Now that the quality of computed tomography (CT) imaging techniques has greatly improved with the introduction of thin section multi-detector row CT scanners, a pulmonary embolism (PE) is increasingly diagnosed on routine CT examinations in the absence of a clinical suspicion. This phenomenon is often referred to as ‘incidental PE’. This increased detection has made incidental PE the topic of recent debate. In particular, physicians have questioned the prognostic relevance of these incidental findings [1,2]. However, scarce data are available to guide clinical decision-making for physicians who are now increasingly being confronted with a diagnosis of an incidental PE . Given the absence of any clinical trial assessing the management of an incidental PE and the paucity of prognostic studies addressing the outcome of these patients, recommendations on the treatment of an incidental PE are almost solely based on expert opinion. To investigate current management practice patterns for an incidental PE, in terms of the initiation and duration of anticoagulant administration, and the need for hospitalization, we undertook a worldwide survey among physicians.
A questionnaire, consisting of two patient cases, was designed and pretested by the investigators (Appendix). The first case described a patient in whom an incidental PE was diagnosed on a CT scan conducted for the follow-up of a malignant disease. In the second case, an incidental PE was detected on a CT scan performed for follow-up of non-malignant disease. The patients did not have any symptoms suggestive of PE. Multiple-choice questions were used to investigate: (i) whether the respondent would initiate anticoagulant treatment and if so, for what duration of time; (ii) the type of anticoagulant prescribed; and (iii) whether the respondent would treat the patient on an out- or inpatient basis. Also, the participants were asked whether a concurrent diagnosis of a deep venous thrombosis (DVT) would influence their management strategy. To assess the impact of the location of PE on the participants’ management approach, both cases were subdivided in a case with a subsegmental PE (SSPE), a segmental PE and a central PE. The questionnaire was published online using Thesis Tools (http://www.thesistools.com/web/?id=264229).
We intended to reach a large number of physicians practicing in various disciplines and in various countries. For this purpose, the memberships list of the International Society on Thrombosis and Haemostasis (ISTH) was used to randomly select a total of 127 physicians practicing in 31 different countries. Criteria used for selection were both being registered as a physician and practicing in a field that potentially encounters patients with PE. Second, 177 members of the Registro Informatizado de Pacientes con Enfermedad TromboEmbólica (RIETE) registry , a multidisciplinary project which aims to register the management and outcome of patients with venous thromboembolism (VTE) were surveyed. Third, 131 members of the Dutch society of Vascular Medicine were asked to participate. Lastly, the survey was distributed among 150 members of the Spanish society of Medical Oncology.
Data were analyzed using descriptive statistics. As not all participants fully completed the survey, proportions were calculated based on the number of respondents per question.
Out of the 585 invitations, a total of 183 (31%) questionnaires were returned and available for analysis. The majority of the group of respondents practiced in the field of internal medicine (45.8%), oncology (14.8%), hematology (14.8%) and pulmonary medicine (13.5%). Of the respondents, 61% practiced in an academic medical center versus 39% in a non-academic hospital.
When a patient was presented with an incidental SSPE, 72% (95% confidence interval [CI]: 64.4–78.8%] of the participants would initiate anticoagulant treatment in case the patient did not have malignant disease, whereas in the presence of a malignancy 89.1% (95% CI: 83.6–93.2%) opted to treat (Table 1). In case an incidental PE was located in a segmental pulmonary artery, 98% (95% CI: 96.4–100%) of the respondents decided to initiate treatment when it concerned a cancer patient, and 97.5% (95% CI: 95.1–99.9%) if the patient did not have malignant disease. Every respondent chose to initiate anticoagulant therapy in case a PE was localized centrally, irrespective of the presence of malignant disease.
|Location of PE||Case 1 – cancer patient||Case 2 – non-cancer patient|
|Initiation of treatment, n (%)||n = 183||n = 175||n = 169||n = 161||n = 157||n = 156|
|No treatment||20 (10.9)||3 (1.7)||0 (0.0)||45 (28.0)||4 (2.5)||0 (0.0)|
|Anticoagulation for 3 months||24 (13.1)||18 (10.3)||9 (5.3)||37 (23.0)||43 (27.4)||24 (15.4)|
|Anticoagulation for 6 months||61 (33.3)||63 (36.0)||54 (32.0)||70 (43.5)||102 (65.0)||118 (75.6)|
|Anticoagulation for indefinite period||78 (42.6)||91 (52.0)||106 (62.7)||9 (5.6)||8 (5.1)||14 (9.0)|
|Type of treatment*, n (%)||n = 161||n = 172||n = 167||n = 115||n = 151||n = 153|
|VKA||14 (8.7)||13 (7.6)||13 (7.8)||50 (43.5)||79 (52.3)||86 (56.2)|
|LMWH – therapeutic dose||99 (61.5)||114 (66.3)||116 (69.5)||50 (43.5)||55 (36.4)||52 (34.0)|
|LMWH – therapeutic dose followed by prophylactic dose||45 (28.0)||41 (23.8)||36 (21.6)||13 (11.3)||15 (9.9)||12 (7.8)|
|LMWH – prophylactic dose||3 (1.9)||4 (2.3)||2 (1.2)||2 (1.7)||2 (1.3)||3 (2.0)|
|Treatment location*, n (%)||n = 163||n = 174||n = 168||n = 115||n = 151||n = 154|
|Outpatient basis||114 (69.9)||107 (61.5)||61 (36.3)||77 (67.0)||94 (62.3)||54 (35.1)|
|Hospitalized||10 (6.1)||17 (9.8)||66 (39.3)||10 (8.7)||21 (13.9)||66 (42.9)|
|Depends on patient preferences||2 (1.2)||2 (1.1)||1 (0.6)||2 (1.7)||3 (2.0)||2 (1.3)|
|Decide on a case-by-case basis||37 (22.7)||48 (27.6)||40 (23.8)||26 (22.6)||33 (21.9)||32 (20.8)|
|Change treatment if proven DVT, n (%)||n = 181||n = 177||n = 168||n = 160||n = 153||n = 156|
|Yes||49 (27.1)||22 (12.4)||12 (7.1)||55 (34.4)||20 (13.1)||14 (9.0)|
|No||132 (72.9)||155 (87.6)||156 (92.9)||105 (65.6)||133 (86.9)||142 (91.0)|
With respect to the duration of treatment, in case an incidental PE was diagnosed in a patient with malignant disease, indefinite treatment was opted by 42.6%, 52% and 62.7% of the respondents if PE was, respectively, localized in a subsegmental, segmental or central pulmonary artery. Most of the remaining respondents choose to treat the patient for 6 months (33.3%, 36% and 32% for subsgemental, segmental and central PE, respectively). In the absence of malignant disease, a treatment period of 6 months was preferred by 43.5% of the respondents in case of a SSPE, 65% in case of a segmental PE and 75.6% for a patient with a central PE.
