Providers had to meet the following inclusion criteria: (1) employee at Baylor Scott & White Health (BSWH) at the time of the intervention; (2) working in the the ED of the Baylor Scott & White Temple hospital; (3) holding a position of resident, attending, or staff physician; and (4) providing informed consent.

To be eligible for inclusion in the electronic health record (EHR) analysis, aggregate de-identified patient-level data must have met the following criteria: (1) patient was evaluated at the ED, and (2) had a primary diagnosis of new-onset VTE as defined by International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes for acute pulmonary embolism (I26.0x, I26.9x), acute lower extremity thrombosis (I80.1x, I80.2x, I80.3x, I80.8x, I80.9x, I82.4x), acute upper extremity thrombosis (I82.60x, I82.62x, I82.A1x, I82.B1x), or acute embolism of other specified veins (I82.890, I82.90). These ICD-10-CM codes have been used in prior publications [9,10,11] and were reviewed by a BSWH clinician. Codes related to superficial thrombosis, chronic thrombosis, and complex diagnoses such as Budd-Chiari syndrome, inferior vena cava thrombus, renal vein thrombus, and intrathoracic thrombus were excluded.

This study was eligible for expedited review as it posed minimal risk to study participants, and it was approved by the BSWH institutional review board (IRB ID: 023-032). All participants provided informed consent to participate in the study. Consent documents and recruitment materials were approved by Pfizer and by the institutional review board/independent ethics committee before use. This study was conducted at the Baylor Scott & White Temple hospital. The evaluation was conducted in accordance with legal and regulatory requirements, as well as with scientific purpose, value, and rigor and follow generally accepted research practices described in Good Practices for Outcomes Research issued by the International Society for Pharmacoeconomics and Outcomes Research and in accordance with the Helsinki Declaration of 1964 and its later amendments.

The pilot program to discharge appropriate patients with acute VTE from the ED with outpatient treatment was initiated on 14 October 2021. For patients considered eligible for discharge, the pilot program consisted of a complimentary 30-day supply of either rivaroxaban or apixaban, a scheduled follow-up appointment within 2 weeks of discharge, and an informal educational program for ED providers. A pre-pilot period from 14 December 2020 to 14 October 2021 (10 months) and a post-pilot period from 15 October 2021 to 14 August 2022 (10 months) were used for the EHR analysis portion of the study (Fig. 1).

Study timeline for the pilot and quality intervention initiatives

Recognizing the favorable prospects of the pilot program, it was determined that the program would be expanded by formally constructing a comprehensive quality intervention. This intervention comprised a robust educational session for ED providers; this quality intervention took place on 16 January 2023. In addition, there were surveys of participating providers to evaluate the impact of the educational session, and a patient-centric infographic focusing on VTE care that took place on 26 January 2023. A pre-quality intervention period from 21 August 2022, to 26 January 2023 (5 months, 5 days) and a post–quality intervention period from 27 January 2023 to 1 July 2023 (5 months, 5 days) were used for the EHR analysis portion of the study (Fig. 1).

The quality intervention comprised of an approximate 10-min pre-provider survey of 56 questions, followed by a 30-min academic presentation, and then an approximately 10-min post-provider survey of 46 questions. The academic presentation reviewed the guidelines for anticoagulant therapy for VTE, a clinical decision framework to guide the outpatient treatment of ED patients diagnosed with acute VTE (DVT and PE), and the resources available. One of the available resources specifically developed for this initiative included a patient-centric infographic, which physicians could utilize when discharging patients from the ED (Electronic Supplementary Material (ESM] Fig. S1). The presentation was delivered by a board-certified hematologist/oncologist with extensive experience in providing care to patients with VTE and over 20 years of experience in educating medical students, physicians, and nursing staff.

Participant surveys were conducted before and after the provider quality intervention education session and comprised four sections: demographics, clinical knowledge, current practice/recall, and SDOH. Respondents’ questions on demographics included provider type, years of experience, and whether they had received the pilot education on 14 October 2021.

Our primary outcomes included providers’ knowledge of and confidence in evaluating factors that determine whether a patient with VTE in the ED is eligible for outpatient treatment. Clinical knowledge was assessed using 17 hypothetical patient scenarios. Each response was categorized as either correct or incorrect, with 1 point given for each correct answer and zero points for each incorrect answer. Respondents who selected “I don’t know” were considered to have answered incorrectly. A final score out of 17 points was calculated for each provider. Quality intervention survey participants rated their overall confidence in evaluating factors that determined the eligibility of a patient with VTE in the ED for outpatient treatment using a 5-point Likert scale (0 = ‘not confident’ to 4 = ‘completely confident’). In addition, participants rated their confidence in assessing clinical criteria (e.g., comorbidities, risk of bleeding, etc.) for patients with DVT and/or PE, selecting an appropriate DOAC, and addressing social health needs for patients with VTE using the same 5-point Likert scale. For questions regarding DVT specifically, the total score ranged from 0 to 20. Each respondent’s score was reported as a percentage of the total possible 20 points and grouped into categories: ≤ 20%, not confident; > 20% to ≤ 40%, slightly confident; > 40 to ≤ 60%, somewhat confident; > 60 and ≤ 80%, fairly confident; and > 80 to ≤ 100%, completely confident. For questions regarding PE specifically, the total score ranged from 0 to 16. Each respondent’s score was reported as a percentage of the total possible 16 points and grouped into the same categories as DVT. Choosing an appropriate DOAC and addressing social health needs for a patient with VTE were each assessed with a single question using a total score ranging from 0 to 4.

Secondary outcomes included assessing the impact of provider-reported barriers and patient SDOH characteristics on the decision to discharge a patient with VTE from the ED, rated on a 3-point Likert scale (0 = not a barrier/not difficult at all to 2 = significant barrier/very difficult). The section on provider-reported barriers included nine barriers, with possible total scores ranging from 0 to 18. The SDOH section included six factors, with possible total scores ranging from 0 to 12. Another secondary outcome was the proportion of patients with VTE discharged from the ED before and after the pilot and quality interventions. Patient admission and discharge information were obtained from the local EHR system, “Epic.”

Descriptive statistics including means with standard deviation (SD) or medians with first and third quartiles (Q1-Q3), were calculated for continuous variables, and frequencies and percentages were calculated for all categorical variables. The Wilcoxon signed-rank test was used to compare providers’ survey responses (i.e., clinical knowledge score, confidence score, and SDOH impact scores) pre-quality and post-quality intervention. A two-proportion z-test was used to compare the differences in the overall proportion of patients with VTE discharged from the Temple ED pre- and post-intervention (pilot and quality intervention). In addition, interrupted time series (ITS) analyses were performed to explore the temporal changes in the ED patient discharge rate taken at a regularly spaced interval (e.g., month) and, more specifically, the changes at the time when an intervention (e.g., the pilot and quality intervention) was implemented.

All statistical analyses were conducted using SAS Enterprise Guide 8.2 (SAS Institute, Cary, NC, USA), R statistical software (version 2023.06.2; R Foundation, Vienna, Austria), and Stata/BE (version 18.0; StataCorp, College Station, TX, USA), with statistical significance set at P < 0.05. No imputation for missing values was performed.