Incorporating Interaction Terms Research Articles

Changes in Case-Mix and Health Outcomes of Medicare Fee-for-Service Beneficiaries and Managed Care Enrollees During the Years 1992-2011.

Recent studies suggest that managed care enrollees (MCEs) and fee-for-service beneficiaries (FFSBs) have become similar in case-mix over time; but comparisons of health outcomes have yielded mixed results. To examine changes in differentials between MCEs and FFSBs both in case-mix and health outcomes over time. Temporal study of the linked Health and Retirement Study (HRS) and Medicare data, comparing case-mix and health outcomes between MCEs and FFSBs across 3 time periods: 1992-1998, 1999-2004, and 2005-2011. We used multivariable analysis, stratified by, and pooled across the study periods. The unit of analysis was the person-wave (n=167,204). HRS participants who were also enrolled in Medicare. Outcome measures included self-reported fair/poor health, 2-year self-rated worse health, and 2-year mortality. Our main covariate was a composite measure of multimorbidity (MM), MM0-MM3, defined as the co-occurrence of chronic conditions, functional limitations, and/or geriatric syndromes. The case-mix differential between MCEs and FFSBs persisted over time. Results from multivariable models on the pooled data and incorporating interaction terms between managed care status and study period indicated that MCEs and FFSBs were as likely to die within 2 years from the HRS interview (P=0.073). This likelihood remained unchanged across the study periods. However, MCEs were more likely than FFSBs to report fair/poor health in the third study period (change in probability for the interaction term: 0.024, P=0.008), but less likely to rate their health worse in the last 2 years, albeit at borderline significance (change in probability: -0.021, P=0.059). Despite the persistence of selection bias, the differential in self-reported fair/poor status between MCEs and FFSBs seems to be closing over time.

Endotoxin and gender modify lung function recovery after occupational organic dust exposure: a 30-year study

ObjectivesThe purpose of this study is to determine the trajectory of lung function change after exposure cessation to occupational organic dust exposure, and to identify factors that modify improvement.MethodsThe Shanghai...

Recapture or precapture? Fallibility of standard capture-recapture methods in the presence of referrals between sources.

Capture-recapture methods, largely developed in ecology, are now commonly used in epidemiology to adjust for incomplete registries and to estimate the size of difficult-to-reach populations such as problem drug users. Overlapping lists of individuals in the target population, taken from administrative data sources, are considered analogous to overlapping “captures” of animals. Log-linear models, incorporating interaction terms to account for dependencies between sources, are used to predict the number of unobserved individuals and, hence, the total population size. A standard assumption to ensure parameter identifiability is that the highest-order interaction term is 0. We demonstrate that, when individuals are referred directly between sources, this assumption will often be violated, and the standard modeling approach may lead to seriously biased estimates. We refer to such individuals as having been “precaptured,” rather than truly recaptured. Although sometimes an alternative identifiable log-linear model could accommodate the referral structure, this will not always be the case. Further, multiple plausible models may fit the data equally well but provide widely varying estimates of the population size. We demonstrate an alternative modeling approach, based on an interpretable parameterization and driven by careful consideration of the relationships between the sources, and we make recommendations for capture-recapture in practice.

Ethnic variation in the association between sleep and body mass among US adolescents

ObjectiveWe investigate whether differences in sleep duration help explain ethnic disparities in body mass index (BMI) among U.S. adolescents. We also evaluate the functional form of the association between sleep duration and BMI, and investigate whether this association varies by sex and ethnicity.Participants and MethodsWe analyzed restricted-use data from the first 2 waves of the National Longitudinal Study of Adolescent Health (n=30 133) to evaluate linear and quadratic associations between sleep duration and BMI. Through a series of models that incorporated interaction terms between sex, ethnicity and sleep duration, we also assessed whether (1) sleep duration mediates associations between ethnicity and BMI, and (2) associations between sleep duration and BMI differ for girls and boys from different ethnic groups.ResultsA linear association between sleep duration and BMI best fits the data in this large sample of U.S. adolescents. We find no evidence that sleep duration contributes substantially to ethnic disparities in BMI. However, we detect significant differences in the association between sleep duration and BMI by sex and ethnicity. Sleep duration is negatively associated with BMI among white, Hispanic and Asian boys, positively associated with BMI among black girls, and not related to BMI among black boys or girls from white, Hispanic or Asian ethnic groups.ConclusionDespite significant associations between sleep duration and BMI for certain groups of adolescents, we find no evidence that ethnic differences in sleep duration exacerbate ethnic disparities in BMI. Future research should explore mechanisms that underlie ethnic differences in the association between sleep and BMI.

