SEARCH

SEARCH BY CITATION

Keywords:

  • cancer;
  • clinical prediction rule;
  • outcome;
  • pulmonary embolism;
  • readmission

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Conflict of interest statement
  9. Reference

Abstract.  Bertoletti L, Le Gal G, Aujesky D, Roy P-M, Sanchez O, Verschuren F, Bounameaux H, Perrier A, Righini M. (University of Geneva, Geneva, Switzerland; Université De Saint-Etienne, Jean Monnet, Saint-Etienne; Brest University Hospital, Brest, France; Bern University Hospital, Bern, Switzerland; Angers University Hospital, Angers; Paris Descartes University, Paris, France; Saint-Luc University Hospital, Bruxelles, Belgium; and University of Geneva, Geneva, Switzerland). Prognostic value of the Geneva prediction rule in patients in whom pulmonary embolism is ruled out. J Intern Med 2011; 269: 433–440.

Objectives.  The prognosis of patients in whom pulmonary embolism (PE) is suspected but ruled out is poorly understood. We evaluated whether the initial assessment of clinical probability of PE could help to predict the prognosis for these patients.

Design.  Retrospective analysis of data obtained during a prospective multicentre management study.

Setting.  Six general and teaching hospitals in Belgium, France and Switzerland.

Subjects.  In 1334 patients in whom PE was ruled out, 3-month mortality data were available (hospital readmission status was unknown for three patients) and clinical probability was evaluated with the revised Geneva score (RGS).

Main outcome measures.  Three-month mortality and readmission rates.

Results.  Three-month mortality and readmissions rates were 3% and 19%, respectively and differed significantly depending on the RGS-determined PE probability group (P < 0.001). When compared with patients presenting with a low probability, the risk of death after 3 months was higher in cases of intermediate or high RGS-based probability {odds ratio: 8.7 [95% confidence interval (CI): 2.7–28.5] and 22.6 (95%CI: 2.1–241.2), respectively}. The readmission risk increased with PE probability group (P < 0.001). The main causes of death were cancer, respiratory failure and cardiovascular failure. In total, 86% of patients with low RGS-based probability were alive and had not been readmitted to hospital, whereas other patients had a twofold increased risk of death or readmission during the 3-month follow-up. The simplified Geneva score, calculated a posteriori, gave similar results.

Conclusions.  Initial assessment of clinical probability may help to stratify prognosis of patients in whom PE has been ruled out. Patients with a low probability of PE have a good prognosis. Whether patients with higher probability might benefit from more vigilant care should be evaluated.


Abbreviations:
COPD

chronic obstructive pulmonary disease

CT

computed tomography

MDCTA

multidetector computed tomography angiography

OR

odds ratio

PE

pulmonary embolism

RGS

revised Geneva score

SGS

simplified Geneva score

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Conflict of interest statement
  9. Reference

Pulmonary embolism (PE) is the third most common cause of cardiovascular mortality following coronary artery disease and stroke. During the past two decades, diagnosis of PE has considerably improved and is currently based mainly on noninvasive strategies including clinical probability assessment, D-dimer measurement and multidetector computed tomography (CT) angiography. Moreover, the index of clinical suspicion of PE has progressively increased, which has resulted in PE being ruled out in at least 80% of patients with suspected PE who are referred to a diagnostic centre [1]. The prognosis of these patients without PE has been poorly studied. Indeed, even if PE has been ruled out, some problems may persist in this large subgroup of patients: (i) an alternative diagnosis is not always nor easily made with certainty; (ii) nearly one-third of these patients will undergo further contrast-enhanced CT for suspicion of a new PE during the subsequent year [2]; and (iii) a high 3-month all-cause mortality has been reported, ranging from 1.5% [3] to 21.5% [4]. Moreover, evidence suggests that these patients might have a similar prognosis compared to patients with objectively proven PE [5].

