Outcomes of pulmonary embolism in patients with psychiatric disorders: a nationwide analysis

Abstract

Background

Pulmonary embolism (PE) is a life-threatening condition with significant morbidity and mortality. The relationship between psychiatric disorders and PE outcomes is complex and not well understood. This study aimed to determine the impact of psychiatric disorders on PE outcomes by comparing patients with and without these conditions.

Methods

Using the National Inpatient Sample database, we analysed 725 725 adult patients hospitalised with PE between 2016 and 2019. Patients were stratified based on the presence or absence of psychiatric disorders. Multivariable logistic regression models were used to examine associations between psychiatric disorders and in-hospital outcomes, adjusting for baseline differences.

Results

Of the patients studied, 26.6% had psychiatric disorders. These patients were younger (59.80 vs 63.91 years, p<0.001), more likely to be female (60.7% vs 48.8%, p<0.001), and had higher rates of smoking and obesity but lower rates of diabetes and atrial fibrillation. After adjustment, patients with psychiatric disorders were less likely to undergo systemic thrombolysis (adjusted OR (aOR) 0.83, 95% CI: 0.80 to 0.85, p<0.001), catheter-directed interventions and placement of inferior vena cava filters (aOR 0.90, 95% CI: 0.88 to 0.92, p<0.001). These patients had lower odds of all-cause mortality (aOR 0.81, 95% CI: 0.78 to 0.84, p<0.001) and major adverse cardiovascular and cerebrovascular events (aOR 0.87, 95% CI: 0.85 to 0.89, p<0.001), with no significant difference in major bleeding risk (aOR 1.01, 95% CI: 0.98 to 1.05, p=0.553).

Conclusions

Psychiatric disorders are associated with distinct management and outcomes in PE. Recognising these unique characteristics may help optimise care for this population; further research is needed to clarify the best management strategies.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Used standardised International Classification of Diseases, 10th Revision, Clinical Modification codes for consistent identification of pulmonary embolism and psychiatric disorders.

• Employed multivariable logistic regression to rigorously adjust for demographic and clinical confounders.

• Relied on retrospective administrative data, which may be subject to coding inaccuracies and lacks clinical detail.

• Classification of psychiatric disorders was based on database diagnostic categories without independent verification, introducing potential misclassification.

• Analysed a large, nationally representative cohort, enhancing the generalisability of findings.

Introduction

Pulmonary embolism (PE) is a life-threatening condition with high morbidity and mortality. It is a major cause of cardiovascular death and has long-term physical and mental health complications. In the USA, the annual incidence of PE is 39–115 cases per 100 000 population¹ and PE leads to approximately 60 000–100 000 deaths per year.² The mortality rates for PE can vary depending on the severity. Short-term mortality rates are especially high for massive PE, with an in-hospital mortality ranging from 20.9% to 37% and 30-day mortality from 22.3% to 39.6%.^{1 3}

The relationship between psychiatric disorders and PE is complex and bidirectional. Patients hospitalised for psychiatric conditions are more likely to develop PE, especially when taking antipsychotics and benzodiazepines, which are associated with increased incidence rates.⁴ Furthermore, a study by Lin et al⁵ found that patients with concurrent depressive, bipolar and schizophrenic disorders have a significantly higher risk of venous thromboembolism, which includes PE. Additionally, psychiatric disorders may develop or worsen after a PE diagnosis. A study by Gkena et al⁶ showed that anxiety, depression and post-traumatic stress disorder can persist for up to 2 years in over 50% of PE survivors, significantly impacting their recovery and quality of life. In addition, patients with psychiatric disorders often face disparities in care for acute cardiovascular conditions. A study by Mohamed et al⁷ found that patients with severe mental illness who experienced acute myocardial infarction were significantly less likely to undergo invasive cardiac procedures like coronary angiography and percutaneous coronary intervention compared with patients without mental illness. This disparity in care may be due in part to concerns about obtaining informed consent from patients with psychiatric conditions.

However, there is an absence of data on the impact of psychiatric conditions on the management and outcome of patients with PE. Therefore, this study aimed to compare the clinical course and outcomes among patients with and without concomitant psychiatric diagnoses who were hospitalised with PE from a large national dataset.

Methods

Data source

The National Inpatient Sample (NIS), which has been available since 1988, is widely regarded as one of the publicly accessible inpatient healthcare databases in the USA. It gathers information from 7 million hospital stays annually through a 20% stratified sample of discharges, from US community hospitals (excluding rehabilitation and long-term acute care hospitals). The NIS, a component of the Healthcare Cost and Utilisation Project, aims to generate estimates for healthcare usage and outcomes at both regional and national levels, in the USA.⁸

Study design and population

In this retrospective study, we conducted a comprehensive analysis of adult patients (aged ≥18 years) hospitalised with a primary diagnosis of PE between 2016 and 2019. Patients were grouped based on the presence or absence of a psychiatric disorder. These patients were chosen based on the International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) diagnosis codes that were implemented in 2016 which provided more granular data as opposed to the previous ICD-9 coding. Demographic details of patients were noted for every hospital discharge, such as age group, gender, ethnicity, day of admission (weekdays or weekends), expected payer and median household income based on ZIP code. Missing data on age, gender, elective, admission type and day, and mortality status were excluded from the analysis (Figure 1 for study flow diagram). Each discharge record contained data on up to 30 diagnoses.

Figure 1. Flow diagram representing the selection of study population from the National Inpatient Sample (NIS) database between 2016 and 2019. ICD, International Classification of Diseases.

Psychiatric disorders were combined into a single variable for this study. This category includes alcohol use disorder, substance use disorder, depressive disorders, bipolar and related disorders, anxiety and fear-related disorders, schizophrenia spectrum and other psychotic disorders. These conditions were selected based on their demonstrated impact on clinical outcomes, particularly in influencing mortality and morbidity rates in patients with complex medical conditions.^{5 9 10} In this study, a ‘high-risk PE’ was defined as PE with cardiogenic shock, mechanical ventilation, mechanical circulatory support or vasopressors.11,13 A full list of ICD-10-CM codes used to identify PE and psychiatric disorders is provided in online supplemental table S1. ICD-10-CM codes were also used to classify procedural information during hospitalisation including systemic thrombolysis, catheter-directed thrombolysis, ultrasound-facilitated catheter-directed thrombolysis, catheter-directed embolectomy, surgical embolectomy/thrombectomy and inferior vena cava (IVC) filter placement.

Patient and public involvement

Patients or the public were not involved in the design, conduct, reporting or dissemination plans of this research.

Outcomes

The primary outcome of interest was the difference in all-cause in-hospital mortality between patients with psychiatric disorder and those without. The study also evaluated secondary outcomes, including in-hospital adverse events such as major adverse cardiovascular and cerebrovascular events (MACCE), all-cause mortality, major bleeding, intracranial haemorrhage (ICH), non-ICH bleeding events, length of stay and cost. MACCE was characterised as a composite of all-cause mortality, acute ischaemic cerebrovascular accident (CVA) or transient ischaemic attack and cardiac complications. Major bleeding was defined as a composite of gastrointestinal, retroperitoneal, intracranial and intracerebral haemorrhage, periprocedural haemorrhage, unspecified haemorrhage or needing blood transfusion. Additionally, the participants’ receipt of invasive management procedures such as systemic thrombolysis, catheter-directed thrombolysis, ultrasound-facilitated catheter-directed thrombolysis, catheter-directed embolectomy, surgical embolectomy/thrombectomy and IVC filter placement was also measured.

Statistical analysis

Statistical analysis was performed on IBM SPSS V.25. Continuous variables were presented as mean, median and IQR, due to skewed data, and categorical data were presented as frequencies and percentages. Categorical variables were compared using Pearson’s χ² test, while continuous variables were compared using the Mann-Whitney U test, as appropriate. Sampling weights were used to calculate the estimated total discharges as specified by AHRQ. Multivariable logistic regression models were used to examine the association between in-hospital outcomes and the presence or absence of psychiatric disorders, expressed as ORs with corresponding 95% CIs. All multivariable models for both primary and secondary outcomes were adjusted for baseline differences between the groups, controlling for the following covariates: age, gender, weekend admission, hospital bed size, region and location/teaching status, cardiogenic shock, ventricular fibrillation, ventricular tachycardia, atrial fibrillation (AF), heart failure (HF), hypertension, valvular heart disease, dyslipidaemia, chronic liver disease, chronic lung disease, chronic kidney disease, anaemia, thrombocytopenia, coagulopathies, diabetes mellitus, malignancies, systemic thrombolysis, catheter-directed thrombolysis, ultrasound-facilitated catheter-directed thrombolysis, catheter-directed embolectomy, surgical embolectomy/thrombectomy and IVC filter placement.

Results

In the study, records of 725 725 patients with a primary diagnosis of acute PE were identified. Among the 192 780 patients with psychiatric disorders, the most common diagnoses were anxiety and fear-related disorders (39%) and depressive disorders (36%), followed by substance use disorder (11%), bipolar and related disorders (6%), alcohol use disorder (4%), and schizophrenia spectrum and other psychotic disorders (4%) (figure 2). Baseline demographic and clinical characteristics of patients, stratified by psychiatric disorder status, are shown in table 1. Baseline demographic and clinical characteristics of patients, stratified by psychiatric disorder status, are shown in table 1. Patients with psychiatric disorders were less likely to experience high-risk PE (4.2% vs 5.2%, p<0.001), cardiac arrest (1.2% vs 1.8%, p<0.001), cardiogenic shock (1.1% vs 1.5%, p<0.001), saddle PE (7.4% vs 9.2%, p<0.001) and acute cor pulmonale (7.0% vs 8.0%, p<0.001). These patients were younger (59.80 vs 63.91 years, p<0.001) and included a higher percentage of females (60.7% vs 48.8%, p<0.001) compared with those without psychiatric disorders.

Figure 2. Distribution of psychiatric disorders among patients with pulmonary embolism.

Table 1. Baseline demographic and clinical characteristics of patients with pulmonary embolism, stratified by psychiatric disorders.

	Without psychiatric disorder	With psychiatric disorder	P value
NIS discharge weight	532 945	192 780	<0.001
Mean age	63.91	59.80	<0.001
Female, %	48.8	60.7	<0.001
Ethnicity			<0.001
White	69.8	76.2
Black	20.3	16.0
Hispanic	6.2	5.0
Asian	1.2	0.6
Native	0.4	0.4
Other	2.2	1.8
Hospital region			<0.001
Northeast	18.3	18.8
Midwest or North Central	24.2	27.0
South	39.5	36.4
West	18.1	17.7
Hospital bed size			<0.001
Small	20.9	20.6
Medium	29.9	29.3
Large	49.2	50.1
Hospital location/teaching status			<0.001
Rural	9.0	8.9
Urban non-teaching	23.1	21.8
Teaching	67.9	69.3
Median ZIP income			<0.001
First quartile	28.0	29.2
Second quartile	26.3	26.7
Third quartile	24.9	25.1
Fourth quartile	20.8	19.0
Primary expected payer			<0.001
Medicare	52.5	50.4
Medicaid	10.1	17.9
Private insurance	30.3	25.3
Self-pay	4.1	3.7
No charge	0.3	0.3
Other	2.7	2.3
Record characteristics
High risk pulmonary embolism	5.2	4.2	<0.001
Cardiac arrest	1.8	1.2	<0.001
Ventricular fibrillation	0.2	0.1	<0.001
Ventricular tachycardia	1.4	1.2	<0.001
Cardiogenic shock	1.5	1.1	<0.001
Saddle PE	9.2	7.4	<0.001
Acute cor pulmonale	8.0	7.0	<0.001
Comorbidities
Heart failure	16.4	17.0	<0.001
Valvular heart disease	6.3	6.3	0.214
Hypertension	61.9	62.2	0.081
Diabetes	23.6	22.6	<0.001
Dyslipidaemia	35.5	37.4	<0.001
Atrial fibrillation	11.4	9.4	<0.001
Smoking	37.1	47.5	<0.001
Dementia	5.6	6.6	<0.001
Chronic kidney disease	13.3	11.5	<0.001
Chronic lung disease	23.2	32.6	<0.001
Chronic liver disease	0.6	0.8	<0.001
Obesity	24.5	28.3	<0.001
Anaemia	22.8	24.9	<0.001
Thrombocytopenia	5.8	5.2	<0.001
Coagulopathy	6.1	6.9	<0.001
Peripheral vascular disease	3.4	3.5	0.001
Cerebrovascular disease	2.7	3.1	<0.001
Ischaemic heart disease	17.6	17.2	<0.001
Homelessness	0.4	1.6	<0.001
Haematologic malignancy	2.4	2.0	<0.001
Solid malignancy	12.4	11.0	<0.001
Metastatic malignancy	7.8	7.3	<0.001

NIS, National Inpatient Sample; PE, pulmonary embolism.

Patients with psychiatric disorders exhibited a distinct comorbidity profile compared with those without psychiatric disorders. They had significantly higher rates of chronic lung disease (32.6% vs 23.2%, p<0.001), obesity (28.3% vs 24.5%, p<0.001) and smoking (47.5% vs 37.1%, p<0.001). Interestingly, these patients also showed higher prevalence of dyslipidaemia (37.4% vs 35.5%, p<0.001) and anaemia (24.9% vs 22.8%, p<0.001). However, they had lower rates of diabetes mellitus (22.6% vs 23.6%, p<0.001), AF (9.4% vs 11.4%, p<0.001) and chronic kidney disease (11.5% vs 13.3%, p<0.001). The prevalence of HF was similar between the two groups (17.0% vs 16.4%, p<0.001), as was hypertension (62.2% vs 61.9%, p=0.081), although the latter difference was not statistically significant. For a comprehensive list of comorbidities and their prevalence, please refer to table 1.

In-hospital procedures and outcomes

Crude rates

Table 2 shows the analysis of in-hospital procedures, complications and outcomes for PE patients between those with and without psychiatric disorders. Patients without psychiatric disorders were more likely to receive invasive treatments such as systemic thrombolysis (3.3% vs 2.7%, p<0.001), catheter-directed thrombolysis (3.8% vs 3.0%, p<0.001), ultrasound-facilitated catheter-directed thrombolysis (1.0% vs 0.8%, p<0.001), catheter-directed embolectomy (1.0% vs 0.7%, p<0.001) and placement of IVC filters (6.2% vs 5.2%, p<0.001).

Table 2. In-hospital procedures, complications and outcomes of patients with pulmonary embolism patients, stratified by psychiatric disorders.

	Without psychiatric disorder	With psychiatric disorder	P value
NIS discharge weight	532 945	192 780	<0.001
High-risk pulmonary embolism	5.2	4.2	<0.001
Management, %
Systemic thrombolysis	3.3	2.7	<0.001
Catheter-directed thrombolysis	3.8	3.0	<0.001
Ultrasound-facilitated catheter-directed thrombolysis	1.0	0.8	<0.001
Catheter-directed embolectomy	1.0	0.7	<0.001
Surgical embolectomy/thrombectomy	0.2	0.2	0.020
IVC filter	6.2	5.2	<0.001
Circulatory and ventilatory support
Vasopressors	0.8	0.6	<0.001
Mechanical ventilation	3.5	2.6	<0.001
ECMO	0.2	0.1	<0.001
Clinical outcomes, %
All-cause mortality	3.4	2.3	<0.001
MACCE	4.8	3.8	<0.001
Major bleeding	2.5	2.3	<0.001
ICH	0.5	0.4	<0.001
Non-ICH
Retroperitoneal	0.2	0.2	0.471
Gastrointestinal	1.8	1.7	0.004
Procedure related	0.1	0.1	0.191
Length of stay, days, mean	4.23	4.61	<0.001
Total charge, US$, mean	48 406.81	48 517.37	0.554

ECMO, Extracorporeal Membrane Oxygenation; ICH, intracranial haemorrhage; IVC, inferior vena cava; MACCE, major adverse cardiovascular and cerebrovascular events; NIS, National Inpatient Sample.

Regarding circulatory and ventilatory support, vasopressor use was slightly higher in patients without psychiatric disorders (0.8% vs 0.6%, p<0.001), as was mechanical ventilation (3.5% vs 2.6%, p<0.001) and Extracorporeal Membrane Oxygenation (ECMO) (0.2% vs 0.1%, p<0.001).

In terms of clinical outcomes, all-cause mortality was higher among patients without psychiatric disorders (3.4% vs 2.3%, p<0.001), as was MACCE (4.8% vs 3.8%, p<0.001) and major bleeding incidents (2.5% vs 2.3%, p<0.001). ICH occurred more frequently in patients without psychiatric disorders (0.5% vs 0.4%, p<0.001).

The mean length of hospital stay was slightly longer for patients with psychiatric disorders (4.61 days vs 4.23 days, p<0.001), while the mean total charge for hospitalisation was marginally higher for patients with psychiatric disorders (US$48 517.37 vs US$48 406.81, p=0.554), although this difference was not statistically significant.

In figure 3, patients with psychiatric disorders were less likely to have been discharged home (62.6% vs 66.3%, p<0.001) and more likely to have been discharged to intermediate care facilities (16.1% vs 13.0%, p<0.001) and to have received home health services (15.6% vs 14.6%, p<0.001).

Figure 3. Disposition of patients with pulmonary embolism, stratified by psychiatric disorder status.

Adjusted analysis

Following adjustments, the multivariate analysis in table 3 revealed several significant findings. Patients with psychiatric disorders were less likely to undergo systemic thrombolysis (adjusted OR, aOR 0.83, 95% CI: 0.80 to 0.85, p<0.001), catheter-directed thrombolysis (aOR 0.80, 95% CI: 0.77 to 0.82, p<0.001), ultrasound-facilitated catheter-directed thrombolysis (aOR 0.78, 95% CI: 0.74 to 0.83, p<0.001), catheter-directed embolectomy (aOR 0.82, 95% CI: 0.77 to 0.87, p<0.001) and placement of IVC filter (aOR 0.90, 95% CI: 0.88 to 0.92, p<0.001). The odds of undergoing surgical embolectomy/thrombectomy were not significantly different between the two groups (aOR 0.90, 95% CI: 0.79 to 1.02, p=0.897).

Table 3. Multivariate analysis showing adjusted OR for in-hospital procedures and complications of patients with pulmonary embolism with psychiatric disorders.

Outcome	aOR (95% CI)	P value
In-hospital procedures
Systemic thrombolysis	0.83 (0.80 to 0.85)	<0.001
Catheter-directed thrombolysis	0.80 (0.77 to 0.82)	<0.001
Ultrasound-facilitated catheter-directed thrombolysis	0.78 (0.74 to 0.83)	<0.001
Catheter-directed embolectomy	0.82 (0.77 to 0.87)	<0.001
Surgical embolectomy/thrombectomy	0.90 (0.79 to 1.02)	0.897
IVC filter	0.90 (0.88 to 0.92)	<0.001
In-hospital complications
MACCE	0.87 (0.85 to 0.89)	<0.001
Mortality	0.81 (0.78 to 0.84)	<0.001
Major bleeding	1.01 (0.98 to 1.05)	0.553
ICH	0.93 (0.86 to 1.02)	0.120

Reference: Without psychiatric disorder; adjusted for age, gender, weekend admission, hospital bed size, region and location/teaching status, cardiogenic shock, VF, VT, AF, HF, hypertension, valvular heart disease, dyslipidaemia, chronic liver disease, chronic lung disease, chronic kidney disease, anaemia, thrombocytopaenia, coagulopathies, diabetes mellitus, malignancies, systemic thrombolysis, catheter-directed therapy, surgical embolectomy/thrombectomy and IVC filter.

AF, atrial fibrillation; aOR, adjusted OR; HF, heart failure; ICH, intracranial haemorrhage; IVC, inferior vena cava; MACCE, major adverse cardiovascular and cerebrovascular events; VF, ventricular fibrillation; VT, ventricular tachycardia.

In terms of in-hospital complications, the analysis showed patients with psychiatric disorders were less likely to have experienced MACCE (aOR 0.87, 95% CI: 0.85 to 0.89, p<0.001) and mortality (aOR 0.81, 95% CI: 0.78 to 0.84, p<0.001). However, there was no significant difference in the odds of major bleeding (aOR 1.01, 95% CI: 0.98 to 1.05, p=0.553) or ICH (aOR 0.93, 95% CI: 0.86 to 1.02, p=0.120) between patients with and without psychiatric disorders.

Sensitivity analysis

Given the heterogeneity of psychiatric disorders included in our analysis, we conducted a sensitivity analysis focusing specifically on high-risk PE cases. The results of this analysis are presented in online supplemental table S2. Briefly, patients with psychiatric disorders were less likely to receive systemic thrombolysis (19.9% vs 22.8%, p<0.001). There were no significant differences in the use of catheter-directed thrombolysis (8.5% vs 8.7%, p=0.713) or IVC filter placement (15.6% vs 16.2%, p=0.225). However, patients with psychiatric disorders were more likely to undergo surgical embolectomy/thrombectomy (3.0% vs 2.5%, p=0.006). Patients with psychiatric disorders were also less likely to receive vasopressors (14.7% vs 15.9%, p=0.006), mechanical ventilation (62.8% vs 66.6%, p<0.001) and ECMO (2.2% vs 3.3%, p<0.001). Notably, patients with psychiatric disorders had significantly lower rates of adverse clinical outcomes, including all-cause mortality (34.2% vs 45.2%, p<0.001), MACCE (37.4% vs 48.4%, p<0.001), major bleeding (10.2% vs 12.7%, p<0.001), ICH (1.9% vs 3.0%, p<0.001) and gastrointestinal bleeding (6.2% vs 7.6%, p<0.001). There was no significant difference in retroperitoneal bleeding (1.8% vs 1.7%, p=0.773) or procedure-related bleeding (1.0% vs 1.0%, p=0.951) between the two groups.

Discussion

The study examined the impact of psychiatric disorders on the management and outcomes in patients with PE, analysing data from 725 725 patients, of whom 192 780 (26.6%) had a psychiatric disorder. The analysis revealed several key insights: First, among PE patients, those with psychiatric disorders were generally younger and included a higher proportion of females compared with those without such conditions. Second, these patients exhibited a distinct profile of comorbidities. They had higher rates of chronic lung disease, obesity, smoking, dyslipidaemia and anaemia. Third, patients with psychiatric disorders experienced slightly longer hospital stays, and their total hospital charges were marginally higher. Fourth, after adjusting for baseline demographics and comorbidities, patients with psychiatric disorders were less likely to undergo invasive procedures such as systemic thrombolysis, catheter-directed interventions and IVC filter placement. Finally, patients with psychiatric disorders had lower adjusted odds of all-cause in-hospital mortality and MACCE. There was no significant difference in the adjusted odds of major bleeding or ICH between patients with and without psychiatric disorders.

Our study found that patients with psychiatric disorders who experienced PE were generally younger and included a higher proportion of females compared with PE patients without psychiatric conditions. This contrasts with the findings reported by Velasco et al,¹⁰ who observed that patients with PE and coexisting psychiatric disorders tended to be older (mean age 72.7 vs 65.6 years) and had a higher prevalence of female gender (71% vs 48%) compared with those without psychiatric disorders. Similarly, Martin Ontiyuelo et al¹⁴ and Kong et al¹⁵ also reported that patients with PE and co-existing psychiatric disorders were older and predominantly female. The younger age in our study might reflect an earlier onset or diagnosis of psychiatric disorder in our population. Additionally, the discrepancy could be due to the differences in how psychiatric disorders were defined and categorised across studies. Furthermore, women may be more likely to seek medical care for both mental health issues and PE symptoms, leading to higher representation.

Our study found lower rates of AF and diabetes mellitus in the psychiatric disorder group. This aligns with a study by Lin et al,⁵ which reported lower rates of hypertension, AF and diabetes in patients with concurrent depressive, bipolar and schizophrenic disorders. However, these findings contrast with several studies which reported higher rates of diabetes mellitus and increased cardiovascular risk in patients with severe mental illness.^{16 17} The lower rates observed in our study may be due to reduced access to medical care for patients with psychiatric disorders, potentially leading to underdiagnosis or delayed diagnosis of these comorbidities.

After adjusting for baseline demographics and comorbidities, we reported that patients with psychiatric disorders were less likely to undergo invasive procedures for PE, which aligns with a previous study.¹⁰ Velasco et al demonstrated significantly lower use of reperfusion therapies (2.5% vs 3.3%, p=0.04) and less frequent insertion of IVC filters (2.0% vs 2.5%, p=0.17) in patients with psychiatric disorders compared with those without.¹⁰ Several factors may contribute to this disparity in treatment. First, patients with psychiatric disorders may face difficulties in accurately reporting their symptoms, which can lead to delayed or missed diagnoses of PE. Second, healthcare providers may have concerns about psychiatric patients' ability to comply with postprocedure care instructions or medication regimens. Third, patients with psychiatric disorders may be more hesitant to undergo invasive procedures due to anxiety, paranoia or other mental health symptoms. Importantly, our data also revealed that patients with psychiatric disorders were less likely to present with high-risk PE. Even among high-risk PE patients, those with psychiatric disorders received less aggressive management, including lower rates of systemic thrombolysis and IVC filter placement. This lower incidence of high-risk PE could partially explain the reduced use of invasive procedures, as these interventions are typically reserved for more severe cases.

Lastly, our study found that patients with psychiatric disorders had lower adjusted odds of all-cause in-hospital mortality compared with those without psychiatric disorders, while there was no significant difference in the adjusted odds of major bleeding between the two groups. In contrast to our findings, Velasco et al¹⁰ demonstrated that patients with psychiatric disorders had increased odds for all-cause mortality and PE-related mortality compared with those without psychiatric disorders. Furthermore, their study found that psychiatric disorders were not significantly associated with increased odds for major bleeds during follow-up.¹⁰

While this study provides valuable insights into the impact of psychiatric disorders on the outcomes of patients with PE, it is important to acknowledge several limitations. First, as a retrospective analysis of administrative data, it is subject to potential coding errors and lacks detailed clinical information that could provide more context to the findings. Our identification of substance use disorders was based on the diagnostic categories available in the database, without independent verification or reclassification of specific ICD-10 codes, which may introduce some misclassification. Second, the study does not differentiate between various types and severities of psychiatric disorders, which may have different impacts on PE outcomes. Third, the analysis does not account for potential confounding factors such as medication use, particularly psychotropic medications that may influence both psychiatric symptoms and thrombotic risk. Finally, our analysis only includes patients who were hospitalised with PE. Patients with psychiatric disorders are at higher risk of sudden death in the community,^{18 19} which may not be captured in our dataset. As a result, the lower in-hospital mortality and lower rates of invasive therapies observed in this group, including among high-risk PE patients, may partly reflect a selection bias, where only less severe cases are admitted, rather than a true difference in outcomes or management.

Several factors could explain why our findings differ from those of previous studies examining the impact of psychiatric disorders on PE outcomes. First, our use of a large, nationally representative inpatient database likely captured a broader and more diverse patient population, whereas prior research often relied on single-centre or regional data, which may not reflect the full spectrum of patients. Differences in how psychiatric disorders are defined and categorised across studies can also contribute to variability in results; our study grouped a wide range of psychiatric conditions based on available diagnostic codes, while others may have focused on specific disorders or used different criteria. Additionally, the extent and method of adjusting for confounding factors—such as comorbidities, socioeconomic status and medication use—vary between studies, potentially influencing observed associations. Changes in clinical practice, management strategies for PE and mental healthcare over time, as well as regional differences in healthcare delivery, may further account for discrepancies. Lastly, previous studies may have included patients with more severe psychiatric illness or those treated in specialised settings, while our study encompassed all hospitalised patients with psychiatric diagnoses, possibly capturing a wider range of illness severity.

These results have implications for clinical practice. The lower rates of certain comorbidities in patients with psychiatric disorders might indicate underdiagnosis. This suggests a need for more comprehensive screening protocols for cardiovascular risk factors in patients with psychiatric disorders. In addition, the lower rates of invasive procedures highlight the need for healthcare providers to be more vigilant in assessing and treating PE in patients with psychiatric disorders. This may involve developing tailored communication strategies to better understand these patients’ symptoms and concerns, as well as implementing protocols to ensure equitable access to necessary interventions.

Conclusions

In conclusion, our study revealed significant insights into the management and outcomes of PE in patients with psychiatric disorders. Patients with psychiatric disorders were generally younger, more likely to be female and had a distinct comorbidity profile. They were less likely to undergo invasive procedures for PE treatment, despite having lower in-hospital mortality rates. However, these patients faced a higher risk of major bleeding events. These findings underscore the importance of recognising the distinct characteristics and needs of patients with psychiatric disorders when managing PE. Further research is warranted to determine whether specific tailored approaches are beneficial in this population.

Data availability statement

Data may be obtained from a third party and are not publicly available.