Suitable habitat model for walleye (Sander vitreus) in Lake Erie: Implications for inter-jurisdictional harvest quota allocations

Models predicting habitat distributions can give insight into species–habitat requirements and anticipate how populations respond to environmental change. Despite the economic and ecological importance of walleye (Sander vitreus) in Lake Erie, no preferred-habitat model exists and the spatial extent of suitable habitat is poorly understood. Empirical species-habitat models for three groups of walleye (juveniles, adults, and all walleye) was developed using records from a long term gill net data base (21years). We examined the degree to which habitat suitability varies with vertical stratum for each group and whether the new model yields different estimates of available walleye habitat when compared to the current depth-based approach. Walleye occurrence in gill nets was positively related to water temperature, negatively related to water depth and water clarity, and unrelated to dissolved oxygen concentration. A model that incorporated interaction terms among the independent variables performed better than the linear, quadratic, and cubic generalized linear models (GLMs) for all three groups. Our results indicate that the extent of suitable habitat varies spatially in Lake Erie and is greatest in the West basin. Weighted Habitat Suitability Areas (WHSA), a combination of habitat quality and quantity, differed significantly among basins and vertical strata in Lake Erie. The current quota allocation strategy for Lake Erie walleye is based on the proportional amount of preferred habitat by jurisdiction. However, the current depth-based definition of preferred habitat may not be an adequate representation of walleye suitable habitat shared by each jurisdiction.

OP69 The use of NHS Routine Data to Analyse The Effects of Healthcare Interventions

BackgroundIn 2007, the strategy ‘Healthcare for London’ planned the establishment of polyclinics in each PCT to help improve the primary care infrastructure. These polyclinics serve as a hub for a...

Depression During Pregnancy Among Young Couples: The Effect of Personal and Partner Experiences of Stressors and the Buffering Effects of Social Relationships

Study ObjectiveTo assess the relationship between personal and romantic partner's experiences of stressful life events and depression during pregnancy, and the social moderators of this relationship, among 296 young couples with low incomes from urban areas. Participants and SettingWe recruited couples who were expecting a baby from four ob/gyn and ultrasound clinics in southern Connecticut; women were ages 14–21 and male partners were 14+. Design and Outcome MeasuresWe analyzed self-reports of stressful events in the previous six months, depression in the past week and current interpersonal social supports. To determine the influence of personal and partner experiences of stressful events on depression, we used multilevel dyadic models and incorporated interaction terms. We also used this model to determine whether social support, family functioning and relationship satisfaction moderated the association between stressful events and depression. ResultsExperiences of stressful life events were common; 91.2% of couples had at least one member report an event. Money, employment problems, and moving were the most common events. Personal experiences of stressful life events had the strongest association with depression among men and women; although partner experiences of stressful life events were also significantly associated with depression among women. Social support, family functioning, and romantic relationship satisfaction significantly buffered the association between personal and partner stressful events and depression. ConclusionInterventions that improve relationships, support systems, and family functioning may reduce the negative impact of stressors, experienced both personally and by a romantic partner, on the emotional well-being of young expectant parents.