The assessment of clinical probability of PE is recommended by international guidelines [6]. The Geneva prediction score is a well-validated diagnostic rule [7, 8]. We studied whether assessment using this diagnostic rule would correlate with the prognosis of patients in whom PE has been ruled out, as some items of the score might have a prognostic value.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Conflict of interest statement
  9. Reference

Patients and setting

Data were collected during a multicentre, randomized, prospective clinical trial to evaluate a diagnostic strategy for PE, combining clinical probability assessment, plasma D-dimer measurement by enzyme-linked immunoassay (ELISA) and multidetector computed tomography angiography (MDCTA) with or without lower limb venous compression ultrasonography [7].

Eligible subjects were consecutive adult outpatients admitted to the emergency department with clinically suspected PE (defined as acute onset of new or worsening shortness of breath or chest pain without any other obvious aetiology) at six university hospitals in Switzerland, France and Belgium between 1 January 2005 and 31 August 2006. The exclusion criteria were contraindication to MDCTA (i.e. allergy to iodine contrast agents, creatinine clearance of <30 mL min−1 or pregnancy), a diagnosis of PE documented before the time of presentation, a terminal illness with an expected survival of <3 months and ongoing anticoagulant therapy at presentation.

All participants first had an assessment of the clinical probability of PE using the revised Geneva score (RGS) [8] (Table 1) followed by the PE diagnostic work-up, as previously described [7]. Patients included in the analysis were those for whom PE was ruled out and 3-month status was known (death or hospital readmission). PE was ruled out: (i) in patients with a low or intermediate clinical probability of PE by a D-dimer level <500 ng mL−1 (rapid ELISA, Vidas DD; BioMérieux, Marcy-l’Etoile, France [9]) or a negative MDCTA result; and (ii) in patients with a high clinical probability PE by a negative MDCTA result combined with either a negative ventilation–perfusion scintigraphy or a negative pulmonary angiography finding. The simplified Geneva score (SGS) [10] (in which each score item has the same weight) was calculated a posteriori for each patient.

Table 1. Revised and simplified versions of the Geneva score
Items of the revised Geneva scorePoints for revised versionPoints for simplified version
Age >65 years1.01.0
History of pulmonary embolism or deep venous thrombosis3.01.0
Surgery or fracture within 1 month2.01.0
Active malignancy2.01.0
Heart rate (beats min−1)
 75–943.01.0
 ≥955.01.0
Pain on deep leg vein palpation4.01.0
Unilateral leg pain3.01.0
Haemoptysis2.01.0
Low probability0–30–1
Intermediate probability4–102–4
High probability≥11≥5

All patients underwent follow-up at 3 months. They were instructed to come back to see the investigators in case of recurrent symptoms of the respiratory tract or legs. At the end of the follow-up, participants were asked by telephone to declare all health-related events (in particular any admission to hospital) during the 3-month period. The family physician was contacted whenever a possible event was reported. Medical data were analysed if a patient was readmitted to hospital for any cause or died during follow-up. Deaths were adjudicated as related, possibly related or unrelated to PE. Death was judged to be related to PE if it was confirmed by autopsy or if it followed a clinically severe PE either initially or after a recurrent event that was objectively confirmed. Sudden or unexpected death was classified as possibly related to PE. Unrelated deaths were because of an obvious cause other than PE. The outcome events were adjudicated by three independent experts who were blinded to the PE probability allocation groups.

Written informed consent was obtained from all patients. The study was approved by the ethics committees of the Geneva and Lausanne University Hospitals for Switzerland, of the Brest University Hospital for France and of Saint-Luc University Hospital for Brussels. This study is registered with ClinicalTrials.gov, number NCT00117169.

Study outcomes

The primary outcome was the 3-month all-cause mortality, according to the PE probability category obtained with the RGS (low/intermediate/high), in patients in whom PE had been ruled out. A secondary outcome was the 3-month all-cause hospital readmission rate (according to probability category obtained with the RGS). Both primary and secondary outcomes were analysed according to each probability category obtained with the SGS. However, as SGS was computed a posteriori, it was not considered as a primary outcome.

Data analysis

Categorical variables are presented as numbers and/or percentages, and continuous variables are presented as means ± standard deviations.

Patients were classified into three probability groups according to their revised score (RGS <4: low; RGS ≥4 to <11: intermediate; RGS ≥11: high). We calculated the SGS a posteriori for each patient, and the proportion of patients classified within each probability group (SGS <2: low; SGS ≥2 to <5: intermediate; SGS ≥5: high). Cut-offs were identical to those used in diagnostic studies.

Comparisons between groups were made using chi-square tests to compare categorical variables and the Student t-test for continuous variables. Three-month all-cause mortality and 3-month rehospitalization amongst the subgroups of clinical probability were compared with logistic regression. Analysis of 3-month all-cause mortality also included Kaplan–Meier analysis and log-rank tests. We then used logistic regression to identify independent score items associated with the occurrence of an event (death or hospital readmission) at 3 months. Finally, potential confounding factors such as sex or presence of chronic obstructive pulmonary disease (COPD) were entered into the model for adjustment. Results were expressed as odds ratio (OR) and 95% confidence interval (CI). A P value <0.05 was considered statistically significant. Statistical analysis was carried out with the Statview 5.0 software (v5; SAS Institute Inc., Cary, NC, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Conflict of interest statement
  9. Reference

Pulmonary embolism was ruled out in 1336 patients, two of whom were lost to follow-up. Thus, our study population comprised 1334 patients. Baseline characteristics are presented in Table 2. Information on hospital readmission was missing for three patients. According to the RGS, the 1334 patients were classified as follows: 545 (41%) in the low-probability group, 780 (59%) in the intermediate-probability group and 9 (1%) in the high-probability group. Despite the fact that very few patients were classified as having a high clinical probability of PE, all groups were maintained, according to the three-group classification of the Geneva prediction rule. The results remained similar if the intermediate- and high-probability groups were pooled.

Table 2. Baseline characteristics
Age, in years (mean ± SD)58 ± 18.5
Age >65, n (%)527/1334 (40%)
Weight, in kg (mean ± SD)72.9 ± 17.6
Sex, male/female (%)46/54
Positive D-dimers, n (%)771/1331 (58%)
Previous VTE, n (%)184/1334 (14%)
Chronic obstructive pulmonary disease, n (%)156/1334 (12%)
Active cancer or remission within 1 year, n (%)94/1334 (7%)
Chronic heart failure, n (%)67/1157 (6%)
Recent surgery or fracture, n (%)83/1334 (6%)
Persistent neurological deficiency, n (%)25/1334 (2%)

Three-month mortality and readmission according to the RGS

Mortality and readmission rates according to RGS-based clinical probability groups are presented in Table 3. The overall 3-month all-cause mortality was 3% (40/1334). Mortality differed significantly according to probability group (P < 0.05, chi-square test). At the end of the 3-month follow-up period, the risk of death was higher in patients presenting with an intermediate [OR = 8.7, 95%CI (2.7–28.5)] or a high RGS-based probability [OR = 22.6, 95%CI (2.1–241.2)] than in patients with a low probability of PE. Kaplan–Meier curves (Fig. 1) differed significantly between the three groups (P < 0.01, log-rank test).

Table 3. Outcomes according to revised Geneva score (RGS)–based probability
OutcomeClinical probability (RGS)
Low n = 545Intermediate n = 780High n = 9All
  1. aIntermediate versus low; bhigh versus intermediate; chigh versus low; all: P < 0.01; dinformation on readmission was missing for three patients in the intermediate-probability group.

Death3 (0.6%)36 (4.6%)a1 (11.1%)bc40/1334 (3.0%)
Readmission76 (13.9%)173 (22.2%)a3 (33.3%)252/1331d (18.9%)
image

Figure 1. Kaplan–Meier curves of 3-month mortality according to revised Geneva score–based probability (P < 0.01, log-rank).

