Dr Beverley Ward and Dr Simon Watkin review MDU cases involving pulmonary embolus.

Medicine is often described as an art, not a science, and it can certainly feel like that when a patient presents with symptoms and signs that could lead to a number of different diagnoses. Patients rarely present with the textbook list of features of a condition, and this is particularly true in the case of a pulmonary embolus (PE).

A missed diagnosis of pulmonary embolus can not only be catastrophic for the patient, but extremely distressing for the clinician involved. Patients or their families may feel let down in these situations, which can then lead to a complaint or claim. The doctor might also be asked to justify their management decisions as part of a coroner's inquest or fatal accident inquiry, or significant event investigation.

The MDU reviewed 108 cases from between 2010 and 2015, where there was an allegation that a pulmonary embolus could have been diagnosed sooner, where the final diagnosis was PE, and where we had sufficient information to answer the study questions. The review did not include cases of a missed diagnosis of deep vein thrombosis (DVT) leading to PE.

In this review we looked at:

  1. the setting of the consultation
  2. the outcome for the patient
  3. the signs and symptoms recorded
  4. the documented working diagnosis
  5. the management.

MDU members had requested assistance with a variety of matters including clinical negligence claims, coroner's inquiries, complaints and significant event investigations.


The majority of cases involved in hours GP consultations, one of which included a practice nurse consultation. Only 16.7% related to GP out of hours consultations.

Table 1: setting


Number (percentage) of cases

In hours GP

69 (63.9%)

Out of hours GP

18 (16.7%)

Hospital ED/MAU

9 (8.3%)

Hospital other

10 (9.3%)


2 (1.9%)

MDU courses 2018

Symptoms and signs

The medical records were reviewed and unsurprisingly, the commonest presenting complaint was of shortness of breath, followed by chest pain and cough. Many patients had more than one symptom, but it was noteworthy that 22 (20.3%) reported shortness of breath alone, without any other recorded symptom, and 16 (14.8%) reported only chest pain, again without any other symptoms. Haemoptysis was only reported in 4.6% of cases. 

In 11 cases (10.2%), there was no record of the patient suffering from any of the typical symptoms of shortness of breath, syncope, chest pain or haemoptysis. These patients presented in a variety of ways, including 'off legs', diarrhoea and vomiting with palpitations, and apparent urinary tract infection.

Chest signs were reported in only 21.3% of cases, and the pulse was raised above 100 beats per minute in only 28 out of the 78 cases (35.8%) where it was recorded.

Oxygen saturations had been recorded in 63.9% of cases.

Table 2: signs and symptoms recorded

Signs/symptoms recorded

Number (percentage) of cases

Shortness of breath

68 (62.9%)

Chest pain

46 (42.6%)


43 (39.8%)


5 (4.6%)

Other symptoms

11 (10.2%)

Chest signs

23 (21.3%)

Pulse >100

28 (25.9%)

Pulse <100

50 (46.3%)

Oxygen saturations recorded

69 (63.9%)

Working diagnosis

In 39 (36.1%) cases, there is evidence in the notes that the clinician considered risk factors for PE, for example, by examining the legs for DVT, or questioning the patient about known risk factors such as immobility, smoking, hormonal contraception, pregnancy, family history or malignancy.

The most common working diagnosis was of chest infection in almost one third of cases, followed by asthma or COPD. Musculoskeletal chest pain was diagnosed in 12 cases, and in five of these cases, the doctor had documented chest wall tenderness.

In 5.6% of cases, the symptoms were attributed to anxiety or low mood.

Table 3: documented working diagnosis

Working diagnosis

Number (percentage) of cases

Chest infection/pleurisy

35 (32.4%)


15 (13.9%)

Musculoskeletal pain

12 (11.1%)

Cardiac cause

4 (3.7%)

Gastrointestinal cause

4 (3.7%)

Anxiety/low mood

6 (5.6%)

The most common working diagnosis was of chest infection in almost one third of cases, followed by asthma or COPD.


Two of the 108 patients were admitted to hospital; one for treatment of leg swelling, the other for a suspected myocardial infarction (MI).

This study focused on cases where there was an allegation of a failure to diagnose PE in a timely fashion. Therefore it is unsurprising that in only one case was the patient investigated for PE as a direct result of the consultation, but there is evidence that PE was considered as a possible diagnosis in another five cases.


In 46 cases (42.6%) the patient died.


This review showed that pulmonary embolus can present in a variety of ways, and in 10% of our cases, the patient did not present with any of the typical symptoms of shortness of breath, chest pain, haemoptysis or syncope.

As well as this, some examination findings that might reassure a clinician - such as a normal heart rate and oxygen saturation - were often present. Of particular note, chest wall tenderness was present in some patients, which may also have falsely reassured the clinician.

Of course, it is also important to remember the context in which the patient was seen. The most common symptoms of PE are extremely common presentations in primary care and emergency medicine, and most of the time will not be associated with a diagnosis of pulmonary embolus. Many of these patients had other co-morbidities, such as chronic lung disease, mental health or cardiac problems, meaning that the symptoms may often be attributed to these.

This means that excluding a diagnosis of pulmonary embolus on a clinical basis has a significant risk of error. In addition, no convenient investigation exists to diagnose or exclude PE, with CTPA being the gold standard, but requiring admission to hospital.

This review demonstrates the high stakes involved in making an alternative diagnosis, with almost half of the patients in the files we reviewed having a fatal outcome.

Dr Simon Watkin, consultant respiratory physician and vice president of the MDU Council, comments on the case analysis.

This examination of the MDU files relating to pulmonary embolism offers a salutary lesson in the difficulties in diagnosis of patients presenting, for the most part to a GP in normal working hours. The presenting symptoms, well known to every medical student and doctor, were nevertheless the commonest, but unfortunately are not specific to pulmonary embolism.

All the patients in this study had proven pulmonary embolism, which was fatal in just under half. It could be argued of course that a fatal outcome is more likely to drive subsequent complaint or litigation, but the symptoms of patients in this study are not materially different from those occurring in patients with non-fatal pulmonary embolism. Furthermore, tachycardia as a possible marker of more extensive embolism was surprisingly uncommon (1 in 4).

Perhaps the important finding from the study was that in only one in three cases was there evidence from the notes that risk factors had been considered, and in none of the cases was there a documented Wells risk score, which is readily available and easy to apply. Consideration of risk factors and documenting a risk score would have made it easier to respond to any challenge that the diagnosis had not properly been considered.

The absence of chest signs in 80% of the cases, combined with the working diagnoses of chest infection, pleurisy, asthma or COPD in a total of 46% of cases, should be no surprise as there are often no physical signs on chest examination in cases of PE, whereas of course there usually would be for all the other diagnoses mentioned.

The hardest signs to detect, but probably the most important in pulmonary emboli likely to be fatal, would be cardiovascular including raised jugular venous pressure, loud pulmonary second sound, right ventricular heave and a tender or enlarged liver. The legs should always be examined by inspection and the findings noted.

Oxygen saturation should be checked if possible, but only if the doctor can interpret and act on the results correctly. Personal casework experience suggests that low readings are often explained incorrectly or attributed to technical factors.

Hospital emergency departments and medical assessment units are used to dealing with suspected PE on a very regular basis, and whilst both under- and over-diagnosis are well documented for clinically diagnosed PE, the best test is a CT pulmonary angiogram, which is only available at hospital. This will only rarely give a false negative result, and only exceptionally will a proven PE not be treated with anticoagulation.

Good history taking, risk assessment, and good physical examination are the key to general practice diagnosis of PE, combined with immediate hospital referral in suspected cases.

This page was correct at publication on 19/12/2017. Any guidance is intended as general guidance for members only. If you are a member and need specific advice relating to your own circumstances, please contact one of our advisers.