Patient journey before and after a formal post-traumatic stress disorder diagnosis in adults in the United States - a retrospective claims study.

OBJECTIVE: To describe post-traumatic stress disorder (PTSD)-related symptoms and frequent psychiatric comorbidities, treatments received, healthcare resource utilization (HRU), and healthcare costs pre- and post-PTSD diagnosis among adults in the United States. METHODS: Adults with PTSD who received a PTSD-related pharmacological treatment (selective serotonin reuptake inhibitor [SSRI], serotonin-norepinephrine reuptake inhibitor [SNRI], atypical antipsychotic [AA]) within 24 months of the first observed PTSD diagnosis (index date) were identified using MarketScan Commercial Database (2015-2020). Study outcomes were assessed during the 6-month pre-diagnosis and 24-month post-diagnosis periods. Subgroup analyses included patients treated or not treated with AAs post-PTSD diagnosis. RESULTS: Of the overall patients (N = 26,306; mean age at diagnosis 39.5 years; 73.3% female), 85.9% had PTSD-related symptoms and frequent psychiatric comorbidities during the 6 months pre-diagnosis. Patients treated with AAs post-PTSD diagnosis (N = 9,298) tended to have higher rates of PTSD-related symptoms and comorbidities at diagnosis than those not treated with AAs (N = 7,011). Following diagnosis, the most commonly observed first-line treatments were SSRI (67.4%), AA (23.4%), and SNRI (22.6%). The rate of PTSD-related symptoms and comorbidities, psychotherapy and pharmacological treatments received, HRU, and healthcare costs increased during the 6 months post-diagnosis relative to the 6 months pre-diagnosis and then declined over time during the 24 months post-diagnosis. CONCLUSIONS: The PTSD diagnosis was associated with increased rates of symptoms and frequent psychiatric comorbidities, psychotherapy and pharmacological treatments received, HRU, and healthcare costs, pointing to increased patient monitoring. Within 6 to 12 months after the PTSD diagnosis, these outcomes tended to reduce, perhaps as patients were obtaining targeted and effective care.

The epidemiology of early deep vein thrombosis in kidney transplant recipients.

BACKGROUND: Because kidney transplant recipients may be at increased risk for deep vein thrombosis (DVT) following transplantation, we investigated the incidence, risk factors, treatments and outcomes of early DVT among kidney transplant recipients. METHODS: An observational, single-centre cohort study was conducted among adult kidney transplant recipients from Jan. 1, 2005, to Dec. 31, 2016 with 1-year followup. Time to DVT was assessed using the Kaplan-Meier method. Cox proportional hazards and linear regression models were used to analyze risk factors for and outcomes of DVT. RESULTS: The cumulative incidence of DVT was 4.25% at 3 months after transplant. In multivariable analysis, the use of depleting induction agents (hazard ratio [HR] 2.13, 95% confidence interval [CI] 1.05-4.35]), white recipient race (HR 1.84. 95% CI 1.08-3.12), the use of kidneys from expanded criteria donors (HR 2.13, 95% CI 1.05-4.32) and lower recipient body mass index (HR 0.95, 95% CI 0.91-1.00) increased the risk for early DVT. Peritransplant DVT prophylaxis was not associated with early DVT. Early DVT was not associated with reduced graft function, death, graft failure or first hospital readmission. CONCLUSION: Risk factors for early DVT in our cohort of kidney transplant recipients included white recipient race, use of depleting agents, lower recipient body mass index and use of expanded criteria donors. As practice patterns of donor and recipient selection in kidney transplantation evolve, the results of this study may aid in perioperative risk assessments and decision-making about the use of DVT prophylaxis.

Treatment patterns among children and adolescents with attention-deficit/hyperactivity disorder in the United States - a retrospective claims analysis.

