Intended for healthcare professionals


Reproduction of chest pain by palpation: diagnostic accuracy in suspected pulmonary embolism

BMJ 2005; 330 doi: (Published 24 February 2005) Cite this as: BMJ 2005;330:452
  1. Grégoire Le Gal ({at}, physician1,
  2. Ariane Testuz, resident2,
  3. Marc Righini, physician2,
  4. Henri Bounameaux, physician2,
  5. Arnaud Perrier, physician2
  1. 1Equipe d'accueil 3878 (GETBO), Brest University Hospital, 29609 Brest, France
  2. 2Division of General Internal Medicine, Department of Internal Medicine, Geneva Faculty of Medicine, Geneva University Hospital, CH-1211 Geneva, Switzerland.
  1. Correspondence to: G Le Gal
  • Accepted 2 December 2004


Chest pain associated with pulmonary embolism is usually sharp and worsens with deep inspiration, cough, and movement, resulting from pleural inflammation in peripheral emboli (pleuritic pain).1 Conversely, chest pain that is reproduced by palpation is thought to be caused by pathology of the musculoskeletal chest wall and may prompt clinicians to discard pulmonary embolism as the cause, although cases of pulmonary embolism with isolated pain in the chest wall have been described.2 Managing patients with chest pain is challenging because signs and symptoms of pulmonary embolism lack specificity, because it requires ruling out other life threatening conditions, and because a sizeable proportion of patients have musculoskeletal or pleural syndromes that require symptomatic treatment only.3 We assessed whether chest pain that can be reproduced by palpation is likely to be more indicative of an absence of pulmonary embolism than chest pain caused by breathing, cough, or movement.

Participants, methods, and results

We analysed a database of consecutive outpatients included in a prospective management study that was designed to validate a diagnostic strategy for suspected pulmonary embolism.4 Suspicion of pulmonary embolism was defined as acute onset of new or worsening shortness of breath or chest pain without another obvious aetiology. The study took place in Geneva and Lausanne University Hospitals, Switzerland, and Angers University Hospital, France, between October 2000 and June 2002. Exclusion criteria (n = 258) were ongoing treatment with coagulants, allergy to contrast iodine agents, creatinine clearance below 30 ml/minute, pregnancy, and life expectancy of less than three months. All patients gave informed consent. Before any test, the doctors in charge used eight variables to assess patients in the emergency ward on the basis of a validated prediction rule (the Geneva score): recent surgery, previous thromboembolism, age, hypocapnia, hypoxaemia, tachycardia, band atelectasis, and hemidiaphragm elevation on chest x ray.5 The doctors completed a standardised data form. Chest pain was recorded, and doctors were asked to specify whether or not it was reproduced by palpation. Pulmonary embolism was ruled out if the patient's D-dimer concentration was below 500 µg/l or if proximal venous ultrasonography and helical computed tomography were both negative. In patients with a high clinical probability of pulmonary embolism, a negative pulmonary angiogram was also required. Follow up of patients was at three months. We used a χ2 test to compare the proportion of confirmed pulmonary embolism in patients with and without chest pain that could be reproduced by palpation.

The average age of the 965 included patients was 61 (SD 19) years; 562 (58%) were women. A negative D-dimer test ruled out pulmonary embolism in 280 patients (29%). The overall prevalence of pulmonary embolism was 23% (222 of 965 patients). The prevalence was not significantly lower in patients with pain reproduced by palpation (19.9% (38/191) v 23.8% (184/774), P = 0.25; 1). The sensitivity and specificity of reproducible chest pain for the diagnosis of pulmonary embolism were 17% (95% confidence interval 13 to 23) and 79% (76 to 82); positive and negative likelihood ratios were 0.83 (0.60 to 1.14) and 1.04 (0.97 to 1.12).

Prevalence of pulmonary embolism according to the presence of reproducible chest pain

View this table:


In patients with suspected pulmonary embolism, chest pain reproduced by palpation is not associated with a lower prevalence of pulmonary embolism. Limitations of our findings are the absence of a standardised definition and evaluation method for eliciting chest pain by palpation. Moreover, these results may not apply to all patients with chest pain, as many patients in the emergency department may have been classified as having another obvious aetiology and were not included in the study.

Elicitation of chest pain is widely used by doctors to assess the clinical likelihood of pulmonary embolism. However, in patients without an obvious aetiology, pain in the chest that is reproduced by palpation is not associated with a lower prevalence of pulmonary embolism. Physicians should take into account that the usefulness of these widespread semiologic descriptions may be limited in this situation.

  • What is already known on this topic

  • Chest pain that is reproduced by palpation is classically thought to be caused by pathology of the musculoskeletal chest wall and may prompt clinicians to discard pulmonary embolism as the cause of pain

  • The diagnostic accuracy of this clinical criterion is unknown

  • What this study adds

  • In patients in whom pulmonary embolism is suspected, chest pain that is reproduced by palpation is not associated with a lower prevalence of pulmonary embolism

This article was posted on on 31 January 2005:


  • Contributors All authors had access to the data, read, and approved the final version of the manuscript. GL and AP are guarantors.

  • Conflict of interest None declared.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract