Every infusion pump keeps its own event log: power on, programme set, rate, volume delivered, alarms, in its own memory with its own clock. Pump 0114 was in a sealed store cupboard for eleven weeks and its log has an infusion in the middle of that.
I am a clinical engineering technician. I look after about nine hundred infusion devices for a trust.
Why the log exists and how much weight it carries.
Every pump keeps an internal event log in non volatile memory. Power on and off, the programme entered, the rate, the volume infused, every alarm and every key press, with a timestamp from its own clock.
That log is downloaded when a device comes in for service and it is downloaded when there is an incident, and it is the evidence. If a patient received too much of something, the pump's log is what tells you what was set and what was delivered, and it is treated in an investigation the way a flight recorder is.
It cannot be edited. There is no user function to clear it and the service software will not write to it.
Pump 0114 came back from a ward in March, went through decontamination, passed its electrical safety test and its performance check, and went into the store as a spare.
The store is a cupboard in our workshop with a lock on it. It held about forty pumps.
It was taken out eleven weeks later, on a Tuesday, and issued to a ward, and as part of that it was connected for a routine log download.
There is an infusion in the log dated the fourteenth of May.
The fourteenth of May is week seven of the eleven.
The entry is a complete sequence and that is what makes it what it is. Power on. Programme entered as a volume over time. Started. Two occlusion alarms, cleared. Volume delivered four point two millilitres. Stopped. Power off.
Eleven minutes from end to end.
What is wrong with that, item by item.
The pump was in a locked cupboard. The key is on a board in the workshop and the workshop is card access and the log shows two entries that week, both mine, both in the afternoon.
The pump had no giving set in it. A set is a single use plastic assembly and you cannot deliver anything without one, and one had not been fitted since decontamination, and there was none fitted when I took it out.
There was nothing to deliver. No bag, no syringe, no fluid.
An occlusion alarm is a pressure measurement. The pump senses the pressure in the line rising against a blocked tube. With no set fitted, the sensor has nothing to press on and the pump will not even start: it tells you to load a set, and it will not accept a programme.
And it delivered four point two millilitres, which the pump measures by counting the revolutions of its own peristaltic mechanism.
What I did about it.
I reported it. This is a device that could go to a patient with a log in it that says it did something it did not do, and if it were ever the pump in an incident, that log would be read by people who trust logs.
It went to the medical device safety officer and to the manufacturer as a formal field report.
The manufacturer's response, over five months, is on file. They downloaded the log themselves, confirmed the entries are consistent with a genuine infusion, confirmed the sequence could not be generated by a software fault they know of, and stated that the log is not writable by any external means.
Their conclusion is that the device was used. Mine is that it was in a cupboard.
We agreed to differ in polite language and they replaced the pump under goodwill.
What happened to it.
The pump was quarantined and returned to the manufacturer. It has not come back and I did not ask for it back.
The trust's device safety officer recorded it and closed it, and the closing note says the device was withdrawn from service as a precaution and that no patient was involved.
That is entirely true.
What I do now.
The store cupboard has a log book on the door and every pump in or out is written in it by hand with a date and a name, which is belt and braces over the card access and the electronic asset system, and I put it there myself in June.
And every pump that leaves that store now gets its log downloaded on the way out as well as on the way in.
I have done about four hundred of those since June.
Six of them have an entry that was not there when they went in. All six are complete infusions, all between two and nine millilitres, all in the small hours, and one of them is on a pump that had been in that cupboard for four days with no set, no bag, and a battery I had taken out myself and left on the bench.