BACKGROUND: Attention-deficit/hyperactivity disorder (ADHD) is a common neurobehavioral disorder affecting approximately 10.0% of children and 6.5% of adolescents in the United States (US). A comprehensive assessment of the current treatment landscape is warranted to highlight potential unmet needs of children and adolescents with ADHD. Therefore, this study described treatment patterns and healthcare costs among commercially insured children and adolescents with ADHD in the US. METHODS: Children and adolescents with ADHD initiating pharmacological treatment indicated for ADHD were identified from IBM MarketScan Commercial Database (2014-2018). A treatment sequence algorithm was used to examine treatment patterns, including discontinuation (≥ 180 days following the last day of supply of any ADHD treatment), switch, add-on, and drop (discontinuation of an agent in combination therapy), during the 12-month study period following the index date (i.e., first observed ADHD treatment). Total adjusted annual healthcare costs were compared between patients with and without treatment changes. RESULTS: Among 49,756 children and 29,093 adolescents included, mean age was 9 and 15 years, respectively, and 31% and 38% were female. As the first treatment regimen observed, 92% of both children and adolescents initiated a stimulant and 11% initiated combination therapy. Over half of the population had a treatment change over 12 months-59% of children and 68% of adolescents. Treatment discontinuation over 12 months was common in both populations-21% of children and 36% of adolescents discontinued treatment. Healthcare costs increased with the number of treatment changes observed; children and adolescents with treatment changes (i.e., 1, 2, or ≥ 3) incurred an incremental annual cost of up to $1,443 and $2,705, respectively, compared to those without a treatment change (p < 0.001). Costs were largely driven by outpatient visits. CONCLUSIONS: Over a 12-month period, treatment changes were commonly observed and were associated with excess costs, highlighting the unmet treatment needs of children and adolescents with ADHD in the US.

Quantifying uncertainty in machine learning classifiers for medical imaging.

PURPOSE: Machine learning (ML) models in medical imaging (MI) can be of great value in computer aided diagnostic systems, but little attention is given to the confidence (alternatively, uncertainty) of such models, which may have significant clinical implications. This paper applied, validated, and explored a technique for assessing uncertainty in convolutional neural networks (CNNs) in the context of MI. MATERIALS AND METHODS: We used two publicly accessible imaging datasets: a chest x-ray dataset (pneumonia vs. control) and a skin cancer imaging dataset (malignant vs. benign) to explore the proposed measure of uncertainty based on experiments with different class imbalance-sample sizes, and experiments with images close to the classification boundary. We also further verified our hypothesis by examining the relationship with other performance metrics and cross-checking CNN predictions and confidence scores with an expert radiologist (available in the Supplementary Information). Additionally, bounds were derived on the uncertainty metric, and recommendations for interpretability were made. RESULTS: With respect to training set class imbalance for the pneumonia MI dataset, the uncertainty metric was minimized when both classes were nearly equal in size (regardless of training set size) and was approximately 17% smaller than the maximum uncertainty resulting from greater imbalance. We found that less-obvious test images (those closer to the classification boundary) produced higher classification uncertainty, about 10-15 times greater than images further from the boundary. Relevant MI performance metrics like accuracy, sensitivity, and sensibility showed seemingly negative linear correlations, though none were statistically significant (p [Formula: see text] 0.05). The expert radiologist and CNN expressed agreement on a small sample of test images, though this finding is only preliminary. CONCLUSIONS: This paper demonstrated the importance of uncertainty reporting alongside predictions in medical imaging. Results demonstrate considerable potential from automatically assessing classifier reliability on each prediction with the proposed uncertainty metric.

Assessing and mitigating the effects of class imbalance in machine learning with application to X-ray imaging.

PURPOSE: Machine learning (ML) algorithms are well known to exhibit variations in prediction accuracy when provided with imbalanced training sets typically seen in medical imaging (MI) due to the imbalanced ratio of pathological and normal cases. This paper presents a thorough investigation of the effects of class imbalance and methods for mitigating class imbalance in ML algorithms applied to MI. METHODS: We first selected five classes from the Image Retrieval in Medical Applications (IRMA) dataset, performed multiclass classification using the random forest model (RFM), and then performed binary classification using convolutional neural network (CNN) on a chest X-ray dataset. An imbalanced class was created in the training set by varying the number of images in that class. Methods tested to mitigate class imbalance included oversampling, undersampling, and changing class weights of the RFM. Model performance was assessed by overall classification accuracy, overall F1 score, and specificity, recall, and precision of the imbalanced class. RESULTS: A close-to-balanced training set resulted in the best model performance, and a large imbalance with overrepresentation was more detrimental to model performance than underrepresentation. Oversampling and undersampling methods were both effective in mitigating class imbalance, and efficacy of oversampling techniques was class specific. CONCLUSION: This study systematically demonstrates the effect of class imbalance on two public X-ray datasets on RFM and CNN, making these findings widely applicable as a reference. Furthermore, the methods employed here can guide researchers in assessing and addressing the effects of class imbalance, while considering the data-specific characteristics to optimize imbalance mitigating methods.