Long short-term memory model - A deep learning approach for medical data with irregularity in cancer predication with tumor markers.

BACKGROUND: Machine learning (ML) has emerged as a superior method for the analysis of large datasets. Application of ML is often hindered by incompleteness of the data which is particularly evident when approaching disease screening data due to varied testing regimens across medical institutions. Here we explored the utility of multiple ML algorithms to predict cancer risk when trained using a large but incomplete real-world dataset of tumor marker (TM) values. METHODS: TM screening data were collected from a large asymptomatic cohort (n = 163,174) at two independent medical centers. The cohort included 785 individuals who were subsequently diagnosed with cancer. Data included levels of up to eight TMs, but for most subjects, only a subset of the biomarkers were tested. In some instances, TM values were available at multiple time points, but intervals between tests varied widely. The data were used to train and test various machine learning models to evaluate their robustness for predicting cancer risk. Multiple methods for data imputation were explored and models were developed for both single time-point as well as time-series data. RESULTS: The ML algorithm, long short-term memory (LSTM), demonstrated superiority over other models for dealing with irregular medical data. A cancer risk prediction tool was trained and validated for a single time-point test of a TM panel including up to four biomarkers (AUROC = 0.831, 95% CI: 0.827-0.835) which outperformed a single threshold method using the same biomarkers. A second model relying on time series data of up to four time-points for 5 TMs had an AUROC of 0.931. CONCLUSIONS: A cancer risk prediction tool was developed by training a LSTM model using a large but incomplete real-world dataset of TM values. The LSTM model was best able to handle irregular data compared to other ML models. The use of time-series TM data can further improve the predictive performance of LSTM models even when the intervals between tests vary widely. These risk prediction tools are useful to direct subjects to further screening sooner, resulting in earlier detection of occult tumors.

Ensemble machine learning to evaluate the in vivo acute oral toxicity and in vitro human acetylcholinesterase inhibitory activity of organophosphates.

Organophosphates (OPs) are hazardous chemicals widely used in industry and agriculture. Distribution of their residues in nature causes serious risks to humans, animals, and plants. To reduce hazards from OPs, quantitative structure-activity relationship (QSAR) models for predicting their acute oral toxicity in rats and mice and inhibition constants concerning human acetylcholinesterase were developed according to the bioactivity data of 456 unique OPs. Based on robust, two-dimensional molecular descriptors and quantum chemical descriptors, which accurately reflect OP electronic structures and reactivities, the influences of eight machine-learning algorithms on the prediction performance of the QSAR models were explored, and consensus QSAR models were constructed. Several strict model validation indices and the results of applicability domain evaluations show that the established consensus QSAR models exhibit good robustness, practical prediction abilities, and wide application scopes. Poor correlation was observed between acute oral toxicity at the mammalian level and the inhibition constants at the molecular level, indicating that the acute toxicity of OPs cannot be evaluated only by the experimental data of enzyme inhibitory activity, their toxicokinetic characteristics must also be considered. The constructed QSAR models described herein provide rapid, theoretical assessment of the bioactivity of unstudied or unknown OPs, as well as guidance for making decisions regarding their regulation.

Shortening a Patient Experiences Survey for Medical Homes.

The Consumer Assessment of Healthcare Providers and Systems-Patient-Centered Medical Home (CAHPS PCMH) Survey assesses patient experiences reflecting domains of care related to general patient experience (access to care, communication with providers, office staff interaction, provider rating) and PCMH-specific aspects of patient care (comprehensiveness of care, self-management support, shared decision making). The current work compares psychometric properties of the current survey and a proposed shortened version of the survey (from 52 to 26 adult survey items, from 66 to 31 child survey items). The revisions were based on initial psychometric analysis and stakeholder input regarding survey length concerns. A total of 268 practices voluntarily submitted adult surveys and 58 submitted child survey data to the National Committee for Quality Assurance in 2013. Mean unadjusted scores, practice-level item and composite reliability, and item-to-scale correlations were calculated. Results show that the shorter adult survey has lower reliability, but still it still meets general definitions of a sound survey for the adult version, and resulted in few changes to mean scores. The impact was more problematic for the pediatric version. Further testing is needed to investigate approaches to improving survey response and the relevance of survey items in informing quality improvement.

Elevated cortisol in older adults with generalized anxiety disorder is reduced by treatment: a placebo-controlled evaluation of escitalopram.

