CLINICAL RESEARCH STUDY. AJM Theme Issue: Pulmonology/Allergy

Transcription

1 The American Journal of Medicine (2006) 119, CLINICAL RESEARCH STUDY Pulmonary Embolism in Patients with Chronic Obstructive Pulmonary Disease or Congestive Heart Failure Manuel Monreal, PhD, a Juan Francisco Sanchez Muñoz-Torrero, MD, b Virjanand S. Naraine, MD, c David Jiménez, MD, d Silvia Soler, MD, e Ramón Rabuñal, MD, f Pedro Gallego, MD; g RIETE Investigators. a Servicio de Medicina Interna, Hospital Universitari Germans Trias i Pujol, Badalona, Spain; b Servicio de Medicina Interna, Hospital San Pedro de Alcántara, Cáceres, Spain; c Division of Respirology, University of Toronto, Canada; d Servicio de Neumología, Hospital Universitario Ramón y Cajal, Madrid, Spain; e Servicio de Medicina Interna, Hospital Sant Jaume, Gerona, Spain; f Servicio de Medicina Interna, Complejo Hospitalario Xeral-Calde, Lugo, Spain; g Servicio de Medicina Interna, Hospital SAS de Jerez, Cádiz, Spain. ABSTRACT AJM Theme Issue: Pulmonology/Allergy BACKGROUND: The diagnosis of pulmonary embolism (PE) is often unreliable in patients with chronic obstructive pulmonary disease (COPD) or congestive heart failure (CHF). SUBJECTS AND METHODS: Registro Informatizado de la Enfermedad TromboEmbólica (RIETE) is an ongoing registry of consecutive patients with acute venous thromboembolism. In this study, the clinical characteristics, laboratory findings, and clinical outcomes of all enrolled patients with acute PE, with or without underlying cardiopulmonary diseases, were compared and contrasted. In addition, the performance of 2 clinical models for the diagnosis of PE was retrospectively evaluated. RESULTS: As of January 2005, 4444 patients with symptomatic PE have been enrolled in RIETE. Of those, 632 patients (14%) had COPD and 422 (9.5%) had CHF. Significant differences were found in clinical presentation and 3-month outcomes among the 3 groups. With the Geneva model, there was a lower percentage of PE patients with COPD (relative risk [RR] 0.82; 95% confidence interval [CI], ) or CHF (RR 0.73; 95% CI, ) who fell into the low pretest probability category, compared with patients with neither. Besides, the percentage of patients with high probability of PE was similar among the 3 patient groups. The frequency of COPD (61%) and CHF (72%) patients with a high pretest probability for PE increased when using the Pisa score, but the percentage of COPD patients into the high probability group was lower (RR 0.60; 95% CI, ). CONCLUSIONS: Significant differences exist in PE patients with and without underlying cardiopulmonary diseases. The performance of the 2 clinical prediction models varied according to the presence or absence of underlying COPD or CHF Elsevier Inc. All rights reserved. KEYWORDS: Pulmonary embolism; Chronic lung disease; Chronic heart failure; Diagnosis Patients with chronic obstructive pulmonary disease (COPD) or congestive heart failure (CHF) are often admitted to the hospital with an exacerbation of their disease that manifests itself with increased dyspnea, chest pain and ankle edema. Both COPD and CHF are considered risk factors Supported by Sanofi-Aventis, Paris, France with an unrestricted educational grant, and partially supported by Red Respira from the Instituto Carlos III, Madrid, Spain (RedRespira-ISCiii-RTIC-03/11). A full list of RIETE investigators is given in the Appendix. Requests for reprints should be addressed Manuel Monreal, PhD, Servicio de Medicina Interna, Hospital Universitari Germans Trias i Pujol, Badalona, Barcelona, Spain. address: mmonreal.germanstrias@gencat.net for pulmonary embolism (PE), but the symptoms of these conditions overlap considerably, and the investigation of PE is often ignored or delayed in these patients. A coexisting COPD or CHF may be assumed to be the cause of the patient s symptoms, and the presence of PE may go undiagnosed in patients who can least tolerate it. Clinical guidelines issued by the American Thoracic Society and the European Society of Cardiology recommend assessing the clinical probability of acute PE as a guide to decision-making and management. 1,2 However, the clinical validity of the 2 most commonly used clinical models in predicting the pretest clinical probability of PE has not been validated in patients with either COPD or CHF /$ -see front matter 2006 Elsevier Inc. All rights reserved. doi: /j.amjmed

