Comprehensive Outpatient Management of Low-Risk Pulmonary Embolism: Can Primary Care Do This? A Narrative Review

Abstract

Introduction

The evidence for outpatient management of hemodynamically stable, low-risk patients with acute symptomatic pulmonary embolism (PE) is mounting. Guidance in identifying patients who are eligible for outpatient (ambulatory) care is available in the literature and society guidelines. Less is known about who can identify patients eligible for outpatient management and in what clinical practice settings.

Objective

To answer the question, “Can primary care do this?” (provide comprehensive outpatient management of low-risk PE).

Methods

We undertook a narrative review of the literature on the outpatient management of acute PE focusing on site of care. We searched the English-language literature in PubMed and Embase from January 1, 1950, through July 15, 2019.

Results

We identified 26 eligible studies. We found no studies that evaluated comprehensive PE management in a primary care clinic or general practice setting. In 19 studies, the site-of-care decision making occurred in the Emergency Department (or after a short period of supplemental observation) and in 7 studies the decision occurred in a specialty clinic. We discuss the components of care involved in the diagnosis, outpatient eligibility assessment, treatment, and follow-up of ambulatory patients with acute PE.

Discussion

We see no formal reason why a trained primary care physician could not provide comprehensive care for select patients with low-risk PE. Leading obstacles include lack of ready access to advanced pulmonary imaging and the time constraints of a busy outpatient clinic.

Conclusion

Until studies establish safe parameters of such a practice, the question “Can primary care do this?” must remain open.

INTRODUCTION

The initial site of care of patients with newly diagnosed, acute, symptomatic pulmonary embolism (PE) is undergoing a transition away from routine hospitalization for select low-risk patients.^1–3 The supporting evidence for outpatient management (without hospitalization) continues to mount and has involved multiple countries and different types of health care systems, including, for example, a multinational randomized controlled trial in academic medical centers and a recent controlled pragmatic trial in community hospitals in the US.^4,5 Outpatient (ambulatory) care for eligible low-risk patients is endorsed by specialty societies around the world.^6–9 The practice improves the health care community’s resource stewardship and spares patients the costs, inconveniences, and risks associated with unnecessary hospitalization.^10,11

However, little is known about who can identify patients eligible for outpatient management and in what clinical settings. A stable, ambulatory patient with PE-related complaints may present to a variety of venues, including the primary care clinic, specialty (or secondary care) clinic, or the Emergency Department (ED). Comprehensive outpatient PE care requires diagnostic confirmation, determination of outpatient eligibility, anticoagulation, patient and family education, and arrangement for close follow-up. This level of care necessitates that the clinician coordinate laboratory, radiology, pharmaceutical, and educational resources (Table 1).

Table 1.

Elements and resources involved in comprehensive outpatient pulmonary embolism care, regardless of clinical setting

Element	Resourcea	Examplesa
Diagnostic evaluation	Laboratory	D-dimer
	Radiology	Chest radiography, computed tomography pulmonary angiography (CTPA); compression ultrasonography
	Nuclear medicine	Ventilation-perfusion imaging
Determination of outpatient eligibility	Oxygen saturation	Peripheral cutaneous oxygen saturation
	Laboratory	Biomarkers of right ventricular dysfunction, eg, troponin
	Radiology	Compression ultrasonography, assessment of right ventricular dysfunction, eg, CTPA or echocardiography if deemed appropriate
	Consultant	Thrombosis specialist (ie, pulmonologist, hematologist, or internal medicine physician), emergency medicine physician
Anticoagulation	Laboratory	Complete blood cell count, creatinine clearance
	In-office medications	Initiate treatment before pulmonary imaging or discharge to pharmacy, depending on pretest risk assessment
	Pharmacy	Direct oral anticoagulants, low-molecular-weight heparinsb
Education of patient and family	Information on the disease and the treatment, including what to expect and what to watch for	Physician or nurse: Conversation supplemented with printed or electronic discharge instructions; pharmacist: Proper medication use (including subcutaneous injections if low-molecular-weight heparins are indicated), adverse effects, complications
Arrangement for close follow-up	Appointment access	Primary or specialty care; telephone-based anticoagulation management services, if available

^aResources for the diagnostic evaluation and the determination of outpatient eligibility assume the performance of a thorough history and physical examination, including basic vital signs. Resource availability and clinical application vary greatly by patient, clinician, and practice setting. We report here only common examples, which may or may not be indicated in every case. See Table 2 and Table 3 (available from: www.thepermanentejournal.org/files/2020/19.163T3.pdf⁴,²⁰–³⁴) for illustrations of variation in practice.

^bFor some patients who are uninsured or have limited pharmacy coverage, additional personnel (eg, social workers) may be needed to help with medication procurement.

Which of the above settings can provide such care? What is the evidence that primary care clinics can marshal the resources needed for outpatient management of acute PE? Or that they have the time and staffing to do so? To address these questions, we undertook a narrative review of the literature.

