Efficacy of Wells score and Pulmonary Embolism Rule Out Criteria(PERC) to exclude pulmonary embolism in a pulmonary emergency ward

Aim: The Pulmonary Embolism Rule-out Criteria (PERC) rule is a diagnostic algorithm to exclude pulmonary embolism (PE). Even it is not widely used in routine practice, has better negative predictive value in low-risk populations among existed clinical assessment tests. We aimed to examine availability of PERC rules combined with Wells score in a pulmonary emergency ward. Material and Methods: A retrospective hospitalized database study was conducted with 163 patients. All patients who underwent radiological investigation due to suspicious for PE were included. We calculated prevalence of PE in group of PERC negative and PERC positive. All PERC parameters were evaluated solely for risk of PE. Results: PERC positivity was 82.4% in patients diagnosed with PE. There were no significant differences in terms of PERC positivity between PE+ and PEgroups (p=0.336). False positivity rate for moderate to high Wells score was 36.1%. Frequency of PERC negativity among PE negative patients was only %36.1 and %17.6 in PE positive group (p=0.75 and r=-0.025). The sensitivity of the PERC test was 82.3%, specificity 19.6%, PPV 63.1%, NPV 40%, false positivity rate 36.8% and false negativity rate 60%. Among patients with moderate to high wells, 61 patients were diagnosed PE and 14 of them were PERC negative. Conclusion: Our study suggests that PERC scoring system even combined with Wells score does not sufficient enough to prevent unnecessary irradiative imaging studies in a pulmonary emergency ward.


Introduction
Pulmonary embolism (PE) can be highly mortal if remains undiagnosed. Referral symptoms are usually non-specific but in cases where hemodynamic instability develops-potential life-threatening condition, the emergency physician does not have enough time to diagnose. In daily practises there are some risk stratification models (Wells score, Geneva score) to estimate clinic probability before diagnostic methods or clinical probability of PE, estimated by the clinician gestalt (unstructured empirical probability) can be preferable. [1] When probability of PE is low and D-dimer level is under 500 µg/mL we tend to exclude PE but higher level of D-dimer can cause further investigation. [2] Computed tomography pulmonary angiogram (CTPA) is first choice in emergency departments (ED) if there is no contraindication. However, by widespread use of CTPA more PE is diagnosed, overall mortality due to PE do not change. [3,4] More over incidence of allergic reactions or nephrotoxicity due to contrast agents are increased.
To avoid unnecessary further diagnostic steps, a scoring system -Pulmonary Embolism Rule-out Criteria (PERC)-was developed to rule out PE by Kline  There are many studies investigating reliability of PERC score in ED. [7][8][9][10][11] But to our knowledge there is not any data of pulmonary hospital about using PERC. So, primary outcome is to present our approach to patients suspected to be PE in ED regarding to both PERC and Wells score. Secondary aim is to demonstrate reliability of this score in patients with all clinical probability.

Study setting and population
This is a retrospective hospitalized database study. All enrolled patients were assessed in ED and who underwent CTPA and ventilation/perfusion scan (VPS) to exclude or verify PE.
When scanning data retrospectively, International Statistical

Classification of Diseases and Related Health Problems 10th
Revision (ICD-10) coding system were used. Patients whose principal or preliminary diagnosis was recorded one of these diagnosis; I.26 (pulmonary embolism), I26.0 (pulmonary embolism with acute cor pulmonale), I.26.9 (pulmonary embolism without cor pulmonale) and I.82 (venous emboli and thrombosis) enrolled to the study. Cut-off value for D-dimer was 500 μg/L.

Statistical Analyses
Analyses were performed using SPSS version 15.0 (SPSS Inc, Chicago, IL). Continuous variables were expressed as mean ± standard deviation and categorical variables were defined as percentages. Other descriptive statistics were presented as median, minimum and maximum values. The categorical variables were compared with Chi-Square and Fisher's Exact test. Comparisons between two independent groups were performed using Student's t-test for normally distributed continuous variables, and Mann -Whitney U test when the distribution was skewed. Correlations were evaluated with Pearson's or Spearman's correlation tests. A p value less than 0.05 was considered as statistically significant.

Results
One hundred sixty-three patients suspected for venous thromboembolism were enrolled study. Among 133 PERC positive group 84 patients were diagnosed PE. However, 18 patients had PE in PERC negative group (n=30, 18.4%). Study algorithm and score results of both groups (PERC positive/ negative) was summarized in Figure 1.

Figure 1. Flowchart for distribution of PE diagnosis according to PERC score
Mean age was 57.3±17.9 (±SD) and gender distribution was almost closed to each other. Chronic obstructive pulmonary disease (COPD) and coronary arterial disease (CAD) were most common co-morbidities (27% vs. 15.2%). Among all patients 102 (62.6%) were diagnosed with PE and deep vein thrombosis (DVT) was accompanied in 19 (11.7%) patients. In diagnostic procedures while 132 (81%) patients underwent CTPA, 31 (19%) patients were performed VPC due to renal failure or contrast allergy. Transthoracic echocardiogram (TE) and chest x-ray findings were shown in (Table 1).
A hundred two patients were diagnosed PE and all were symptomatic. In PE group D-dimer level, Wells score were higher than PE negative patients. (p=0.336 vs p=0.003). As expected patients diagnosed with PE had PERC positivity with 82.4% ratio. But even number of PERC positive patients were lower in PE negative group (n=61), there were no significant differences in terms of PERC positivity between PE+ and PE -groups (PE -: 80.3%, PE +: 82.4% and p=0.336). False positivity rate for moderate to high Wells score was 36.1% (Table 2).   is not applicable to exclude in a pulmonary emergency ward.

Conclusion
Our results revealed that even combined with clinical probability tests, PERC score is not sufficient enough to exclude PE. Additionally, co-morbidities are important confounding factors so even in low risk patients PERC scoring system does not seem to be sufficient enough to prevent unnecessary irradiative imaging studies.