Download figure to PowerPoint

The overall 3-month readmission rate was 19% (252 of 1331 patients). The rate of readmission differed significantly according to RGS-determined probability group (P < 0.001, chi-square test). The risk of readmission at the end of the 3-month follow-up was significantly higher in patients presenting with an intermediate RGS-determined probability [OR = 1.8, 95%CI (1.3–2.4)] than in patients with a low probability, but failed to reach significance in the high-probability group [OR = 3.7, 95%CI (0.9–12.6), P = 0.08].

Three-month mortality and readmission according to the SGS

According to a posteriori calculated SGS, 540 (41%) patients were classified as low probability of PE, 781 (59%) as intermediate probability and 13 (1%) as high probability. The main results are illustrated in Table 4.

Table 4. Outcomes according to simplified Geneva score (SGS)–based probability
OutcomeClinical probability (SGS)
Low n = 540Intermediate n = 781High n = 13All
  1. aIntermediate versus low; bhigh versus intermediate; chigh versus low; all: P < 0.01; dinformation on readmission was missing for three patients.

Death3 (0.6%)32 (4.1%)a3 (23.1%)bc38/1334 (2.9%)
Readmission71 (13.2%)174 (22.3%)a7 (53.9%)c252/1331d (18.9%)

Rate of death or hospital readmission differed significantly according to SGS-based probability (P < 0.001 for both, chi-square test). At the end of the 3 months of follow-up, the risk of death was higher in patients presenting with an intermediate [OR = 8.1, 95%CI (2.5–26.7)] or a high SGS-based probability [OR = 53.7, 95%CI (9.6–299.4)] than in patients presenting with a low probability.

The risk of readmission during the 3-month follow-up was significantly higher in patients presenting with an intermediate [OR = 1.9, 95%CI (1.4–2.6)] or a high SGS-determined probability [OR = 7.7, 95%CI (2.5–23.6)] than in patients presenting with a low probability.

Death and readmission analysis

The three main causes of death were cancer (19 patients), respiratory failure (seven patients) and cardiovascular (six patients) failure. A known active malignant condition [OR = 8.3, 95%CI (4–17)], heart rate above 94 beats min−1 [OR = 4.8, 95%CI (1.6–14.7)] and age above 65 years [OR = 3.7, 95%CI (1.8–7.8)] were the score items that were independently associated with death. When baseline characteristics were included in the analysis (Table 5), cancer, heart rate >94 beats min−1 and age >65 years remained independently associated with the 3-month risk of death. Male gender and positive D-dimer test results were also independently associated with death, whereas the presence of COPD was associated with death in univariate analysis but failed to reach significance in multivariate analysis [OR = 2.0, 95%CI (0.9–4.5), P = 0.075]. The same score items were associated with 3-month risk of hospital readmission. The main reasons for readmission were cardiovascular (48 cases, 19%), respiratory (except cancer and acute infection, 43 cases, 17%), abdominal (except cancer, 38 cases, 15%), cancer (27 cases, 11%) and acute infection (mostly bronchopneumonia, 26 cases, 10%). The initial diagnosis (for the index event) and the reason for readmission were similar in two-thirds of the patients readmitted for cardiovascular or respiratory reasons.

Table 5. Univariate and multivariate analysis of death according to the baseline characteristics and items of the score
 Univariate analysisMultivariate analysis
Odds ratio (95%CI)P- valueOdds ratio (95%CI)P -value
  1. Only baseline characteristics with a P-value <0.2 in univariate analysis are shown.

Sex
 Female1 10.005
 Male2.9 (1.4–5.7)0.0032.9 (1.4–6.1) 
D-Dimers
 Negative1 10.008
 Positive14.5 (3.5–60.2)<0.0017.4 (1.7–32.2) 
Chronic obstructive pulmonary disease
 No1 10.075
 Yes3.4 (1.7–6.9)<0.0012.0 (0.9–4.5) 
Items of the Geneva rule
Age >65 years
 No1 10.015
 Yes4.2 (2.1–8.5)<0.0012.6 (1.2–5.5) 
Previous deep venous thrombosis or pulmonary embolism
 No1 10.97
 Yes0.9 (0.3–2.3)0.811.0 (0.4–2.7) 
Surgery or fracture within 1 month
 No1 10.62
 Yes1.7 (0.6–4.9)0.321.3 (0.4–4.2) 
Active malignancy
 No1 1<0.001
 Yes9.2 (4.7–18.2)<0.0017.0 (3.3–14.8) 
Unilateral leg pain
 No1 10.53
 Yes0.2 (0.0–1.7)0.160.5 (0.1–4.2) 
Haemoptysis
 No1 10.54
 Yes2.1 (0.7–6.1)0.171.5 (0.4–5.0) 
Pain on deep leg vein palpation and unilateral oedema
 No1 10.21
 Yes1.7 (0.2–13.2)0.604.3 (0.4–42.2) 
Heart rate
 ≤741 1 
 >74 and <952.6 (0.8–8.0)0.102.4 (0.7–7.6)0.15
 ≥956.0 (2.1–17.5)0.0014.7 (1.5–14.6)0.007

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Conflict of interest statement
  9. Reference

The results of our study show that, in patients in whom PE has been ruled out, the clinical outcomes differ markedly amongst the three groups of different clinical probability of PE. In particular, 3-month mortality and hospital readmission rates increased stepwise from the low to the high clinical probability groups, as assessed by the RGS.

Clinical impact

Most outcome studies have focused on patients with confirmed PE. Nevertheless, patients in whom PE was suspected but ruled out (i.e. the majority of patients) also experience adverse outcomes such as death and readmission during the 3 months following PE suspicion. Moreover, once PE is ruled out, and without any additional procedure than returning to the initial clinical probability assessment, clinicians may have some indication of their patient’s prognosis. For example, 86% of patients classified as having a low clinical probability were alive and had not been readmitted to hospital 3 months later. Conversely, patients with an intermediate or high clinical probability of PE have a twofold higher risk (P < 0.001) of death or readmission to hospital at 3 months than patients with a low clinical probability.

Strengths and limitations

The main limitation of this study is that it is a retrospective analysis. However, it was based on predefined and prospectively collected data (items of the Geneva prediction rule, death and readmission during the 3-month follow-up), limiting the possibility of bias. Moreover, the RGS is a well-validated clinical prediction rule and data arise from a multicentre multinational trial [7]. Another limitation of the study is the small number of patients of high clinical probability, which is because of the high prevalence of PE in this subgroup. The results remained similar if the intermediate- and high-probability groups were pooled.

Items of the score independently associated with death or readmission were cancer, heart rate ≥95 beats min−1 and age above 65 years. It could be hypothesized that our results were predictable as the worse prognosis of patients having a high clinical probability is related to a greater prevalence of both advanced age and cancer in this subgroup. Forty patients died during the 3-month follow-up, and cause of death was cancer in 19 patients (48%). However, cancer was known and included in the score in only 12 of these 19 cases. Therefore, the presence of cancer had no influence on the clinical probability assessment in a third of this sample. Moreover, our results indicate that the performance of the Geneva score depends less on the presence or absence of one item than on the combination of items. As an example, none of the patients with a known active malignant condition and classified in the low-probability group died during the 3-month period. Conversely, the 12 patients who died because of a previously known cancer were all classified as having a nonlow (intermediate or high) clinical probability. In addition, our findings remain consistent when determining PE probability using the SGS (Table 1), in which the weight of each item of the rule is identical. Therefore, we are confident that integrating all items in a score provides a more accurate classification of a patient’s prognosis than considering each item separately, and independently of the weight of each item, as demonstrated in other disease areas (e.g. the CHADS2 score for atrial fibrillation [11]).

Strengths and weaknesses in relation to other studies

To our knowledge, this is the first study to report the prognostic significance of the initial assessment of clinical probability for patients in whom PE was suspected and ruled out after a well-validated diagnostic work-up.

Only one retrospective study has provided information on clinical probability, and 3-month mortality in patients in whom PE was ruled out. In a monocentric study, Bourriot et al. [12] followed a cohort of 117 inpatients hospitalized in cardiology and pulmonology units in whom PE was ruled out on the basis of a negative single-detector CT angiography. They reported a 19.6% mortality rate after a mean follow-up of 21.0 ± 11.5 months. Patients were retrospectively classified into three PE probability subgroups, with an empirical assessment of clinical probability [13]. The rate of death was higher in the intermediate- (10/37, 27%) and high-probability (10/40, 25%) groups than in the low-probability group (2/35, 6%). However, the studied population was heterogeneous. First, two-thirds of included patients were inpatients admitted for a condition other than PE at the time PE was suspected. Secondly, the duration of the follow-up period was not identical for all patients. Finally, 39% of patients received anticoagulant therapy at discharge (in similar proportions for deep venous thrombosis or cardiac reasons).

Pérez de Llano et al. [14] conducted a small monocentric study and also reported the distribution of probability group (evaluated with the original Geneva prediction rule [15]) in 87 patients in whom PE was ruled out on the basis of a negative single-detector CT angiogram. Despite an older version of the score, the distribution of PE probability was very similar to that found in this study: PE probability was classified as low, intermediate and high in 35 (40%), 47 (54%) and 5 (6%) patients, respectively.

More generally, the prognosis of patients in whom PE is ruled out is poorly known. According to published reports [3, 4, 12, 14, 16–19], the 3-month mortality ranges from 1.4% [3] to 21.5% [4]; in our study, the 3-month mortality was 3%. This relatively low mortality rate is probably due to the inclusion of outpatients alone. Despite this low 3-month mortality, differences amongst groups could be detected with the Geneva prediction rule. Prognosis amongst the groups mostly depends on comorbidities [17, 18] and alternative diagnoses. Investigators of the ANTELOPE study reported that 3-month mortality of patients in whom PE has been ruled out was associated with CT results. Indeed, the 3-month mortality was 4.1% in patients with normal CT findings and increased to 21.5% in patients with an alternative diagnosis according to the CT findings [4, 20]. In the former subgroup of patients with an alternative diagnosis based on the CT results, the 3-month readmission rate was 20% [4], which is similar to our results. However, evaluation of clinical probability was not part of the diagnostic work-up.

Unanswered questions and future research

Three questions arise from this study and deserve further examination. First, our data suggest that patients in whom PE is ruled out and who are initially classified as having a low clinical probability of PE have a good prognosis. These patients would probably benefit from an early discharge. Second, higher clinical probability classes are associated with less favourable outcomes, and patients in these classes might benefit from more careful medical management such as active monitoring for cancer or tighter control of comorbidities including COPD and chronic heart failure. Third, D-dimer levels may be increased by numerous factors (for example cancer or infection), which may have an important prognostic impact. It is possible that that a negative D-dimer result may reinforce the probability of a good prognosis in patients with a low RGS. Conversely, a positive D-dimer result may be associated with a poorer prognosis in patients with a high RGS.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Conflict of interest statement
  9. Reference

In diagnostic studies of venous thromboembolism, attention has mainly been paid to patients with confirmed PE. This study shows that in patients in whom PE was ruled out, the initial assessment of clinical probability may be used as a readily available prognostic tool. In particular, patients classified as having a low clinical probability of PE have a statistically significant better prognosis. Conversely, some patients in whom PE was suspected and ruled out may have a poor prognosis, which may be predicted by the initial assessment of clinical probability. This poor prognosis is likely to be under-recognized at present but might deserve more attention in the future because this patient population represents the vast majority of those referred to emergency departments with a clinically suspected PE.

Reference

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Conflict of interest statement
  9. Reference
  • 1
    Le Gal G, Bounameaux H. Diagnosing pulmonary embolism: running after the decreasing prevalence of cases among suspected patients. J Thromb Haemost 2004; 2: 12446.
  • 2
    Kline JA, Courtney DM, Beam DM, King MC, Steuerwald M. Incidence and predictors of repeated computed tomographic pulmonary angiography in emergency department patients. Ann Emerg Med 2009; 54: 418.
  • 3
    Wells PS, Anderson DR, Rodger M et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med 2001; 135: 98107.
  • 4
    van Strijen MJ, Bloem JL, de MW et al. Helical computed tomography and alternative diagnosis in patients with excluded pulmonary embolism. J Thromb Haemost 2005; 3: 244956.
  • 5
    Akram AR, Cowell GW, Logan LJA et al. Clinically suspected acute pulmonary embolism: a comparison of presentation, radiological features and outcome in patients with and without PE. QJM 2009; 102: 40714.
  • 6
    Torbicki A, Perrier A, Konstantinides S et al. Guidelines on the diagnosis and management of acute pulmonary embolism: the Task Force for the Diagnosis and Management of Acute Pulmonary Embolism of the European Society of Cardiology (ESC). Eur Heart J 2008; 29: 2276315.
  • 7
    Righini M, Le GG, Aujesky D et al. Diagnosis of pulmonary embolism by multidetector CT alone or combined with venous ultrasonography of the leg: a randomised non-inferiority trial. Lancet 2008; %19;371: 134352.
  • 8
    Le Gal G, Righini M, Roy PM et al. Prediction of pulmonary embolism in the emergency department: the revised Geneva score. Ann Intern Med 2006; 144: 16571.
  • 9
    De Moerloose P, Desmarais S, Bounameaux H et al. Contribution of a new, rapid, individual and quantitative automated D-dimer ELISA to exclude pulmonary embolism. Thromb Haemost 1996; 75: 113.
  • 10
    Klok FA, Mos IC, Nijkeuter M et al. Simplification of the revised Geneva score for assessing clinical probability of pulmonary embolism. Arch Intern Med 2008; 168: 21316.
  • 11
    Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA 2001; 285: 286470.
  • 12
    Bourriot K, Couffinhal T, Bernard V, Montaudon M, Bonnet J, Laurent F. Clinical outcome after a negative spiral CT pulmonary angiographic finding in an inpatient population from cardiology and pneumology wards. Chest 2003; 123: 35965.
  • 13
    Perrier A, Bounameaux H, Morabia A et al. Diagnosis of pulmonary embolism by a decision analysis-based strategy including clinical probability, D-dimer levels, and ultrasonography: a management study. Arch Intern Med 1996; 156: 5316.
  • 14
    Perez de Llano LA, Veres RA, Ortiz PM et al. Safety of withholding anticoagulant therapy in patients who have clinically suspected pulmonary embolism and negative results on helical computed tomography. Respiration 2006; 73: 5149.
  • 15
    Wicki J, Perneger TV, Junod AF, Bounameaux H, Perrier A. Assessing clinical probability of pulmonary embolism in the emergency ward: a simple score. Arch Intern Med 2001; 161: 927.
  • 16
    Donato AA, Scheirer JJ, Atwell MS, Gramp J, Duszak R Jr. Clinical outcomes in patients with suspected acute pulmonary embolism and negative helical computed tomographic results in whom anticoagulation was withheld. Arch Intern Med 2003; 163: 20338.
  • 17
    van Beek EJ, Kuijer PM, Buller HR, Brandjes DP, Bossuyt PM, ten Cate JW. The clinical course of patients with suspected pulmonary embolism. Arch Intern Med 1997; 157: 25938.
  • 18
    Poulsen SH, Noer I, Moller JE, Knudsen TE, Frandsen JL. Clinical outcome of patients with suspected pulmonary embolism. A follow-up study of 588 consecutive patients. J Intern Med 2001; 250: 13743.
  • 19
    Swensen SJ, Sheedy PF, Ryu JH et al. Outcomes after withholding anticoagulation from patients with suspected acute pulmonary embolism and negative computed tomographic findings: a cohort study. Mayo Clin Proc 2002; 77: 1308.
  • 20
    van Strijen MJ, de MW, Schiereck J et al. Single-detector helical computed tomography as the primary diagnostic test in suspected pulmonary embolism: a multicenter clinical management study of 510 patients. Ann Intern Med 2003; 138: 30714.