We did what most of you recommended: Reduce the calorie intake, maintain the infusions (they were already going at the maximum rate recommended by the BNF) and discuss the platelet scenario with the haematologist.

Reducing the calorie intake is recommended if electrolytes get deranged in refeeding syndrome, as reducing the insulin surges should help prevent the intracellular shift.

Although the electrolytes were low, one concept I have come to appreciate is that you can do more harm by correcting electrolytes too rapidly than leaving the original imbalance. For example, magnesium infusions cause flushing, hypotension and potentially respiratory paralysis. You would certainly want to spot these problems early and stop the infusion if necessary. Testing the patellar reflex is a quick way of screening for magnesium toxicity (reflex absent in toxicity). I would still err on correcting the electrolytes at the maximum recommended rate, even at these levels.

I was actually on another ward when I got a bleep about this patient. I was told she had fainted whilst on the toilet, passing a PV clot. When I returned, her BP was fine and I had requested a repeat ECG and FBC, U&E, Bone Profile and Coagulation. The ECG showed sinus tachy at 120. She felt a little dizzy. I told her to lie on the bed and gave Hartmann’s 500ml over 20 mins as a fluid challenge to see the response.

The consultant haematologist was at a golf club when I rang. I explained about her bleeding episodes, and that this was a presumed new platelet problem. I gave him the obs and the latest Hb (10.1, from 10.7 the day before). He was not concerned, and asked me to ring if the patient bled again. I have since looked at the guidelines and could not classify where exactly my patient would fit – can you?

The next morning (Sunday), the patient became very confused and paranoid. She insisted that we were holding her against her will. She refused all IV medication and would only take tablets.

I politely asked her if she would be happy to take some extra vitamins. I gave her an extra 200mg of thiamine and asked the nurses to avoid feeding her until she seems calmer, or until lunch, whichever is the earlier. Looking back, this may have been an alcohol withdrawal episode. I suspect she was drinking more than a bottle of vodka a week prior to admission. Delirium tremens is not all about Liliputian hallucinations. It is a hyperadrenergic states, with sweating, tachycardia and cardiovascular collapse. My patient had many of these signs. Whilst gradually realising this and writing up a chlordiazepoxide regime, I was tapped on the shoulder by a concerned nurse.

“She’s just made a 999 call that we are holding her against her will.”

30 minutes later, she self discharged.

This was probably an acute confusional state. She was so grateful for our care the night before. Medico-lego-ethico wise, were we right to just let her go?

(In case you are worried, I have liased with her GP, and found out she has visited him since and that her bloods have improved markedly).