BACKGROUND: Generalized anxiety disorder (GAD) is a common disorder in older adults, which has been linked to hyperactivity of the hypothalamic-pituitary-adrenal (HPA) axis in this age group. The authors examined whether treatment of GAD in older adults with a selective serotonin reuptake inhibitor (SSRI) corrects this HPA axis hyperactivity. METHODS: The authors examined adults aged 60 years and older with GAD in a 12-week randomized controlled trial comparing the SSRI escitalopram with placebo. The authors collected salivary cortisol at six daily time points for 2 consecutive days to assess peak and total (area under the curve) cortisol, both at baseline and posttreatment. RESULTS: Compared with placebo-treated patients, SSRI-treated patients had a significantly greater reduction in both peak and total cortisol. This reduction in cortisol was limited to patients with elevated (above the median) baseline cortisol, in whom SSRI-treated patients showed substantially greater reduction in cortisol than did placebo-treated patients. Reductions in cortisol were associated with improvements in anxiety. Additionally, genetic variability at the serotonin transporter promoter predicted cortisol changes. CONCLUSIONS: SSRI treatment of GAD in older adults reduces HPA axis hyperactivity. Further research should determine whether these treatment-attributable changes are sustained and beneficial.

Relation of serotonin transporter genetic variation to efficacy of escitalopram for generalized anxiety disorder in older adults.

OBJECTIVE: Generalized anxiety disorder (GAD) is common in older adults and can be treated with selective serotonin reuptake inhibitors (SSRIs). Genetic variation in the serotonin transporter gene promoter region is posited to be associated with SSRI efficacy: 2 polymorphisms (5-HTTLPR S/L and rs25531 g/a) form a haplotype with the La combination having higher transcription activity than other haplotypes. We hypothesized that GAD patients with no La haplotypes (La⁻) have lower SSRI treatment efficacy than those with 1 to 2 La haplotypes (La+). METHOD: The study enrolled subjects aged 60 years or older with a principal diagnosis of GAD into a 12-week, randomized trial of escitalopram versus placebo. One hundred fifty subjects were genotyped for the serotonin transporter promoter region haplotype and were divided into La⁻ and La+ genotype groups; the primary analyses were done in European Americans only (n = 125; n = 59 with escitalopram and n = 66 with placebo). RESULTS: Escitalopram had no efficacy in the La⁻ group versus moderate efficacy in the La+ group. This genetic moderation of SSRI efficacy was due to a higher placebo response in La⁻ subjects, compared with La+ subjects. Drug concentration did not affect the genetic results. Exploratory analyses suggest that La⁻ subjects had greater variability of anxiety symptoms unrelated to treatment. CONCLUSIONS: The serotonin transporter promoter haplotype is associated with variability in SSRI efficacy for late-life GAD. The variability may result from a genetic effect on anxiety symptom variability unrelated to treatment, rather than a pharmacodynamic effect that has been previously assumed. Further research is needed to understand the pharmacogenetic mechanism of this haplotype.

Child Welfare Workers' Connectivity to Resources and Youth's Receipt of Services.

Youth involved in the child welfare system are at high risk for mental illness, substance abuse, and other behavioral health issues, which child welfare workers are expected to address through referrals. Child welfare workers (N=27) who participated in Project IMPROVE (Intervention for Multisector Provider Enhancement) reported on services they provided to youth (N=307) in their caseloads. Using survey and administrative data, this paper examines workers' service actions on behalf of youth. Results were consistent with the Gateway Provider Model and showed that youth received help from a greater variety of service sectors when their workers were able to identify behavioral health problems, and were familiar with and connected to other providers in the community. Improving service delivery to youth in child welfare may be accomplished by training workers in the signs and symptoms of behavioral health problems and familiarizing them with providers in the community.

Towards DSM-V: considering other withdrawal-like symptoms of pathological gambling disorder.

Despite clinical reports of other withdrawal-like symptoms, the DSM-IV considers only restlessness/irritability as a withdrawal-like criterion comprising pathological gambling disorder (PGD). We explored whether this criterion should be broadened to include other gambling withdrawal-like symptoms.Community-recruited adult gamblers (n = 312) participated in telephone interviews about gambling and related behaviors as a part of a larger psychometric study. Frequency and chi-square analyses described the association of gambling withdrawal-like symptoms by gambling disorder status. Multinomial forward selection logistic regression obtained a multivariate model describing the simultaneous relationship between these symptoms and gambling disorder status.One-quarter of the sample experienced the DSM-IV PGD criterion of restlessness/irritability. However, 41% experienced additional gambling withdrawal-like symptoms when attempting to quit or control gambling. A model including restlessness/irritability and three additional non-DSM-IV withdrawal-like symptoms (i.e. feelings of anger, guilt, and disappointment) is a stronger model of gambling disorder (chi(2) = 217.488; df = 8, p < 0.0001; R(2) = 0.5428; p < 0.0001) than restlessness/irritability alone (chi(2) = 151.278; df = 2, p < 0.0001; R(2) = 0.4133). The overlap of gambling withdrawal-like symptoms with substance use withdrawal (11%) and depressive symptoms (34%) failed to fully account for these associations with gambling disorder status.Future PGD conceptualization and potential criteria revisions for DSM-V may warrant a broader inclusion of gambling withdrawal-like symptoms.

