of sepsis during late pregnancy and labour." Upon close inspection, one can discern that the majority of these problems are equaly applicable to non-pregnant patients, as they are problems of the obsolete definition of sepsis and SIRS.

• Neutrophilia is normal in pregnancy
• Inflammatory markers are normally slightly elevated
• The patient is normally slightly tachycardic, and may be slightly hypotensive, with a decreased SVRI. The circulation is normally hyperdynamic. They appear to be in early stages of vasodilated shock at the best of times.
• Tachypnoea is normal in pregnancy and labour (chronic respiratory alkalosis)
• The whole process of labour tends to obscure the findings one would normally associate with sepsis.
• Then, suddenly, they collapse (as they tend to compensate right up until the last minute, and then drop everything, given their increased susceptibility to lipopolysaccharide)

The new Sepsis III definitions are somewhat better, as they call for organ dysfunction. Also, to make septic shock the recipe now calls for vasopressors and/or raised lactate, neither of which are features of normal pregnancy and labour. Which is good news: unlike SIRS and pregnancy, "Sepsis III" and pregnancy now appear to be distinct entities by criteria. One should never again be in a position where one is confused about which is which. 

On a serious note, the LITFL entry on this topic brings up several additional points which this author is not above stealing:

• Pneumonia may masquerade as the normal shortness of breath in pregnancy
• Abdominal pain due to visceral pathology may masquerade as contractions (or - more  confusingly - may co-present with labour)
• Appendicitis may be more difficult to diagnose, as the gravid uterus gets in the way of palpation.
• Nausea, vomiting, constipation - all of these may be routine for the pregnant woman, and a bowel obstruction may present disastrously late (eg. this case report by Khan et al, 2012).
• Abdominal pathology may be impossible to image, as everybody is going to freak out about the radiation exposure associated with a CT abdo/pelvis. This is in fact quite irrational, as the risks to a late-term foetus are quite small  (Chen et al, 2008). Instead, people should think about the foetal risk associated with remaining inside the septic abdomen of a critically ill mother whose source control is compromised by indecision about imaging. (In case anybody is interested, the dose from such a CT will total about 5 rads, which maxes out the recommended safe total foetal cumulative dose of 5-10 rads).
• Urinary sepsis may be asymptomatic (urinary frequency is pretty normal)

Typical sources of sepsis in the pregnant patient:

• Chorioamnionitis
• Urinary tract infections
• Pyelonephritis
• Endometritis
• Septic abortion

also...

• Episiotomy infections
• Necrotising fasciitis of the ceasarian wound
• Septic thrombophlebitis - especially pelvic vein thrombophlebitis
• Aspiration pneumonia

If you like comprehensive lists and grey boxes, you're going to love Box 1 from Barton et al (2012)

Acute pyelonephritis Retained products of conception Septic abortion Conservative management of placenta accreta or percreta Neglected chorioamnionitis or endomyometritis Uterine microabscess or necrotizing myometritis Gas gangrene Pelvic abscess Unrecognized or inadequately treated necrotizing fasciitis Abdominal incision Episiotomy Perineal laceration

Bacterial pneumonia Bacterial examples Staphylococcus Pneumococcus Mycoplasma Legionella Viral pneumonia Influenza H1N1 Herpes Varicella Intraperitoneal etiology (nonobstetric) Ruptured appendix or acute appendicitis Bowel infarction Acute cholecystitis Necrotizing pancreatitis

It is more appealing by virtue of the fact that surgical neglect and the consequences of "conservative management" are included in the list of differentials.

Risk factors for maternal sepsis

• Caesarean section is the single most important risk factor.
• Home birth in unhygienic conditions
• Low socioeconomic status
• Poor nutrition
• Primiparity
• Anaemia
• Prolonged rupture of membranes
• Prolonged labour
• Multiple vaginal examinations in labour (more than five)
• Multiple pregnancy
• Artificial reproductive techniques
• Obesity
• Obstetric manoeuvres eg. forceps delivery

The common pathogens encountered in pregnancy are:

Oh's Manual quotes the following bugs:

• Gram-negatives
• Streptococci (especially Group A, i.e. S.pyogenes)
• Bacteroides

To this, the Sanford Guide adds a few:

• Enterobacteriaceae
• Chlamydia trachomatis
• Ureaplasma urealyticum

(these are specific to chorioamnionitis and septic abortion)

About antibiotic use in pregnancy

Oh's Manual recommends ampicillin, gentamicin and clindamicin as an empiric therapy (with meropenem and vanc as a backup second tier). The Sanford guide agrees, replacing ampicillin with ceftriaxone, and suggesting that Tazocin could be used as a sole agent.

A full list of antibiotics which are contraindicated in pregnancy is also offered here for completeness (paraphrased from Mylonas, 2011 and Pilmis et al, 2015). It is not essential to memorise the entire list. Time-poor candidate should at least be able to recite the top five:

• vancomycin (renal toxicity)
• sulphonamides (neontatal jaundice)
• tetracycline (abnormal bone and teeth development)
• chloramphenicol (grey baby syndrome)
• Antifungals (fluconazole, itraconazole, ketoconazole, griseofulvin) - pretty much the only antifungal thought to be safe for use in pregnancy is amphotericin.

Antibiotics which are contraindicated in pregnancy or lactation, and Why

Contraindicated antibiotics	Why
Aminoglycosides in high doses	Increased uptake by neonatal kidney leads to increased nephrotoxicity (but apparently gentamicin is still relativey safe)
Streptomycin	8th cranial nerve damage
Sulfonamides	Kernicterus in the newborn due to displacement of bilirubin off albumin, particularly if used shortly before birth. The specific culprit is sulfamethoxazole. Trimpethoprim appears to be relatively safer.
Tetracyclines	Tetracyclines have nightmarish dental and bony effects.  In fact, these drugs are contraindicated from 16th week of gestation all the way until the 7th year of extrauterine life.
Quinolones	Quinolones cause birth defects (though it seems fluoroquinolones are safe, and it was really mainly nalidixic acid that was the culprit).
Rifampicin	Seems to be somewhat teratogenic, mainly in animal studies (spina bifida and impaired osteogenesis seem to be the major consequences)- but in humans one may overlook this if rifampicin is strongly indicated (eg. treatment of lifethreatening tuberculosis)
Fusidic acid	Like sulfonamides, causes displacement of bilirubin by competing with it for albumin binding
Chloramphenicol	This "Grey baby syndrome" is the consequence of disrupted mitochondrial function, when chloramphenicol metabolites interfere with the electron transport chain. Foetal failure to properly metabolise chloraphenicol by glucourinidation seems to be to blame.
Azole antifungals	Teratigenic and embryotoxic. The specific dangerous ones are ketoconazole  fluconazole and voriconazole, earning a D classification.
Echinocandins	Hard to discuss humans in the absence of real data, but in animals using normal treatment doses caspofungin and anidulafungin cause skeletal abnormalities, reduction of litter size, ossification and rib malformations.
Flucytosine	Embryotoxic and in fact abortificant in the first trimester, which is hardly surprising given that its main metabolite is 5-fluorouracil
Albendazole	Teratogenic and embryotoxic. If the pregnant lady has some sort of hideous helminthic parasitosis which cannot wait until after delivery, ivermectin is probably a safer alternative
Foscarnet	Seems to be teratogenic in animals, but if your CV infection is so severe and resistant that you've failed primary therapy, the chances are that your foetus is already significantly damaged by congenital CMV.

The college answer for Question 3.1 from the first paper of 2014 also lists nitrofurantoin, isoniazid and macrolides as drugs which are relatively contraindicated. This in fact, is not true:

• Nitrofurantoin seems safe and has been used extensively in pregnancy (David et al, 1995). 
• Macrolides likewise (Kueiyu et al, 2013) - even the new ones (Bar-Oz et al, 2008)
• All first line anti-TB drugs are reasonably safe. Isoniazid in fact was not associated with any major foetal malformations at up to 60 times the treatment dose (Ormerod et al, 2001). Exceptions are rifampicin, which may have a slight risk of  teratogenicity,  and streptomycin which causes nephrotoxicity and deafness.