1735 DIFFERENTIAL ADOPTION OF LASER PROSTATECTOMY FOR TREATMENT OF BENIGN PROSTATIC HYPERPLASIA

You have accessJournal of UrologyBenign Prostatic Hyperplasia: Epidemiology and Natural History/Evaluation and Markers1 Apr 20121735 DIFFERENTIAL ADOPTION OF LASER PROSTATECTOMY FOR TREATMENT OF BENIGN PROSTATIC HYPERPLASIA Florian R. Schroeck, John M. Hollingsworth, Samuel R. Kaufman, Rodney L. Dunn, Brent K. Hollenbeck, and John T. Wei Florian R. SchroeckFlorian R. Schroeck Ann Arbor, MI More articles by this author , John M. HollingsworthJohn M. Hollingsworth Ann Arbor, MI More articles by this author , Samuel R. KaufmanSamuel R. Kaufman Ann Arbor, MI More articles by this author , Rodney L. DunnRodney L. Dunn Ann Arbor, MI More articles by this author , Brent K. HollenbeckBrent K. Hollenbeck Ann Arbor, MI More articles by this author , and John T. WeiJohn T. Wei Ann Arbor, MI More articles by this author View All Author Informationhttps://doi.org/10.1016/j.juro.2012.02.1696AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail INTRODUCTION AND OBJECTIVES Laser prostatectomy is thought to be less invasive than transurethral resection of the prostate (TURP). However, adopting this new technology requires upfront investments, which may lead to differential adoption in socioeconomically privileged versus disadvantaged regions. We therefore examined whether differences in regional socioeconomic status are related to adoption of laser prostatectomy. METHODS Using the State Ambulatory Surgery and Inpatient Databases for Florida from 2001-2009, we calculated the proportion of hospital service areas (n=114) offering laser prostatectomy or TURP. These data were linked to the Area Resource File to obtain regional characteristics, including a summary score of socioeconomic status. We then used random-effects logistic regression models to identify regional factors associated with offering laser prostatectomy or TURP. We examined the association of socioeconomic score with time trends by incorporating interaction terms into our models. RESULTS Overall, the proportion of regions offering laser prostatectomy increased from 34% to 82% (p<0.01), while the proportion of regions offering TURP decreased from 96% to 86% (p<0.01). Regions were less likely to offer either surgery if they had a higher percentage of African Americans (p<0.01), higher poverty rates (p<0.01), lower household incomes (p≤0.03), less urban population (p<0.01), less educated population (p<0.01), and fewer urologists (p≤0.01). Regions in the highest versus lowest socioeconomic tertile adopted laser prostatectomy more rapidly (p<0.01 for interaction, Figure). There was no association of socioeconomic score with time trends for TURP (p=0.48 for interaction, Figure). CONCLUSIONS Regional availability of laser prostatectomy increased significantly over the last decade. However, socioeconomically disadvantaged regions were less likely to adopt laser prostatectomy, which may lead to disparities in access to this new technology. © 2012 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 187Issue 4SApril 2012Page: e699 Advertisement Copyright & Permissions© 2012 by American Urological Association Education and Research, Inc.MetricsAuthor Information Florian R. Schroeck Ann Arbor, MI More articles by this author John M. Hollingsworth Ann Arbor, MI More articles by this author Samuel R. Kaufman Ann Arbor, MI More articles by this author Rodney L. Dunn Ann Arbor, MI More articles by this author Brent K. Hollenbeck Ann Arbor, MI More articles by this author John T. Wei Ann Arbor, MI More articles by this author Expand All Advertisement Advertisement PDF downloadLoading ...