2 852 The American Journal of Medicine, Vol 119, No 10, October 2006 The Registro Informatizado de la Enfermedad Trombo- Embólica (RIETE) was initiated in March 2001 to record current clinical management of venous thromboembolism (VTE) within Spanish hospitals. It is a multicenter, observational registry designed to gather and analyze data on treatment patterns and clinical outcomes in consecutive patients with symptomatic, objectively confirmed, acute venous thrombosis (DVT) or PE. 3-6 In this analysis, the clinical characteristics, laboratory findings, and 3-month clinical outcomes of all enrolled patients with acute PE, with or without underlying cardiopulmonary diseases, were compared and contrasted. In addition, the performance of 2 structural models 7,8 in predicting the pretest clinical probability of PE in patients with or without underlying COPD or CHF was evaluated, within the limits of the registry s data. CLINICAL SIGNIFICANCE PATIENTS AND METHODS Inclusion and Exclusion Criteria Patients with symptomatic, acute DVT or PE, confirmed by objective tests (ie, contrast venography, ultrasonography, or impedance plethysmography for suspected DVT; pulmonary angiography, lung scintigraphy, or helical computed tomography [CT] scan for suspected PE) were consecutively enrolled in RIETE. Patients were excluded if they were participating in a therapeutic clinical trial or unavailable for follow-up. For this analysis, only patients with PE were considered. Patient Population Patients were divided into 3 groups: those with underlying COPD, those with CHF, and those without COPD or CHF. All patients provided oral consent to their participation in the registry, in accordance with the requirements of the ethics committee within each hospital. Study Parameters The parameters recorded by the registry comprise details of each patient s baseline characteristics; clinical status including any coexisting or underlying conditions such as chronic heart or lung disease; risk factors; clinical characteristics of the thrombotic event; laboratory findings including data on the electrocardiogram, chest radiograph, arterial blood gases, d-dimer levels, and other diagnostic tests; treatment received upon PE diagnosis; and clinical outcome during the first 3 months of therapy. Data were obtained from medical records and recorded on case report forms by a study coordinator. Coexisting medical conditions or comorbidities were specified according to a prespecified list. Significant differences exist in the clinical presentation of pulmonary embolism patients with and without underlying cardio-pulmonary diseases. The performance of the 2 clinical prediction models varied according to the presence or absence of underlying chronic obstructive pulmonary disease or congestive heart failure. Validation of PE Diagnosis All patients had acute respiratory symptoms suggesting PE. The diagnosis was considered definitive if patients also had a positive helical CT scan, a high-probability ventilation-perfusion lung scintigraphy, a positive pulmonary angiography, visualization of thrombus on echocardiogram, or indeterminate-probability lung scan plus evidence of DVT in the lower limbs (by either, compression ultrasonograpy or contrast venography). Follow-up All patients were followed-up for at least 3 months after hospital discharge. During each visit, any signs or symptoms suggesting recurrences of DVT or PE, or bleeding complications were noted. Each episode of clinically suspected recurrent DVT or PE was documented by repeat compression ultrasonography, venography, lung scanning, helical CT scan, or pulmonary angiography. Fatal PE was defined as death shortly after PE diagnosis and in the absence of any alternative cause of death. Bleeding complications were classified as major if they were overt and were either associated with a decrease in the hemoglobin level of 2.0 g/dl (20 g/l) or more, required a transfusion of 2 units of blood or more, or were retroperitoneal or intracranial. Any other clinically relevant bleeding events were considered minor. Data Collection Data were recorded onto a computer-based case report form by a RIETE registry coordinator at each participating hospital and submitted to a centralized coordinating center through a secure website. The coordinators also ensured that eligible patients were consecutively enrolled. Patient identities remained confidential because they were identified by a unique