Venous Thromboembolism (VTE), including pulmonary embolism (PE) and deep vein thrombosis (DVT), is one of the most severe and common cardiovascular pathologies worldwide, with up to 23% of patients dying within 30 days of diagnosis. Timely identification and treatment of the condition are critically important, but diagnosis is complicated in primary care due to non-specific symptoms. Until now, there were no standardized methods in outpatient settings for the effective monitoring of diagnostic delays.
DOVE eCQM
A multidisciplinary group from Mass General Brigham (MGB) and Penn State Health (PSH) created the DOVE eCQM (Diagnostic Delay in Venous Thromboembolism electronic Clinical Quality Measure). This innovative tool uses data retrieved from Electronic Health Records (EHR) to comprehensively identify all delays existing in VTE diagnostics.
The DOVE eCQM works in two main stages:
Structured Data Analysis: A specialized EHR algorithm processes coded data (e.g., ICD disease codes, CPT scanning orders, and medication records, specifically anticoagulants) to accurately identify VTE cases.
Unstructured Data Analysis (NLP): This specialized program examines the ordinary, textual notes of primary care visits (which are not marked with codes). The program searches for 29 symptoms characteristic of VTE (e.g., leg pain or swelling) that are documented by the doctor in free text.
By combining these two methods (structured codes + free text), the DOVE eCQM can determine the “index visit”—the first time a patient consulted a doctor with VTE symptoms. The system then measures how much time elapsed from this consultation to the official diagnosis. A diagnosis is considered delayed if it was made within a period of 24 hours to 30 days after the index visit. Since the tool also takes into account un-coded text, it is more accurate than traditional systems, which relied only on diagnosis codes.
Alarming Clinical Outcomes
The DOVE eCQM was tested on two completely different clinical systems (Epic at MGB and Oracle/Cerner at PSH—different EHR programs), which confirmed its possibility for widespread use.
The results are alarming: the rates of diagnostic delay were high—79.4% at MGB and 82.4% at PSH (if a delay is counted as more than 24 hours). The main reason for the delay was found to be the incorrect interpretation of symptoms by the practicing physician, specifically attributing VTE signs to other chronic diseases.
Delayed diagnosis directly increases mortality: According to MGB data, in cases of diagnostic delay, patients had a 3-fold increased risk of lethality within 30 days. The most dangerous form of VTE, pulmonary embolism (PE), was frequently recorded in deceased patients.
Despite the study’s limitations (such as dependence on EHR data and the exclusion of asymptomatic cases), this innovative instrument represents a significant achievement in ensuring patient safety in outpatient settings. Future plans include its use in emergency and urgent care departments.
Source: JAMA
