Article Text

Download PDFPDF

Hyperventilation: cause or effect?
  1. T A Mehta,
  2. J G Sutherland,
  3. D W Hodgkinson
  1. Accident and Emergency Department, The Ipswich Hospital NHS Trust, Heath Road, Ipswich IP4 5PD
  1. Correspondence to: Mr Hodgkinson, Consultant


A young person presenting with shortness of breath is common to the accident and emergency department. Usually this hyperventilation is anxiety related or a panic attack, but sometimes it can be caused by a serious underlying condition like pulmonary embolus. Acute shortness of breath in any patient should never be dismissed lightly. It is important to realise that pulmonary embolus can present without chest pain and with shortness of breath as the major symptom. Such patients can be distinguished by close attention to history and examination, risk factors for thromboembolic disease and the use of basic investigations (electrocardiogram, chest radiography and arterial blood gas analysis). A serious cause for shortness of breath must be excluded before labelling it as “hysteria” or “panic”.

  • hyperventilation

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Doctors in accident and emergency (A&E) medicine are confronted with a large number of patients that have the label of a “panic attack”. Typically these are young patients who are hyperventilating. Not all hyperventilation in the A&E department is “cause”; some of it is “effect”. The difficulty is spotting the one when hyperventilation is effect and not cause. It is important for emergency physicians, who are often junior doctors to consider any history of shortness of breath as a major symptom. Such patients require detailed assessment to exclude a serious cause. The following case discussion highlights the factors that differentiate hyperventilation “cause” from “effect”.

Case report

A 26 year old woman presented to the A&E department of a district general hospital at 1402 with a history of sudden onset “shortness of breath, fast heart beat and intermittent dizzy spells to the point of near collapse” lasting 30 minutes. Her symptoms had almost resolved at the time of arrival in the A&E department. She had a similar attack in the past and was admitted to a medical ward of the same hospital eight months previously, from where she took her own discharge against medical advice within 24 hours of admission. She had undergone outpatient investigations including a 24 hour cardiac tape recording, which was reportedly normal. Past medical history revealed that she had undergone surgery for varicose veins at age 16 years. She was taking oral contraceptives and smoked 10 cigarettes per day. On examination, she had a pulse rate of 125 beats/min and a blood pressure of 140/90 mm Hg. Notably her oxygen saturation on air was recorded as 88% in the ambulance before arrival, but subsequently was normal on 10 l/min of oxygen and remained normal even after the oxygen was discontinued. Examination of the respiratory and cardiovascular system was unremarkable. An electrocardiogram (ECG) was performed, but not commented upon. It was subsequently reported as showing a right ventricular strain pattern with S1Q3T3 and was considered to be abnormal (fig 1). Arterial blood gas analysis and chest radiography were not performed. After observing the patient for 30 minutes, a reassessment of her condition revealed a pulse rate of 90 beats/min, blood pressure of 140/90 mm Hg and no hypoxia on air. She was discharged with a presumptive diagnosis of “anxiety related hyperventilation” and asked to see her general practitioner if her symptoms recurred. Two days later she was admitted to hospital after another episode of acute shortness of breath. On this occasion she was unwell, tachycardic, hypotensive and hypoxic. ECG was unchanged. Arterial blood gas analysis showed a Po2 of 8.0 kPa on 40% oxygen. A Krypton ventilation perfusion lung scan revealed multiple defects on the perfusion scan (fig 2) with a high probability of pulmonary embolus. She was given heparin and responded well. She was discharged after nine days and prescribed warfarin.

Figure 1

ECG showing right ventricular strain with S1Q3T3.

Figure 2

V/Q scan showing multiple perfusion defects.


A young woman presenting with shortness of breath, fast pulse and respiratory rate and dizziness is quite common in emergency medicine. In many cases anxiety and panic are the cause of this shortness of breath, but sometimes it may be a symptom of a more serious underlying condition like pulmonary embolus. It is important for emergency physicians to exclude such pathology before labelling shortness of breath as anxiety related. This can be done by close attention to history and examination, pre-existing risk factors and the use of some basic investigations attainable in the A&E department.

Regarding the clinical history, too little was made of her having shortness of breath to the point of near collapse. A detailed history would have highlighted this.

Examination revealed a tachycardia, tachypnoea and a documented period of hypoxia. This should have indicated the likelihood of a serious underlying cause. No reassurance should have been gained by the fact that her physiology was seen to normalise.

Risk factors for thromboembolic disease include major abdominal/pelvic surgery, immobilisation, recent lower limb trauma, pregnancy or post-partum, major medical illness and previous proven deep vein thrombosis or pulmonary embolus. Her risk factors for thromboembolism were minor, but did include smoking, oral contraceptives and previous lower limb varicose vein surgery (10 years ago).

Any abnormality in ECG and chest radiograph will suggest a serious underlying cause for “hyperventilation”. Such patients should be referred for urgent investigations to exclude pulmonary embolism. Her ECG was abnormal and that in itself should have dictated a referral.

The main discriminating factors in this case were dizziness to the point of near collapse, shortness of breath, respiratory rate > 20 breaths/min, hypoxia recorded by pulse oximetry in the ambulance and an abnormal ECG.

A district general hospital with a catchment population of 200 000 may expect to diagnose 50 cases of pulmonary embolism annually.1 As some of these only become apparent at necropsy, the true incidence of pulmonary embolism is probably much higher at 1% of all admissions.2 The incidence in A&E attendances is unknown, but is likely to be only a small proportion of young patients presenting with hyperventilation. Risk factors are important and 80–90% of patients with pulmonary embolus have predisposing factors. The clinical patterns of pulmonary embolus include sudden collapse (faintness and/or hypotension), pulmonary haemorrhage syndrome (pleuritic chest pain and/or hemoptysis) and isolated dyspnoea. This last category of patients presenting with shortness of breath without chest pain contributes up to 25% of all those diagnosed with pulmonary embolus.3 Most patients (>97%) with pulmonary embolus are short of breath with a respiratory rate > 20 breaths/min. ECG, arterial blood gas analysis and chest radiography are the basic investigations for diagnosing pulmonary embolus. d-dimer assays can be used to exclude the presence of thromboembolic disease.4 Newer d-dimer assays show promise as a tool for use in the A&E department. The diagnosis of pulmonary embolus can be confirmed by lung imaging using ventilation/perfusion isotope scanning, pulmonary angiography or spiral computed tomography. The guidelines for initial assessment and action are outlined by the British Thoracic Society, Standards of Care Committee Report.3 (At present the British Thoracic Society guidelines do not mention the use of d-dimer assays).

The lesson that should be learnt from our case is that pulmonary embolism can present without chest pain and consequently a symptom of shortness of breath should always be taken seriously, particularly if the features previously mentioned are present.



Tapan Mehta contributed in data collection, literature search, illustrations and writing of the paper. John Sutherland contributed in literature search and writing of the paper. David Hodgkinson edited the paper and is also the guarantor.



  • Funding: none.

  • Conflicts of interest: none.