Escitalopram for older adults with generalized anxiety disorder: a randomized controlled trial.

CONTEXT: Generalized anxiety disorder (GAD) is one of the most common psychiatric disorders in older adults; however, few data exist to guide clinicians in efficacious and safe treatment. Selective serotonin reuptake inhibitors (SSRIs) are efficacious for younger adults with GAD, but benefits and risks may be different in older adults. OBJECTIVE: To examine the efficacy, safety, and tolerability of the SSRI escitalopram in older adults with GAD. DESIGN, SETTING, AND PARTICIPANTS: A randomized controlled trial in primary care practices and related specialty clinics in Pittsburgh, Pennsylvania, of 177 participants aged 60 years or older with a principal diagnosis of GAD randomized to receive either escitalopram or placebo and conducted between January 2005 and January 2008. INTERVENTIONS: Twelve weeks of 10 to 20 mg/d of escitalopram (n = 85) or matching placebo (n = 92). MAIN OUTCOME MEASURES: Cumulative response defined by Clinical Global Impressions-Improvement score of much or very much improved; time to response; and anxiety and role functioning changes measured by the Clinical Global Impressions-Improvement scale, Hamilton Anxiety Rating Scale, Penn State Worry Questionnaire, Late-Life Function and Disability Instrument activity limitations subscale, and the role-emotional impairment and social function subscales of the Medical Outcome Survey 36-item Short Form. RESULTS: In the primary analytic strategy in which participants (n = 33) were censored at the time of dropout, mean cumulative response rate for escitalopram was 69% (95% confidence interval [CI], 58%-80%) vs 51% (95% CI, 40%-62%) for placebo (P = .03). A conservative intention-to-treat analysis showed no difference in mean cumulative response rate between escitalopram and placebo (57%; 95% CI, 46%-67%; vs 45%; 95% CI, 35%-55%; P = .11). Participants treated with escitalopram showed greater improvement than with placebo in anxiety symptoms and role functioning (Clinical Global Impressions-Improvement scale: effect size, 0.93; 95% CI, 0.50-1.36; P < .001; Penn State Worry Questionnaire: 0.30; 95% CI, 0.23-0.48; P = .01; activity limitations: 0.32; 95% CI, 0.01-0.63; P = .04; and the role-emotional impairment and social function: 0.96; 95% CI, 0.03-1.90; P = .04). Adverse effects of escitalopram (P < .05 vs placebo) were fatigue or somnolence (35 patients [41.1%]), sleep disturbance (12 [14.1%]), and urinary symptoms (8 [9.4%]). CONCLUSIONS: Older adults with GAD randomized to escitalopram had a higher cumulative response rate for improvement vs placebo over 12 weeks; however, response rates were not significantly different using an intention-to-treat analysis. Further study is required to assess efficacy and safety over longer treatment durations. TRIAL REGISTRATION: clinicaltrials.gov Identifier: NCT00105586.

Health insurance discontinuities among adolescents leaving foster care.

PURPOSE: To determine whether adolescents who lose Medicaid entitlements when they leave foster care are subsequently able to secure employer-sponsored or student health insurance coverage. METHODS: This was a 2-year follow-up study of a cohort of 404 adolescents leaving foster care in eight counties in a midwestern state. We conducted survival analysis to study predictors of time to first insurance loss, and logistic regression analysis to determine factors associated with insurance reacquisition, among these youth. RESULTS: A total of 206 adolescents (51%) left foster care during follow up, of whom 138 (67%) lost health insurance coverage within a mean of 3 months of leaving foster care. Those who regained coverage (34; 17% of those leaving foster care) did so after a mean period of 8 months spent without insurance. Hazard of insurance loss was lower for employed adolescents (HR=.5; 95% CI=.4-.7; p < .0001), but only half of all adolescents leaving foster care reported being able to secure employment. Student health insurance did not reduce hazard of insurance loss. Boys had significantly lower odds of regaining insurance compared with girls (OR=.2, SE=.5, p=.003). CONCLUSIONS: Most youth leaving the child welfare system seem unable to transition to other forms of health insurance coverage. Even those that do acquire coverage, do so after an inordinate period of time. Enacting existing extensions of Medicaid coverage until age 21 for foster care youth is necessary to provide the resources to address the considerable health and mental health needs among these youth.