A 48-year-old driver, Mr W, was sent to hospital by his GP with a one-week history of unremitting back pain and associated mild shortness of breath. On direct questioning, he also reported non-specific malaise for at least three months with 3kg weight loss but no symptoms of fever. There was no previous history of cardiac problems and no recent dental or other invasive procedures.
Initial investigations demonstrated a mild leucocytosis with normal biochemistry. The ECG and chest x-ray were normal and there was no elevation of troponin, BNP or D-dimers. There was some concern about the possibility of an aortic dissection but a CT scan of the chest was also normal. Inflammatory markers were not measured.
The consultant cardiologist Dr H saw Mr W only once – on the post-take ward round after being admitted – and requested an echocardiogram after hearing “an aortic murmur”. The medical records indicate that he did not see Mr W again during his in-patient stay – nor was he consulted about his progress, results of investigations or plans for discharge or follow-up.
Mr W’s temperature was recorded once daily. The echocardiogram demonstrated a bicuspid aortic valve with moderate aortic regurgitation and no other abnormality. The template report included the statement: “endocarditis is not excluded”.
He was discharged directly from the medical assessment unit without senior review, with a diagnosis of musculoskeletal back pain and possible atypical pneumonia, with a plan for outpatient follow-up in four weeks’ time to assess progress and review the results of the echocardiogram. The GP received only an interim discharge summary, which did not show an appointment had been arranged.
Eight weeks later, Mr W was readmitted to hospital with a high temperature, further weight loss, and shortness of breath secondary to pulmonary oedema. He was anaemic with an ESR of 104mm/hr and six out of six blood cultures were positive for Streptococcus mutans. A clinical diagnosis of infective endocarditis was made and confirmed by echocardiography, which demonstrated a large vegetation on the aortic valve with destruction of the non-coronary cusp and severe aortic regurgitation. He was treated appropriately after microbiological consultation with intravenous benzylpenicillin and gentamicin and his case discussed with the local cardiothoracic surgical centre.
Unfortunately, within 24 hours, and before he could be transferred, Mr W deteriorated acutely with hypotension and pulmonary oedema refractory to diuretics and could not be resuscitated. The postmortem showed large vegetations on the aortic valve and extensive destruction of both leaflets of the bicuspid aortic valve secondary to bacterial endocarditis. The case was settled for a moderate amount.
- The diagnosis of infective endocarditis is difficult and depends upon a low threshold of suspicion (see Beynon R, Bahl VK, Prendergast BD, Infective endocarditis, BMJ 333:334-339(2006)). The disease may present in a variety of forms to a variety of clinical specialties.
- Senior medical input to the care of seriously ill patients is important.
- There is little purpose in requesting investigations if the results are not carefully reviewed and acted upon at an appropriately early stage. There were several diagnostic clues in this particular case, which should have alerted the clinical team to the earlier diagnosis and management of infective endocarditis.
- The pressure to discharge patients and create beds for further admissions means that the results of important investigations are easily overlooked.
- Clear and comprehensive communication with the patient and GP is essential.
- Team working and fractured continuity of care can easily create “I thought you did it” situations where required tasks are not completed. The outcome for this unfortunate patient may have been different had an early follow-up appointment been arranged.
- Safe systems should be in place to check that outpatient clinics are arranged. It is worthwhile telling the patient that they should get in touch if plans are not confirmed.