METHODS

One of us (PMR) recently published a narrative review of outpatient PE management¹ that we in this current review have adopted, modified, and expanded. The original search was a systematic review from January 1950 to December 2016 using PubMed and Embase, with a manual search of references used in the main studies. We used the search terms pulmonary embol* or pulmonary thromboembol* and outpatient* or ambulatory care or home care or home treatment. Studies were included only if they were published in English and explicitly mentioned the outpatient treatment setting or early hospital discharge of patients with acute, symptomatic, objectively proven PE. We excluded abstracts, editorials, and reviews.¹

For this current narrative review, we ran a second search from January 1, 2017, through July 15, 2019, using the same sources, search terms, and eligibility criteria. From the expanded collection of studies, we excluded those not reporting outpatient management (defined here as discharge to home from the ambulatory clinic, the ED or specialty unit, or within 48 hours [≤ 2 nights] of hospitalization for observation), not reporting PE-specific clinical outcomes for patients with nonincidental PE, not specifying venues of care (ED vs clinic), discharging patients to a patient hotel, and those with secondary analyses of datasets already included in the review.

RESULTS

We identified 26 eligible studies.^4,5,11–34 As of July 15, 2019, we found no studies of comprehensive PE management provided in a primary care or general practice clinic. In 19 studies the site-of-care decision making occurred in the hospital-based ED (or ambulatory care unit) or after a short period of supplemental outpatient or inpatient observation.^4,5,11–27 In 7 studies, site-of-care decision making occurred in a specialty clinic.^28–34 The research on this topic has been recently accelerating, because 10 of the 19 ED studies were published since January 1, 2017.^5,11–19 We report findings from the included studies in Table 2 (studies published on or after January 1, 2017) and Table 3^a (studies published before January 1, 2017). Both Tables 2 and 3^a are organized by patient care setting (ED/ambulatory care unit and specialty clinic). Seventeen studies are prospective in nature, and 16 include more than 100 outpatients (range = 30–544 outpatients). The research on outpatient PE management is an international endeavor, because the 26 studies were conducted in 16 countries.

Table 2 and Table 3^a illustrate the variety encompassed under the broad umbrella of outpatient PE management. Variation is evident across the spectrum of care: Who arrives for evaluation (walk-ins such as to the ED vs a referred population, as seen in many specialty clinics); how patients are identified as eligible for home care (physician discretion vs explicit criteria, which also vary widely; one study used a separate risk stratification score for patients with cancer-associated PE^23,35); whether observation is required and, if so, for how long; pharmacotherapy (eg, low-molecular-weight heparin [LMWH], warfarin, or a direct oral anticoagulant [DOAC]); the nature of postdischarge follow-up care; and the timing of outcome metrics. Differences continue beyond the variables reported in Table 2 and Table 3^a, such as extent and content of patient education.

Despite the diversity of approaches to outpatient PE management illustrated here, the clinical outcomes are reassuring. The combination of careful patient selection, appropriate treatment, attentive patient and family education, and close follow-up (Table 1) facilitates favorable outcomes, as attested by the low incidence of adverse outcomes across the studies.

DISCUSSION

Two Sites of Outpatient Pulmonary Embolism Care Described in the Literature

1. Emergency Department (and Ambulatory Care Unit)

Much of the research on comprehensive outpatient PE management that we identified in our literature search has been undertaken in traditional hospital-based EDs (Table 2 and Table 3^a). The ED is a natural venue for outpatient PE research because many patients with suspected or newly diagnosed PE present themselves (or are directed) to its doors, which are conveniently open 24/7. The acceptance of all-comers includes patients with PE arriving by ambulance, who are a higher-acuity population and can constitute in some settings approximately 20% of the entire PE population in the ED.¹⁵ The ED has easy access to laboratory, radiology, and nuclear medicine studies to pursue and secure a PE diagnosis (Table 1).³⁶ Continuous cardiopulmonary monitoring is readily available if needed. If a 12-hour to 24-hour period of formal observation is indicated, some EDs just extend the patient stay, whereas others transfer care to an affiliated outpatient observation or clinical decision unit.³⁷ Some studies of outpatient PE management include up to a 24-hour observation period in their definition of outpatient care. The meaning of outpatient itself varies, as there is no established definition. In some PE studies, outpatient care includes a stay in the inpatient setting. We note those studies of expanded ED care in Table 2 and Table 3^a.

When the time for disposition arrives, the ED can easily risk-stratify their patients with PE to identify those eligible for discharge to home (more on this later in this section).³⁸ On the treatment side, the ED can initiate anticoagulation therapy and begin patient education, which can continue when the patient is introduced to the pharmacy before or just after discharge. Thrombosis specialists are often available at all hours for consultation. Facilitating postdischarge follow-up care is the 1 element of comprehensive PE care that can be difficult for some EDs to achieve.^39,40 Post-ED follow-up can include more than just general practitioner or specialty clinic appointments; some health care organizations also provide a pharmacy-led, telephone-based outpatient anticoagulation team (anticoagulation management services) that follows-up with these patients, whatever their anticoagulant.^39,41,42

A variation of the traditional ED care delivery model is the UK’s hospital-based ambulatory emergency care unit.⁴³ Patients are accepted into the unit by clinician referral only and are limited to those who are likely manageable as outpatients,¹² including patients transferred in via ambulance. Most of these units are not open around-the-clock. Proximity to the affiliated medical center gives these ambulatory care units ready access to the laboratory and advanced imaging resources needed for the diagnosis and risk stratification of patients with acute PE.

2. Specialty Clinic Setting

In some countries outside the US, such as Canada, it is not the ED to which patients with diagnosed or suspected PE are referred. Specialty-run thrombosis clinics have featured prominently in the literature on outpatient PE management (Table 3^a). The specialty that manages these “clot clinics” varies and includes internal medicine, pulmonology, hematology, and vascular medicine. Oncology clinics can also provide comprehensive care for their stable, outpatient care-eligible patients with PE, and sometimes share tasks with pharmacists.⁴⁴ These secondary care thrombosis clinics, like the ED, have the skill set and resources to provide care from diagnosis to treatment, risk stratification, and discharge, and, contrary to the ED, specialty clinics can provide their own follow-up care. The disadvantages compared with the ED is that these clinics often do not receive ambulance traffic, nor are they always open around-the-clock. Another difference is that specialty-run clinics are not usually equipped with continuous cardiopulmonary monitoring, although the importance of this component of care in assessing outpatient eligibility is not known. For many hemodynamically stable patients with low-risk PE, 1 or 2 sets of vital signs may be sufficient to confirm stability.

Several society guidelines address criteria for outpatient PE site-of-care decision making without specifying the training and experience of the decision maker.^6,9 The British Thoracic Society, however, is more explicit: If PE is diagnosed by a general practitioner in the outpatient setting in the UK, the patient should be transferred to the ED or an ambulatory care unit,^12,45,46 as explained earlier, where they can be evaluated by a consultant or a clinician “designated to undertake this role within the department with consultant advice available.”⁷

Paving the Way for Comprehensive Primary Care-based Pulmonary Embolism Management

The growing literature on the safety and effectiveness of outpatient management of PE in the ED and specialty clinic setting have set the stage for management of select patients with low-risk PE in the primary care setting. Two other factors have helped pave the way for primary care physicians to expand their role in PE management: Decentralization of management of deep vein thrombosis (DVT) and simplification of pharmacotherapy.

Decentralizing Deep Vein Thrombosis Management

For select patients with DVT, a similar shift in site of care—from the ED to the primary care clinic—began years ago in the US and is now well established in some countries, such as France. After the advent of LMWH, one of our medical centers in the US, part of a large integrated health care system, developed an outpatient clinical care pathway for select patients with DVT.⁴⁷ Initially, all patients with newly diagnosed DVT were directed to the hospital-based ED for risk stratification to inform site-of-care decision making. Over time, it was realized that for some low-risk patients the temporary transfer of care to the ED was superfluous—the referring primary care clinician was just as capable of identifying which patients were eligible for outpatient treatment and managing these patients without recourse to the ED. Our medical center then pulled together a multidisciplinary team to design, implement, and monitor a clinical care pathway to enable general practitioners to provide comprehensive outpatient DVT management.⁴⁸ Today such practice has become more common in multiple settings around the world.⁴⁹ Perhaps such a change is on the horizon for select patients in the right practice settings with acute symptomatic PE.

Simplifying Pharmacotherapy

A more recent shift in pharmacotherapy away from vitamin K antagonists, such as warfarin, might facilitate the provision of comprehensive PE care in the primary care setting.⁵⁰ Recent society guidelines recommend DOACs, also known as nonvitamin K (or novel) oral anticoagulants, as the preferred agents for most patients with acute PE.^6,7,9,51 The DOACs avoid some of the complexities associated with vitamin K antagonists, such as regular laboratory monitoring and dose adjustments, as well as many food and drug interactions.⁵² Even greater ease of administration is achieved with DOACs that are approved as monotherapy for PE (eg, rivaroxaban and apixaban), obviating the need for a 5- to 10-day lead-in period of subcutaneous LMWH therapy required with some DOACs (eg, dabigatran and edoxaban). The acquisition costs of DOACs, however, are an ongoing concern, particularly among socioeconomically disadvantaged populations, for whom out-of-pocket costs might be prohibitive.⁵³ The efficacy and safety of DOACs in patients with cancer-associated PE are currently under investigation.^54–56 Because DOACs have been associated with an increased risk of gastrointestinal and possible genitourinary tract bleeding, they should be used with caution in patients with malignancies in these regions.⁵⁷ The 2019 European Society of Cardiology guidelines recommend that in “patients with an anticipated low risk of bleeding and without gastrointestinal tumours, the choice between LMWH and edoxaban or rivaroxaban is left to the discretion of the physician, and the patient’s preference.”⁹ Access to DOACs alone, however, is insufficient to facilitate outpatient PE care without concurrent implementation of the structural processes of care needed to support ambulatory PE management.⁵⁸

Exploring the Primary Care Setting for Comprehensive Pulmonary Embolism Management

In this review of the literature we failed to identify any studies meeting our eligibility criteria that describe PE management contained entirely in the primary care clinic setting, that is, comprehensive primary care clinic-based management. The lack of literature on this model of care delivery does not mean that such care is not being provided—we know anecdotally that it is. Lack of a published description of this care model, however, prevents a critical understanding and analysis of its execution by the medical community at large and impedes its expansion and adaptation to other clinics. In advance of such literature, we introduce the 4 key elements required for comprehensive care of patients with acute PE in the primary care setting: 1) outpatient diagnosis, 2) identification of patients eligible for outpatient care, 3) patient education, and 4) timely follow-up.

