Indoor allergen exposure and its association to upper respiratory infections and pulmonary outcomes among children with asthma.

BACKGROUND: Certain environmental allergen exposures are more common in disadvantaged communities and may contribute to differences in susceptibility to upper respiratory infections (URIs). OBJECTIVES: We examined associations between indoor allergens and: (1) URI; (2) URI + cold symptoms; (3) URI + cold symptoms + pulmonary eosinophilic inflammation (fraction of exhaled nitric oxide ≥20 ppb); and (4) URI + cold symptoms + reduced lung function (percent predicted forced expiratory volume in 1 second of <80%). METHODS: We used data from the Environmental Control as Add-on Therapy for Childhood Asthma (ECATCh) study. Allergen concentrations were measured in air (mouse) and settled dust (mouse, cockroach, dog, and cat). URI was determined by testing nasal mucus for upper respiratory viruses. We evaluated associations between allergen concentrations and URI-associated outcomes accounting for age, sex, study month, season, health insurance, and household size. RESULTS: Ninety participants (92% Black, 92% public insurance) with 192 observations were included; 52 (27%) of observations were positive for URI. A doubling in cockroach allergen concentration increased the odds of a URI with cold symptoms by 18% (odds ratio [OR] = 1.18, 95% confidence interval [CI], 0.99-1.40), the odds of a URI + cold symptoms + pulmonary eosinophilic inflammation by 31% (OR = 1.31, 95% CI, 1.10-1.57), and the odds of a URI + cold symptoms + reduced lung function by 45% (OR = 1.45, 95% CI, 1.13-1.85). Mouse allergen concentrations were positively associated with all outcomes. Associations were suggestively stronger among children sensitized to pest allergens. CONCLUSIONS: Cockroach and mouse, but not dog or cat, allergen exposure may predispose children with asthma to URIs with colds and lower respiratory outcomes.

Racial and Ethnic Identity and Vulnerability to Upper Respiratory Viral Infections Among US Children.

BACKGROUND: It is unclear whether there are racial/ethnic disparities in the risk of upper respiratory viral infection acquisition and/or lower respiratory manifestations. METHODS: We studied all children and children with asthma aged 6 to 17 years in the National Health and Nutrition Examination Survey (2007-2012) to evaluate (1) the association between race/ethnicity and upper respiratory infection (URI) and (2) whether race/ethnicity is a risk factor for URI-associated pulmonary eosinophilic inflammation or decreased lung function. RESULTS: Children who identified as Black (adjusted odds ratio [aOR], 1.38; 95% CI, 1.10-1.75) and Mexican American (aOR, 1.50; 95% CI, 1.16-1.94) were more likely to report a URI than those who identified as White. Among those with asthma, Black children were more than twice as likely to report a URI than White children (aOR, 2.28; 95% CI, 1.31-3.95). Associations between URI and pulmonary eosinophilic inflammation or lung function did not differ by race/ethnicity. CONCLUSIONS: Findings suggest that there may be racial and ethnic disparities in acquiring a URI but not in the severity of infection. Given that upper respiratory viral infection is tightly linked to asthma exacerbations in children, differences in the risk of infection among children with asthma may contribute to disparities in asthma exacerbations.

Comparing diagnostic criteria for posttraumatic stress disorder in a diverse sample of trauma-exposed youth.

