Development and validation of a machine-learning prediction model to improve abdominal aortic aneurysm screening.

OBJECTIVE: Despite recommendations by the United States Preventive Services Task Force and the Society for Vascular Surgery, adoption of screening for abdominal aortic aneurysms (AAAs) remains low. One challenge is the low prevalence of AAAs in the unscreened population, and therefore a low detection rate for AAA screenings. We sought to use machine learning to identify factors associated with the presence of AAAs and create a model to identify individuals at highest risk for AAAs, with the aim of increasing the detection rate of AAA screenings. METHODS: A machine-learning model was trained using longitudinal medical records containing lab results, medications, and other data from our institutional database. A retrospective cohort study was performed identifying current or past smoking in patients aged 65 to 75 years and stratifying the patients by sex and smoking status as well as determining which patients had a confirmed diagnosis of AAA. The model was then adjusted to maximize fairness between sexes without significantly reducing precision and validated using six-fold cross validation. RESULTS: Validation of the algorithm on the single-center institutional data utilized 18,660 selected patients over 2 years and identified 314 AAAs. There were 41 factors identified in the medical record included in the machine-learning algorithm, with several factors never having been previously identified to be associated with AAAs. With an estimated 100 screening ultrasounds completed monthly, detection of AAAs is increased with a lift of 200% using the algorithm as compared with screening based on guidelines. The increased detection of AAAs in the model-selected individuals is statistically significant across all cutoff points. CONCLUSIONS: By utilizing a machine-learning model, we created a novel algorithm to detect patients who are at high risk for AAAs. By selecting individuals at greatest risk for targeted screening, this algorithm resulted in a 200% lift in the detection of AAAs when compared with standard screening guidelines. Using machine learning, we also identified several new factors associated with the presence of AAAs. This automated process has been integrated into our current workflows to improve screening rates and yield of high-risk individuals for AAAs.

Intramural hematoma of the thoracic aorta: A single-institution, 12-year experience.

OBJECTIVE: The natural history and management of intramural hematoma (IMH) has varied significantly worldwide. From the present retrospective analysis of our institutional database, we have reported the long-term results from medical and surgical management of types A and B IMH. METHODS: Computed tomography reports completed at our tertiary care hospital from July 2007 to July 2020 were used to identify patients with IMH with a thickness of ≥7 mm. Those with IMH directly related to trauma, previous aortic surgery, penetrating atheromatous ulcer, dissection flap, or an iatrogenic source and those who had never received any treatment of IMH at presentation were excluded. RESULTS: A total of 54 patients with IMH had met the inclusion and exclusion criteria. Of the 54 patients, 24 had presented with Stanford type A. Of these 24 patients, 10 had initially undergone surgery and 14 had initially received medical treatment. Two patients in the medical group had subsequently undergone surgery. In addition, 30 patients had presented with type B IMH and had initially received medical treatment, with 3 eventually requiring surgical intervention. In-hospital survival was 90% for type A IMH treated surgically, 93% for type A IMH treated medically, and 97% for type B IMH treated medically. At the last follow-up imaging study of the medically treated patients, 36% of those with type A IMH and 31% of those with type B IMH had experienced complete resolution of IMH at 3.7 and 31.5 months respectively, without surgical intervention. The development of an aortic aneurysm at the site of a previous IMH had occurred in 18% (2 of 11) and 12% (3 of 26) of the type A medical and type B medical cohorts. The overall rate of aortic aneurysm formation in the region of IMH or in another segment was 50%. No difference was found in long-term survival between the three cohorts at a mean follow-up of 22.8 months. CONCLUSIONS: A role appears to exist for medical treatment with anti-impulse therapy for appropriately selected patients with type A IMH. These patients must be followed up closely clinically and radiographically for signs of deterioration in the short- and long-term phases of their care. They can achieve long-term survival similar to that of surgically treated type A IMH and medically treated type B IMH patients using this algorithm. However, they might require late surgical intervention, especially for aneurysmal disease.

Reoperation for Recurrent Intervertebral Disc Herniation in the Spine Patient Outcomes Research Trial: Analysis of Rate, Risk Factors, and Outcome.

STUDY DESIGN: This study was a post-hoc subgroup analysis of prospectively collected data in the Spine Patient Outcomes Research Trial (SPORT). OBJECTIVE: The aim of this study was to determine the risk factors for and to compare the outcomes of patients undergoing revision disc excision surgery in SPORT. SUMMARY OF BACKGROUND DATA: Risk factors for reherniation and outcomes after revision surgery have not been well-studied. This information is critical for proper patient counseling and decision-making. METHODS: Patients who underwent primary discectomy in the SPORT intervertebral disc herniation cohort were analyzed to determine risk factors for undergoing revision surgery. Risk factors for undergoing revision surgery for reherniation were evaluated using univariate and multivariate analysis. Primary outcome measures consisted of Oswestry Disability Index (ODI), the Sciatica Bothersomeness index (SBI), and the Short Form 36 (SF-36) at 6 weeks, 3 months, 6 months, and yearly to 4 years. RESULTS: Of 810 surgical patients, 74 (9.1%) received revision surgery for reherniation. Risk factors for reherniation included: younger age (hazard ratio [HR] 0.96 [0.94-0.99]), lack of a sensory deficit (HR 0.61 [0.37-0.99]) lack of motor deficit (HR 0.54 [0.32-0.91]), and higher baseline ODI score (HR 1.02 [1.01-1.03]). The time-adjusted mean improvement from baseline to 4 years was less for the reherniation group on all outcome measures (Bodily Pain Index [BP] 39.5 vs. 44.9, P = 0.001; Physical Function Index [PF] 37.1 vs. 44.5, P < 0.001; ODI 33.9 vs. 38.3, P < 0.001; SBI 8.7 vs. 10.5, P < 0.001). At 4 years, only SBI (-9 vs. -11.4, P = 0.002) was significantly lower in the reherniation group. CONCLUSION: Younger patients with higher baseline disability without neurological deficit are at increased risk of undergoing revision surgery for reherniation. Those considering revision surgery for reherniation will likely improve significantly following surgery, but possibly not as much as with primary discectomy. LEVEL OF EVIDENCE: 3.