Cardiorespiratory fitness and risk of Alzheimer's disease and related dementias among American veterans.

INTRODUCTION: Cardiorespiratory fitness (CRF) is associated with improved health and survival. Less is known about its association with Alzheimer's disease and related dementias (ADRD). METHODS: We identified 649,605 US veterans 30 to 95 years of age and free of ADRD who completed a standardized exercise tolerance test between 2000 and 2017 with no evidence of ischemia. We examined the association between five age- and sex-specific CRF categories and ADRD incidence using multivariate Cox regression models. RESULTS: During up to 20 (median 8.3) years of follow-up, incident ADRD occurred in 44,105 (6.8%) participants, with an incidence rate of 7.7/1000 person-years. Compared to the least-fit, multivariable-adjusted hazard ratios (95% confidence intervals) for incident ADRD were: 0.87 (0.85-0.90), 0.80 (0.78-0.83), 0.74 (0.72-0.76), and 0.67 (0.65-0.70), for low-fit, moderate-fit, fit, and high-fit individuals, respectively. DISSCUSSION: These findings demonstrate an independent, inverse, and graded association between CRF and incident ADRD. Future studies may determine the amount and duration of physical activity needed to optimize ADRD risk reduction.

Criticality: A New Concept of Severity of Illness for Hospitalized Children.

OBJECTIVES: To validate the conceptual framework of "criticality," a new pediatric inpatient severity measure based on physiology, therapy, and therapeutic intensity calibrated to care intensity, operationalized as ICU care. DESIGN: Deep neural network analysis of a pediatric cohort from the Health Facts (Cerner Corporation, Kansas City, MO) national database. SETTING: Hospitals with pediatric routine inpatient and ICU care. PATIENTS: Children cared for in the ICU (n = 20,014) and in routine care units without an ICU admission (n = 20,130) from 2009 to 2016. All patients had laboratory, vital sign, and medication data. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A calibrated, deep neural network used physiology (laboratory tests and vital signs), therapy (medications), and therapeutic intensity (number of physiology tests and medications) to model care intensity, operationalized as ICU (versus routine) care every 6 hours of a patient's hospital course. The probability of ICU care is termed the Criticality Index. First, the model demonstrated excellent separation of criticality distributions from a severity hierarchy of five patient groups: routine care, routine care for those who also received ICU care, transition from routine to ICU care, ICU care, and high-intensity ICU care. Second, model performance assessed with statistical metrics was excellent with an area under the curve for the receiver operating characteristic of 0.95 for 327,189 6-hour time periods, excellent calibration, sensitivity of 0.817, specificity of 0.892, accuracy of 0.866, and precision of 0.799. Third, the performance in individual patients with greater than one care designation indicated as 88.03% (95% CI, 87.72-88.34) of the Criticality Indices in the more intensive locations was higher than the less intense locations. CONCLUSIONS: The Criticality Index is a quantification of severity of illness for hospitalized children using physiology, therapy, and care intensity. This new conceptual model is applicable to clinical investigations and predicting future care needs.

Detecting Secular Trends in Clinical Treatment through Temporal Analysis.

Medical treatments change over time for multiple reasons, including introduction of new treatments, availability of new scientific evidence, change in institutional guidelines, and market efforts by pharmaceutical and medical device companies. Monitoring and analyzing these secular trends will also inform the evaluation of evidence based practice as well as outcome research. Using a large national clinical dataset from the United States Veterans Health Administration (VHA), we measured the change in prevalence of all diseases, medications, and procedures by year from 2001 to 2014. To assess statistical significance, we used a generalized linear model. Among the large number of changes that were observed, multiple significant changes were related to diabetes mellitus type II (DM2). Prevalence of DM2 in the VHA increased after 2001 but plateaued by 2008; blood sugar testing by glycosylated hemoglobin increased consistently while glucose testing decreased; and the trend of insulin and metformin use was consistent with the trend in DM2 prevalence, while glyburide and rosiglitazone use dropped sharply.

Patient Dietary Supplements Use: Do Results from Natural Language Processing of Clinical Notes Agree with Survey Data?

There is widespread use of dietary supplements, some prescribed but many taken without a physician's guidance. There are many potential interactions between supplements and both over-the-counter and prescription medications in ways that are unknown to patients. Structured medical records do not adequately document supplement use; however, unstructured clinical notes often contain extra information on supplements. We studied a group of 377 patients from three healthcare facilities and developed a natural language processing (NLP) tool to detect supplement use. Using surveys of these patients, we investigated the correlation between self-reported supplement use and NLP extractions from the clinical notes. Our model achieved an F1 score of 0.914 for detecting all supplements. Individual supplement detection had a variable correlation with survey responses, ranging from an F1 of 0.83 for calcium to an F1 of 0.39 for folic acid. Our study demonstrated good NLP performance while also finding that self-reported supplement use is not always consistent with the documented use in clinical records.

Identification of colorectal cancer using structured and free text clinical data.

Colorectal cancer incidence has continually fallen among those 50 years old and over. However, the incidence has increased in those under 50. Even with the recent screening guidelines recommending that screening begins at age 45, nearly half of all early-onset colorectal cancer will be missed. Methods are needed to identify high-risk individuals in this age group for targeted screening. Colorectal cancer studies, as with other clinical studies, have required labor intensive chart review for the identification of those affected and risk factors. Natural language processing and machine learning can be used to automate the process and enable the screening of large numbers of patients. This study developed and compared four machine learning and statistical models: logistic regression, support vector machine, random forest, and deep neural network, in their performance in classifying colorectal cancer patients. Excellent classification performance is achieved with AUCs over 97%.

Cardiovascular Safety of Hydroxychloroquine in US Veterans With Rheumatoid Arthritis.

OBJECTIVE: Hydroxychloroquine (HCQ) may prolong the QT interval, a risk factor for torsade de pointes, a potentially fatal ventricular arrhythmia. This study was undertaken to examine the cardiovascular safety of HCQ in patients with rheumatoid arthritis (RA). METHODS: We conducted an active comparator safety study of HCQ in a propensity score-matched cohort of 8,852 US veterans newly diagnosed as having RA between October 1, 2001 and December 31, 2017. Patients were started on HCQ (n = 4,426) or another nonbiologic disease-modifying antirheumatic drug (DMARD; n = 4,426) after RA diagnosis, up to December 31, 2018, and followed up for 12 months after therapy initiation, up to December 31, 2019. RESULTS: Patients had a mean ± SD age of 64 ± 12 years, 14% were women, and 28% were African American. The treatment groups were balanced with regard to 87 baseline characteristics. There were 3 long QT syndrome events (0.03%), 2 of which occurred in patients receiving HCQ. Of the 56 arrhythmia-related hospitalizations (0.63%), 30 occurred in patients in the HCQ group (hazard ratio [HR] associated with HCQ 1.16 [95% confidence interval (95% CI) 0.68-1.95]). All-cause mortality occurred in 144 (3.25%) and 136 (3.07%) of the patients in the HCQ and non-HCQ groups, respectively (HR associated with HCQ 1.06 [95% CI, 0.84-1.34]). During the first 30 days of follow-up, there were no long QT syndrome events, 2 arrhythmia-related hospitalizations (none in the HCQ group), and 13 deaths (6 in the HCQ group). CONCLUSION: Our findings indicate that the incidence of long QT syndrome and arrhythmia-related hospitalization is low in patients with RA during the first year after the initiation of HCQ or another nonbiologic DMARD. We found no evidence that HCQ therapy is associated with a higher risk of adverse cardiovascular events or death.

Regular Expression-Based Learning for METs Value Extraction.

Functional status as measured by exercise capacity is an important clinical variable in the care of patients with cardiovascular diseases. Exercise capacity is commonly reported in terms of Metabolic Equivalents (METs). In the medical records, METs can often be found in a variety of clinical notes. To extract METs values, we adapted a machine-learning algorithm called REDEx to automatically generate regular expressions. Trained and tested on a set of 2701 manually annotated text snippets (i.e. short pieces of text), the regular expressions were able to achieve good accuracy and F-measure of 0.89 and 0.86. This extraction tool will allow us to process the notes of millions of cardiovascular patients and extract METs value for use by researchers and clinicians.

The effect of simulated narratives that leverage EMR data on shared decision-making: a pilot study.