number assigned by the study coordinator center, which was responsible for all data management. Study endpoints were adjudicated by the RIETE registry coordinators. At regular intervals, data quality was monitored and documented electronically to detect inconsistencies or errors, which were resolved by the coordinators. Data quality was also monitored by periodic visits to participating hospitals by contract research organizations that compare the medical records with the data on the secure website, as is the case for most clinical trials. In the event of substantial or unjustifiable inconsistencies from a particular center, patients enrolled from that center were not included in the database. A full data audit was performed at periodic intervals. Clinical Models The clinical probability of PE was retrospectively estimated according to 2 logistic regression models. 7,8 A third model,

3 Monreal et al PE in Patients with COPD or CHF 853 Table 1 Diagnostic Tests Performed in 4444 Consecutive Patients with PE, According to the Existence or Absence of Underlying COPD or CHF Helical CT scan n 2009 (60%) n 356 (57%) n 202 (49%) Pulmonary embolism 1869 (93%) 324 (91%) 181 (90%) Lung scan n 1709 (51%) n 328 (53%) n 234 (56%) Normal 31 (1.8%) 4 (1.2%) 2 (0.9%) Low probability 120 (7.0%) 27 (8.2%) 16 (6.8%) Intermediate probability 111 (6.5%) 69 (21%) 31 (13%) High probability 1446 (85%) 228 (70%) 185 (79%) Echocardiogram n 456 (30%) n 78 (28%) n 65 (38%) Visualization of thrombus 14 (3.1%) 6 (7.8%) 2 (3.1%) Right atrial dilatation 150 (34%) 30 (40%) 30 (47%) Right ventricle 113 (26%) 18 (25%) 17 (27%) hypokinesia Pulmonary angiography n 45 (1.4%) n 16 (2.6%) n 0 Pulmonary embolism 28 (88%) 14 (93%) Ultrasonography n 2144 (70%) n 394 (62%) n 237 (65%) Venous thrombosis 1352 (63%) 259 (66%) 145 (61%) Contrast venography n 204 (6.7%) n 45 (8.2%) n 21 (5.9%) Venous thrombosis 135 (67%) 27 (60%) 15 (71%) Impedance plethysmography n 37 (1.2%) n 7 (1.3%) n 7 (2.0%) Venous thrombosis 27 (75%) 4 (57%) 4 (57%) Lower-limb CT scan n 138 (11%) n 38 (18%) n 9 (6.9%) Venous thrombosis 103 (77%) 29 (78%) 8 (89%) D-dimer levels n 2406 (71%) n 384 (61%) n 300 (71%) Increased 2290 (95%) 357 (93%) 280 (93%) COPD chronic obstructive pulmonary disease; CHF congestive heart failure. the Wells score, 9 was not used because the RIETE registry did not identify the presence of an alternative diagnosis more likely than PE. The following items were not recorded in the RIETE database: elevated hemidiaphragm for the Geneva model and oligemia or amputation of the hilar artery for the Pisa model. Accordingly, we did not consider the presence of an elevated hemidiaphragm, and we hypothesized that all patients with vascular redistribution signs in their chest radiograph had both oligemia and amputation of the hilar artery. Statistical Analysis The clinical characteristics, laboratory data, and clinical outcomes of PE patients, with or without underlying COPD or CHF, were compared. A commercial software package (SPSS version 11.5; SPSS Inc., Chicago, Ill) was used to calculate relative risks and corresponding 95% confidence intervals (CI), and a P value.01 was considered to be statistically significant. RESULTS As of January 2005, 4444 patients with acute, symptomatic, objectively confirmed PE have been enrolled in RIETE: 2066 males and 2378 females aged 15 to 99 years (mean 69 years). Of them, 632 patients (14%) had COPD, and 422 (9.5%) had CHF. Diagnostic Methods The diagnostic methods used to confirm the diagnosis of PE are depicted in Table 1. PE diagnosis was confirmed in 2374 patients with a positive CT scan, 1859 with a high-probability ventilation-perfusion lung scan, 42 with a positive angiogram, 22 with visualization of a thrombus on the echocardiogram, and 147 patients with intermediate-probability lung scan plus evidence of DVT in the lower limbs. There were no significant differences in the rates of positive CT scan among the three patient groups: it was positive in 91% of patients with underlying COPD, in 90% of those with CHF, and in 93% of patients with no COPD or CHF. By contrast, lung scan revealed highprobability defects in 85% of patients without COPD or CHF, but in only 70% of those with COPD (relative risk [RR] 0.5; 95% CI, ; P.001); and 79% in patients with CHF (RR 0.7; 95% CI, ; P.03). There were no differences among groups in the rates of objectively confirmed DVT or in the d-dimer levels. Clinical Characteristics and Laboratory Findings Patients with COPD were significantly older, were more commonly males, and had recent immobility more often than those with no COPD or CHF (Table 2). Dyspnea and cough were more often present, whereas syncope was less common. Electrocardiographic abnormalities such as

4 854 The American Journal of Medicine, Vol 119, No 10, October 2006 Table 2 Clinical Characteristics, Risk Factors for Venous Thromboembolism, and Laboratory Data of the 4444 Patients with PE RR (95% CI) COPD RR (95% CI) CHF Clinical characteristics n 3390 n 632 n 422 Sex (males) 1501 (44%) 435 (69%) 130 (31%) 2.4 ( ) 0.6 ( ) Age 70 years 1687 (50%) 408 (65%) 360 (85%) 1.7 ( ) 5.0 ( ) Outpatients 2403 (73%) 434 (71%) 281 (68%) 0.9 ( ) 0.8 ( )* Risk factors for VTE n 3390 n 632 n 422 Previous VTE 527 (16%) 100 (16%) 61 (14%) 1.0 ( ) 0.9 ( ) Cancer 708 (21%) 109 (17%) 44 (10%) 0.8 ( )* 0.5 ( ) Surgery 550 (16%) 70 (11%) 35 (8.3%) 0.7 ( ) 0.5 ( ) Immobility 3 days 735 (22%) 210 (33%) 172 (41%) 1.6 ( ) 2.2 ( ) VTE symptoms n 3390 n 632 n 422 Dyspnea 2761 (82%) 559 (89%) 381 (90%) 1.7 ( ) 2.0 ( ) Chest pain 1830 (54%) 317 (51%) 181 (43%) 0.9 ( ) 0.7 ( ) Hemoptysis 226 (6.7%) 54 (8.6%) 24 (5.7%) 1.3 ( ) 0.9 ( ) Cough 584 (17%) 203 (32%) 96 (23%) 1.9 ( ) 1.4 ( ) Syncope 513 (15%) 64 (10%) 67 (16%) 0.7 ( ) 1.0 ( ) DVT signs 1289 (38%) 233 (37%) 139 (33%) 1.0 ( ) 0.8 ( )* Fever 38 C 416 (12%) 78 (12%) 36 (8.6%) 1.0 ( ) 0.7 ( )* ECG findings n 3067 n 548 n 381 Atrial fibrillation 228 (7.4%) 67 (12%) 123 (32%) 1.6 ( ) 4.2 ( ) S 1 Q 3 T 3 pattern 536 (18%) 67 (12%) 57 (15%) 0.7 ( ) 0.8 ( ) Right bundle branch block 474 (18%) 127 (23%) 81 (21%) 1.3 ( ) 1.2 ( ) Abnormal repolarization 631 (21%) 102 (19%) 109 (29%) 0.9 ( ) 1.5 ( ) Any of the above 1394 (46%) 269 (49%) 248 (65%) 1.2 ( ) 2.1 ( ) Chest radiograph findings n 3098 n 555 n 380 Atelectasis 356 (12%) 73 (13%) 29 (7.8%) 1.1 ( ) 0.7 ( )* Cardiomegaly 640 (21%) 166 (30%) 236 (62%) 1.5 ( ) 4.8 ( ) Pleural effusion 624 (20%) 106 (19%) 118 (31%) 0.95 ( ) 1.7 ( ) Pulmonary infarction 213 (7.0%) 28 (5.1%) 19 (5.0%) 0.8 ( ) 0.7 ( ) Pulmonary infiltrate 566 (19%) 129 (23%) 90 (24%) 1.3 ( ) 1.3 ( )* Vascular redistribution 263 (8.6%) 86 (16%) 95 (25%) 1.7 ( ) 2.9 ( ) Any of the above 1684 (54%) 365 (66%) 308 (81%) 1.5 ( ) 3.2 ( ) Arterial blood gases n 2716 n 543 n 363 PO 2 60 mm Hg 1103 (41%) 281 (52%) 183 (50%) 1.5 ( ) 1.4 ( ) PCO 2 49 mm Hg 78 (2.9%) 65 (12%) 34 (9.3%) 3.0 ( ) 2.7 ( ) COPD chronic obstructive pulmonary disease; CHF congestive heart failure; VTE venous thromboembolism; DVT deep vein thrombosis; BBB bundle branch block; RR relative risk; CI confidence intervals. Comparisons among patients: *P.05; P.01; P.001. atrial fibrillation or right bundle branch block appeared more often, but the S 1 Q 3 T 3 pattern appeared less often than in patients with no COPD or CHF. Chest radiograph was more often abnormal in patients with COPD, with both enlarged cardiac size and vascular redistribution signs appearing more often. They also had a higher frequency of hypoxemia and hypercapnia. Patients with CHF also were significantly older than those with no COPD or CHF and were more likely to have had recent immobilization for 3 days. However, they were more commonly females and had cancer or recent surgery less often. Dyspnea and cough were again significantly more frequent, but they had chest pain less often. Electrocardiographic and chest radiograph abnormalities were more commonly found, as were hypoxemia and hypercapnia. Clinical Outcomes In total, 491 patients (11%) died during the 3-month study period. Overall mortality and fatal PE rates were significantly higher in patients with CHF (17% and 6.6%, respectively) than in the other two groups (12% and 4.6%, respectively, in those with COPD; 10% and 3.5%, respectively, in those with no COPD or CHF), as shown in Table 3. Bleeding complications were significantly more common in patients with COPD. There were no differences in the recurrence rate among the 3 patient groups. Clinical Prediction Models With the Geneva score, there was a lower percentage of PE patients with COPD (RR 0.82, 95% CI, ; P.072)

5 Monreal et al PE in Patients with COPD or CHF 855 Table 3 Clinical Outcomes of the Patients during the First 3 Months of Anticoagulant Therapy* RR (95% CI) COPD RR (95% CI) CHF Patients, n Total bleeding 204 (6.0%) 59 (9.3) 32 (7.6%) 1.5 ( ) 1.2 ( ) Fatal bleeding 16 (0.5%) 5 (0.8%) 4 (0.9%) 1.5 ( ) 1.8 ( ) Major bleeding 105 (3.1%) 27 (4.3%) 13 (3.1%) 1.3 ( ) 1.0 ( ) Minor bleeding 99 (2.9%) 32 (5.1%) 19 (4.5%) 1.6 ( ) 1.5 ( ) Fatal PE 117 (3.5%) 29 (4.6%) 28 (6.6%) 1.3 ( ) 1.8 ( ) Fatal initial PE 97 (2.9%) 24 (3.8%) 24 (5.7%) 1.3 ( ) 1.8 ( ) Fatal recurrent PE 20 (0.6%) 5 (0.8%) 4 (0.9%) 1.3 ( ) 1.5 ( ) Recurrent VTE 92 (2.7%) 17 (2.7%) 14 (3.3%) 1.0 ( ) 1.2 ( ) Recurrent DVT 42 (1.2%) 7 (1.1%) 6 (1.4%) 0.9 ( ) 1.1 ( ) Recurrent PE 50 (1.5%) 10 (1.6%) 8 (1.9%) 1.1 ( ) 1.3 ( ) Overall death 340 (10%) 78 (12%) 73 (17%) 1.2 ( ) 1.7 ( ) COPD chronic obstructive pulmonary disease; CHF congestive heart failure; PE pulmonary embolism; VTE venous thromboembolism; RR relative risk; CI confidence intervals. Comparisons among patients: *P.05; P.01; P.001. or CHF (RR 0.73, 95% CI, ; P.019) who fell into the low pretest probability category, compared with patients with neither (Table 4). Besides, the percentage of patients with high-probability of PE was similar among the 3 patient groups. The increased frequency of patients aged 60 years or with hypoxemia among those with COPD or CHF was counterbalanced by their lower frequency of recent surgery or hypocapnia. With the Pisa model, no patients in the 3 groups were categorized as having low probability for PE (Table 5). In addition, the percentage of COPD patients in the high probability group was significantly lower than in the other 2 groups, mostly due to the increased frequency of males. DISCUSSION Our data, obtained from a large prospective series of consecutive patients presenting with acute, symptomatic PE, revealed that there were significant differences in the clinical presentation for patients with or without underlying COPD or CHF. Although the differences among the 3 groups appeared to be quite striking, we cannot exclude that many of these differences in clinical signs or symptoms, arterial blood gas results, or abnormalities in chest radiographs or electrocardiogram may, in fact, be attributed to the underlying disease processes rather than the acute PE. To complicate matters further, both illnesses lead to prolonged bed rest and are independent risk factors for PE. Indeed, approximately 5% to 10% of patients with an acute COPD exacerbation have a concomitant PE. 10,11 However, PE has been more frequently found (up to 30% of cases) in autopsic series that included patients who died from acute exacerbation of COPD. 12 The discrepancy in prevalence rates may reflect an underestimation in patients with COPD (or other cardiopulmonary diseases) who have their decompensation incorrectly attributed to progression of their underlying disease. Not surprisingly, overall mortality was higher in the COPD (12%) and CHF (17%) patients compared with those without the two diagnoses (10%). Patients with COPD or CHF have poor reserve and are less likely to tolerate the PE event. It is noteworthy that 5.7% of the CHF sub-group had an initial fatal PE (odds ratio 2.9; 95% CI, ). In our opinion, our findings add to our understanding of the PE process and the variations of presentations in different populations. Patients with COPD or CHF often present diagnostic challenges in the workup of PE Although other observational studies have looked at patients with COPD or CHF and presentation or validity of common tests for PE, most studies have been small. The exception is the ANTELOPE study, which looked at the validity of lung scans and angiograms, but it did not compare clinical presentations or outcomes of those with or without COPD or CHF. 18 Given the lack of a single diagnostic test or clinical finding with adequate sensitivity and specificity, the diagnosis of PE generally involves combined interpretation of multiple data points to obtain an estimated pretest probability. Clinical and laboratory data can be combined to stratify patients into three categories (low, intermediate, and high clinical probability) reflecting increasing likelihood of PE. In our experience, the 2 clinical models for PE diagnosis performed quite differently with regard to their predictive accuracy. In the present study, the Geneva score turned out to have low sensitivity for the diagnosis of PE: only 11% of PE patients with COPD, and 9.1% of patients with CHF fell into the high pretest category. The relative inaccuracy of this model may be due to the fact that it was derived from a database of outpatients presenting to the emergency de-

6 856 The American Journal of Medicine, Vol 119, No 10, October 2006 Table 4 Application of the Geneva Score for the 2785 PE Patients with All the Data RR (95% CI) COPD RR (95% CI) CHF Patients, n n 2101 n 393 n 297 Age, years (50%) 240 (61%) 137 (46%) 1.5 ( ) 0.9 ( ) (24%) 125 (32%) 149 (50%) 1.4 ( ) 2.7 ( ) Previous DVT or PE 344 (16%) 59 (15%) 42 (14%) 0.9 ( ) 0.9 ( ) Recent surgery 331 (16%) 44 (11%) 26 (8.8%) 0.7 ( )* 0.5 ( ) Pulse rate 100 /min 639 (30%) 121 (31%) 80 (27%) 1.0 ( ) 0.9 ( ) PaCO 2 36 mm Hg 1321 (63%) 198 (50%) 154 (52%) 0.7 ( ) 0.7 ( ) mm Hg 389 (19%) 55 (14%) 43 (14%) 0.8 ( )* 0.8 ( ) PaO 2 49 mm Hg 311 (15%) 74 (19%) 50 (17%) 1.3 ( )* 1.1 ( ) mm Hg 545 (26%) 132 (34%) 106 (36%) 1.4 ( ) 1.5 ( ) mm Hg 661 (32%) 111 (28%) 73 (25%) 0.9 ( ) 0.7 ( )* mm Hg 343 (16%) 51 (13%) 40 (14%) 0.8 ( ) 0.8 ( ) Chest radiograph Platelike atelectasis 250 (12%) 56 (14%) 19 (6.4%) 1.2 ( ) 0.5 ( ) Elevation of hemidiaphragm Clinical probability Low 573 (27%) 90 (23%) 62 (21%) 0.8 ( ) 0.7 ( )* Intermediate 1304 (62%) 259 (66%) 208 (70%) 1.2 ( ) 1.4 ( ) High 224 (11%) 44 (11%) 27 (9.1%) 1.0 ( ) 0.9 ( ) COPD chronic obstructive pulmonary disease; CHF congestive heart failure; PE pulmonary embolism; VTE venous thromboembolism; RR relative risk; CI confidence intervals. Comparisons among patients: *P.05; P.01; P.001. partment with suspected PE. In the original report, patients with PE featured older age, higher prevalence of predisposing risk factors, and more severe arterial blood gas abnormalities than those without PE. 8 The Geneva model therefore seems better suited for evaluating outpatients who have a low prevalence of risk factors, such as recent surgery or immobility, and of comorbid conditions that may affect pulmonary gas exchange. The Pisa score had a higher proportion of PE patients who fell into the high probability group. However, the 61% percentage of COPD patients who fell into the high probability group was significantly lower than the 74% of patients without underlying COPD or CHF, or the 72% in patients with CHF. This has to be considered when suspecting PE in patients with COPD. The main limitation of the present study is its design, which contains several sources of potential bias. First, RIETE is a registry, and all patients had symptomatic, objectively confirmed PE. Accordingly, it is not possible to assess specificity of the scoring system. Although the Pisa system had the highest proportion of PE patients in the high probability group, it might be that similar COPD and CHF patients, as well as patients with neither, who did not have PE, might have similar scores. Second, patients were not assigned to follow a strict protocol but underwent the diagnostic tests of their doctor s choice. Third, the diagnosis of COPD and CHF was based on clinical information, not on objective methods, so it is possible that the diagnosis had been over- or underestimated in some patients. Nevertheless, similar rates have been reported in a recent registry performed in the United States. 20 Fourth, the Registry does not allow one to fill both COPD and CHF diagnoses simultaneously, so the treating physician had to select which of them was most relevant. Finally, some of the variables included in the prediction models (ie, elevated hemidiaphragm; oligemia or amputation of the hilar artery on chest radiograph) are not recorded in our database, making it impossible to strictly compare different subgroups. Furthermore, our assumption that all patients with vascular redistribution signs in their chest radiograph had both oligemia and amputation of the hilar artery may be wrong. This is the reason why we cannot conclude from this study that one specific score is better than the other. However, the strength of this report is the prospective collection of data from actual practice, from a very large number of consecutive patients with objectively confirmed PE, and by strictly applying objective criteria for diagnosis of PE. The goal of RIETE ( is to improve the treatment of PE patients through a better understanding of demographics, management, and in-hospital and postdischarge outcomes. Starting as a Spanish initiative in March 2001, it has since been opened to other countries, thus becoming an international registry. Data captured and reported in the registry will therefore reflect real-world approaches and outcomes in the treatment of PE.

7 Monreal et al PE in Patients with COPD or CHF 857 Table 5 Application of the Pisa Model for the 3700 PE Patients with All the Data RR (95% CI) COPD RR (95% CI) CHF Patients, n n 2834 n 504 n 362 Male sex 1248 (44%) 355 (70%) 113 (31%) 2.5 ( ) 0.6 ( ) Age, years (24%) 153 (30%) 55 (15%) 1.3 ( ) 0.6 ( ) (45%) 286 (56%) 290 (80%) 1.5 ( ) 4.2 ( ) Prior cardiovascular disease (100%) - - Prior pulmonary disease (100%) Prior DVT or PE 449 (16%) 84 (17%) 49 (13%) 1.0 ( ) 0.8 ( ) Dyspnea 2334 (82%) 456 (90%) 330 (91%) 1.8 ( ) 2.0 ( ) Chest pain 1564 (55%) 262 (52%) 162 (45%) 0.9 ( ) 0.7 ( ) Hemoptysis 193 (6.8%) 41 (8.1%) 21 (5.8%) 1.2 ( ) 0.9 ( ) Fever 38 C 337 (12%) 56 (11%) 30 (8.3%) 0.9 ( ) 0.7 ( )* ECG: acute RV overload 1164 (41%) 207 (41%) 177 (49%) 1.0 ( ) 1.3 ( ) Chest radiograph Oligemia amputation hilar 229 (8.1%) 74 (15%) 89 (25%) 1.7 ( ) 2.9 ( ) artery Consolidation (infarction) 186 (6.5%) 27 (5.3%) 19 (5.2%) 0.8 ( ) 0.8 ( ) Consolidation (no 486 (17%) 113 (22%) 83 (23%) 1.3 ( ) 1.4 ( ) infarction) Pulmonary edema N.A. N.A. N.A. N.A. N.A. Clinical probability Low 0 (0%) 0 (0%) 0 (0%) - - Intermediate 728 (26%) 195 (39%) 102 (28%) 1.7 ( ) 1.1 ( ) High 2106 (74%) 309 (61%) 260 (72%) 0.6 ( ) 0.9 ( ) COPD chronic obstructive pulmonary disease; CHF congestive heart failure; PE pulmonary embolism; DVT deep venous thrombosis; ECG electrocardiogram; RV right ventricle; N.A. not available; RR relative risk; CI confidence intervals. Comparisons among patients: *P.05; P.01; P.001. In summary, significant differences exist in the clinical presentation of PE patients with and without underlying cardiopulmonary diseases. The performance of the two clinical prediction models varied according to the presence or absence of underlying COPD or CHF. ACKNOWLEDGMENTS We would like to thank Salvador Ortíz, Professor Universidad Autónoma de Madrid and Statistical Advisor, S & H Medical Science Service for the statistical analysis of the data presented in this article. We thank the Registry Coordinating Center and S&H Medical Science Service for their logistic and administrative support. APPENDIX Members of the RIETE Group: J.C. Alvárez, M.R. Gutiérrez, E. Laserna, R. Otero (Sevilla); J.I. Arcelus, I. Casado (Granada); M. Barrón (La Rioja); J.L. Beato (Albacete); J. Bugés, C. Falgá, M. Monreal, F.J. Muñóz, A. Raventós, C. Tolosa (Barcelona); R. Barba, J. del Toro, C. Fernández- Capitán, J. Gutiérrez, D. Jiménez, P. Rondón, C. Suárez (Madrid); A. Blanco, L. López, R. Tirado (Córdoba); J. Bosco, P. Gallego, M.J. Soto (Cádiz); M.A. Cabezudo, I. López, (Asturias); J.M. Calvo (Badajoz); F. Conget (Zaragoza); J.A. Escobedo, J.L. Pérez-Burkhardt (Tenerife); F. Gabriel, E. Grau, F. López, M.D. Naufall, P. Román, J.A. Todolí (Valencia); F. García Bragado, A. Grau, S. Soler (Girona); R. Guijarro, M. Guil, J.J. Martín, (Málaga); L. Hernández, A. Maestre, R. Sánchez (Alicante); R. Lecumberri, M.T. Orue, A.L. Sampériz, G. Tiberio (Navarra); J.L. Lobo (Vitoria); J. Montes, M.J. Núñez (Vigo); J.A. Nieto (Cuenca); M.A. Page, J. Trujillo (Murcia); J. Portillo (Ciudad Real); R. Rabuñal (Lugo); J.F. Sánchez (Cáceres); J.A. Torre, B. Varela (A Coruña); F. Uresandi (Bilbao); R. Valle (Cantabria); and X. Llobet (Medical Department, Sanofi Aventis). References 1. Tapson V, Carroll BA, Davidson BL, et al. The diagnostic approach to acute venous thromboembolism: clinical practice guideline. Am J Respir Crit Care Med. 1999;160: Guidelines on diagnosis and management of acute pulmonary embolism. Eur Heart J. 2000;21: Arcelus JI, Monreal M, Caprini JA, Suárez C, González-Fajardo JA, and the RIETE investigators. The management and outcome of acute venous thromboembolism: a prospective registry including 4011 Patients. J Vasc Surg. 2003;38:

8 858 The American Journal of Medicine, Vol 119, No 10, October Monreal M, Kakkar AK, Caprini JA, et al, the RIETE Investigators. The outcome after treatment of venous thromboembolism is different in surgical and acutely ill medical patients. Findings from the RIETE Registry. J Thromb Haemost. 2004,2: Monreal M, Suárez C, González Fajardo JA, et al, the RIETE investigators. Management of patients with acute venous thromboembolism: findings from the RIETE Registry. J Pathophysiol Haemost Thromb. 2003/2004;33: Nieto JA, Díaz de Tuesta A, Marchena PJ, et al, the RIETE investigators. Clinical outcome of patients with venous thromboembolism and recent major bleeding: findings from a prospective registry (RIETE). J Thromb Haemost. 2005;3: 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: Miniati M, Monti S, Bottai M. A structured clinical model for predicting the probability of pulmonary embolism. Am J Med. 2003;114: Wells PS, Anderson DR, Rodger M, et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the simplered D-dimer. Thromb Haemost. 2000;83: Perrier A, Perneger T, Cornuz J, Jounieaux V, Bounameaux H. Édude BPCO-EP: prévalence et prédiction de l embolie pulmonaire dans = exacerbation de la broncho-pneumopathie obstructive chronique. Rev Mal Respir. 2004;21: Erelel M, Çuhadaroglu Ç, Ece T, Arseven O. The frequency of deep venous thrombosis and pulmonary embolus in acute exacerbation of chronic obstructive pulmonary disease. Respir Med. 2002;96: Prescott SM, Richards KL, Tikoff G, Armstrong JD Jr, Shigeoka JW. Venous thromboembolism in decompensated chronic obstructive pulmonary disease. A prospective study. Am Rev Respir Dis. 1981;123: Fedullo PF, Tapson VF. The evaluation of suspected pulmonary embolism. N Engl J Med. 2003;349: Stein PD, Terrin ML, Hales CA, et al. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no preexisting cardiac or pulmonary disease. Chest. 1991;100: Laack TA, Goyal DG. Pulmonary embolism: an unsuspected killer. Emerg Med Clin Am. 2004;22: Stebbings AEL, Lim TK. A patient with exacerbation of COPD who did not respond to conventional treatment. Chest. 1998;114: Tamariz LJ, Eng J, Segal JB, et al. Usefulness of clinical prediction rules for the diagnosis of venous thromboembolism: a systematic review. Am J Med. 2004;117: Hartmann IJC, Hagen PJ, Melissant CF, Postmus PE, Prins MH on behalf of the ANTELOPE Study Group. Diagnosing acute pulmonary embolism. Effect of chronic obstructive pulmonary disease on the performance of d-dimer testing, ventilation/perfusion scintigraphy, spiral computed tomographic angiography, and conventional angiography. Am J Respir Crit Care Med. 2000;162: Lesser BA, Leeper KV, Stein PD, et al. The diagnosis of acute pulmonary embolism in patients with chronic obstructive pulmonary disease. Chest. 1992;102: Goldhaber SZ, Tapson VF, for the DVT FREE Steering Committee. A prospective Registry of 5,451 patients with ultrasound-confirmed deep vein thrombosis. Am J Cardiol. 2004;93: