Volume 29, Issue 1 , Pages 26-32, January 2011
The role of risk factors in delayed diagnosis of pulmonary embolism
Article Outline
Abstract
Background
Despite diagnostic advances, delays in the diagnosis of pulmonary embolism (PE) are common.
Objective
In this study, we aimed to investigate the relationship between delays in the diagnosis of PE and underlying risk factors for PE.
Methods
We retrospectively evaluated the records of 408 patients with acute PE. Patients were divided into 2 groups, surgical or medical, based on risk factors leading to the embolism. Analysis involved demographic characteristics of the patients, dates of symptom onset, first medical evaluation, and confirmatory diagnostic tests. Diagnostic delay was described as diagnosis of PE more than 1 week after symptom onset.
Results
The mean time to diagnosis for all patients was 6.95 ± 8.5 days (median, 3 days; range, 0-45 days). Of the total number of patients, 29.6% had presented within the first 24 hours and 72.3% within the first week. The mean time to diagnosis was 4.4 ± 7.6 days (median, 2 days; range, 0-45 days) in the surgical group and 8.0 ± 8.6 days (median, 4 days; range, 0-45 days) in the medical group (P = .000). The mean time to diagnosis in the medical group was approximately 4 times greater than that of the surgical group on univariate analysis. Early or delayed diagnosis had no significant impact on mortality in either group.
Conclusion
Delay in the diagnosis of PE is an important issue, particularly in medical patients. We suggest that a public health and educational initiative is needed to improve efficiency in PE diagnosis.
1. Introduction
Despite advances in prophylaxis, diagnosis, and therapeutic options, delays between onset of symptoms and diagnosis of pulmonary embolism (PE) are still common, and PE remains a lethal entity. Pulmonary embolism has been determined in 15% of the autopsy series, whereas the antemortem diagnosis of fatal PE has not changed appreciably over time, remaining fixed at approximately 30% [1], [2]. Most patients who die of PE do so within hours of the event [2], [3]. Patients with PE often have nonspecific symptoms, and, as a result, the diagnosis is often delayed. Earlier diagnoses of deep vein thrombosis (DVT) and PE may reduce the morbidity and mortality associated with venous thromboembolism [4]. However, there are only a limited number of studies about delays in the diagnosis of PE. A recent study showed that patients with delayed diagnosis of PE had worse outcomes, such as endotracheal intubation, shock in the emergency department (ED), and hospital death [5]. Some studies reported a correlation between disease severity and early presentation but none between comorbidity, thrombus localization, and presentation [6], [7]. The relationship between delays in the diagnosis of PE and causal risk factors has not been well documented.
In this study, we investigated the effect of risk factors on the delay in the diagnosis of PE and also the relationship between delays and mortality in PE.
2. Materials and methods
2.1. Study design
We retrospectively analyzed the records of all patients with PE who were diagnosed between January 2001 and December 2008. The study was conducted at Karadeniz Technical University, Farabi Hospital, which is a tertiary care hospital that serves as a primary referral center for patients with suspected PE. The study was approved by the local ethical committee.
2.2. Patients and setting
The records of all patients diagnosed in our hospital were analyzed in terms of age, sex, symptoms, clinical and laboratory findings, risk factors for PE, diagnostic tests, and the dates of symptom onset, hospital admission, and definitive diagnosis. The data were collected using standard anamnesis forms including template questions. Because all PE patients in our hospital are treated and followed up by the pulmonology department, patients were retrieved from our clinic registry. In case of a concomitant disease, the day on which the symptoms became severe was accepted as the date of symptom onset. The time from the onset of symptoms to diagnosis was defined as “time to diagnosis,” and the delay from the first medical attention to diagnosis was defined as “presentation to diagnosis.” No effort was made to subgroup data by the hour of the day. For example, if a patient's first symptom occurred on May 12, 2004, and they were evaluated for medical attention on May 15 and diagnosed on May 16, then presentation to diagnosis was 1 day, and the time to diagnosis was 4 days. This information is recorded routinely for all patients diagnosed with PE. Diagnostic delay was defined as diagnosis of PE more than 1 week after symptom onset. Patients were divided into 2 groups, namely, surgical and medical, based on risk factors leading to the embolism. Patients with a history of surgery or trauma within the previous 2 months were considered as the surgical group and the remaining cases as the medical group.
