No abstract

Objective: Phenotyping algorithms applied to electronic health record (EHR) data enable investigators to identify large cohorts for clinical and genomic research. Algorithm development is often iterative, depends on fallible investigator intuition, and is time-and labor-intensive. We developed and evaluated 4 types of phenotyping algorithms and categories of EHR information to identify hypertensive individuals and controls and provide a portable module for implementation at other sites. Materials and Methods: We reviewed the EHRs of 631 individuals followed at Vanderbilt for hypertension status. We developed features and phenotyping algorithms of increasing complexity. Input categories included International Classification of Diseases, Ninth Revision (ICD9) codes, medications, vital signs, narrative-text search results, and Unified Medical Language System (UMLS) concepts extracted using natural language processing (NLP). We developed a module and tested portability by replicating 10 of the best-performing algorithms at the Marshfield Clinic. Results: Random forests using billing codes, medications, vitals, and concepts had the best performance with a median area under the receiver operator characteristic curve (AUC) of 0.976. Normalized sums of all 4 categories also performed well (0.959 AUC). The best non-NLP algorithm combined normalized ICD9 codes, medications, and blood pressure readings with a median AUC of 0.948. Blood pressure cutoffs or ICD9 code counts alone had AUCs of 0.854 and 0.908, respectively. Marshfield Clinic results were similar. Conclusion: This work shows that billing codes or blood pressure readings alone yield good hypertension classification performance. However, even simple combinations of input categories improve performance. The most complex algorithms classified hypertension with excellent recall and precision.

This study demonstrated that, although all the models tested had good discrimination, performance characteristics varied between methods, and the random forests models did not calibrate as well as the lasso or logistic regression models. In addition, novel modifiable risk factors were explored and found to be significant.

Prognostic Level III. See Instructions for Authors for a complete.

Background:Deep venous thrombosis (DVT) and pulmonary embolism (PE) are recognized as major causes of morbidity and mortality in orthopaedic trauma patients. Despite the high incidence of these complications following orthopaedic trauma, there is a paucity of literature investigating the clinical risk factors for DVT in this specific population. As our healthcare system increasingly emphasizes quality measures, it is critical for orthopaedic surgeons to understand the clinical factors that increase the risk of DVT following orthopaedic trauma.Objectives:Utilizing the ACS-NSQIP database, we sought to determine the incidence and identify independent risk factors for DVT following orthopaedic trauma.Patients and Methods:Using current procedural terminology (CPT) codes for orthopaedic trauma procedures, we identified a prospective cohort of patients from the 2006 to 2013 ACS-NSQIP database. Using Wilcoxon-Mann-Whitney and chi-square tests where appropriate, patient demographics, comorbidities, and operative factors were compared between patients who developed a DVT within 30 days of surgery and those who did not. A multivariate logistic regression analysis was conducted to calculate odds ratios (ORs) and identify independent risk factors for DVT. Significance was set at P < 0.05.Results:56,299 orthopaedic trauma patients were included in the analysis, of which 473 (0.84%) developed a DVT within 30 days. In univariate analysis, twenty-five variables were significantly associated with the development of a DVT, including age (P < 0.0001), BMI (P = 0.037), diabetes (P = 0.01), ASA score (P < 0.0001) and anatomic region injured (P < 0.0001). Multivariate analysis identified several independent risk factors for development of a DVT including use of a ventilator (OR = 43.67, P = 0.039), ascites (OR = 41.61, P = 0.0038), steroid use (OR = 4.00, P < 0.001), and alcohol use (OR = 2.98, P = 0.0370). Compared to patients with upper extremity trauma, those with lower extremity injuries had significantly increased odds of developing a DVT (OR = 7.55, P = 0.006). The trend toward increased odds of DVT among patients with injuries to the hip/pelvis did not reach statistical significance (OR = 4.51, P = 0.22). Smoking was not found to be an independent risk factor for developing a DVT (P = 0.1217).Conclusions:This is the largest study to date using the NSQIP database to identify risk factors for DVT in orthopaedic trauma patients. Although the incidence of DVT was low in our cohort, the presence of certain risk factors significantly increased the odds of developing a DVT following orthopaedic trauma. These findings will enable orthopaedic surgeons to target at-risk patients and implement post-operative care protocols aimed at reducing the morbidity and mortality associated with DVT in orthopaedic trauma patients.