Home Treatment of Patients with Pulmonary Embolism: A Single Center 10-Year Experience from Ljubljana Registry

Abstract

Current guidelines suggest careful risk stratification using a structured clinical approach when selecting patients with pulmonary embolism (PE) for home treatment. The aim of our study was to assess whether PE patients referred to home treatment are appropriately risk-stratified according to guidelines prior to referral and what the real-life course of the disease in these patients is. We included patients with confirmed PE referred to outpatient management and treated with anticoagulants between 2010 and 2019, whose data were collected in a prospective management registry. Using simplified PE severity index and/or signs of right ventricular strain, we classified patients to either appropriate or inappropriate low-risk classes for outpatient management. We compared 30-day mortality, overall mortality, and rates of recurrent thromboembolism or major bleeding between both classes. Among 278 patients, 188 (67.6%) and 90 (32.4%) were classified as appropriate or inappropriate class, respectively. In total, 30-day mortality was low in both groups: 0% in appropriate class and 1.1% in inappropriate class. The overall mortality rate was higher in the inappropriate than in the appropriate class (12.1 vs 0.9/100 patient-years, respectively, P < .001). Rates of recurrent thromboembolism and major bleeding were similar for both classes. We conclude that in real-life, a significant proportion of inappropriate low-risk class PE patients are referred to outpatient management. However, with careful follow-up, early mortality is low, even in home-treated patients inappropriately classified as low-risk.

Keywords

pulmonary embolism home management prognosis risk stratification

Venous thromboembolism (VTE)—encompassing deep vein thrombosis and pulmonary embolism (PE)—is a relatively common disease, affecting approximately 1 in 1000 individuals.¹ It is still considered a significant cause of cardiovascular morbidity and mortality.^2,3 To this day, anticoagulant treatment represents the mainstay in managing patients with VTE. Traditionally, patients had to be hospitalized for intravenous anticoagulant treatment, which used to be the only way to ensure instant anticoagulation. However, low molecular weight heparins (LMWH) and later direct oral anticoagulants enabled safe and effective outpatient treatment of patients with VTE.⁴ Yet, in patients with PE, the spectrum of clinical presentations is wide, and the risk of severe adverse outcomes is significant. PE severe enough to produce hemodynamic instability, for instance, is associated with a 30% mortality.⁵

Two meta-analyses have shown that home treatment of patients with PE might be safe, albeit still associated with a tangible 2% mortality. Also, the included trials were heterogeneous in design, the number of included patients was relatively low, and the criteria for selecting patients for home treatment were mostly not standardized, so the results had to be interpreted carefully.^6,7 Several attempts have been made to assess mortality risk in patients with PE. PE severity index (PESI) has been well validated for stratifying patients according to their 30 day mortality risk.⁸ A simplified version was later derived—ie, the simplified PESI (s-PESI)—which has been shown to successfully discriminate between patients with low and high 30-day mortality risk.⁹ Moreover, robust clinical HESTIA criteria have been shown to successfully help select PE patients eligible for home treatment.¹⁰ Recent European guidelines suggest considering patients with PE for home treatment only if PE is not severe as assessed by PESI, s-PESI, or HESTIA criteria, have no signs of right ventricular (RV) dysfunction as assessed by imaging methods or biomarkers, and also have no additional conditions that would require hospitalization.¹¹ However, it is not clear how often this approach is followed in clinical practice.

Home treatment of PE has been promoted as a possible management strategy, provided adequate medical follow-up (ie, in the form of day hospitals) is available. In Slovenia, the University medical center ofLjubljana is the largest hospital, catering to about 600 000 inhabitants (ie, 1/3 of the country's population). The vast majority of patients with VTE are treated at the Department of Vascular Diseases, which includes a specialized outpatient VTE clinic. In the present analysis, we wanted to determine (i) whether patients referred for home treatment of PE are adequately risk-stratified according to disease severity and early mortality risk, and (ii) what is the real-life course of the disease in patients with PE who are managed at home. We hypothesized that a significant proportion of patients referred to outpatient management actually do not fulfill low-risk criteria and have higher risk than the minimum risk recommended by the guidelines for the selection of candidates for home treatment. Furthermore, we wanted to check whether this is reflected also in the increased mortality and adverse event rates in home-treated patients inappropriately classified as low-risk.