2.3. Diagnosis and treatment of PE
Spiral chest computed tomography pulmonary angiography (CTPA) was used most frequently to confirm acute PE (n = 362, 88.7%) followed by lung scans (n = 34, 8.3%) and clinical (n = 12, 2.9%) diagnosis.
Computed tomography pulmonary angiography was performed using a 4- and 16-channel multislice scanner (Somatom Volume Zoom and Sensation 16; Siemens, Erlangen, Germany). A 100-mL dose of iopromide (Ultravist 370; Schering, Berlin, Germany) was then injected intravenously using a power injector via the right antecubital vein at a rate of 2 mL/s. Commencement of acquisition was determined using the bolus tracking method. The technical parameters used were detector section collimation 4 × 1 mm, section thickness 1.25 mm, collimation 4 × 1, table speed 56 mm/rotation, and a rotation time of 0.5 seconds, 80 kw, 120 kVp. Confirmation of PE was established by intraluminal partial or total filling defect on CTPA [8].
A macro field-of-view gamma camera (Siemens E-cam Dual-Head; Siemens) equipped with a low-energy parallel-hole high-resolution collimator was used to confirm lung perfusion imaging. Comment criteria for perfusion scanning (without ventilation imaging) were those proposed by the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis. These were evaluated as normal, near-normal, abnormal and compatible with PE (PE+, single or multiple wedge-shaped perfusion defects), or abnormal and not compatible with PE (PE−, perfusion defects other than wedge-shaped perfusion defects) [9].
Diagnosis of PE in 12 patients in whom these 2 techniques had not been used was confirmed by joint analysis of clinical scoring (Wells score), DVT, and d-dimer levels [10]. Doppler ultrasonography, CT venography, or clinical characteristics (pain, fever, swelling, or cyanosis of the affected leg) were considered for lower-extremity DVT diagnosis. Diagnoses were reviewed according to the British Thoracic Society guidelines [11].
Patients with nonmassive PE had received standard anticoagulant treatment (unfractionated heparin or low-molecular-weight heparin [LMWH]), and those with massive embolism had received thrombolytic treatment (recombinant tissue plasminogen activator) after diagnosis of PE. In the initial treatment of PE, an intravenous bolus of 80 U/kg was administered followed by a continuous infusion of 18 U/kg intravenously per hour. The target activated partial thromboplastin time is 1.5 to 2.5 times that of a normal control sample. For stable patients, LMWH 100 IU/kg was used twice daily. Patients with massive PE were treated with recombinant tissue plasminogen activator 100 mg/2 h followed by unfractionated heparin. Vitamin K antagonists were given for a period of at least 3 months with the aim of an international normalized ratio of 2.0:3.0 during follow-up. In the same patients, LMWH 100 IU/kg twice daily was used at least 3 months during follow-up.
2.4. Statistical analysis
The Mann-Whitney U and χ2 tests were used for the statistical analysis of the correlations between delays, mortality, and clinical data. Results are given as mean ± SD (median, minimum-maximum). The relation between final analysis of early diagnosis and factors associated with PE was evaluated using univariate logistic regression; those parameters with P < .1 were then subjected to multivariate logistic regression analysis. Final analysis mortality associated with risk factors was analyzed using univariate logistic regression; those parameters with a P < .2 were then subjected to multivariate logistic regression analysis. Results are given as odds ratio (OR) (95% confidence interval [CI]), and a P value less than .05 was considered statistically significant. Data were analyzed using SPSS statistical software (version 13.01, serial number 9069728; SPSS Inc, Chicago, Ill).
3. Results
3.1. Patient characteristics
We retrospectively enrolled a total of 454 consecutive patients with acute PE. Forty-six patients were excluded from the study because the data were incomplete, leaving a study population of 408 patients. After the onset of symptoms, 282 (69.1%) patients had presented to the ED and 61 (15%) to outpatient polyclinics, whereas 65 (15.9%) patients were diagnosed during hospitalization (Table 1).