Lrrk2 S1647T and BDNF V66M interact with environmental factors to increase risk of Parkinson’s disease

Parkinson’s disease (PD) is a complex neurodegenerative disorder contributed by both environmental and genetic factors. The inconsistent findings in genetic association studies may be due to unrecognized interactions with other genetic or environmental factors. Therefore, we assessed the combined effects of genetic variants of three candidate genes of familial PD and environmental exposure on risk of PD in Taiwanese population. A total of 744 study subjects, 453 PD cases and 291 ethnicity-matched controls, were included. The genetic variants on the PINK1, BDNF, and LRRK2 genes were sequenced accordingly. We used a regression approach based on a generalized linear model to evaluate single-locus genotype effects and detection of gene–environment interaction by incorporating interaction terms in the model. We found a significant difference of LRRK2 G2385R and R1628P between PD patients and controls, which confirmed our previous findings. A logistic regression model which included gene–environment interactions was applied. Notably, we identified the variant of LRRK2 T4939A (S1647T) (TT, OR = −0.36, p = 0.03) is associated with increased PD risk, after considering the interaction effects with environmental factors in the model. Additionally, two novel interactions were detected: pesticide exposure with BDNF (OR = −0.85, p = 0.01) and to a lesser extent, with PINK1 (OR = 1.99, p = 0.07). Our findings reinforced the importance that PD risk is modulated by both genetic and environmental exposures. LRRK2 S1647T may be another risk factor for PD development in our ethnicity while considering the joint interaction effects with environmental factors.

An optimal cooperation in a team of agents subject to partial information

The objective of this work is to design a controller for a team of agents to accomplish cohesive motion with consensus on an agreed upon trajectory in both leaderless and modified leader–follower structures. The agent's dynamical model is considered to be governed by a general linear representation instead of a point-mass model which is commonly used in the literature. A semi-decentralised optimal control strategy is designed based on minimisation of individual cost functions over a finite horizon using local information. Decentralisation is achieved through incorporating interaction terms in the team members model. Minimisation of our proposed cost function results in a modified version of the consensus algorithm. Corresponding to the leader–follower structure, the utilisation of a corrective feedback from followers to the leader provides potential advantages in terms of improving robustness of the team to faults in individual agents. Simulation results are presented to demonstrate the effectiveness and capabilities of our proposed method in achieving user predefined requirements and specifications.

Interaction of gender, hypertension, and the angiotensinogen gene haplotypes on the risk of coronary artery disease in a large angiographic cohort

There is increasing evidence suggesting the importance of evaluating gene–environment interactions in the genetic study of coronary artery disease (CAD). We investigated the association of multiple single nucleotide polymorphisms in the angiotensinogen (AGT) gene with CAD, considering the interaction between the genetic and non-genetic factors, using a larger and ethnically homogeneous angiographic cohort. A total of 1254 consecutive patients who underwent cardiac catheterization (735 with CAD and 519 without) were recruited. T174M (rs4762), M235T (rs699), G-6A, A-20C, G-152A, and G-217A polymorphisms of the AGT gene were genotyped. We used a regression approach based on a generalized linear model to evaluate haplotype effects defined by the multilocus data and detection of gene–environment interaction by incorporating interaction terms in the model. We found significant differences in global AGT gene haplotype profile between patients with and without CAD (the global score statistic=25.411, P=0.008). Significant interactions between AGT gene haplotypes, gender and hypertension were detected. We also used haplotype counting to directly estimate the odds ratio of each AGT gene haplotype, and found that the effects of haplotypes were markedly different in subgroups defined by gender and hypertension, providing strong evidence of gene–environment interaction. Female gender synergistically enhances (or male gender reverses) the effects of AGT gene haplotypes on the risk of CAD in the presence of hypertension. In conclusion, the effect of AGT gene haplotypes on the risk of CAD was significantly increased in women with hypertension, which highlights the importance of evaluating gene–environment interactions in the genetic study of CAD.