Divergent conceptualization of posttraumatic stress disorder (PTSD) within the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5) and International Statistical Classification of Diseases and Related Health Problems (11th ed..; ICD-11) significantly confounds both research and practice. Using a diverse sample of trauma-exposed youth (N = 1,542, age range: 8-20 years), we compared these two diagnostic approaches along with an expanded version of the ICD-11 PTSD criteria that included three additional reexperiencing symptoms (ICD-11+). Within the sample, PTSD was more prevalent using the DSM-5 criteria (25.7%) compared to the ICD-11 criteria (16.0%), with moderate agreement between these diagnostic systems, κ = .57. The inclusion of additional reexperiencing symptoms (i.e., ICD-11+) reduced this discrepancy in prevalence (24.7%) and increased concordance with DSM-5 criteria, κ = .73. All three PTSD classification systems exhibited similar comorbidity rates with major depressive episode (MDE) or generalized anxiety disorder (GAD; 78.0%-83.6%). Most youths who met the DSM-5 PTSD criteria also met the criteria for ICD-11 PTSD, MDE, or GAD (88.4%), and this proportion increased when applying the ICD-11+ criteria (95.5%). Symptom-level analyses identified reexperiencing/intrusions and negative alterations in cognition and mood symptoms as primary sources of discrepancy between the DSM-5 and ICD-11 PTSD diagnostic systems. Overall, these results challenge assertions that nonspecific distress and diagnostically overlapping symptoms within DSM-5 PTSD inflate comorbidity with depressive and anxiety disorders. Further, they support the argument that the DSM-5 PTSD criteria can be refined and simplified without reducing the overall prevalence of psychiatric diagnoses in youth.

Do upper respiratory viruses contribute to racial and ethnic disparities in emergency department visits for asthma?

BACKGROUND: There are marked disparities in asthma-related emergency department (ED) visit rates among children by race and ethnicity. Following the implementation of coronavirus disease 2019 (COVID-19) prevention measures, asthma-related ED visits rates declined substantially. The decline has been attributed to the reduced circulation of upper respiratory viruses, a common trigger of asthma exacerbations in children. OBJECTIVES: To better understand the contribution of respiratory viruses to racial and ethnic disparities in ED visit rates, we investigated whether the reduction in ED visit rates affected Black, Latinx, and White children with asthma equally. METHODS: Asthma-related ED visits were extracted from electronic medical records at Dell Children's Medical Center in Travis County, Texas. ED visit rates among children with asthma were derived by race/ethnicity. Incidence rate ratios (IRRs) and 95% CIs were estimated by year (2019-2021) and season. RESULTS: In spring 2019, the ED visit IRRs comparing Black children with White children and Latinx children with White children were 6.67 (95% CI = 4.92-9.05) and 2.10 (95% CI = 1.57-2.80), respectively. In spring 2020, when infection prevention measures were implemented, the corresponding IRRs decreased to 1.73 (95% CI = 0.90-3.32) and 0.68 (95% CI = 0.38-1.23), respectively. CONCLUSIONS: The striking reduction of disparities in ED visits suggests that during nonpandemic periods, respiratory viruses contribute to the excess burden of asthma-related ED visits among Black and Latinx children with asthma. Although further investigation is needed to test this hypothesis, our findings raise the question of whether Black and Latinx children with asthma are more vulnerable to upper respiratory viral infections.

Indoor environmental exposures and obstructive lung disease phenotypes among children with asthma living in poor urban neighborhoods.

BACKGROUND: Air trapping is an obstructive phenotype that has been associated with more severe and unstable asthma in children. Air trapping has been defined using pre- and postbronchodilator spirometry. The causes of air trapping are not completely understood. It is possible that environmental exposures could be implicated in air trapping in children with asthma. OBJECTIVE: We investigated the association between indoor exposures and air trapping in urban children with asthma. METHODS: Children with asthma aged 5 to 17 years living in Baltimore and enrolled onto the Environmental Control as Add-on Therapy for Childhood Asthma study were evaluated for air trapping using spirometry. Aeroallergen sensitization was assessed at baseline, and spirometry was performed at 0, 3, and 6 months. Air trapping was defined as an FVC z score of less than -1.64 or a change in FVC with bronchodilation of ≥10% predicted. Logistic normal random effects models were used to evaluate associations of air trapping and indoor exposures. RESULTS: Airborne and bedroom floor mouse allergen concentrations were associated with air trapping but not airflow limitation (odds ratio 1.19, 95% confidence interval 1.02-1.37, P = .02 per 2-fold increase in airborne mouse allergen; odds ratio 1.23, 95% confidence interval 1.07-1.41, P = .003 per 2-fold increase in bedroom floor mouse allergen). Other indoor exposures (cockroach, cat, dog, dust mite, particulate matter, and nicotine) were not associated with air trapping or airflow limitation. CONCLUSION: Mouse allergen exposure, but not other indoor exposure, was associated with air trapping in urban children with asthma.

