Validation of natural language processing to determine the presence and size of abdominal aortic aneurysms in a large integrated health system.

OBJECTIVE: Previous studies of the natural history of abdominal aortic aneurysms (AAAs) have been limited by small cohort sizes or heterogeneous analyses of pooled data. By quickly and efficiently extracting imaging data from the health records, natural language processing (NLP) has the potential to substantially improve how we study and care for patients with AAAs. The aim of the present study was to test the ability of an NLP tool to accurately identify the presence or absence of AAAs and detect the maximal abdominal aortic diameter in a large dataset of imaging study reports. METHODS: Relevant imaging study reports (n = 230,660) from 2003 to 2017 were obtained for 32,778 patients followed up in a prospective aneurysm surveillance registry within a large, diverse, integrated healthcare system. A commercially available NLP algorithm was used to assess the presence of AAAs, confirm the absence of AAAs, and extract the maximal diameter of the abdominal aorta, if stated. A blinded expert manual review of 18,000 randomly selected imaging reports was used as the reference standard. The positive predictive value (PPV or precision), sensitivity (recall), and the kappa statistics were calculated. RESULTS: Of the randomly selected 18,000 studies that underwent expert manual review, 48.7% were positive for AAAs. In confirming the presence of an AAA, the interrater reliability of the NLP compared with the expert review showed a kappa value of 0.84 (95% confidence interval [CI], 0.83-0.85), with a PPV of 95% and sensitivity of 88.5%. The NLP algorithm showed similar results for confirming the absence of an AAA, with a kappa of 0.79 (95% CI, 0.799-0.80), PPV of 77.7%, and sensitivity of 91.9%. The kappa, PPV, and sensitivity of the NLP for correctly identifying the maximal aortic diameter was 0.88 (95% CI, 0.87-0.89), 88.8%, and 88.2% respectively. CONCLUSIONS: The use of NLP software can accurately analyze large volumes of radiology report data to detect AAA disease and assemble a contemporary aortic diameter-based cohort of patients for longitudinal analysis to guide surveillance, medical management, and operative decision making. It can also potentially be used to identify from the electronic medical records pre- and postoperative AAA patients "lost to follow-up," leverage human resources engaged in the ongoing surveillance of patients with AAAs, and facilitate the construction and implementation of AAA screening programs.

Early and delayed rupture after endovascular abdominal aortic aneurysm repair in a 10-year multicenter registry.

OBJECTIVE: Rupture after abdominal endovascular aortic aneurysm repair (EVAR) is a function of graft maintenance of the seal and fixation. We describe our 10-year experience with rupture after EVAR. METHODS: From 2000 to 2010, 1736 patients with abdominal aortic aneurysm (AAA) from 17 medical centers underwent EVAR in a large, regional integrated health care system. Preoperative demographic and clinical data of interest were collected and stored in our registry. We retrospectively identified patients with postoperative rupture, characterized as "early" and "delayed" rupture (≤30 days and >30 days after the initial EVAR, respectively), and identified predictors associated with delayed rupture. RESULTS: The overall follow-up rate was 92%, and the median follow-up was 2.7 years (interquartile range, 1.2-4.4 years) in these 1736 EVAR patients. We identified 20 patients with ruptures; 70% were male, the mean age was 79 years, and mean AAA size at the initial EVAR was 6.3 cm. Six patients underwent initial EVAR for rupture (n = 2) or symptomatic presentation (n = 4). Of the 20 post-EVAR ruptures, 25% (five of 20) were early, all occurring within 2 days after the initial EVAR. Of these five patients, four had intraoperative adverse events leading directly to rupture, with one type I and one type III endoleak. Of the five early ruptures, four patients underwent endovascular repair and one received repair with open surgery, resulting in two perioperative deaths. Among the remaining 15 patients, the median time from initial EVAR to rupture was 31.1 months (interquartile range, 13.8-57.3 months). Most of these delayed ruptures (10 of 15) were preceded by AAA sac increases, including three patients with known endoleaks who underwent reintervention. At the time of delayed rupture, nine of 15 patients had new endoleaks. Among all 20 patients, six patients did not undergo repair (all delayed patients) and died, nine underwent repeated EVAR, and five had open repair. For patients who underwent repair for delayed rupture, mortality at 30 days and 1 year were 44.4% and 66.7%, respectively. Multivariable Cox regression analysis identified age 80 to 89 (hazard ratio, 3.3; 95% confidence interval, 1.1-9.4; P = .03), and symptomatic or ruptured initial indication for EVAR (hazard ratio, 7.4; 95% confidence interval, 2.2-24.8; P < .01) as significant predictors of delayed rupture. CONCLUSIONS: Rupture after EVAR is a rare but devastating event, and mortality after repair exceeds 60% at 1 year. Most delayed cases showed late AAA expansion, thereby implicating late loss of seal and increased endoleaks as the cause of rupture in these patients and mandating vigilant surveillance.

Ten-year results of endovascular abdominal aortic aneurysm repair from a large multicenter registry.

OBJECTIVE: To assess outcomes after endovascular abdominal aortic aneurysm repair (EVAR) in an integrated health care system. METHODS: Between 2000 and 2010, 1736 patients underwent EVAR at 17 centers. Demographic data, comorbidities, and outcomes of interest were collected. EVAR in patients presenting with ruptured or symptomatic aneurysms was categorized as urgent; otherwise, it was considered elective. Primary outcomes were mortality and aneurysm-related mortality (ARM). Secondary outcomes were change in aneurysm sac size, endoleak status, major adverse events, and reintervention. RESULTS: Overall, the median age was 76 years (interquartile range, 70-81 years), 86% were male, and 82% were Caucasian. Most cases (93.8%) were elective, but urgent use of EVAR increased from 4% in the first 5 years to 7.3% in the last 5 years of the study period. Mean aneurysm size was 5.8 cm. Patients were followed for an average of 3 years (range, 1-11 years); 8% were lost to follow-up. Intraoperatively, 4.5% of patients required adjunctive maneuvers for endoleak, fixation, or flow-limiting issues. The 30-day mortality rate was 1.2%, and the perioperative morbidity rate was 6.6%. Intraoperative type I and II endoleaks were uncommon (2.3% and 9.3%, respectively). Life-table analysis at 5 years demonstrated excellent overall survival (66%) and freedom from ARM (97%). Postoperative endoleak was seen in 30% of patients and was associated with an increase in sac size over time. Finally, the total reintervention rate was 15%, including 91 instances (5%) of revisional EVAR. The overall major adverse event rate was 7.9% and decreased significantly from 12.3% in the first 5 years to 5.6% in the second 5 years of the study period (P < .001). Overall ARM was worse in patients with postoperative endoleak (4.1% vs 1.8%; P < .01) or in those who underwent reintervention (7.6% vs 1.6%; P < .001). CONCLUSIONS: Results from a contemporary EVAR registry in an integrated health care system demonstrate favorable perioperative outcomes and excellent clinical efficacy. However, postoperative endoleak and the need for reintervention continue to be challenging problems for patients after EVAR.