Case 26 All eggs in one basket

Jade Nelson is a 34-year-old woman who self-presents in the emergency department because over the last two days she has experienced dull chest discomfort and feels out of breath when going upstairs, both of which are new symptoms for her. She is normally in reasonable health and with no significant previous medical history and no family history of note. She is overweight and smokes 20 cigarettes a day and her only regular medication is the combined oral contraceptive pill that she has taken for the last two years.

What diagnoses are in your mind?

Mrs Nelson is seen by a CT1 in emergency medicine, Dr Fentiman, who finds her to be a little anxious, but he elicits no abnormalities on physical examination apart from a resting pulse rate of 90 and an oxygen saturation of 95%. The electrocardiogram confirms a sinus tachycardia and Mrs Nelson’s chest X-ray is normal. The junior doctor requests a full blood count, basic biochemistry, and a D-dimer to help decide if she needs further imaging to rule out a pulmonary embolus. He calculates her pre-test probability for pulmonary embolism to be low using the Wells criteria.

Just after the blood samples have been sent, the hospital laboratory computer system fails with the effect that no blood results are available online. The problem persists for some hours and the delay in obtaining results to investigations slows the flow of patients through the Emergency Department, making a busy day more fraught for staff and patients alike.

What should Dr Fentiman do?

After three hours, the laboratory phones through the results of Mrs Nelson’s investigations and these are written down by a nurse who happens to be by the phone but who does not know Mrs Nelson. Dr Fentiman reviews the results and sees that the D-dimer is 398 and he concludes that Mrs Nelson is very low risk for having had a pulmonary embolism and requires no further investigation. By this time, some five hours after arrival in hospital, Mrs Nelson is feeling slightly better yet very frustrated. She is sent home with a putative diagnosis of a ‘viral chest infection’.

The next day Dr Fentiman is again on duty and is called to attend to a patient in the resuscitation room who has suffered an out-of-hospital cardiac arrest. The resuscitation attempts fail and Dr Fentiman is distressed to find that the patient is Mrs Nelson who he sent home the day before.

A review of the previous day’s investigations (now available online again) show the D-dimer was in fact 3980, not 398 as recorded following the telephone conversation. A coroner’s post mortem examination later confirms that Mrs Nelson died following massive pulmonary embolism.

Expert opinion

This case highlights a number of problem areas:

Dr Fentiman appears not to have considered what the mechanism of increased breathlessness and chest discomfort might have been in this young woman. There was no evidence in the notes of a logical differential diagnosis or a clear management plan. He used the Wells score to calculate a low pre-test probability of pulmonary embolism (prevalence <4%) but arguably without a clear alternative and more likely diagnosis for Mrs Nelson’s symptoms he should have scored her at moderate risk (i.e. 21% prevalence).

The test result was communicated incorrectly – it is not clear whether the laboratory technician made an error or if the nurse who took the telephone call simply recorded the result incorrectly. Errors of this nature are worryingly common and in one study 3.5% of laboratory data communicated over the telephone were recorded incorrectly. The use of read-back to confirm that the information had been properly and accurately received abolished all of these errors.

The failure of the hospital results server during a busy time caused delay in processing patients and increased the pressure on staff. Delay also may have made patients more frustrated and could have made Mrs Nelson more likely to play down her problems and agree to go home. When systems do fail and staff have to revert to paper-based results and telephone communication it is all the more important that particular care is taken to ensure that results are correctly passed on.

When under pressure to process patients in a busy environment with the added strain of a system failure, it is easy to see how Mrs Nelson’s symptoms and signs might be explained away. She is anxious and so the heart rate of 90 is dismissed, she is a smoker and the oxygen saturations of 95% were similarly ignored. It is always important to ask: ‘what has changed to cause this otherwise well woman to attend the emergency department today’? In the absence of any infective symptoms, how likely is this presentation to be caused by a viral infection? How do viruses make you breathless (pneumonitis or unmasking of obstructive airways disease)?

Legal comment

A Trust-wide failure of laboratory results systems would trigger the Serious Incident Requiring Investigation (SIRI) process, given the number of patients whose safety could be adversely affected.

The coroner seeking the factual cause of death would need to consider the systems failures that entailed no test results for a number of hours, and jeopardizing patient safety. The coroner may consider a Rule 43 letter about not only the failure of the results system but also the poor practice that resulted in an incorrect test result being written down. It would help to reassure the coroner that a recurrence is unlikely if a completed clinical risk action plan/root cause analysis from the SIRI was provided to him, along with evidence of improved systems and back up contingencies to improve patient safety.

Simple changes in individual practice (for example, read-back) would reduce the danger of flawed clinical decision-making, improve patient safety and reduce litigation risk. The quality of documentation is crucial to the Trust’s ability to defend itself against litigation.

As it is, since there is a three-year limitation period from the date of death, following the coroner’s inquest, there could then be a civil action which would not be defensible.

Whilst it is important for the Trust to identify root causes and learning, it is equally important to ensure that support is offered to Dr Fentiman following the shock of recognizing Mrs Nelson during resuscitation. Effective risk management places an emphasis on openness rather than blame.

Stay updated, free articles. Join our Telegram channel

Join