Balancing risks and priorities: Achieving right care in diagnosing pulmonary embolism during pregnancy

Pulmonary embolism (PE) is arguably one of the most diagnostically challenging conditions in emergency medicine, particularly in pregnant women. Pregnancy, with its hormonal changes and the venous stasis caused by the gravid uterus, is a significant risk factor for venous thromboembolic disease and a leading cause of maternal mortality in industrialized countries.1 The challenge arises from the nonspecific nature of PE symptoms, which often overlap with normal physiological changes of pregnancy, such as shortness of breath, palpitations, lower limb pain or swelling, and syncope. These symptoms frequently raise suspicion for PE, yet pregnancy—despite being a recognized thromboembolic risk factor—is associated with a low prevalence of confirmed PE among suspected cases in emergency settings.2 Emergency physicians thus face a critical balance between underdiagnosing a potentially fatal condition and overusing diagnostic procedures that may expose both mother and fetus to unnecessary harm. The difficulty of this balance has led to diagnostic inconsistencies. A 2003 study involving 117 emergency departments (EDs) identified pregnancy as an independent risk factor for noncompliance with diagnostic guidelines, with inappropriate exclusions correlating with significantly higher risks of thromboembolic events or sudden death during the 3-month follow-up.3 While specific diagnostic strategies have been validated and scientific recommendations developed for pregnant women,4, 5 the APED study by Vinson et al.,6 published in this issue of AEM, reveals that diagnostic practices remain suboptimal even two decades later. This underscores the importance and complexity of delivering “right care” in emergency settings. “Right care” aims to provide necessary high-quality care that addresses the underuse, overuse, and misuse of medical interventions.7 The APED study revealed a concerning degree of underuse and overuse in diagnostic testing for suspected PE. Underuse was evident as one in five pregnant patients declined recommended diagnostic testing, exposing themselves to the risk of missed diagnoses. While serious adverse outcomes were rare, this result should not be considered reassuring, given the low prevalence of PE in the study (1.6%). Conversely, overuse was seen in nearly one-third of patients who underwent chest imaging, mainly computed tomography pulmonary angiography (CTPA), without prior D-dimer testing or consideration of results. In many cases, D-dimer testing alone could have excluded PE, sparing patients from unnecessary radiation exposure.4, 5 Overtesting with CTPA also increases the risk of overdiagnosis and overtreatment, particularly given the higher likelihood of false-positive results when pretest probability is low or imaging is suboptimal—scenarios often encountered in pregnant women.8, 9 Beyond these observations, the study explores the complexities of the physician–patient relationship and the factors influencing patient refusal of proposed tests. “Right care” embodies delivering the appropriate intervention to the right person at the right time.10 For emergency physicians, providing the right care entails adhering to peer-developed, evidence-based guidelines, such as integrating D-dimer testing and ultrasonography into diagnostic protocols while accounting for radiation risks associated with imaging. Tailoring care to the right person requires consideration of patient-specific factors, including the clinical probability of PE and gestational age in discussions of risk. Importantly, it also means respecting the patient's preferences and values through honest communication and shared decision making. The notion of right time introduces a critical reflection on the role of time in emergency medicine. In emergency settings, time is not merely a parameter but a defining feature of care. Physicians face constant pressure to act swiftly, as delays in diagnosing and treating conditions like PE can have severe consequences. Rapid diagnostic testing facilitates earlier therapeutic decisions, shortens ED stays, and optimizes resource use. However, this need for speed may conflict with the patient's perspective. Many patients need time to process information, weigh risks and benefits, and arrive at informed decisions. The APED study highlights this disparity: patients who initially refused diagnostic imaging later consented during follow-up, likely due to additional time to reflect and clearer communication with health care providers. These findings underscore the ethical importance of allowing patients sufficient time to deliberate, even in the high-pressure environment of emergency care. The timing of care delivery also interacts with the trust dynamics between patients and providers. The APED study revealed higher acceptance of diagnostic tests in obstetric settings compared to EDs. This likely reflects the preexisting relationships patients often have with obstetric providers, built on continuity of care and mutual understanding. In contrast, EDs can feel impersonal, with interactions often limited to brief, high-pressure exchanges. Strengthening communication and trust, even in time-constrained settings, is therefore crucial. Addressing these challenges requires anticipatory strategies. Physicians must be equipped with up-to-date knowledge to provide appropriate care. Although PE is a frequent concern in emergency medicine, its occurrence in pregnancy is less common, requiring specific expertise. Local protocols and decision aids, such as smartphone applications or AI-driven tools integrated into patient management systems, can support adherence to evidence-based practices.11 Communication is equally essential. Transparent, empathetic discussions about diagnostic uncertainties, risks, and evidence-based approaches promote shared decision making and patient engagement.12 Patient information tools and including others, such as family members, primary care providers, or obstetricians, can further build trust and facilitate consensus.12 Communication should not be seen as a discrete task but as a continuous process that begins with explaining initial diagnostic hypotheses and evolves with the care plan. Training in effective communication should therefore be a core element of emergency medicine education. The APED study also brings the notion of error in emergency medicine to the forefront. Errors in this context extend beyond diagnostic failures or delays in treatment. They include overuse of diagnostic tests, communication breakdowns, and failures to involve patients in shared decision making. Such errors often arise from a misalignment between the physician's urgency to act and the patient's need for reflection. They also reflect systemic inefficiencies, such as overreliance on imaging without proper risk stratification. To minimize these errors, emergency care must move beyond rigid, protocol-driven models to embrace a patient-centered approach that integrates clinical guidelines with individual circumstances and values. In conclusion, the APED study illustrates the challenges of achieving “right care” in emergency medicine. It highlights how underuse and overuse of diagnostic resources stem from the inherent tensions of emergency care—balancing immediacy with patient-centered decision making. It also underscores the broader implications of error, which extend beyond technical or diagnostic failures to encompass failures in communication, empathy, and patient engagement. By redefining errors in this broader context, emergency medicine can embrace a paradigm shift: from viewing care delivery as a series of isolated, time-sensitive decisions to understanding it as a dynamic process that integrates evidence, patient values, and clinical expertise. Such an approach not only improves outcomes but also fosters trust and collaboration between practitioners and the communities they serve. Pierre-Marie Roy drafted the manuscript and Thomas Moumneh performed critical revision for important intellectual content. The authors declare no conflicts of interest. We used ChatGPT to review and correct this manuscript as English is not our native language. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.