Table 1. Characteristics of the patients with PE
| Characteristics | Medical group (n = 288) | Surgery group (n = 120) | Total (N = 408) | P |
|---|---|---|---|---|
| Age (y) | 63.35 (±15.6) | 59.17 (±16.4) | 62.12 (±16.2) | .018 |
| Sex (female/male) Female | 163/125 (56.6) | 73/47 (60.8) | 234/172 (57.8) | NS |
| Pulse ≥100 beats/min | 66 (22.9) | 30 (25) | 96 (23.5) | NS |
| Hypotension (SBP/DBP) (<90/60 mm Hg) | 38 (13.2) | 19 (15.8) | 57 (14) | NS |
| Po2 ≤60 mm Hg | 81 (28.1) | 30 (25) | 111 (27.2) | NS |
| DVT (389 patients) | 126 (45.6%) | 40 (35.3%) | 166 (42.7) | NS |
| Comorbidity | 196 (68.8%) | 48 (40%) | 244 (60%) | .000 |
| Symptoms | ||||
 Dyspnea | 231 (78.3) | 94 (80.2) | 325 (79.7) | NS |
| Chest/pleuritic pain | 141 (48.9) | 70 (58.3) | 211 (51.7) | NS |
| Hemoptysis | 42 (14.6) | 20 (16.7) | 62 (15.2) | NS |
| Syncope | 24 (8.3) | 15 (12.5) | 39 (9.6) | NS |
| Leg swelling | 36 (12.5) | 11 (9.2) | 47 (11.5) | NS |
| Time to diagnosis (d) | 8.0 (±8.6) | 4.4 (±7.6) | 6.9 (±8.5) | .000 |
| Early, ≤7 | 190 (66) | 105 (87.5) | 295 (72.3) | .000 |
| Delayed, >7 | 98 (34) | 15 (12.5) | 113 (27.7) | .000 |
| Presentation to diagnosis (d) | 2.9 (±5.5 ) | 1.4 (±4.2) | 2.4 (±5.2) | .002 |
| Diagnostic methods | NS | |||
| CTPA | 251 (87.2) | 111 (92.5) | 362 (88.7) | |
| Lung perfusion scans | 28 (9.7) | 6 (5) | 34 (8.3) | |
| Clinical | 9 (3.1) | 3 (2.5) | 12 (2.9) | |
| Mortality | 28 (9.7) | 12 (10) | 40 (9.8) | NS |
The average age of the patients was 62.12 (±16.2) years, and 234 (57.8%) were females. The most frequent presentation symptom was dyspnea (79.7%), whereas 91.4% had at least one of the following respiratory symptoms: dyspnea, chest pain, pleuritic pain, cough, hemoptysis, and syncope. On admission, 23.5% of the patients had tachycardia, 27.2% had hypoxia on arterial blood gases analyses (≤60 mm Hg), 14% had massive embolism (defined by systolic blood pressure <90 mm Hg and/or diastolic blood pressure <60 mm Hg), and 42.7% had DVT (Table 1). Sixty percent of the patients had one or more comorbid diseases, such as cancer, hypertension, heart disease, diabetes mellitus, collagen vascular disease, tuberculosis, chronic obstructive pulmonary disease, and neurologic disease. Of the total number of patients, 70.5% (288 patients) had medical risk factors, and the remaining (120 patients) had surgical risk factors. On the other hand, 25% of the patients in the surgical group also had a medical risk factor. The distribution of risk factors is summarized in Table 2.
Table 2. Inpatients risk factors for PE
| Medical conditions (n = 288) | Patients (%) |
| Immobility | 64 (22) |
| Cancer | 41 (14) |
| Heart failure | 42 (14) |
| Stroke | 25 (9) |
| Prior VTE | 21 (7) |
| Chronic venous failure | 12 (4) |
| Oral contraceptive drugs | 9 (3) |
| Rarely causes | 43 (15) |
| Unknown etiologya | 31 (11) |
| Type of surgery(n = 120) | |
| Orthopedic surgery | 49 (41) |
| Hip replacement | 14 (29) |
| Hip fracture surgery | 8 (16) |
| Knee replacement | 6 (12) |
| Tibia-fibula fracture surgery | 6 (12) |
| Arthroscopy | 5 (10) |
| Others | 10 (20) |
| Abdominal surgery | 17 (14) |
| Inguinal hernia | 5 (29) |
| Hepatobiliary | 5 (29) |
| Splenectomy | 3 (18) |
| Others | 4 (24) |
| Urologic surgery | 11 (9) |
| Gynecologic surgery | 11 (9) |
| Neurosurgery | 15 (12) |
| Thoracolombar | 10 (67) |
| Intracranial | 5 (33) |
| Vascular surgery | 9 (8) |
| Other surgery | 2 (2) |
| Trauma but surgery not done | 6 (5) |
| Additional medical risk factors in this group | |
| Cancer | 17 (14.2) |
| Heart failure | 5 (4.2) |
| Cerebrovascular disease | 5 (4.2) |
| Previous VTE | 2 (1.7) |
| Chronic obstructive pulmonary disease | 1 (0.8) |
aPatients who have no surgical history but also no determined medical risk factor. |
3.2. Time to diagnosis
The mean time to diagnosis in all patients was 6.95 ± 8.5 days (median, 3 days; range, 0-45 days). Of the total patients, 29.6% had presented within the first 24 hours and 72.3% within the first week. The mean time to diagnosis was 4.39 ± 7.6 days (median, 2 days; range, 0-45 days) in the surgical group and 8.0 ± 8.6 days (median, 4 days; range, 0-45 days) in the medical group (P = .000; Fig. 1). Time to diagnosis was longest in PE developing because of oral contraceptive drug use in the medical risk group (13.3 ± 12 days) and shortest in PE resulting from stroke (4.1 ± 4.8 days). The percentage of cases diagnosed within the first week (87.5% of the surgical group, 66% of the medical group) was significantly higher in the surgical group (P = .000; Table 1). Pulmonary embolism was diagnosed more than 3 weeks after the onset of symptoms in 3.3% of the patients in the surgical group and in 8% of the patients in the medical group (P > .05).
Fig. 1.
Delays in the diagnosis of acute PE (N= 408) showed as a frequency distribution in the medical and surgical groups. The mean time to diagnosis was 4.4 ± 7.6 days, and presentation to diagnosis was 1.4 ± 4.2 days in the surgical group. The mean time to diagnosis was 8.0 ± 8.6 days, and presentation to diagnosis was 2.9 ± 5.5 days in the medical group.
According to the symptoms on first presentation, the mean time to diagnosis was 4.1 ± 6.4 days (median, 2 days; range, 0-30 days) in patients who presented with syncope when compared with those without it (7.25 ± 8.7 days; median, 3 days; range, 0-45 days; P = .036). In terms of methods used in the diagnosis, CTPA use provided an early diagnostic advantage (P = .013). Parameters likely to influence early diagnosis (medical risk, comorbidity, surgery, etc) were analyzed using univariate logistic regression (Table 3). Parameters likely to influence early diagnosis (medical risk subgroup, surgery, syncope, hypoxia, use of spiral CT in diagnosis) were finally analyzed using multivariate logistic regression (Table 4). Surgery, presence of cancer, and stroke were found to be related with early diagnosis in PE.
Table 3. Univariate logistic regression analysis of variables associated with early presentation
| Variable | Early presentation, ≤7 d | ||
|---|---|---|---|
| OR | 95% CI | P | |
| ≤65 y | 1.041 | 0.674-1.607 | .855 |
| Comorbidity | 0.762 | 0.488-1.192 | .234 |
| ≥3 Respiratory symptoms | 0.814 | 0.395-1.673 | .575 |
| Syncope | 2.814 | 1.072-7.387 | .036a |
| Tachycardia (>100/min) | 0.835 | 0.495-1.409 | .500 |
| Hypotension (<90/60 mm Hg) | 1.207 | 0.632-2.303 | .569 |
| Hypoxia (Po2 ≤60 mm Hg) | 0.712 | 0.427-1.189 | .195 |
| DVT | 0.823 | 0.523-1.296 | .401 |
| Diagnosis with CTPA | 2.225 | 1.186-4.173 | .013a |
| Surgery | 3.611 | 1.995-6.535 | .000a |
| Medical | 0.970 | 0.475-1.976 | .931 |
| Heart failure | 0.734 | 0.371-1.452 | .374 |
| Stroke | 2.939 | 0.862-10.020 | .085 |
