The addition of depression to the Framingham Risk Equation model for predicting coronary heart disease risk in women

BACKGROUND: Depression is widely considered to be an independent and robust predictor of Coronary Heart Disease (CHD), however is seldom considered in the context of formal risk assessment. We assessed whether the addition of depression to the Framingham Risk Equation (FRE) improved accuracy for predicting 10-year CHD in a sample of women.

DESIGN: A prospective, longitudinal design comprising an age-stratified, population-based sample of Australian women collected between 1993 and 2011 (n=862).

METHODS: Clinical depressive disorder was assessed using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders (SCID-I/NP), using retrospective age-of-onset data. A composite measure of CHD included non-fatal myocardial infarction, unstable angina coronary intervention or cardiac death. Cox proportional-hazards regression models were conducted and overall accuracy assessed using area under receiver operating characteristic (ROC) curve analysis.

RESULTS: ROC curve analyses revealed that the addition of baseline depression status to the FRE model improved its overall accuracy (AUC:0.77, Specificity:0.70, Sensitivity:0.75) when compared to the original FRE model (AUC:0.75, Specificity:0.73, Sensitivity:0.67). However, when calibrated against the original model, the predicted number of events generated by the augmented version marginally over-estimated the true number observed.

CONCLUSIONS: The addition of a depression variable to the FRE equation improves the overall accuracy of the model for predicting 10-year CHD events in women, however may over-estimate the number of events that actually occur. This model now requires validation in larger samples as it could form a new CHD risk equation for women.

Depression is a risk factor for incident coronary heart disease in women: an 18-year longitudinal study

BACKGROUND: According to a recent position paper by the American Heart Association, it remains unclear whether depression is a risk factor for incident Coronary Heart Disease (CHD). We assessed whether a depressive disorder independently predicts 18-year incident CHD in women. METHOD: A prospective longitudinal study of 860 women enrolled in the Geelong Osteoporosis Study (1993-2011) was conducted. Participants were derived from an age-stratified, representative sample of women (20-94 years) randomly selected from electoral rolls in South-Eastern Australia. The exposure was a diagnosis of a depressive disorder using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders. Outcomes data were collected from hospital medical records: (1) Primary outcome: a composite measure of cardiac death, non-fatal Myocardial Infarction or coronary intervention. (2) Secondary outcome: any cardiac event (un/stable angina, cardiac event not otherwise defined) occurring over the study period. RESULTS: Seven participants were excluded based on CHD history. Eighty-three participants (9.6%) recorded ≥1 cardiac event over the study period; 47 had a diagnosis that met criteria for inclusion in the primary analysis. Baseline depression predicted 18-year incidence, adjusting for (1) anxiety (adj. OR:2.39; 95% CIs:1.19-4.82), plus (2) typical risk factors (adj. OR:3.22; 95% CIs:1.45-6.93), plus (3) atypical risk factors (adj. OR:3.28; 95% CIs:1.36-7.90). This relationship held when including all cardiac events. No relationship was observed between depression and recurrent cardiac events. CONCLUSION: The results of this study support the contention that depression is an independent risk factor for CHD incidence in women. Moreover, the strength of association between depression and CHD incidence was of a greater magnitude than any typical and atypical risk factor.

Prediciton of emergency events: a multi-task multi-label learning approach

Prediction of patient outcomes is critical to plan resources in an hospital emergency department. We present a method to exploit longitudinal data from Electronic Medical Records (EMR), whilst exploiting multiple patient outcomes. We divide the EMR data into segments where each segment is a task, and all tasks are associated with multiple patient outcomes over a 3, 6 and 12 month period. We propose a model that learns a prediction function for each task-label pair, interacting through two subspaces: the first subspace is used to impose sharing across all tasks for a given label. The second subspace captures the task-specific variations and is shared across all the labels for a given task. The proposed model is formulated as an iterative optimization problems and solved using a scalable and efficient Block co-ordinate descent (BCD) method. We apply the proposed model on two hospital cohorts - Cancer and Acute Myocardial Infarction (AMI) patients collected over a two year period from a large hospital emergency department. We show that the predictive performance of our proposed models is significantly better than those of several state-of-the-art multi-task and multi-label learning methods.

Hierarchical Bayesian nonparametric models for knowledge discovery from electronic medical records

Electronic Medical Record (EMR) has established itself as a valuable resource for large scale analysis of health data. A hospital EMR dataset typically consists of medical records of hospitalized patients. A medical record contains diagnostic information (diagnosis codes), procedures performed (procedure codes) and admission details. Traditional topic models, such as latent Dirichlet allocation (LDA) and hierarchical Dirichlet process (HDP), can be employed to discover disease topics from EMR data by treating patients as documents and diagnosis codes as words. This topic modeling helps to understand the constitution of patient diseases and offers a tool for better planning of treatment. In this paper, we propose a novel and flexible hierarchical Bayesian nonparametric model, the word distance dependent Chinese restaurant franchise (wddCRF), which incorporates word-to-word distances to discover semantically-coherent disease topics. We are motivated by the fact that diagnosis codes are connected in the form of ICD-10 tree structure which presents semantic relationships between codes. We exploit a decay function to incorporate distances between words at the bottom level of wddCRF. Efficient inference is derived for the wddCRF by using MCMC technique. Furthermore, since procedure codes are often correlated with diagnosis codes, we develop the correspondence wddCRF (Corr-wddCRF) to explore conditional relationships of procedure codes for a given disease pattern. Efficient collapsed Gibbs sampling is derived for the Corr-wddCRF. We evaluate the proposed models on two real-world medical datasets - PolyVascular disease and Acute Myocardial Infarction disease. We demonstrate that the Corr-wddCRF model discovers more coherent topics than the Corr-HDP. We also use disease topic proportions as new features and show that using features from the Corr-wddCRF outperforms the baselines on 14-days readmission prediction. Beside these, the prediction for procedure codes based on the Corr-wddCRF also shows considerable accuracy.