Renin-Angiotensin System Gene Polymorphisms and Atrial Fibrillation: A Regression Approach for the Detection of Gene-Gene Interactions in a Large Hospitalized Population

Objectives: To test the association between renin-angiotensin system gene variants and atrial fibrillation (AF) using a regression approach. Methods: A total of 1,236 consecutive patients (227 with AF and 1,009 with normal sinus rhythm as controls) were recruited. Angiotensin-converting enzyme (ACE) gene I/D polymorphism; T174M, M235T, G–6A, A–20C, G–152A and G–217A polymorphisms of the angiotensinogen (AGT) gene, and A1166C polymorphism of the angiotensin II type I receptor (AT1R) gene were genotyped. We used a regression approach based on a generalized linear model to evaluate haplotype effects and to detect gene-gene interactions by incorporating interaction terms in the model. Results: In single-locus analyses, no locus was associated with AF. After adjustment for AF risk factors, we found significant differences in the global AGT gene haplotype profile (the global score statistic = 30.364, p = 0.001) and individual haplotype frequencies between AF patients and controls. Furthermore, significant 2-way gene-gene interactions between ACE I/D polymorphism and AGT gene haplotypes and between AT1R A1166C polymorphism and AGT gene haplotypes, and 3-way interaction between ACE I/D, AT1R A1166C and AGT gene haplotypes were detected. Conclusions: These results are compatible with the concept of multilocus and multigene effects in determining the risk of complex diseases such as AF, which would be missed with conventional single-locus approaches.

Bayesian networks with a logistic regression model for the conditional probabilities

Logistic regression techniques can be used to restrict the conditional probabilities of a Bayesian network for discrete variables. More specifically, each variable of the network can be modeled through a logistic regression model, in which the parents of the variable define the covariates. When all main effects and interactions between the parent variables are incorporated as covariates, the conditional probabilities are estimated without restrictions, as in a traditional Bayesian network. By incorporating interaction terms up to a specific order only, the number of parameters can be drastically reduced. Furthermore, ordered logistic regression can be used when the categories of a variable are ordered, resulting in even more parsimonious models. Parameters are estimated by a modified junction tree algorithm. The approach is illustrated with the Alarm network.

Chemotherapy use for hormone receptor-positive, node-negative breast cancer

11004 Background: High quality evidence has shown that adjuvant chemotherapy (chemo) improves survival for hormone receptor- positive, node negative breast cancer measuring &gt;1 cm (H+/N-/T&gt;1cm BC). Since 2002 National Comprehensive Cancer Network (NCCN) guidelines have recommended chemo for this group. Because the benefit from chemo in this situation is modest there is debate regarding which women, if any, can forego chemo. We describe trends in chemo use and identify factors associated with not receiving chemo for H+/N-/T&gt;1cm BC. Methods: We identified women less than 70 years old with H+/N-/T&gt;1cm BC diagnosed 1997–2004 and treated at 8 NCCN institutions. We analyzed the frequency of chemo use on a yearly basis (chi square). A multivariable logistic regression model assessed the relationship between likelihood of not receiving chemo and year of diagnosis, institution, tumor size, histologic characteristics and socio- demographic variables. We incorporated interaction terms to explore how chemo use varied over time and between institutions. Results: Of 3190 women with H+/N-/T&gt;1cm BC, 56% received chemo. Chemo use was less common for 1.1–2 cm than for &gt;2 cm tumors (48 vs. 83%; p&lt;.01). There was no significant change in chemo use over time, but there was significant variability among institutions (range 47–65%; p&lt;.01). On multivariable analysis, predictors independently associated with not receiving chemo included lower histologic/nuclear grade, non-ductal/lobular histology, absence of lymphovascular invasion or HER2 overexpression, and older age (p&lt;.01 for all). The interaction between institution and year of diagnosis was also significant (p&lt;.01). Over the 2 years following the NCCN guideline’s definitive recommendation for chemo in 2002, the relative change in institutional chemo use ranged from a 25% decrease to a 28% increase. Conclusions: Many women did not receive chemo despite clinical trial evidence. Providers considered molecular features, tumor size and patient age when deciding who should receive chemo. While overall trends in chemo use did not change, use appeared to increase at some institutions and decreased at others after the 2002 guideline recommendation for chemo. This suggests a lack of consensus at NCCN institutions regarding when to administer chemo for H+/N-/T&gt;1cm BC. No significant financial relationships to disclose.