BACKGROUND: Shared decision-making can improve patient satisfaction and outcomes. To participate in shared decision-making, patients need information about the potential risks and benefits of treatment options. Our team has developed a novel prototype tool for shared decision-making called hearts like mine (HLM) that leverages EHR data to provide personalized information to patients regarding potential outcomes of different treatments. These potential outcomes are presented through an Icon array and/or simulated narratives for each "person" in the display. In this pilot project we sought to determine whether the inclusion of simulated narratives in the display affects individuals' decision-making. Thirty subjects participated in this block-randomized study in which they used a version of HLM with simulated narratives and a version without (or in the opposite order) to make a hypothetical therapeutic decision. After each decision, participants completed a questionnaire that measured decisional confidence. We used Chi square tests to compare decisions across conditions and Mann-Whitney U tests to examine the effects of narratives on decisional confidence. Finally, we calculated the mean of subjects' post-experiment rating of whether narratives were helpful in their decision-making. RESULTS: In this study, there was no effect of simulated narratives on treatment decisions (decision 1: Chi squared = 0, p = 1.0; decision 2: Chi squared = 0.574, p = 0.44) or Decisional confidence (decision 1, w = 105.5, p = 0.78; decision 2, w = 86.5, p = 0.28). Post-experiment, participants reported that narratives helped them to make decisions (mean = 3.3/4). CONCLUSIONS: We found that simulated narratives had no measurable effect on decisional confidence or decisions and most participants felt that the narratives were helpful to them in making therapeutic decisions. The use of simulated stories holds promise for promoting shared decision-making while minimizing their potential biasing effect.

Identification and Use of Frailty Indicators from Text to Examine Associations with Clinical Outcomes Among Patients with Heart Failure.

Frailty is an important health outcomes indicator and valuable for guiding healthcare decisions in older adults, but is rarely collected in a quantitative, systematic fashion in routine healthcare. Using a cohort of 12,000 Veterans with heart failure, we investigated the feasibility of topic modeling to identify frailty topics in clinical notes. Topics were generated through unsupervised learning and then manually reviewed by an expert. A total of 53 frailty topics were identified from 100,000 notes. We further examined associations of frailty with age-, sex-, and Charlson Comorbidity Index-adjusted 1-year hospitalizations and mortality (composite outcome) using logistic regression. Frailty (≤ 4 topics versus <4) was associated with twice the risk of the composite outcome, Odds Ratio: 2.2, 95% Confidence Interval: (2.0-2.4). This study demonstrates the feasibility of identifying frailty indicators from clinical notes and linking these to clinically relevant outcomes. Future work includes integrating frailty indicators into validated predictive tools.

Automated Learning of Temporal Expressions.

Clinical notes contain important temporal information that are critical for making clinical diagnosis and treatment as well as for retrospective analyses. Manually created regular expressions are commonly used for the extraction of temporal information; however, this can be a time consuming and brittle approach. We describe a novel algorithm for automatic learning of regular expressions in recognizing temporal expressions. Five classes of temporal expressions are identified. Keywords specific to those classes are used to retrieve snippets of text representing the same keywords in context. Those snippets are used for Regular Expression Discovery Extraction (REDEx). These learned regular expressions are then evaluated using 10-fold cross validation. Precision and recall are very high, above 0.95 for most classes.

Pediatric hepatopulmonary syndrome is seen with polysplenia/interrupted inferior vena cava and without cirrhosis.

Hepatopulmonary syndrome (HPS) is a triad of liver dysfunction, hypoxemia, and intrapulmonary vascular dilatation. We describe the prevalence and clinical features of HPS at a pediatric liver transplant center. Patients referred to Children's Healthcare of Atlanta/Emory University transplant program from February 1999 to May 2005 were reviewed. Oxygen saturation in room air was screened by percutaneous pulse oximetry. HPS cases were compared with similar age non-HPS recipients (n = 38) to determine differences in clinical characteristics, Pediatric End-Stage Liver Disease (PELD) scores, and posttransplantation survival. Of 211 patients referred and 114 patients transplanted, 7 met criteria for HPS (3.3% and 6.1%, respectively). Patients with HPS had lower PELD score (-0.4 +/- 5.9 vs. 11 +/- 11; P = 0.01) and total bilirubin (1.7 +/- 1.1 vs. 11.2 +/- 10.1; P = 0.02) at the time of transplantation. Four of 7 patients with HPS had polysplenia/interrupted inferior vena cava (PS/IVC) compared with 0 of 38 age-matched controls (P = 0.0002). Three patients with HPS did not have cirrhosis; 2 of these 3 had PS/IVC. All HPS cases normalized room air oxygen saturation by 6 months, and survival after transplantation in HPS cases was 100%. Marked hepatic synthetic or biochemical dysfunction may not be present, and cirrhosis is not a requirement for the development of HPS in children. HPS in children is frequently associated with PS/IVC. Histologic evidence of abnormal intrahepatic portal vein flow and the demonstration of portosystemic communications at any level should be sought in children presenting with unexplained intrapulmonary vascular dilatation. Liver transplantation for HPS in childhood may be appropriate even in the absence of cirrhosis.

Renal artery pseudoaneurysm secondary to blunt trauma treated with microcoil embolization.

BACKGROUND: Renal artery pseudoaneurysms are rarely described sequelae of blunt abdominal trauma. Interventional radiological advances have allowed such lesions to be managed nonoperatively. METHODS: The authors review the presentation, diagnostic evaluation, and hospital course of an 11-year-old girl who developed a right renal artery pseudoaneurysm 14 days after blunt abdominal trauma. RESULTS: An 11-year-old girl fell off a horse onto her right flank. She sustained multiple right hepatic lobe lacerations and a complex fracture of the upper pole of the right kidney. Her initial hospital course was uncomplicated, and she was discharged after an uneventful 6-day course. The child did well for 2 weeks, until she developed right back pain and gross hematuria. A computed tomography scan revealed a right renal artery pseudoaneurysm. Angiography confirmed the presence of a pseudoaneurysm, which was fed by a single segmental branch originating from the renal artery. The artery was successfully occluded with a single platinum microcoil, which was demonstrated by the absence of contrast flow into the pseudoaneurysm. The patient recovered and was discharged shortly after the procedure. She initially had intermittent pain and hematuria, which resolved. Follow-up computed tomography scans have shown resolution of both the renal and hepatic lesions. CONCLUSIONS: Renal artery pseudoaneurysms that arise after blunt abdominal trauma in the pediatric population may be safely and effectively managed with arterial embolization, thereby avoiding extensive surgical interventions.

Endoluminal versus open treatment of descending thoracic aortic aneurysms.

PURPOSE: This report describes the authors' initial experience with the Excluder thoracic endoprosthesis (W. L. Gore and Associates, Inc, Flagstaff, Ariz) and the thoracic Talent endoprosthesis (Medtronic AVE, Sunrise, Fla) and their safety and efficacy in the primary endovascular repair of descending thoracic aortic aneurysms (TAAs). In addition, comparison with a historic nonrandomized cohort of patients that had undergone open repair of descending TAAs is reported. PATIENTS AND METHODS: Repair of TAA (mean diameter, 68 +/- 22 mm) was attempted in 19 patients with the Excluder (n = 14) and the Talent (n = 5) endoprostheses between March 1999 and January 2000. This group was compared with a historic nonrandomized cohort of 10 patients that had undergone open repair of anatomically similar descending TAA (mean diameter, 74 +/- 22 mm) between January 1996 and January 1998. The mean age in the endovascular group was 70.6 +/- 5.3 years versus 70.1 +/- 4.5 years in the historic open group. All the procedures were performed in a standard operating room with angiographic capabilities. In the historic open group, each standard tube graft repair of descending TAA was performed by one of three staff surgeons. RESULTS: Endograft deployment was successful in 18 patients (95%). The procedure was aborted in one patient (Excluder) because of small iliac arteries and access difficulty. The average operative time was 155 +/- 62 minutes, with a mean blood loss of 325 +/- 353 mL (versus 256 +/- 102 minutes and 1205 +/- 1493 mL, respectively, in the open group). Eight patients needed the planned use of more than one component for enhanced sealing or additional length in the endovascular group. No type I endoleaks were identified on the intraoperative completion angiography. One perioperative mortality occurred in the endovascular group and the open group. In the endovascular group, other complications included retroperitoneal hematoma and external iliac artery dissection (n = 1), lymphocele (n = 1), and common femoral artery pseudoaneurysm (n = 1). In the open group, other complications included ischemic colitis (n = 1), severe renal insufficiency (n = 2), wound infection (n = 1), and stroke (n = 1). In the endovascular group, the length of stay was 6.2 +/- 3.3 days (range, 1 to 13 days), with only nine patients needing intensive care, whereas in the open group, the length of stay was 16.3 +/- 6.7 days, with all patients needing intensive care. Endoleaks, graft migrations, or ruptures were not seen on the 1-month, 6-month, and 12-month follow-up computed tomographic scans in the endovascular group. On the average, aneurysm size decreased from 68 +/- 22 mm to 58 +/- 13 mm, to 51 +/- 14 mm, and to 49 +/- 12 mm at 1, 6, and 12 months after endovascular repair, respectively. No spinal cord ischemia was seen in either group. CONCLUSION: The endoluminal repair was effective in exclusion of descending TAAs from the systemic circulation in this selected group of patients. In this short-term follow-up, compared with the nonrandomized historic cohort of open descending TAA repair, the endovascular group had significantly shorter operating times and hospital and intensive care unit stays and lower operative blood loss. Further follow-up and continued assessment of the long-term durability of these devices in elective and emergency circumstances are warranted.