Background Diagnosing pulmonary embolism (PE) can be a difficult process for junior doctors, because it relies on clinical probability scoring and knowledge of how to apply and interpret D-dimer, VQ and CT scanning. The THREAD study aimed to assess the potential diagnostic role of novel biomarkers, in attempt to identify a future, more simple test. The aim of this analysis is to assess the potential role of C reactive protein (CRP) in the diagnosis of PE.

Methods The prospective diagnostic study was conducted at a single general hospital in the UK, between September 2008 and June 2009. All patients investigated for PE were eligible. Exclusions were age <16, lack of capacity and refusal. All patients underwent an evidence based protocol to diagnose or exclude PE, along with a 3 month clinical follow-up period. CRP was not conducted as a blinded research test, instead, the study documented the initial CRP result when the investigating physician ordered the test. Potential for diagnostic use was assessed by constructing receiver operating characteristic (ROC) curves for all patients investigated for PE and emergency department patients investigated for PE.

Results 411/414 patients investigated for PE were approached for consent and 354 patients investigated for PE were enrolled to the study. PE was diagnosed in 19.2% patients. Of this cohort, 269 patients had CRP testing ordered by their physician. The area under the ROC curve (AUC) for CRP in the diagnosis of PE was 0.72 (95% CI 0.65 to 0.78). The AUC for PE was 0.77 (95% CI 0.69 to 0.84) in patients presenting to the emergency department (n=199).

Discussion This is a limited exploratory analysis. Results will be available for the full cohort of patients (n=354) by the conference.

Conclusions CRP has a moderate association with PE in emergency department patients, however could not diagnose PE alone.