| Prior VTE | 2.370 | 0.684-8.206 | .173 |
| Oral contraceptive use | 0.297 | 0.078-1.126 | .074 |
| Unknown etiology | 0.325 | 0.155-0.682 | .003⁎ |
| Cancer | 1.962 | 0.843-4.564 | .118 |
| Immobility | 0.532 | 0.305-0.930 | .027⁎ |
aStatistically significant. |
Table 4. Multivariate logistic regression analysis of risk factors relation to earlier presentation
| Variable | Early presentation, ≤7 d | ||
|---|---|---|---|
| OR | 95% CI | P | |
| Diagnosis with CTPA | 1.66 | 0.711-3.870 | .242 |
| Hypoxia(Po2 ≤60 mm Hg) | 0.72 | 0.409-1.259 | .247 |
| Syncope | 2.50 | 0.899-6.949 | .079 |
| Surgery | 4.54 | 2.186-9.443 | .000a |
| Medical | |||
| Stroke | 6.23 | 1.357-28.641 | .019a |
| Prior VTE | 3.21 | 0.855-12.110 | .084 |
| Oral contraceptive use | 0.65 | 0.160-2.690 | .559 |
| Unknown etiology | 0.90 | 0.368-2.240 | .834 |
| Cancer | 2.61 | 1.020-6.708 | .045⁎ |
| Immobility | 0.82 | 0.410-1.665 | .592 |
aStatistically significant. |
3.3. Presentation to diagnosis
The mean time from first medical attention to diagnosis was 2.4 ± 5.2 days (median, 0 days; range, 0-30 days). This time was 2.9 ± 5.5 days (median, 3 days; range, 0-30 days) in the medical risk group and 1.4 ± 4.2 days (median, 0 days; range, 0-30 days) in the surgical patients (P = .002; Fig. 1). On the first presentation, 99 patients were misdiagnosed and mistreated, and this was significantly higher in the medical group (82.8% in the medical group vs 17.2% in the surgical group; P = .002). Accompanying comorbid conditions were found in 68.8% of the medical risk group and in 40% of the surgical risk group patients (P = .000; Table 1). No significant correlation was determined between other clinical parameters of the patients (tachycardia, hypotension, DVT, hypoxia, sex, diagnostic technique, previous PE, cancer, or fainting) or between the initial medical evaluation and confirmatory diagnosis.
3.4. Relation between delays and mortality
A total of 40 (10%) patients died (28 in the medical group and 12 in the surgical group). There were no statistical differences between the 2 groups in terms of mortality. Delay in diagnosis was not different between those who died and those who survived (P > .05). The mean time to diagnosis in the 40 patients who died was 5.7 ± 6.7 days (median, 3 days; range, 0-30 days) and 7.1 ± 8.7 days (median, 3 days; range, 0-45 days) in those who survived. Earlier diagnosis had no impact on mortality (P > .05).
The average age of the patients who died was 67.78 ± 14.1 years and of those who survived was 61.51 ± 16.4 years (P = .017). Factors likely to affect mortality were analyzed using multivariate logistic regression analysis, and a clear correlation was found between mortality and the presence of massive embolism, cancer, or heart failure (Table 5).
Table 5. Multivariate logistic regression analysis of clinical variables related to mortality
| Variable | OR | 95% CI | P |
|---|---|---|---|
| Massive | 13.5 | 6.12-29.67 | .000a |
| Cancer | 7.4 | 2.85-19.28 | .000a |
| Heart failure | 2.9 | 1.09-7.66 | .033a |
| Age >65 y | 2.1 | 0.96-4.60 | .064 |
aStatistically significant. |
Among the risk factors, the longest delay in diagnosis was detected among oral contraceptive users (13 days), and the shortest delay was detected in patients with stroke (4 days) in the medical risk group, but there was no statistically significant difference in mortality.
4. Discussion
Pulmonary embolism is a lethal disease associated with various risk factors and requires treatment to be started as soon as a definitive diagnosis is reached. However, because symptoms and findings in PE are nonspecific, diagnosis is difficult at first presentation. Various clinical parameters for determining probability have been established to facilitate a diagnosis of PE. Within the Wells and Geneva scoring system, which is widely used today, a history of surgery is an important factor that increases PE probability [12], [13]. On the other hand, the presence of medical disease is a risk factor for PE and can also mask PE symptoms, thus delaying diagnosis. Previous studies investigating delayed diagnosis in PE have not focused on the effect of risk factors on delayed diagnosis.
