An honest account from a junior doctor on moving from paper to electronic observation charts and why user testing should be done before rolling it out in hospitals.
Paper observation charts are now a thing of the past where I work. Gone are the days of charting your patients’ blood pressure and pulse in the tiniest of boxes. So small you could barely see the date and time of day at the top. Often, the chart looked as if it had been filled in by a spider with inky feet, sometimes it was sticky from medication that had been spilt on it (or sometimes worse). It would be passed from one clinician to another, a little ragged round the edges. Nurses had to remember when to do the next set of observations according to the National Early Warning Score (NEWS). As for auditing observations to ensure we were adhering to national guidance for the whole hospital… forget it.
We had been use to this for years. But now we have a new chart in town… e-obs. This is going to solve all our problems.
At the click of a button you have a clean, legible, fully completed observation chart. Each patient would have followed the NEWS escalation as the ‘electronic system’ would remind the nurse to complete the next set of observations at the correct time. Auditing would be a few clicks away. How many patients are scoring 5 or more? Who and where are the sickest patients? Which wards are not adhering to national policy? It is all there.
This is a terminal case of ‘work as imagined’.
Firstly, lets clear this up. Just because a patient is scoring a NEWS score of over 5 does not mean they are the sickest of patients. Many patients who are deteriorating, especially the younger population, score lower than 5. Patients in acute kidney injury often do not score at all but may require a trip to the intensive care unit. Do not be fooled by the NEWS score. NEWS is but a number. We must look holistically at our patients and not rely on looking at just numbers.
I would like to share something that happened the other day that highlights some of the pitfalls of using an electronic observation system.
I am a junior doctor on an elderly care ward. One of my patients became acutely unwell at 10pm on a Sunday evening. He couldn’t breathe, his NEWS was 9, he looked and sounded awful. I thought he was going to die. The medial emergency team came. They gave him suction, the chest physiotherapist came, they changed his antibiotics. He got a little better. His NEWS went down to 4.
How did this happen?
Surely, he didn’t suddenly get this unwell. He was doing well the day before. I looked at his observations. They were documented beautifully on the screen. Very clear.
However, he hadn’t had his obs taken for 12 hours despite his last NEWS score was 3 (this means obs need to be taken again 4–6 hours later). Why didn’t the electronic system alert the nurses to take the obs? This is a forceable function of the system? This is why we changed to an electronic system in the first place… to prevent this type of harm from happening.
So, what happened?
The patient had been scoring 0 for the last few days. This means that obs can be taken every 12 hours. The patient then scored 3. His oxygen saturations had dropped. As he was ‘stable’ the nurse then changed the profile of when the next set of observations were taken. Instead of the default setting of 4–6 hourly, they had set it for 12 hourly again. This is against national guidance.
Profile changes are taught to the nurses and doctors when being inducted to the e-obs system. This is important to know especially if the patient is dying, off the ward or having a blood transfusion, post op etc... This means that the patient will get their observations taken at the right time depending on what is going on.
Instead, the nurse had changed the profile so that the patient received less monitoring. What was the reason for this? Was it because he had been so stable before, that they thought he didn’t require more frequent observations? Was it due to ward pressures – they didn’t have time to do that frequency of obs?
What ever the reason. Its against national guidance and this ‘safe system’ has allowed us to do so.
There is also another problem at play here. Take another example.
A patient is receiving 12 hourly observations. They are stable. What happens when the patient may ‘look unwell’ or they complain of pain or breathlessness? You take another set of observations. The trouble is. They are not due. The system won’t let you ‘log them’.
Not only is this frustrating, it also takes away intuition and assessing your patient from the bedside. We cannot be complacent. Looking at numbers on a screen is not an indicator on how well your patient is.
Paperless, automated systems are brilliant. They will revolutionise healthcare, it will make care safer. We have to be mindful that these are early stages. There will be problems along the way.
I just wish that there was some user testing before they rolled e-obs out. Healthcare staff will take short cuts, will do unexpected things, won't always realise these consequences. Yes, it would have cost money, it would have taken time but, if they had user tested this with real staff, perhaps this man may not have suffered?