1. Pursuing the Outpatient Diagnosis of Pulmonary Embolism

The most difficult and challenging aspect for securing the diagnosis of PE in primary care is identifying which patients with PE-related complaints warrant diagnostic evaluation. Both underimaging and overimaging may cause harm; the former contributes to a delay in diagnosing a potentially fatal condition, and the latter, in the case of computed tomography pulmonary angiography (CTPA), can lead to needless complications from intravenous contrast medium (eg, allergic reaction and contrast agent-induced acute kidney injury) and exposure to radiation (eg, breast cancer), not to mention poor resource utilization. Nevertheless, once a suspicion is clear and ruling out PE becomes imperative, the management of patients with suspected PE typically relies on the combination of pretest probability (ie, the clinical assessment based on historical and examination findings) and selective D-dimer testing, both readily available in primary care.³⁶ We will address these separately.

Assessing pretest probability

Owing to the frequency and lack of specificity of the signs and symptoms of PE, the clinical decision to investigate appears to be mainly subjective. A promising starting point in the evaluation of a patient with possible PE is the PE rule-out criteria.^59–61 When applied to patients with a low pretest probability of PE as judged by physician gestalt, these criteria can exclude PE solely on clinical grounds, without the need for laboratory or radiology testing. A randomized trial found that ED patients with very low pretest probability who had none of the specified 8 criteria could safely forgo additional diagnostic evaluation, including a D-dimer test, with reassuring outcomes.^62,63 The PE rule-out criteria are advocated by the American College of Physicians for use by outpatient physicians,³⁶ but they may not be ready for broad application in primary care until they are validated in this setting.

Patients who have 1 or more of the PE rule-out criteria or for whom the criteria are not applicable (because patients are not low risk by gestalt) need additional pretest probability stratification using one of several evidence-based clinical prediction rules widely endorsed by society guidelines.^6,8,9,36 Five of these prediction tools for PE diagnosis have been validated in primary care and are easily applied in this setting: The original Wells, modified Wells, simplified Wells, revised Geneva, and simplified revised Geneva models.⁶⁴ Whereas efficiency was comparable for all 5, the Wells rules demonstrated the best performance in terms of lower failure rates, that is, the lowest risk of missed PE when imaging was withheld.⁶⁴ Performance of these rules can vary considerably depending on differences in disease prevalence and practice environment, where both case mix and physician experience vary.^65,66

Using D-dimer in the assessment

Patients with low to moderate pretest probability of PE should receive D-dimer testing. A low D-dimer value in this population safely excludes PE. Specifically for primary care, a meta-analysis found this to be true also for the use of rapid point-of-care D-dimer assays.⁶⁷ Results of a prospective study in Dutch primary care settings confirmed that the combination of the Wells score with a qualitative point-of-care D-dimer assay safely excluded the diagnosis in patients with suspected PE, comparing favorably with similar studies performed in secondary and tertiary care settings.⁶⁸ D-dimer values show improved efficiency when interpreted in light of age as well as pretest clinical probability.^69–71 A structured diagnostic approach that is built around a simplified Wells rule is the YEARS algorithm, which has demonstrated good performance in the ED and inpatient settings.⁷⁰ A large prospective study of the YEARS algorithm is under way to validate a risk-stratified use of D-dimer (rather than a 1-size-fits-all approach) in the primary care setting.⁷²

Obtaining advanced pulmonary imaging

The probability assessment crosses the threshold for advanced imaging if the patient has a high pretest probability for PE or a low to moderate pretest probability with an elevated D-dimer value.³⁶ Research findings have established the effectiveness and safety of validated strategies for the diagnosis of acute PE in the ambulatory care setting.^64,68,73 Multidetector CTPA is the imaging method of choice in most patients with suspected PE. A ventilation-perfusion scan is preferred for patients with severe renal failure.⁹ Which physician specialty orders advanced imaging, however, varies considerably across practice settings and may be subject to established local (or national) patterns of care as well as physician schedule, staffing, and time of day. In some practice settings, the primary care physician has ready access to timely pulmonary imaging and radiology interpretation and can proceed with imaging if indicated. We see this in action in one of our own practice settings (DRV). For example, in a real-world study of outpatient PE management in the US, 14.5% of 1703 ED patients arrived with a diagnosis in hand, thanks to a pulmonary imaging study ordered by an outpatient clinician, most commonly primary care physicians.⁵ However, timely and convenient advanced imaging services are not available to all primary care clinics. In these cases, patients may need to be referred to the ED, ambulatory care unit, or specialty clinic for reassessment and ordering of diagnostic imaging if indicated. In some countries, such as the Netherlands and the UK, primary care physicians who identify patients in need of advanced PE imaging customarily transfer them to a higher level of care to confirm the diagnosis.^7,73

2. Identifying Patients with Pulmonary Embolism Who are Eligible for Ambulatory Care

If a primary care physician sought to provide comprehensive care for select patients with newly diagnosed acute PE, the next step would be determining eligibility for outpatient management. The broader topic of outpatient PE care has been much studied, as the results in Table 2 and Table 3^a attest, although none of these studies speak directly to the primary care setting. The CHEST criteria to determine outpatient eligibility are simple and sensible. The patient should be “clinically stable with good cardiopulmonary reserve; no contraindications such as recent bleeding, severe renal or liver disease, or severe thrombocytopenia (ie, <70,000/mm³); expected to be compliant with treatment; and the patient feels well enough to be treated at home.”⁶ Treatment compliance requires a certain level of health literacy, motivation, and psychosocial stability, factors commonly included in the eligibility criteria of outpatient PE studies (Table 2 and Table 3^a).⁷⁴

Numerous prognostic models are available to aid the physician in identifying low-risk patients who may be eligible for outpatient management.⁷⁵ The validated instruments most well studied to guide the disposition decision are the PE Severity Index and its shortened counterpart, the simplified PE Severity Index (Table 4).^76,77 Both indexes provide estimates of 30-day all-cause mortality.^4,78,79 The simplified PE Severity Index identifies fewer patients who are eligible for outpatient care than the original.^75,80 It is, however, easier to remember than the original, a distinction less meaningful in this day of autopopulating electronic clinical decision-support tools.⁸⁰ The European Society of Cardiology has incorporated the PE Severity Index into its risk stratification system.⁹ When used in site-of-care decision making, short-term mortality estimates are combined with commonsense contraindications to ambulatory care, as several studies have done (Table 2 and Table 3^a).^5,81

Table 4.

Pulmonary embolism severity indexes

Parameter	Original scorea77	Simplified scoreb76
Demographic characteristics
Age/y	+1
Age > 80 y	—	+1
Male sex	+10	—
Comorbid illness
Cancer (active or history of)	+30	+1
Heart failure (systolic or diastolic)	+10	+1c
Chronic lung disease (includes asthma)	+10
Clinical findingsd
Pulse ≥ 110/min	+20	+1
Systolic blood pressure < 100 mmHg	+30	+1
Respiratory rate ≥ 30/min	+20	—
Temperature < 36°C	+20	—
Arterial oxygen saturation < 90%e	+20	+1
Altered mental statusf	+60	—

^aA total point score for a given patient is obtained by summing the patient’s age in years and the points for each applicable prognostic variable. Point scores correspond with the following classes that estimate escalating risks of 30-day all-cause mortality: ≤ 65 points, class I; 66–85 points, class II; 86–105 points, class III; 106–125 points, class IV; > 125 points, class V. Patients with 85 points or less (classes I and II) are considered low risk and eligible for ambulatory care consideration.⁴

^bA total point score for a given patient is obtained by summing the points for each applicable prognostic variable. Patients with 0 points are considered low risk.

^cThe 2 variables were combined into a single category of chronic cardiopulmonary disease, that is, a patient is awarded 1 point for having either heart failure or chronic lung disease.

^dThe most abnormal vital signs in the direction of interest were used. Some studies include prearrival findings from emergency medicine services or the referring clinic.⁵,¹⁵

^eWith or without supplemental oxygenation.

^FAcute or preexisting disorientation, lethargy, stupor, or coma.

Index scores can be used in a strict fashion; for example, only patients with lower-risk class I or II scores on the PE Severity Index are considered for ambulatory care,^4,18 or in a looser, informative fashion, in which mortality estimates contribute to the decision-making process but do not categorically govern it.^5,12,82

The American College of Chest Physicians endorses this more flexible use of the PE Severity Index in their recent PE guideline, stating, “We consider clinical prediction rules as aids to decision making and do not require patients to have a predefined score (eg, low-risk PE Severity Index score) to be considered for treatment at home.”⁶ This approach of using prognostic rules as an adjunct to clinical judgment has been adopted by other guideline committees in site-of-care recommendations for other clinical conditions. For example, the UK’s National Institute for Health and Care Excellence (NICE) guideline for adult pneumonia recommends that physicians “use clinical judgement in conjunction with the CRB65 score[83] to inform decisions about whether patients need hospital assessment.”⁸⁴ Clinicians are advised to “consider” hospitalization for patients with higher-risk scores.

A second, validated, commonly used approach to identify patients with PE who are eligible for home discharge focuses on outpatient management exclusion criteria (Table 5). The first such list originated in Canada, where it has been safely used for decades.^34,85,86 These were expanded to form the Hestia criteria (Table 5), which also perform well in varied settings (Table 2 and Table 3^a).^20,24 A similar list of outpatient exclusion criteria was employed in a large multinational outpatient PE trial that identified home eligibility on the basis of low-risk classification by the PE Severity Index (Table 5).⁴ How the 2 overall strategies (mortality estimates plus exclusion criteria vs exclusion criteria alone) compare in terms of safety and efficiency has not been well studied. An international randomized controlled trial of the 2 approaches recently completed enrollment (clinicaltrials.gov identifier: NCT02811237).⁸⁷ This and similar studies will help define the role these tools can play in assisting site-of-care decision making.

Table 5.

Criteria used to exclude patients with acute pulmonary embolism (PE) from outpatient management

Categorizationa	Criteria used in randomized controlled trial of PE Severity Index4	Criteria used in Hestia Studyb24
PE factor
Pain	Chest pain necessitating parenteral opioids	Severe pain needing intravenous pain medication > 24 h
Hemodynamics	SBP < 100 mmHg	(SBP < 100 mmHg + pulse > 100/min) or unstable by clinical judgment or requiring ICU care
O2 saturation	Hypoxemia	> 24 h of O2 supply needed to maintain O2 saturation > 90%
Prearrival anticoagulation	Therapeutic oral anticoagulation	PE diagnosed during anticoagulation therapy
Treatment	Not included	Requiring thrombolysis or embolectomy for reasons other than hemodynamic instability
Comorbid condition
Bleeding or risk thereof	Active bleeding or high risk of bleeding	Active bleeding or high risk of bleeding: GI bleeding or surgery ≤ 2 wk ago, stroke ≤ 1 mo ago, bleeding disorder or platelet count < 75 × 109/L, uncontrolled hypertension (SBP > 180 mmHg or DBP > 110 mmHg), or by clinician judgment
Renal function	Severe renal failure	Creatinine clearance < 30 mL/min according to Cockroft-Gault formula
Liver function	Not included	Severe liver impairment by physician judgment
Pregnancy	Pregnant	Pregnant
Heparin intolerance	Not included	Documented history of heparin-induced thrombocytopenia
Psychosocial factor
Psychosocial factor	Barriers to adherence or follow-up; imprisonment	Medical or social reason for admission > 24 h (infection, malignancy, no support system)

^aThe tripartite categorization of PE factors, comorbid conditions, and psychosocial factors has been used elsewhere.⁷⁴

^bThe 11 Hestia criteria were originally framed as questions; if any were answered in the affirmative, outpatient treatment was contraindicated.

DBP = diastolic blood pressure; GI = gastrointestinal; ICU = intensive care unit; O₂ = oxygen; SBP = systolic blood pressure.

Most of the above patient identification strategies do not require routine evaluation of right ventricular function in hemodynamically stable, low-risk patients. Selective use of echocardiography and serum biomarkers, such as troponin, accords with the recommendation of leading society guidelines.^6,7 The 2019 PE guidelines of the European Society of Cardiology, however, are the exception, calling for routine imaging of the right ventricle, even in otherwise low-risk patients, using CTPA or echocardiography.⁹ Some evidence suggests that such testing may add clinically useful prognostic value even in normotensive patients with low-risk PE, although this is still being worked out.^88–91 Routine testing of right ventricular function has been incorporated into some clinical pathways to identify patients with PE who are eligible for outpatient care (Tables 2 and 3^a).^18,92 However, adding N-terminal B-type natriuretic peptide to the Hestia rule does not appear to improve risk stratification for outpatient PE treatment.²⁰ What role the assessment of right ventricular function will play in the determination of primary care clinic-based outpatient eligibility is unclear.

If outpatient PE management is a viable option for the primary care patient with acute PE, the physician should work together with the patient to arrive at a mutually agreed-on site-of-care treatment plan (transfer of care vs home discharge).^93,94 Who better to take into account a patient’s values and preferences in shared decision making than the physician who knows the patient best? Few studies have evaluated shared decision making in any aspect of venous thromboembolic disease management; site-of-care decision making from the primary care clinic is not among them.⁹⁵ Also given the paucity of literature on comprehensive primary care-based PE management, the evidence used in the shared decision-making discussion would have to be drawn from the broader outpatient PE literature performed in the ED and specialty clinic settings (Tables 2 and 3^a).

3. Patient Education

Once the diagnosis of PE is established and eligibility for outpatient care is confirmed, additional responsibilities fall on the clinic that is entertaining comprehensive outpatient management (Table 1). The first among these is patient education. Topics here include at a minimum the risk factors, course, complications and prevention of PE; anticoagulation dosing, duration, medication interactions and adverse effects; and when and where to seek medical evaluation for new or worsening symptoms. Society guidelines in both Europe and the US recommend DOACs as the drugs of choice in the treatment of acute PE.^6,9 Some DOACs, however, such as dabigatran and edoxaban, require a 5- to 10-day lead-in with a LMWH, in which case instruction on subcutaneous medication administration will be necessary. In some practice settings, patient education of this sort lies principally with the nursing staff.

Currently, most society guidelines recommend at least 3 months of anticoagulation in the treatment of a first episode of acute PE, barring major contraindications.^6,9,51 The decision to extend anticoagulation therapy beyond 3 months depends on weighing the risks of venous thromboembolic recurrence with the risk of bleeding and can be a complex calculation in which patient preference and consultation with a thrombosis specialist factor prominently.⁹

4. Timely Follow-up

Timely follow-up after initial home discharge is important to assess symptom control; evaluate for the effectiveness of anticoagulation therapy and its adverse effects; and continue patient education on the disease, its treatment, and the prevention of recurrence and complications. The optimal timing and frequency of initial postdischarge follow-up has not been established, as the variation in Tables 2 and 3^a attests. Most outpatient PE studies and clinical care pathways include an initial outpatient clinic appointment within 7 days.³⁹ Follow-up thereafter can be tailored to the patient’s needs. An additional feature of long-term management of patients with a history of PE is to monitor for recurrence as well for the development of chronic thromboembolic pulmonary hypertension.⁹⁶ The aspects of long-term outpatient PE management that typically follow discharge from the ED or hospital are well within the established purview of primary care in the countries in which we practice.

Case Example

We include a hypothetical case example in the Sidebar: Case Example to illustrate the components of comprehensive primary care-based PE management that we have discussed in this narrative review (Table 1).

Case Example.

A 32-year-old woman presents to your primary care office in the morning with a 3-day history of intermittent, mild, lateral right-sided pleuritic chest pain and mild dyspnea with moderate exertion. She says she has no fever, coryza, hemoptysis, or leg complaints. She returned home last week to San Francisco, CA, after a 10-day family vacation in Auckland, New Zealand. She has no abnormalities in her medical history. Her only medication is an oral estrogen-progestin contraceptive, which she began 3 months ago. Neither she nor her relatives have a history of thrombophilia nor venous thromboembolism. You are working in a multispecialty group that has an established outpatient pulmonary embolism (PE) diagnostic algorithm, based on the American College of Physicians Best Practice Advice,¹ and a disposition pathway, based on the American College of Chest Physicians CHEST guideline and expert panel report.²

Her vital signs are normal, including a cutaneous peripheral oxygen saturation. The results of her heart, lung, and limb examinations are also normal, as is a 12-lead electrocardiogram and chest radiograph. Using the original Wells criteria, you calculate that she has a moderate pretest probability for acute PE, so you send her to the on-site laboratory for a serum D-dimer, which returns a result at 1075 ng/dL (normal value for her age is < 500 ng/dL). Her complete blood cell count, renal function, and aspartate aminotransferase level are normal. Because of her moderate pretest probability, you administer an initial dose of a direct oral anticoagulant (DOAC) approved as monotherapy for PE and arrange for a computed tomography pulmonary angiography (CTPA) early that afternoon at the affiliated radiology suite across town. They call to inform you that she has a right-sided lobar embolism and no evidence of right ventricular enlargement or dysfunction.

She and her partner return to your office. Her second set of vital signs are relatively unchanged. You explain to them the diagnosis, the need to begin a 3-month course of a DOAC, and the options they have for when to discontinue oral contraceptive treatment.^3–5 In evaluating her site-of-care options, you calculate her PE Severity Index score.⁶ Her 32-point score places her in the lowest mortality-risk category (class I) and, because of her lack of contraindications to outpatient care, including the Hestia criteria, she is eligible to safely forgo hospitalization.^2,7–9 She also meets the American College of Chest Physicians criteria for outpatient care.²

Using a shared decision-making model, you discuss the benefits and risks of the next site-of-care options. For the first option, they can drive to the local affiliated Emergency Department, which has access to her electronic health records, including your note and today’s laboratory and radiology results. She then may be discharged home from the Emergency Department or observed overnight. Alternatively, they can go to the pharmacy down the hall, pick up her anticoagulation and analgesic medications, then go home. She and her partner prefer the second option. You write her a note for 1 week off work. While the clinic nurse completes the patient education that you began, the office staff schedules her for a telephone appointment with you in 2 days and an in-office visit in 1 week. The couple decides to continue oral contraception for the next 8 weeks with plans to switch to an intrauterine device 1 month before discontinuing anticoagulation.

Her 3-month PE treatment course is uneventful. No additional venous thromboembolism or other complications develop over the subsequent 2 years.

Advantages and Risks of Comprehensive Primary Care-based Pulmonary Embolism Management

Advantages of comprehensive primary care-based outpatient PE management are expected at the patient level. These include maintaining continuity of care throughout the course of PE management by reducing the care transitions that can jeopardize patient safety.⁹⁷ Maximizing home time (ie, time alive and out of a health care institution) and minimizing ED and hospital visits are additional important patient-centered outcomes.^98,99 In the US it also will save patients out-of-pocket costs, which can be substantial. These patient-level factors may contribute to improvements in patient satisfaction and quality of life. Benefits may also be seen at the public health level, with reductions in overall health care expenses and a better stewardship of hospital resources.^10,11,100

How the risks of this newer model—in terms of unplanned ED visits and hospitalization, and short-term major hemorrhage, recurrent venous thromboembolism, and mortality—compare with transferring care to the ED or specialty care clinic, however, is unknown. As our literature review findings demonstrate, little research has been undertaken on comprehensive PE care in the primary care setting. To begin to address this deficit, one of us (DRV) has a retrospective cohort study under way that will shed some light on this new model of PE care delivery, at least as practiced in a community-based, integrated health care system in the US.¹⁰¹ Far more research, however, will be needed before this novel approach to PE management is well understood in its varied settings and optimized in terms of operations and outcomes.

Limitations

We acknowledge several limitations of this narrative review. First, our search methods were limited by pragmatic constraints and excluded studies not in the English language, not cited in PubMed or Embase, and not referenced in the included studies or leading systematic reviews of outpatient PE management. Nevertheless, it is unlikely that our principal finding—that there is little research on comprehensive primary care-based PE—will be overturned by a more thorough search. Second, we did not address the management of acute PE in pregnancy, as it requires special considerations with diagnosis and treatment.⁹ Third, the lack of research on primary care-based PE management precluded a more formal systematic review and left us to draw inferences about the requirements of primary care-based management from outpatient care in other settings, particularly hospital-based ED and specialty clinics. Pulmonary embolism research in these 2 settings may not be immediately translatable to the primary care clinic setting, given differences in case mix, disease prevalence, physician training and experience, and access to testing resources. Future studies emerging directly from the primary care setting will help fill the many gaps currently in the literature. Last, our limited experiences prevent us from speaking to the breadth of diversity encompassed under the banner of primary care, although we have published broadly on PE diagnostics and treatment and represent 3 specialties—primary care, internal medicine, and emergency medicine—and different practice settings in 4 countries. We look to other authors to supplement this initial foray into a what is sure to be a broad subject of investigation.

CONCLUSION

To the larger research question, “Can primary care do this?,” that is, provide comprehensive outpatient management for low-risk patients with acute PE, we have 3 answers, which address the topic from skill-based, logistical, and evidence-based perspectives. The first answer arises from a general knowledge about the training, skills, and experience characteristic of primary care clinicians. (Two of the authors of this review are board-certified primary care physicians, in the US and the Netherlands, respectively.) General practitioners are skilled in risk stratification, frequently sorting out which patients with headache need cranial imaging, which patients with epigastric pain would benefit from laboratory testing, which patients with pneumonia can safely forgo hospitalization, and so on. With a little guidance, these clinicians could become just as adept at identifying which stable patients with acute PE may be eligible to bypass the hospital, and even forgo ED transfer. We anticipate that trained general practitioners, with direction from specialty guidelines, treatment pathways, or clinical decision-support systems, and ready access to on-call thrombosis specialists, can be capable of providing comprehensive outpatient PE management. Our first answer, then, is yes, absolutely; primary care physicians have the risk-stratification capabilities and informational resources to manage select low-risk patients with acute PE without needing to routinely transfer care.

The physician’s knowledge base and diagnostic skills, however, are not the only variables in the equation, as there are several logistical considerations that must be addressed. For example, how accessible are the necessary laboratory and radiology services? Is advanced pulmonary imaging located nearby, and are timely appointments and radiology interpretations available? Are clinical staff available to assist with patient education? Does the physician have the extra time to coordinate this complex operation, time that is sure to exceed a routine appointment duration? Some care delivery systems may be more conducive to comprehensive outpatient PE management than others. Even if the primary care physician can provide comprehensive management of select low-risk patients with acute PE (answer 1), they cannot provide such care if their practice location, setting, staffing, or operational constraints do not accommodate the requirements of this new model of PE care delivery (answer 2). Primary care physicians who believe that their practice is already overburdened may not welcome a resource-intense expansion of responsibilities. The additional burden of PE care may be attenuated by designing clinical care pathways that lighten the cognitive load on physicians, share responsibilities, and streamline patient flow.

Our third approach to our research question is not as amenable to an answer as the first 2, for it looks to the literature for primary care specific evidence-based guidance. As we found in this narrative review, little has been published that describes and analyzes the practice of primary care-based comprehensive PE management. There is much we do not understand. What characteristics of primary care clinicians are associated with outpatient care? How are primary care clinicians selecting their patients for outpatient care? In what patients is screening for right ventricular dysfunction necessary? Should routine assessment of right ventricular dysfunction be required of the primary care risk stratification protocol? What are the risk profiles, treatment, and outcomes of patients managed exclusively in the primary care setting? Is the practice safe? Is it efficient? How can it be improved? What is its impact on the patient care experience and the clinician’s experience? On a comparative level, do the selection criteria need to be more conservative than those used in the ED or specialty clinic? Are the outcomes similar to those of patients sent home from the ED or specialty clinic? There is a sizable gap in the literature that needs to be filled if we hope to understand this yet unexplored facet of outpatient PE management. Until then, our third answer to the question, “Can primary care do this?” must be that we do not know for certain yet. We look forward to what we will learn as this field of research expands.