In our study, diagnostic delay in PE related to medical risk factors (immobility, cancer, heart failure, etc) was approximately 4 times greater compared with that in PE developed postoperatively on univariate analysis. Similarly, the medical group patients were diagnosed later after admission to the hospital. It seems that a history of surgery is a positive predictive factor in earlier diagnosis. First of all, diagnosis might be easier in surgical cases because PE develops while most are still hospitalized during the event. Another point is that individuals with medical diseases such as heart failure probably underestimate their complaints and delay seeking medical care. This is exacerbated by the treating physician often attributing the patient's symptoms to his underlying medical condition. In this study, about 24% of patients had been misdiagnosed before definitive diagnosis of PE, and 82.8% of these cases were within the medical group.
Risk factors were separately considered in 2 previous studies, although the effects of all risk factors on the delay to diagnosis were not investigated [7], [14]. In the MASTER study, presence of transient risk factors (recent surgery, severe medical diseases, immobilization, pregnancy, etc) had been found significantly associated with earlier diagnosis [7]. Castro et al [14] established early diagnosis in patients with surgical risk factors. In that study, however, the effects of other risk factors on the delay to diagnosis were not reported in half of the patients. A study by Elliott et al [15] only investigated delays in the diagnosis without detecting the factors affecting the diagnostic delay. In another study conducted in Turkey, Bulbul et al [6] showed that hypotension and tachypnea were correlated with early diagnosis. However, no data regarding the effect of risk factors on delays in diagnosis were provided in that study.
In several studies, 16% to 30.4% of the patients with acute PE were diagnosed 1 week or longer after the symptom onset, and the mean time to diagnosis was 4.8 to 8.4 days. It has been reported that 31.6% to 50% of the patients presented within 24 hours of symptom onset [6], [14], [15], [16]. In our study, 29.6% of the patients had presented within the first 24 hours and 72.3% within the first week, with an average of 6.9 days between onset of symptoms and diagnosis. In the 2 previous studies, delays from first medical attention to diagnosis were found to range from 0.9 to 2 days [6], [15]. In another study, PE was diagnosed in 12% of the patients more than 48 hours after admission to the ED [5]. In our study, the mean time to diagnosis was found as 2.4 days. Our findings were in agreement with those previous studies.
Although early diagnosis of PE is thought to reduce mortality [5], sufficient data on the subject are still lacking. Although the diagnosis of PE has increased 2 to 3.6 times with the use of CT and associated use of enoxaparin has risen 2.6 times, mortality rates were not reported to fall in parallel [17], [18]. The increased number of patients with PE might be because of the detection of subsegmental silent peripheral embolisms by CT. On the other hand, a 0% to 5% mortality rate has been reported even in patients monitored without treatment [19], [20]. Considering this fact, the most important point in the early diagnosis might be the detection of massive embolisms, which are the major cause of deaths in PE. In our study, in agreement with that by Castro et al [14], we observed that early or late diagnosis had no effect on PE mortality. Although the presence of surgical risk factors and of cancer in the medical risk group was observed to shorten diagnostic delay by 3 to 4 times, it had no effect on mortality. In our study, we found that massive embolism, presence of cancer, and heart failure were independent risk factors for mortality. Therefore, because many factors other than diagnostic delay affect mortality, a comparison of PE patients with different delays but with similar characteristics might disclose the most objective factor related to mortality.
There are a number of limitations to this study. First, the number of the cases undergoing surgery who received adequate or inadequate prophylaxis during the preoperative and/or postoperative period is not clear. Second, our results do not provide any information on the long-term effects of diagnostic delay. For example, the relation between delay and post-thrombophlebitic syndrome or long-term mortality is uncertain. In the study of Castro et al [14], however, late diagnosis was not observed to have any effect on reembolism in a 3-month period. Third, if we had analyzed cardiac biomarkers and/or right ventricular dysfunction, which are independent risk factors for mortality, we might have more correctly predicted the impact of delayed diagnosis on mortality.
Pulmonary embolism is an important issue in hospitalized medical patients. Most PE cases among hospitalized patients are in medical services, and PE is still a major cause of death in this group [21], [22]. Thus, the American College of Chest Physicians guidelines have been recommended regarding venous thromboembolism prophylaxis in hospitalized medical patients, especially in those with heart failure, active cancer, or sepsis [23]. In fact, only 30% to 50% of at-risk patients received prophylactic treatment, and the proportion of at-risk patients who should receive prophylaxis globally remains unknown [21], [24], [25]. A metaanalysis showed that anticoagulant prophylaxis reduced the relative risk for symptomatic PE during treatment by 58% [26]; thus, we want to emphasize again the importance of implementing more intensive prophylaxis.
