Clinical Case 136: the Diagnosis and Discussion
Last week I posted Clinical Case 136: A Maternity Mystery
If you have not had a read – then please go back and ponder the scenario before reading the rest of this post.
If you have been following along on the comments or social media then you will know that this is a tragic case.
Kelli became increasingly hypertensive, tachycardia, tachypnoeic and hypoxic over the initial hours after giving birth.
She suffered a cardiorespiratory collapse about 6 hours postpartum and died after a prolonged resuscitation.
The clinical picture in the peri-arrest period and after intubation was that of severe type 1 respiratory failure secondary to cardiogenic pulmonary oedema. This did not respond to non-invasive ventilation or vasopressors, catecholamine support or diuresis.
The diagnosis was not clear until autopsy. Though it was suspected after a careful review of the chart and all the data that we usually do not have time to contemplate in the heat of the Resus.
Tom and Leah were both right in the comments after the initial post. Kelli had an undiagnosed adrenal phaeochromocytoma.

Radiopaedia.org. From the case rID: 11765
This is an impossible diagnosis to make in the labour ward or in the ensuing resuscitation. Only with the benefit of retrospect can one see how it all fits together. My reason for sharing this case is not to see how smart you all are, or to scare you with horrific & undiagnosable demons. I believe that this case contains a number of very important lessons. It illustrates how we can do better as clinicians working in teams with shared ideas and respect for one another’s strengths and weaknesses. The case also demonstrates the folly of “risk management” when it comes to our practice. Obstetric practice is especially demonstrative of our inability to predict risk with alarming regularity and disastrous consequences.
This is an incredibly rare scenario. It is estimated to occur in 1 in 54,000 pregnancies. It is basically clinically indistinguishable from pre-eclampsia in the second half of pregnancy. Hypertension and other symptoms are more distinctive in the earlier stages of gestation. Without a strong clinical suspicion it would be very difficult to arrive at this diagnosis unless it is discovered incidentally on imaging or it the vasoactive symptoms are witnessed but a trained eye. Unfortunately the symptoms are transient and sometimes subtle.
It is estimated that there is about a 50% maternal and foetal mortality if undiagnosed during pregnancy with most succumbing in the peri-partum period.
If you want to read more about the presentation and diagnosis of Pheochromocytoma in Pregnancy then have a read through this summary in European Journ Endocrinology, 2012 by Lender et al. If you know about it – then you may think about it. However, I suspect most of us have a vague recall of this diagnosis from undergrad physiology and pathology exam papers. It is rare to see in practice.
So lets go back and review the case – with the optically flawless retrospectoscope.
If we review the only historical features that are prominent in the notes we get the following list:
- Panic-attacks
- Palpitations
- Epistaxis – review of the ED notes showed a BP of 180/100 during this episode
- Intermittent severe headaches, self-resolving
- There were no specific mentions of symptoms such as profuse sweating or flushing – although these are usually symptoms we need to ask about in order to unmask.
- Weight change, frequent urination, vague abdominal pains and constipation are so prevalent in pregnancy that they become useless to help in this scenario.
With the wisdom of hindsight – all the holes line up! The symptoms are reasonably classical if one managed to arrange all the pieces of the puzzle in the right order. Alas, there is always a but…
Each of the symptoms that might have been used to triangulate this esoteric diagnosis had been “referred and reviewed”. The Cardiologist saw the tachycardia, the Psychiatrist – the anxiety and the Neurologist had screened the headaches… it is a perfect storm of explanationitis.
I certainly do not believe that any single specialist would have been in a better position to clinch the diagnosis than another. This is the nature of specialism – or “single-organ doctors” as Dr Leeuwenburg sometimes calls them. There is a tendency to investigate symptoms and exclude causes that originate in the organ to which that doctor is devoted. Thinking outside the box (or heart, cranium or psyche..) doesn’t come naturally. This is really the lot of the generalist. In retrospect I think that Kelli would probably have been most likely to be diagnosed if she had seen a crusty old General Physician or very enthusiastic, experienced GP. For me this is one of the most important lessons to learn from this case:
- Generalists are masters of the bigger picture. We should absolutely take specialist advice and appreciate it. However, we need to be aware of the biases and limitations that specialism brings to the table. It is our role to bring all the pieces together either in the ED or clinic. We are the ones with knowledge of the patient’s environment and family dynamics. We understand the symptoms within the context of a life. So we need to be circumspect when we incorporate our partialist colleagues’ advice into our care.
Then there is of course imaging…
Review of the chart revealed an incidental finding of a 4.5 cm right adrenal mass on the 19 week anatomy scan done by the local sonographer. This was noted and a plan to investigate with CT in the post-partum was discussed with the reporting radiologist. This report was summarised as “normal foetal anatomy & growth. Incidental adrenal mass on the right.”
[Ed: I did not include this in the initial case notes as the information was not available on the day.]
I can completely understand how a busy clinician in an antenatal clinic can see this report and, in the absence of any clinical suspicion, put it to one side until after the baby is out. Ultrasounds in pregnancy throw up all sorts of oddities – maternal and foetal. Most are purely incidental and require no action. Indeed ignoring them can be perfectly anxiolytic for expectant mothers-to-be.
The next big lesson that I have taken away from this case is about being “empirically aggressive” with treatment in the face of an enigmatic, malignant processes. This is essentially the core tenant of RESUSCITATION: empirical treatment in the absence of a diagnosis. Playing the odds that doing something will have a benefit over doing nothing.
That may seem like an obvious concept. Most of us are perfectly happy to perform CPR or a jaw thrust without knowing what is “going on”. However there are many shades of grey. This case demonstrates perfectly the paralysis that can arise when we think we know what is happening and are wrong, or when we are perplexed and frozen by diagnostic indecision. Certainly, after dissecting this case backwards I am sure that the wrong moves were made as a result of early closure and fixation biases. The clinical evolution just did not fit pre-eclampsia, and yet that was the pathway followed.
So my take away lesson is this: always keep an open mind and be prepared to abandon your “beliefs” in favour of aggressive treatment based on empirical data – even if it doesn’t seem to make sense. The degree of “aggression” should match the virulence of the clinical picture. I have spent many days, hours and moments in my career fixing physiology whilst remaining in the “diagnostic dark” as a result of being a long way from a surgeon, CT, MRI or fancy biochemical test. This is not ideal medicine, but it works. We keep the patient alive until the clouds clear and the way forward becomes clear.
In the case presented the resuscitation was aimed purely at Obstetric issues, specifically treating eclampsia… however the picture was clinically identical to a sympathetic, crashing pulmonary oedema. In retrospect, it was of course an acute, decompensated cardiomyopathy in the face of a massive adrenergic crisis – maybe precipitated by the labour, anti-emetics or the physical process of pushing. My “automatic” treatment of this situation is to attempt to kill afterload and preload simultaneously with GTN and PEEP… who knows, it may have helped.
Thanks to all of you who have read and commented upon this case.
Hope these lessons come in handy for you one day.
Casey
Thanks
great case and awesome discussion we all have to learn about the whole case.
Thank you Casey. #FOAMED and #POCUS FOREVER¡¡¡¡¡
Wow. Thanks for sharing this fascinating and tragic case. Definitely a reminder of the pitfalls of dealing with multiple symptoms in isolation.
Thanks Casey.
Hey Casey – long time no see! Why the proteinuria though? (Sorry, you know nephrologists are obsessed with it!) I could attribute it to a hypertensive crisis except that she didn’t really have a hypertensive crisis at the time she was noted to have proteinuria….second, having literally just been to a talk on phaeochromocytomas this afternoon at the ASN meeting, did she have any genetic testing done? I have now learned that up to 40% of patients with phaeo have a detectable mutations, which might have implications for her baby and other first degree relatives…interesting and awful case in equal measures…hope you’re well!
Hi Rebecca
Proteinuria – maybe an unknown unknown. I imagine that there was an unwitnessed hypertensive period on the day which may have caused proteinuria?
You are the nephrologist – do you know?
C
HI Casey
Excellent case and review.
I am wondering if I can use your paragraph here for our introduction in the GP training at WAGPET? : Generalists are masters of the bigger picture. We should absolutely take specialist advice and appreciate it. However, we need to be aware of the biases and limitations that specialism brings to the table. It is our role to bring all the pieces together either in the ED or clinic. We are the ones with knowledge of the patient’s environment and family dynamics. We understand the symptoms within the context of a life. So we need to be circumspect when we incorporate our partialist colleagues’ advice into our care.
Thanks
Sure. Cheque is in the mail?
😉 of course!
Thanks for this very interesting case yet again Casey.
Thanks Casey. Very interesting case and insightful discussion. Obviously agree re generalists ( i hope to include big city EM at least a bit in that category).
I suppose one of the challenges is that many generalists also spend their days in Type 1/ pattern recognition thinking (to survive !) and this is exactly the sort of case that needed to trigger a step back into system 2.
On reflection do you think there was a point(s) where the ‘metacognitive flip’ could have happened for anyone involved?
Hi Victoria
Yes, there does need to be a moment where somebody / the team realise that “we are not in Kansas anymore”!!
That realisation is easy to make when we walk into a MET or have a severely injured patient delivered by ambulance.
The flipping moment is tougher to recognise when there is a gradual slide down the slope of sickness…. especially if the transition requires a change of speciality midway.(tribes)
I guess this is why we have ORC charts and MET teams, however this scenario played out in a time well before those coloured charts in a small hospital where the 2 or 3 clinicians involved would have also constituted the whole MET team – so hitting the help button may not have been useful?
Having said that I still see plenty of sick folk on the wards be missed through the “explanationism” – the scenarios where the wildly abnormal OBS are all explained away with benign reasons and nobody stands back to see the bigger picture / puts it all together.
Solutions are hard
I’d like to coin the term “not in Kansas anymore” (NIKA) which any team member can declare – this is a trigger to force a cognitive STOP and reframe / overview of the situation
You are the Sim guru – do you think it might help as a shared cognitive aide ?
Casey
I love the idea of NIKA… although the popular cultural connection required may be lost of some of our younger colleagues (or those with less Hollywood influenced upbringings…)
I’ve never thought sim (in the sense of team based scenarios) was much good for training diagnostic processes, and possibly even less so when cases require that reversion to system 2 thinking . In fact sim probably at its best training reflex responses to solid, common ‘patterns’
That said i’ve been peripherally involved in ‘forced error’ sims where the objective is to recognise that a pattern wasn’t right and to speak up when occurred. Not easy to simulate realistically, but there is probably a format waiting to be discovered that might be useful.
Probably this is back to re-inforcing need for ‘metacognitive moment’ when ‘something’s not right’ and having the communication skills and processes to involve multiple generalist and specialist perspectives synchronously to the problem – a Utopian ideal for my day job !