In everyday clinical practice, computed tomography (CT) is increasingly used to rule out pulmonary embolism (PE), as clinical scoring systems incl. D-dimers are often too inaccurate. This unnecessarily exposes patients to high levels of stress in many cases. The recently developed 4PEPS score could reduce the need for CT angiography to rule out PE in the future.
Patients with suspected pulmonary embolism (PE) are the "bread and butter" of emergency medicine. However, the symptoms of PE are very unspecific, so that scoring systems such as the Geneva, Wells or YEARS score and the determination of D-dimers are usually used. However, the scores often do not provide a reliable statement and elevated D-dimers do not always help in clinical decision-making due to their low specificity. Therefore, CT angiography is performed in many cases to rule out PE, also for forensic confirmation. This procedure is readily available and extremely sensitive. However, affected patients may be exposed to unnecessary radiation and may develop an allergic reaction and kidney failure due to the contrast agent. An international research team has therefore developed a new scoring system to reduce unnecessary CT angiography.
To do this, the team used retrospective data from over 12,000 patients from three different emergency departments in France, Belgium and the USA. All of them were suspected of having PE. The researchers divided the population into a derivation cohort (about 5,500 patients) with which the new score was created, an internal validation cohort (about 3,700 patients) to test the score in the same population as the derivation cohort, and two external validation cohorts (about 3,000 patients) in which the score was tested in a new population from another study centre. In the derivation cohort, 62% of the subjects were female, the mean age was 52 years and 11% were diagnosed with PE. The two external validation cohorts differed in PE prevalence (21.5% and 11.7%).
To develop the score, the researchers collected clinical and demographic parameters of the patients in the derivation cohort that could be related to the diagnosis of PE. They then performed a regression analysis and identified a total of 13 parameters that they integrated into the score: these included young or older age, gender, a pulse below 80/min, the presence of chronic respiratory disease, chest pain and dyspnoea, oestrogen therapy, a history of deep vein thrombosis, syncope, immobility in the last 4 weeks, an oxygen saturation < 95%, calf pain or leg oedema and PE as the most likely diagnosis. Each parameter was weighted differently according to its regression coefficient on PE probability. For example, age below 50 years lowers the score by 2 points, while calf pain or leg oedema increases the score by 3 points. The lowest score was -6 points, while the highest score was 22 points. The score was called the Pulmonary Embolism Clinical Probability Score (4PEPS).
To make it easier to use in clinical practice, the researchers divided the score into four categories: A score of less than 0 points can rule out PE without further diagnosis (clinical pretest probability < 2%). A score of 0 to 5 points can exclude PE if the D-dimers are below 1 µg/ml (clinical pretest probability 2 to 20%). With a score of 6 to 12 points, PE can be ruled out if the D-dimers are below 0.5 µg/ml or slightly higher in an age-adjusted manner (clinical pretest probability 20 to 65%). A score above 12 points requires further imaging in any case, even with inconspicuous D-dimers (clinical pretest probability above 65%).
The researchers check the accuracy of their score using the statistical model of the area under the ROC curve (ROC = Receiver Operating Characteristics). The score gave a good result of 0.78 and 0.79 (1.00 is the optimum) in both external validation cohorts with high and low PE prevalence respectively. False negatives occurred in 0.71% and 0.89% of cases in both cohorts. This is comparable to common scores. However, the score reduced CT angiographies in both validation cohorts by 21% and 19% compared to the Geneva score.
In conclusion, the 4PEPS score could be a safe and effective way to diagnose PE in the emergency department in the future, saving unnecessary CT imaging at the same time. However, the score still has weaknesses. So far, it has only been tested on one patient collective in the emergency department and is therefore not applicable in other situations (e.g. in everyday ward life). Furthermore, it is very complex with 13 parameters and therefore cannot be calculated "quickly". Rather, in most cases a PC with an input mask is required. Before the 4PEPS score can be used in everyday clinical practice, it still has to prove its relevance in a large-scale prospective cohort study.
Reference:
Roy et al. Derivation and Validation of a 4-Level Clinical Pretest Probability Score for Suspected Pulmonary Embolism to Safely Decrease Imaging Testing. JAMA Cardiology. March 3, 2021.