In conclusion, our study showed that there was an important delay in the diagnosis of PE, which was more prominent in patients with medical risk factors. Public and professional education represents a critical step for establishing an early diagnosis and treatment in these patients particularly. We suggest that there is a need for new strategies, which should include public health outreach models, and for increased awareness, especially in the population with medical risk factors.
References
- . Venous thromboses and pulmonary embolisms in post-mortem series: probable causes by correlations of clinical data and basic diseases. Pathol Res Pract. 1994;190:394–399
- . Prevalence of acute pulmonary embolism among patients in a general hospital and at autopsy. Chest. 1995;108:978–981
- . Major pulmonary embolism: review of a pathophysiologic approach to the golden hour of hemodynamically significant pulmonary embolism. Chest. 2002;121:877–905
- . Pulmonary embolism: what have we learned since Virchow?. Chest. 2002;122:1440–1446
- Prospective study of the clinical features and outcomes of emergency department patients with delayed diagnosis of pulmonary embolism. Acad Emerg Med. 2007;14:592–598
- Bulbul Y, Ozsu S, Kosucu P, et al. Time delay between onset of symptoms and diagnosis in pulmonary thromboembolism. Respiration DOI: 10.1159/000167409.
- Factors associated with the timing of diagnosis of venous thromboembolism: results from the MASTER registry. Thromb Res. 2008;121:751–756
- Central pulmonary thromboembolism: diagnosis with spiral volumetric CT with the single-breath-hold-technique—comparison with pulmonary angiography. Radiology. 1992;185:381–387
- Value of perfusion lung scan in the diagnosis of pulmonary embolism: results of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISA-PED). Am J Respir Crit Care Med. 1996;154:1387–1393
- Derivation of a simple clinical model to categorize patients' probability of pulmonary embolism: increasing the model's utility with the SimpliRED d-dimer. Thromb Haemost. 2000;83:416–420
- . British Thoracic Society guidelines for the management of suspected acute pulmonary embolism. Thorax. 2003;58:470–484
- Use of a clinical model for safe management of patients with suspected pulmonary embolism. Ann Intern Med. 1998;129:997–1005
- Assessing clinical probability of pulmonary embolism in the emergency ward. Arch Intern Med. 2001;161:92–97
- Prognostic significance of delays in diagnosis of pulmonary embolism. Thromb Res. 2007;121:153–158
- . Delays in diagnosis of deep vein thrombosis and pulmonary embolism. Chest. 2005;128:3372–3376
- Pulmonary embolism: differences in presentation between older and younger patients. Age Ageing. 2003;32:601–605
- Pulmonary embolism incidence is increasing with use of spiral computed tomography. Am J Med. 2008;121:611–617
- Increased diagnosis of pulmonary embolism without a corresponding decline in mortality during the CT era. Clin Radiol. 2008;63:381–386
- . Untreated patients with pulmonary embolism. Chest. 1995;107:931–935
- Silent pulmonary embolism in patients with deep venous thrombosis. Incidence and fate in a randomized, controlled trial of anticoagulation versus no anticoagulation. J Intern Med. 1994;235:457–461
- . New onset of venous thromboembolism among hospitalized patients at Brigham and Women's Hospital is caused more often by prophylaxis failure than by withholding treatment. Chest. 2000;118:1680–1684
- . Autopsy proven pulmonary embolism in hospital patients: are we detecting enough deep vein thrombosis?. J R Soc Med. 1989;82:203–205
- Prevention of venous thromboembolism. Chest. 2004;126:338S–400S
- . Thromboprophylaxis rates in US medical centers: success or failure?. J Thromb Haemost. 2007;5:1610–1616
- Venous thromboembolism risk and prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-sectional study. Lancet. 2008;371:387–394
- Meta-analysis: anticoagulant prophylaxis to prevent symptomatic venous thromboembolism in hospitalized medical patients. Ann Intern Med. 2007;146:278–288
PII: S0735-6757(09)00374-X
doi:10.1016/j.ajem.2009.07.005
Crown Copyright © 2011. Published by Elsevier Inc. All rights reserved.
Volume 29, Issue 1 , Pages 26-32, January 2011
