The twilight shift agrees with my circadian rhythm like no other. The hours are 1600-0000, which is 8 hours just like your regular 9-5. But OMG, it just works so much better.
Wake up whenever you feel ready, go to the gym, prepare some OSCE teaching for medical students, come in around 2pm, find a patient and go through the examination properly with no risk of being bleeped. Meet all the other teams between 1600 and 1700 to check for sick patients and that they have done all the jobs which you should not be doing e.g. deciding if this patient is for the LCP this evening or whether or not to re-dress the wound this evening.
Such a straightforward shift compared to the weekend shift, where we do a ward round of the entire surgical inpatient population. This ward round goes from 8am till 2pm. Then do the jobs from that, plus all the weekend bloods for the entire surgical inpatient population, whilst getting bleeped for the usual cannulas/TTOs/fluids etc. And then you have the sick patients.
Teach others to teach yourself
A few medical students hung around to do the twilight shift with me last week. It was definitely a different experience having to explain my actions as I did them, and it probably made me more thorough than normal. I try not to offload cannulas and bloods onto them, but when it is busy then I justify it on the grounds that we can go through more teaching the quicker the mundane jobs are done.
Last Thursday seemed like it was going to be a disappointing collection of piddly jobs, without much for the student to learn from. You never want a patient to be seriously sick, but you do want the opportunity to go through the ABC assessment of a real patient with the student. At around 10pm, when we were thinking of letting the student go home, we got called to see a patient who had pulled out his NG tube and cannula.
This patient was a man in his early 60s who was day 4 of acute severe pancreatitis, scoring 6 on the Glasgow criteria. He had been declared not suitable for HDU/ITU, and was not for resuscitation. I had been warned by the outgoing day team at 5pm that he was ‘impossible’ to cannulate. I have recently discovered this savior vein between the thumb and first finger, about 3cm distal to the anatomical snuffbox. This vein is remarkably well tethered, and thankfully ignored by other cannulators.
At 6pm, I decided to give it a go. Careful not to burst the vein by pausing as soon as I sight blood in a cannula to retreat the needle, I was treated to extensive flashback and a satisfying flush. I taped the cannula down with extra MicroPore tape, and walked away with the satisfaction of a job well done. I couldn’t wait to let the day team know I got it in first time. During this time, I conversed with the patient and he was well orientated.
ABC is always the right approach
For the same patient to now pull out everything a few hours later was significant. His sats were 92% on 2L of oxygen according to the saturation probe. This was where he had been for the past few days. However, his respiratory rate was 28, and his work of breathing was substantial, with use of accessory muscles and nasal flaring. He was barely able to talk in sentences, which was different. For the patient to need to breathe like that to maintain his sats was clearly abnormal.
I hunted where I had success before. The area was a bit swollen, but I could just about feel the ruins of the vein. The medical student had a different approach. She went straight for the right antecubital fossa, and found something there. I asked her to go for a pink cannula if she’s confident. She got it in. Boom.
During this time, we first fully appreciated his confusion and incredible work of breathing. I asked the student to go through an ABC approach.
His airway was patent. Now, although his O2 sats were where they were normally, he was clearly in respiratory distress. This meant O2 was necessary to reduce the work of breathing before he fatigues out. On examination of the chest, there were bilateral crackles.
Acute severe pancreatitis is a known risk factor for ARDS. ARDS occurs in about 10-20% of episodes of acute pancreatitis, and has a mortality of about 30-40%. It is strongly associated with pancreatic necrosis. This was my clinical impression at the time. I realised the patient was not a candidate for HDU/ITU, and ward based NIV does not happen in our hospital. This meant that the best ward based care would be oxygen. The fluid regime for this patient would also become tricky, as ARDS is treated with a fluid conservative approach, whereas pancreatitis is a state of ongoing third space losses.
The patient did not have any risk factors for hypercapnic respiratory failure, so we really should have gone full whack with the O2 at that stage.
Circulation wise, I decided to give a fluid challenge despite the bibasal crackles. He had worsening mental status and this was pancreatitis after all. His cap refill and urine output were fine, and his pulse was 112 from 100 at 5pm and his BP had dropped to 120/80 from 150/90 at 5pm. He had no history of cardiac problems, and I suspected a non cardiac etiology for the bibasal crackles (ARDS). I wanted to see if this made him better or worse, and it was only 250ml.
The response was a slight improvement in BP to 130/84 and pulse of 106. This suggested he was fluid deplete and could do with extra fluids.
We took an ABG on 2L of oxygen, which returned:
We put the O2 up to 5L / min via a facemask, and saw the saturation probe return values of around 94-96% which we were happy with.
I asked for capillary ketones, which came back as 1.4.
With our only IV line, I asked for an insulin sliding scale to get glucose under control, as the high glucose would dehydrate the patient further. I then informed the registrar, who would review the patient.
Bloods were taken, which showed LDH 1560, lactate 2.5, WCC 20.8 (similar to earlier), CRP 250 (similar to earlier) and unchanged eGFR of about 60. Corrected calcium was normal. LFTs were remarkably normal.
There was no baseline LDH level to compare to. The reason I took this was not just to reassess the severity according to the Glasgow criteria but to look for a rise, which may go with pancreas necrosis. This is an area of controversy, but it makes sense that the release of a ubiquitous intracellular enzyme is a decent marker for cell death somewhere.
The registar reviewed the patient later, and started antibiotics. The evidence for antibiotics in pancreatitis is not definitive, but imipenam (good pancreas penetration) may be used when there is strong clinical suspicion of infection and/or more than 30% pancreatic necrosis.
Until then, the British Society of Gastroenterology has recent guidance on pancreatitis. And I must be more vigilant to go back to ABC when there is the slightest suspicion, and not let anything come in the way of a full assessment.
Eidt 10/4/13 – I have recently learnt that most of the time hypoxia in pancreatitis is not caused by ARDS. Have a look at an SBA on this.