Methods

Study Design

This was a registry-based, nonconcurrent prospective cohort study. We included patients who were treated for PE (ICD diagnoses I26.0 to I26.9) at the Department of Vascular Diseases, University Medical Centre Ljubljana, Slovenia, between January 1, 2010 and December 31, 2019. All patients had PE objectively confirmed with computed tomographic (CT) angiography, as per local imaging protocol. We excluded patients who had been admitted for in-hospital management for >24 h. Home-treated PE patients were managed through a registry-based mandatory clinical pathway, which included: (i) Initial evaluation (clinical examination, medical records overview, laboratory analysis, screening diagnostics for VTE etiology), (ii) therapy initiation (oral anticoagulation and/or LMWH), (iii) patient education, and (iv) structured follow up (at 1 week, at 1 month, at 3 months, at 6 months, and yearly thereafter for patients requiring long-term anticoagulation; as needed if patients’ condition required or in patients receiving vitamin K antagonists for the purposes of INR monitoring). The fill-in clinical pathway for VTE management is connected to the registry with built-in decision-making support software for the management of anticoagulant therapy.

Data Source and Outcomes

Data were obtained from the anticoagulation management registry (Ljubljana registry), which includes data on all outpatients receiving anticoagulant therapy, including patients managed at the VTE outpatient clinic. The registry was described previously.¹² For the purposes of the Registry, Trombo software with built-in decision-making support was developed (Magas, d.o.o., Slovenia) both for day-to-day management as well as for quality analysis of anticoagulant treatment. Since patients are managed according to current standards of best medical practice with no experimental interventions, verbal consent was obtained and recorded in patients’ files. The Registry has been approved by the Committee for Medical Ethics of the Republic of Slovenia (decision letter No. 150/10/12 of November 24, 2012). We extracted demographic data and data for retroactive calculation of s-PESI, namely: Sex, age, presence of cancer, pulmonary and/or cardiovascular diseases, heart rate, systolic blood pressure, and blood oxygen saturation. Patients’ documentation was reviewed for reported use of PE risk scores, imaging data on RV strain, and biomarker levels (troponin and NT-proBNP). As per local protocols, RV strain was defined as any of the following: CT signs of RV enlargement (RV dimensions, septal bowing, and RV/left ventricular ratio)¹³ or echocardiographic signs of RV strain (as proposed by the PEITHO criteria).¹⁴ s-PESI score was calculated,⁹ and patients were divided according to risk stratification: Appropriate low-risk class (s-PESI 0 and no signs of RV strain) and inappropriate low-risk class (s-PESI > 0 and/or signs of RV strain).

Follow-up data included event adjudication by 2 VTE specialists (with a third VTE specialist consulted in case of disagreement). Events included all-cause mortality, recurrent thromboembolism, or major bleeding. Recurrent thromboembolism was defined as any symptomatic, objectively confirmed arterial or venous thromboembolic event. Major bleeding was defined as either central nervous system, retroperitoneal, or vitreal hemorrhage, a decrease in hemoglobin >20 g/L, need for ≥2 units of blood transfusion, hospital admission, or endoscopic or surgical procedure to stop the bleeding.

Statistical Analysis

Baseline characteristics of patients are presented as mean (± standard deviation) for normally distributed continuous variables, median (interquartile range) for non-normally distributed continuous variables, and frequency (percentage) for categorical variables. Differences between groups were compared using the Student t-test for normally distributed continuous variables, Mann-Whitney U test for non-normally distributed continuous variables, and Fisher's exact test to compare categorical variables (including for evaluating differences in 30-day mortality rates). Event rates were expressed as number of deaths, VTE recurrences, or major bleedings, respectively, per 100 patient-years. Kaplan–Meier curves and log-rank tests were used to evaluate event-free survival. A 2-tailed P value <.05 was established as the level of statistical significance for all tests. For statistical analysis, the program SPSS® Statistics 27.0.1.0 (International Business Machines Corp., New York) was used.