Marginal modeling of cluster-period means and intraclass correlations in stepped wedge designs with binary outcomes.

Stepped wedge cluster randomized trials (SW-CRTs) with binary outcomes are increasingly used in prevention and implementation studies. Marginal models represent a flexible tool for analyzing SW-CRTs with population-averaged interpretations, but the joint estimation of the mean and intraclass correlation coefficients (ICCs) can be computationally intensive due to large cluster-period sizes. Motivated by the need for marginal inference in SW-CRTs, we propose a simple and efficient estimating equations approach to analyze cluster-period means. We show that the quasi-score for the marginal mean defined from individual-level observations can be reformulated as the quasi-score for the same marginal mean defined from the cluster-period means. An additional mapping of the individual-level ICCs into correlations for the cluster-period means further provides a rigorous justification for the cluster-period approach. The proposed approach addresses a long-recognized computational burden associated with estimating equations defined based on individual-level observations, and enables fast point and interval estimation of the intervention effect and correlations. We further propose matrix-adjusted estimating equations to improve the finite-sample inference for ICCs. By providing a valid approach to estimate ICCs within the class of generalized linear models for correlated binary outcomes, this article operationalizes key recommendations from the CONSORT extension to SW-CRTs, including the reporting of ICCs.

Generalized case-control sampling under generalized linear models.

A generalized case-control (GCC) study, like the standard case-control study, leverages outcome-dependent sampling (ODS) to extend to nonbinary responses. We develop a novel, unifying approach for analyzing GCC study data using the recently developed semiparametric extension of the generalized linear model (GLM), which is substantially more robust to model misspecification than existing approaches based on parametric GLMs. For valid estimation and inference, we use a conditional likelihood to account for the biased sampling design. We describe analysis procedures for estimation and inference for the semiparametric GLM under a conditional likelihood, and we discuss problems with estimation and inference under a conditional likelihood when the response distribution is misspecified. We demonstrate the flexibility of our approach over existing ones through extensive simulation studies, and we apply the methodology to an analysis of the Asset and Health Dynamics Among the Oldest Old study, which motives our research. The proposed approach yields a simple yet versatile solution for handling ODS in a wide variety of possible response distributions and sampling schemes encountered in practice.

Model misspecification and robust analysis for outcome-dependent sampling designs under generalized linear models.

Outcome-dependent sampling (ODS) is a commonly used class of sampling designs to increase estimation efficiency in settings where response information (and possibly adjuster covariates) is available, but the exposure is expensive and/or cumbersome to collect. We focus on ODS within the context of a two-phase study, where in Phase One the response and adjuster covariate information is collected on a large cohort that is representative of the target population, but the expensive exposure variable is not yet measured. In Phase Two, using response information from Phase One, we selectively oversample a subset of informative subjects in whom we collect expensive exposure information. Importantly, the Phase Two sample is no longer representative, and we must use ascertainment-correcting analysis procedures for valid inferences. In this paper, we focus on likelihood-based analysis procedures, particularly a conditional-likelihood approach and a full-likelihood approach. Whereas the full-likelihood retains incomplete Phase One data for subjects not selected into Phase Two, the conditional-likelihood explicitly conditions on Phase Two sample selection (ie, it is a "complete case" analysis procedure). These designs and analysis procedures are typically implemented assuming a known, parametric model for the response distribution. However, in this paper, we approach analyses implementing a novel semi-parametric extension to generalized linear models (SPGLM) to develop likelihood-based procedures with improved robustness to misspecification of distributional assumptions. We specifically focus on the common setting where standard GLM distributional assumptions are not satisfied (eg, misspecified mean/variance relationship). We aim to provide practical design guidance and flexible tools for practitioners in these settings.

Prenatal Weight Change Trajectories and Perinatal Outcomes among Twin Gestations.

OBJECTIVE: Despite an increase in twin pregnancies in recent decades, the Institute of Medicine twin weight gain recommendations remain provisional and provide no guidance for the pattern or timing of weight change. We sought to characterize gestational weight change trajectory patterns and examine associations with birth outcomes in a cohort of twin pregnancies. STUDY DESIGN: Prenatal and delivery records were examined for 320 twin pregnancies from a maternal-fetal medicine practice in Austin, TX 2011-2019. Prenatal weights for those with >1 measured weight in the first trimester and ≥3 prenatal weights were included in analyses. Trajectories were estimated to 32 weeks (mean delivery: 33.7 ± 3.3 weeks) using flexible latent class mixed models with low-rank thin-plate splines. Associations between trajectory classes and infant outcomes were analyzed using multivariable Poisson or linear regression. RESULTS: Weight change from prepregnancy to delivery was 15.4 ± 6.3 kg for people with an underweight body mass index, 15.4 ± 5.8 kg for healthy weight, 14.7 ± 6.9 kg for overweight, and 12.5 ± 6.4 kg for obesity. Three trajectory classes were identified: low (Class 1), moderate (Class 2), or high gain (Class 3). Class 1 (24.7%) maintained weight for 15 weeks and then gained an estimated 6.6 kg at 32 weeks. Class 2 (60.9%) exhibited steady gain with 13.5 kg predicted total gain, and Class 3 (14.4%) showed rapid gain across pregnancy with 21.3 kg predicted gain. Compared to Class 1, Class 3 was associated with higher birth weight z-score (ß = 0.63, 95% confidence interval [CI]: 0.31,0.96), increased risk for large for gestational age (IRR = 5.60, 95% CI: 1.59, 19.67), and birth <32 weeks (IRR = 2.44, 95%CI: 1.10, 5.4) that was attenuated in sensitivity analyses. Class 2 was associated with moderately elevated birth weight z-score (ß = 0.24, 95%CI: 0.00, 0.48, p = 0.050). CONCLUSION: Gestational weight change followed a low, moderate, or high trajectory; both moderate and high gain patterns were associated with increased infant size outcomes. Optimal patterns of weight change that balance risk during the prenatal, perinatal, and neonatal periods require further investigation, particularly in high-risk twin pregnancies. KEY POINTS: · A majority gained weight below IOM twin recommendations.. · Three patterns of GWC across pregnancy were identified.. · Moderate or high GWC was associated with infant size..

Model-assisted analyses of longitudinal, ordinal outcomes with absorbing states.

Studies of critically ill, hospitalized patients often follow participants and characterize daily health status using an ordinal outcome variable. Statistically, longitudinal proportional odds models are a natural choice in these settings since such models can parsimoniously summarize differences across patient groups and over time. However, when one or more of the outcome states is absorbing, the proportional odds assumption for the follow-up time parameter will likely be violated, and more flexible longitudinal models are needed. Motivated by the VIOLET Study (Ginde et al), a parallel-arm, randomized clinical trial of Vitamin D3 in critically ill patients, we discuss and contrast several treatment effect estimands based on time-dependent odds ratio parameters, and we detail contemporary modeling approaches. In VIOLET, the outcome is a four-level ordinal variable where the lowest "not alive" state is absorbing and the highest "at-home" state is nearly absorbing. We discuss flexible extensions of the proportional odds model for longitudinal data that can be used for either model-based inference, where the odds ratio estimator is taken directly from the model fit, or for model-assisted inferences, where heterogeneity across cumulative log odds dichotomizations is modeled and results are summarized to obtain an overall odds ratio estimator. We focus on direct estimation of cumulative probability model (CPM) parameters using likelihood-based analysis procedures that naturally handle absorbing states. We illustrate the modeling procedures, the relative precision of model-based and model-assisted estimators, and the possible differences in the values for which the estimators are consistent through simulations and analysis of the VIOLET Study data.