A large group of participants decided to treat a patient with an incidental SSPE on an outpatient basis: 69.9% (95% CI: 62.9–76.9%) in the presence and 67% (95% CI: 58.4–75.6%) in the absence of malignant disease. These proportions slightly decreased in the case of a segmental PE: 61.5% (95% CI: 54.3–68.7%) for a patient with and 62.3% (95% CI: 54.6–70%) for a patient without malignancy. When it concerned a patient with a central PE, hospitalization became the most prevalent treatment option chosen by the respondents: 39.3% (95% CI: 31.9–46.7%) in the case of a cancer patient and 42.9 (95% CI: 35.1%–50.7%) in the case of a patient without cancer.
In the case of a cancer patient with SSPE, 16 of the 18 (89%) respondents who initially choose to withhold treatment, decided to initiate treatment once a DVT was proven concurrently. When a non-cancer patient with SSPE was presented, 37 of the 41 (90%) of the respondents who initially opted not to treat, choose to start anticoagulant therapy after a diagnosis of a DVT.
This study demonstrates that the majority of physicians would initiate anticoagulant therapy once an incidental PE is diagnosed, in particular if a PE is localized proximally. Of importance, least consensus appears to exist regarding the need for anticoagulant therapy in patients with an incidental SSPE. In particular in the absence of malignant disease, a large proportion of participants (28%) opted not to treat. Also, there seemed to be a tendency to treat more a centrally located incidental PE for a longer period of time. These findings may reflect the uncertainty in recent literature about the clinical relevance of small, subsegmental emboli [5,6]. A survey specifically addressing the management of a symptomatic SSPE found similar high rates of respondents who would not initiate treatment in such patients, or at least not before performing additional diagnostic tests . The issue concerning the clinical significance of a SSPE is of particular relevance for patients with an incidental PE, as incidentally detected PE are likely to be smaller emboli, with a reported rate of SSPE in 27% of the cases .
Our survey also addressed the issue of outpatient management. Three recent cohort studies have demonstrated that selecting PE patients for outpatient treatment solely on a clinical basis, primarily focusing on hemodynamic stable PE patients who do not require an oxygen supply, yields safe results in terms of low risks of recurrent VTE and mortality [9–11]. Given that patients with an asymptomatic PE are clearly able to compensate for the hemodynamic and respiratory consequences of the PE and unlikely to have decreased right ventricular function, these patients may be deemed good candidates for outpatient management, in particular if a PE is detected on elective CT scans in an outpatient setting. Indeed, the present study indicates that the majority of physicians would feel comfortable to treat patients with an incidental PE on an outpatient basis.
It should be stated that this survey specifically addressed the management of incidental PE patients in whom symptoms suggestive of PE were absent. It has previously been demonstrated that 75% of the cancer patients diagnosed with an incidental PE, actually were symptomatic at the time of diagnosis . In fact, the presence of symptoms in patients with an incidental PE may be associated with a worse outcome .
The results of this survey may be compromised by the relatively low response rate (31%). However, it was our intention to reflect the practice patterns of a wide variety of randomly selected clinicians. Distributing the survey among a selected group of experts would possibly have yielded a higher response rate; however, this would have made the results less generalizable to clinical practice. A second limitation is that the survey did not assess the respondents’ country of practice. We were therefore unable to investigate inter-country differences in the management approach to incidental PE. For instance, the choice whether to manage patients on an outpatient basis may well be influenced by local health care systems.
In summary, this study reveals most physicians decide to treat a patient with an incidental PE. Uncertainty exists about the need for anticoagulant treatment in patients with an incidental SSPE, particularly in the absence of cancer and a DVT. The majority of physicians would manage patients with a small, asymptomatic incidental PE on an outpatient basis. Further studies are needed to clarify the risk-benefit ratio of anticoagulant therapy in an incidental PE, with a particular focus on the safety of withholding treatment in patients with an incidental SSPE and the safety of treating incidental PE patients as outpatients.