Results

Baseline Data

Between 2010 and 2019, 2241 patients with diagnosis of PE were registered, of whom 282 were treated at home. Four patients were lost to follow-up, leaving a total of 278 fully registered home-treated PE patients with adequate follow-up data for final analysis. All patients had full information on demographical characteristics, heart rate, oxygen saturation, and blood pressure measurement (ECG and vitals as per ER intake triage protocol). Malignant disease was adjudicated in case of recorded and/or available history of malignant disease anywhere within the medical records (otherwise, the patient was categorized as malignancy-free). RV strain evaluation was based primarily on CT scans with RV strain reporting which was available for all patients. Troponin levels were available for 175 (62.9%) patients, and NT-proBNP levels were available for 83 (29.8%) patients.

The mean age was 62 ± 18 years, 58.6% were women; according to s-PESI criteria, 188 patients (67.6%) were retroactively categorized as an appropriate low-risk class and 90 patients (32.4%) as inappropriate low-risk class. The majority of patients were classified as inappropriate low-risk class because of cancer (52.2%), followed by age > 80 years (47.8%) and cardiopulmonary disease (23.3%). 66 patients had 1 risk factor, and 24 had 2 or more risk factors. Patients in the appropriate low-risk class were younger and more often men. See Table 1. During the observed period, both the number of patients referred to outpatient management as well as the percentage of inappropriately classified patients steadily increased (the number of referred patients and the percentage of inappropriately classified patients were 31% and 19% in 2010-2011, 52% and 29% in 2012-2013, 69% and 30% in 2014-2015, 52% and 33% in 2016-2017, and 74% and 42% in 2018-2019). The data is shown in Supplemental Table 1; P-value for chi-square test for trends: P = .021.

Table 1.

Appropriate and Inappropriate Low-Risk Class Study Population.

	Appropriate low-risk class	Inappropriate low-risk class	P
Number of patients	188	90	/
Age; median (interquartile range)	55 (44-70)	80 (67-84)	<.001
Female	97 (51.6%)	66 (73.3%)	<.001
Chronic lung or heart disease	0 (0%)	21 (23.3%)	<.001
Cancer	0 (0%)	47 (52.2%)	<.001
Elevated troponin	0 (0%)	1 (1.1%)	.102
Elevated NT pro-BNP	11 (5.9%)	10 (11.1%)	.176
Morphologic signs of right ventricular strain	0 (0%)	3 (3.3%)	.028

CT scans with RV strain reporting were available for all patients, troponin levels were available for 175 (62.9%) patients, and NT-proBNP levels were available for 83 (29.8%) patients. Abbreviations: CT: computed tomography; RV, right ventricular.

Early and Long-term Mortality

Thirty-day mortality was low in both risk classes—1 patient (1.1%) died in the inappropriate low-risk class and none in the appropriate low-risk class. Over a median follow-up time of 280 (176-412) days, the overall mortality rate was higher in the inappropriate low-risk class (12.1 vs 0.9/100 patient-years). Median survival time in the inappropriate low-risk class was 186 (110-418) days and 1022 (691-1352) in the appropriate low-risk class, P < .001. Figure 1.

Figure 1.

All-cause mortality according to risk class.

Thromboembolism Recurrence

Thromboembolism recurrence rate was 3.1/100 patient-years in the inappropriate low-risk class and 2.3/100 patient-years in the appropriate low-risk class, respectively (P = .671) (Figure 2).

Figure 2.

Thromboembolism recurrence rate according to risk class.

Major Bleeding

Major bleeding rate was 1.0/100 patient-years in the inappropriate low-risk class and 0.9/100 patient-years in the appropriate low-risk class (P = .583). Figure 3.

Figure 3.

Major bleeding rate according to risk class.

Discussion

Current guidelines suggest early discharge and home treatment of low-risk PE patients upon structured prognostic assessment for serious adverse outcomes.¹¹ Validated prognostic scores can effectively be used for risk stratification.¹⁵ In real-life clinical practice, however, even patients at low risk are seldom treated fully at home.¹⁶ In the present analysis, we have shown that real-life PE home treatment does not necessarily follow a structured risk-assessment prereferral approach. Almost one-third of patients with PE referred for outpatient management were at high 30-day mortality risk (>0 s-PESI score) and should therefore have been considered for hospital admission, had a structured prereferral risk assessment been undertaken. Nonetheless, our findings also suggest that despite a retroactively calculated high early mortality risk, the real-life 30-day mortality was very low and comparable in both groups of patients (appropriate and inappropriate low-risk class).

Several risk scores have been proposed to assess early mortality in PE. s-PESI is recommended by recent guidelines as a simple and effective tool for assessing the risk of early adverse events in patients with PE requiring hospital admission.¹¹ While the simplicity of s-PESI makes it an attractive risk-scoring system, real-life data suggest that physicians more often override s-PESI based risk assessment as compared to more clinically oriented scoring systems. In the HOME-PE trial, risk stratification using HESTIA criteria was less likely to be overridden by the treating physician, and patients with negative HESTIA criteria were more likely to be treated at home than patients stratified using s-PESI. However, regardless of the scoring system used, non-low-risk patients were very rarely discharged.¹⁷ Our analysis shows that, as opposed to randomized clinical trials that use strict clinical protocols, in everyday clinical practice, clinicians don’t routinely resort to standardized scores but rather tend to base their decision on individual clinical judgment. Clinical judgment may encompass wider determinants of health and prognosis beyond those captured by risk scores—such as age, education, and health literacy, or socioeconomic status.¹⁸ This may have yielded a biased selection of our patient populations, but it may also reflect a real-life physician decision-making process. However, as these determinants were not measured in our study, such hypotheses remain highly speculative.

Our post hoc risk stratification showed that the primary criteria for yielding inappropriate low-risk classification were demographic (age) and clinical characteristics (cancer or cardiopulmonary disease presence), rather than alarming hemodynamic signs (ie, heart rate or saturation). While demographic and clinical characteristics may predict high fatality rate of early VTE recurrences in patients with poor cardiopulmonary reserve, hemodynamic criteria point to a more severe form of VTE event, which clinicians may be more prone to admit and monitor intensively.

Importantly, 30-day mortality was lower than previously reported^9,17,19 and was comparable between s-PESI-derived low-risk (appropriate low-risk class) and non-low-risk (inappropriate low-risk class) patients. s-PESI and other scoring systems were primarily devised to help identify at-risk patients for early mortality, which may benefit from hospital admission and intensive monitoring during the early phase of VTE management. Conversely, our patient population—irrespective of s-PESI score—had low early mortality and was safely managed at home despite a lack of structured prereferral risk assessment.²⁰ On the one hand, referring physicians may have chosen to base their clinical judgment on other—non-PESI—criteria, given the accessibility of VTE outpatient clinic in the region (ie, walk-in day hospital in our case). On the other hand, VTE outpatient clinics provide intensive monitoring and follow-up for early detection and management of possible complications, potentially minimizing mortality even in inappropriate low-risk class PE patients.²¹ Because an increasing trend in the proportion of patients inappropriately classified as low-risk cannot be excluded and—while no firm conclusions can be drawn because of power limitations—in combination with the observed clinically relevant outcomes, there may be considerable margins of improvement in the accuracy of risk stratification in the outpatient setting.

In our analysis, we also appraised the performance of s-PESI for the prediction of long-term events. s-PESI emerged as a significant predictor of long-term mortality, but not VTE recurrence and/or major bleeding. Patients in the inappropriate low-risk class had roughly a 12-fold higher long-term mortality rate as compared to appropriate low-risk class patients, confirming that s-PESI criteria identify VTE populations at risk for overall mortality. Although a long-term mortality risk of PE patients beyond 30 days of diagnosis does not influence the decision about hospitalization in the acute phase, s-PESI might help identify patients who require closest possible long-term follow-up.

Strengths and Limitations

To our knowledge, the present analysis represents one of the longest real-life registry-based studies of outpatient management of PE. Nonetheless, we have identified some limitations. Firstly, this was a single-center study at a tertiary/university center with a functional dedicated, and specialized VTE outpatient clinic. Both availability and local expertise should be taken into account when generalizing our findings. Moreover, patients were primarily selected from anticoagulation registry, which may not fully encompass all PE patients who may be eligible for home treatment, thus rendering the generalizability of our results unclear. Secondly, our patient population was relatively small—although comparable with similar reports^10,22,23 and reflecting the fairly recent shift from hospital to home treatment of VTE. Thirdly, ours was an observational study. Some information (eg, troponin and NT-proBNP) was missing/unavailable, possibly resulting in classification bias. In our analysis, however, biomarkers were only recorded as supporting information and no primary observation was categorized on the basis of either troponin or NT-proBNP levels. Another limitation of our analysis might represent the fact that we based our RV strain analysis on CT scans. While echocardiography still represents the gold standard for RV strain evaluation, echocardiography was unfortunately rarely performed at the time of diagnosis. Although concordance between CT and echocardiographic evaluation of RV strain is relatively low, CT is still highly sensitive, albeit not specific, for detecting RV strain.²⁴ Moreover, no patient was classified as high-risk based solely on RV strain, but all primarily met s-PESI criteria. Lastly, although our registry was prospectively designed, the specific analysis of PE home treatment was conceived post hoc and may have therefore also incurred bias (eg, referral bias).

In conclusion, our study has shown that real-life PE home treatment does not necessarily follow a structured risk-assessment prereferral approach, with almost one-third of patients referred to outpatient services possibly being at high early mortality risk. Nonetheless, the mortality of patients was low, suggesting other risk-scoring tools should be developed.

Supplemental Material

sj-docx-1-cat-10.1177_10760296231203209 - Supplemental material for Home Treatment of Patients with Pulmonary Embolism: A Single Center 10-Year Experience from Ljubljana Registry

Supplemental material, sj-docx-1-cat-10.1177_10760296231203209 for Home Treatment of Patients with Pulmonary Embolism: A Single Center 10-Year Experience from Ljubljana Registry by Gregor Tratar, Anteja Batič and Klara Svetina in Clinical and Applied Thrombosis/Hemostasis

Footnotes

Acknowledgments

The authors would like to thank the head of the Department of Vascular Diseases, professor Borut Jug, MD, PhD for his extensive help in preparing the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Gregor Tratar

Supplemental Material

Supplemental material for this article is available online.

References

White

. The epidemiology of venous thromboembolism. Circulation. 2003;107(23 Suppl 1):I4-I8. doi:10.1161/01.CIR.0000078468.11849.66

Barco

Mahmoudpour

Valerio

, et al. Trends in mortality related to pulmonary embolism in the European region, 2000-15: analysis of vital registration data from the WHO mortality database. Lancet Respir Med. 2020;8(3):277-287. doi:10.1016/s2213-2600(19)30354-6.

Barco

Valerio

Gallo

, et al. Global reporting of pulmonary embolism-related deaths in the World Health Organization mortality database: Vital registration data from 123 countries. Res Pract Thromb Haemost. 2021;5(5):e12520. doi:10.1002/rth2.12520.

Othieno

Okpo

Forster

. Home versus in-patient treatment for deep vein thrombosis. Cochrane Database Syst Rev. 2018;1:CD003076. doi:10.1002/14651858.CD003076.pub3.

Wood

. Major pulmonary embolism: review of a pathophysiologic approach to the golden hour of hemodynamically significant pulmonary embolism. Chest. 2002;121(3):877-905. doi:10.1378/chest.121.3.877.

Piran

Le Gal

Wells

, et al. Outpatient treatment of symptomatic pulmonary embolism: A systematic review and meta-analysis. Thromb Res. 2013;132(5):515-519. doi:10.1016/j.thromres.2013.08.012.

Zondag

Kooiman

Klok

Dekkers

Huisman

. Outpatient versus inpatient treatment in patients with pulmonary embolism: A meta-analysis. Eur Respir J. 2013;42(1):134-144. doi:10.1183/09031936.00093712.

Aujesky

Obrosky

Stone

, et al. Derivation and validation of a prognostic model for pulmonary embolism. Am J Respir Crit Care Med. 2005;172(8):1041-1046. doi:10.1164/rccm.200506-862OC.

Jimenez

Aujesky

Moores

, et al. Simplification of the pulmonary embolism severity index for prognostication in patients with acute symptomatic pulmonary embolism. Arch Intern Med. 2010;170(15):1383-1389. doi:10.1001/archinternmed.2010.199.

10.

Zondag

Mos

Creemers-Schild

, et al. Outpatient treatment in patients with acute pulmonary embolism: the Hestia study. J Thromb Haemost. 2011;9(8):1500-1507. doi:10.1111/j.1538-7836.2011.04388.x.

11.

Konstantinides

Meyer

Becattini

, et al. 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS). Eur Heart J. 2019. doi:10.1093/eurheartj/ehz405

12.

Vene

Mavri

Gubensek

, et al. Risk of thromboembolic events in patients with non-valvular atrial fibrillation after dabigatran or rivaroxaban discontinuation – data from the Ljubljana registry. PLoS One. 2016;11(6):e0156943. doi:10.1371/journal.pone.0156943.

13.

Chornenki

NLJ

Poorzargar

Shanjer

, et al. Detection of right ventricular dysfunction in acute pulmonary embolism by computed tomography or echocardiography: A systematic review and meta-analysis. J Thromb Haemost. 2021;19(10):2504-2513. doi:10.1111/jth.15453.

14.

Kokalj

Kozak

Jug

. Post-acute pre-discharge echocardiography in the long-term prognostic assessment of pulmonary thrombembolism. Sci Rep. 2021;11(1):2450. doi:10.1038/s41598-021-82038-1.

15.

Palas

Silva

Jorge

Almeida

Pinto

Caldeira

. The accuracy of Hestia and simplified PESI to predict the prognosis in pulmonary embolism: systematic review with meta-analysis. TH Open. 2022;6(4):e347-e353. doi:10.1055/a-1942-2526.

16.

Mastroiacovo

Dentali

di Micco

, et al. Rate and duration of hospitalisation for acute pulmonary embolism in the real-world clinical practice of different countries: analysis from the RIETE registry. Eur Respir J. 2019;53(2). doi:10.1183/13993003.01677-2018.

17.

Roy

Penaloza

Hugli

, et al. Triaging acute pulmonary embolism for home treatment by Hestia or simplified PESI criteria: The HOME-PE randomized trial. Eur Heart J. 2021;42(33):3146-3157. doi:10.1093/eurheartj/ehab373.

18.

Wadhera

Secemsky

Wang

Yeh

Goldhaber

. Association of socioeconomic disadvantage with mortality and readmissions among older adults hospitalized for pulmonary embolism in the United States. J Am Heart Assoc. 2021;10(13):e021117. doi:10.1161/jaha.121.021117.

19.

Yamashita

Morimoto

Amano

, et al. Validation of simplified PESI score for identification of low-risk patients with pulmonary embolism: From the COMMAND VTE registry. Eur Heart J Acute Cardiovasc Care. 2020;9(4):262-270. doi:10.1177/2048872618799993.

20.

Elias

Schmidt

Bellou

, et al. Opinion and practice survey about the use of prognostic models in acute pulmonary embolism. Thromb Res. 2021;198:40-48. doi:10.1016/j.thromres.2020.10.027.

21.

Wan

Rahmani

Hanakova

, et al. Reducing emergency department visits in patients with deep vein thrombosis: Introducing a standardised outpatient treatment pathway. BMJ Open Qual. 2021;10(2). doi:10.1136/bmjoq-2020-001123.

22.

Otero

Uresandi

Jimenez

, et al. Home treatment in pulmonary embolism. Thromb Res. 2010;126(1):e1-e5. doi:10.1016/j.thromres.2009.09.026.

23.

Elf

Jogi

Bajc

. Home treatment of patients with small to medium sized acute pulmonary embolism. J Thromb Thrombolysis. 2015;39(2):166-172. doi:10.1007/s11239-014-1097-y.

24.

Lyhne

Giordano

Dudzinski

, et al. Low concordance between CTPA and echocardiography in identification of right ventricular strain in PERT patients with acute pulmonary embolism. Emerg Radiol. 2023;30(3):325-331. doi:10.1007/s10140-023-02130-z.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB