A Mobile App to Support Self-management of Chronic Kidney Disease: Development Study.

BACKGROUND: Chronic kidney disease (CKD) is a common and costly condition that is usually accompanied by multiple comorbidities including type 2 diabetes, hypertension, and obesity. Proper management of CKD can delay or prevent kidney failure and help mitigate cardiovascular disease risk, which increases as kidney function declines. Smart device apps hold potential to enhance patient self-management of chronic conditions including CKD. OBJECTIVE: The objective of this study was to develop a mobile app to facilitate self-management of nondialysis-dependent CKD. METHODS: Our stakeholder team included 4 patients with stage 3-4 nondialysis-dependent CKD; a kidney transplant recipient; a caretaker; CKD care providers (pharmacists, a nurse, primary care physicians, a nephrologist, and a cardiologist); 2 health services and CKD researchers; a researcher in biomedical informatics, nutrition, and obesity; a system developer; and 2 programmers. Focus groups and in-person interviews with the patients and providers were conducted using a focus group and interview guide based on existing literature on CKD self-management and the mobile app quality criteria from the Mobile App Rating Scale. Qualitative analytic methods including the constant comparative method were used to analyze the focus group and interview data. RESULTS: Patients and providers identified and discussed a list of requirements and preferences regarding the content, features, and technical aspects of the mobile app, which are unique for CKD self-management. Requirements and preferences centered along themes of communication between patients and caregivers, partnership in care, self-care activities, adherence to treatment regimens, and self-care self-efficacy. These identified themes informed the features and content of our mobile app. The mobile app user can enter health data including blood pressure, weight, and blood glucose levels. Symptoms and their severity can also be entered, and users are prompted to contact a physician as indicated by the symptom and its severity. Next, mobile app users can select biweekly goals from a set of predetermined goals with the option to enter customized goals. The user can also keep a list of medications and track medication use. Our app includes feedback mechanisms where in-range values for health data are depicted in green and out-of-range values are depicted in red. We ensured that data entered by patients could be downloaded into a user-friendly report, which could be emailed or uploaded to an electronic health record. The mobile app also includes a mechanism that allows either group or individualized video chat meetings with a provider to facilitate either group support, education, or even virtual clinic visits. The CKD app also includes educational material on CKD and its symptoms. CONCLUSIONS: Patients with CKD and CKD care providers believe that a mobile app can enhance CKD self-management by facilitating patient-provider communication and enabling self-care activities including treatment adherence.

Application of machine learning to the prediction of postoperative sepsis after appendectomy.

BACKGROUND: We applied various machine learning algorithms to a large national dataset to model the risk of postoperative sepsis after appendectomy to evaluate utility of such methods and identify factors associated with postoperative sepsis in these patients. METHODS: The National Surgery Quality Improvement Program database was used to identify patients undergoing appendectomy between 2005 and 2017. Logistic regression, support vector machines, random forest decision trees, and extreme gradient boosting machines were used to model the occurrence of postoperative sepsis. RESULTS: In the study, 223,214 appendectomies were identified; 2,143 (0.96%) were indicated as having postoperative sepsis. Logistic regression (area under the curve 0.70; 95% confidence interval, 0.68-0.73), random forest decision trees (area under the curve 0.70; 95% confidence interval, 0.68-0.73), and extreme gradient boosting (area under the curve 0.70; 95% confidence interval, 0.68-0.73) afforded similar performance, while support vector machines (area under the curve 0.51; 95% confidence interval, 0.50-0.52) had worse performance. Variable importance analyses identified preoperative congestive heart failure, transfusion, and acute renal failure as predictors of postoperative sepsis. CONCLUSION: Machine learning methods can be used to predict the development of sepsis after appendectomy with moderate accuracy. Such predictive modeling has potential to ultimately allow for preoperative recognition of patients at risk for developing postoperative sepsis after appendectomy thus facilitating early intervention and reducing morbidity.

Development and application of a high throughput natural language processing architecture to convert all clinical documents in a clinical data warehouse into standardized medical vocabularies.

OBJECTIVE: Natural language processing (NLP) engines such as the clinical Text Analysis and Knowledge Extraction System are a solution for processing notes for research, but optimizing their performance for a clinical data warehouse remains a challenge. We aim to develop a high throughput NLP architecture using the clinical Text Analysis and Knowledge Extraction System and present a predictive model use case. MATERIALS AND METHODS: The CDW was comprised of 1 103 038 patients across 10 years. The architecture was constructed using the Hadoop data repository for source data and 3 large-scale symmetric processing servers for NLP. Each named entity mention in a clinical document was mapped to the Unified Medical Language System concept unique identifier (CUI). RESULTS: The NLP architecture processed 83 867 802 clinical documents in 13.33 days and produced 37 721 886 606 CUIs across 8 standardized medical vocabularies. Performance of the architecture exceeded 500 000 documents per hour across 30 parallel instances of the clinical Text Analysis and Knowledge Extraction System including 10 instances dedicated to documents greater than 20 000 bytes. In a use-case example for predicting 30-day hospital readmission, a CUI-based model had similar discrimination to n-grams with an area under the curve receiver operating characteristic of 0.75 (95% CI, 0.74-0.76). DISCUSSION AND CONCLUSION: Our health system's high throughput NLP architecture may serve as a benchmark for large-scale clinical research using a CUI-based approach.