SARS-CoV-2 viral load is associated with risk of transmission to household and community contacts.

BACKGROUND: Factors that lead to successful SARS-CoV-2 transmission are still not well described. We investigated the association between a case's viral load and the risk of transmission to contacts in the context of other exposure-related factors. METHODS: Data were generated through routine testing and contact tracing at a large university. Case viral loads were obtained from cycle threshold values associated with a positive polymerase chain reaction test result from October 1, 2020 to April 15, 2021. Cases were included if they had at least one contact who tested 3-14 days after the exposure. Case-contact pairs were formed by linking index cases with contacts. Chi-square tests were used to evaluate differences in proportions of contacts testing positive. Generalized estimating equation models with a log link were used to evaluate whether viral load and other exposure-related factors were associated with a contact testing positive. RESULTS: Median viral load among the 212 cases included in the study was 5.6 (1.8-10.4) log10 RNA copies per mL of saliva. Among 365 contacts, 70 (19%) tested positive following their exposure; 36 (51%) were exposed to a case that was asymptomatic or pre-symptomatic on the day of exposure. The proportion of contacts that tested positive increased monotonically with index case viral load (12%, 23% and 25% corresponding to < 5, 5-8 and > 8 log10 copies per mL, respectively; X2 = 7.18, df = 2, p = 0.03). Adjusting for cough, time between test and exposure, and physical contact, the risk of transmission to a close contact was significantly associated with viral load (RR = 1.27, 95% CI 1.22-1.32). CONCLUSIONS: Further research is needed to understand whether these relationships persist for newer variants. For those variants whose transmission advantage is mediated through a high viral load, public health measures could be scaled accordingly. Index cases with higher viral loads could be prioritized for contact tracing and recommendations to quarantine contacts could be made according to the likelihood of transmission based on risk factors such as viral load.

Longitudinal change in speech classification between 4 and 10 years in children with cerebral palsy.

AIM: To examine speech impairment severity classification over time in a longitudinal cohort of children with cerebral palsy (CP). METHOD: A total of 101 children (58 males, 43 females) between the ages of 4 and 10 years with CP participated in this longitudinal study. Speech severity was rated using the Viking Speech Scale (VSS), a four-level classification rating scale, at 4, 6, 8, and 10 years (age 4 years: mean = 52 months [3 SD]; age 6 years: mean = 75 months [2 SD]; age 8 years: mean = 100 months [4 SD]; age 10 years: mean = 125 months [5 SD]). We used Bayesian mixed-effects ordinal logistic regression to model (1) the extent to which speech severity changed over time and (2) patterns of change across age groups and classification rating group levels. RESULTS: VSS ratings decreased (speech severity became less severe) between 4 and 10 years of age. Children who were first classified in VSS levels I, II, or III at age 4 years had a high probability of staying at, or improving to, VSS level I by 10 years. Children who were first classified in VSS level IV at 4 years had a high probability of remaining in VSS level IV at 10 years. INTERPRETATION: Early speech performance is highly predictive of later childhood speech abilities. Children with any level of speech impairment at age 4 years should be receiving speech therapy. Those with more severe speech impairments should be introduced to augmentative and alternative communication as soon as possible. WHAT THIS PAPER ADDS: Children with early Viking Speech Scale (VSS) ratings below level IV have a good prognosis for speech improvement. Children with early VSS level IV ratings are likely to remain at VSS level IV over time. Children did not show worsening of VSS level between the ages of 4 and 10 years.

Cambio longitudinal en la clasificación del habla entre los 4 y 10 años en niños con parálisis cerebral OBJETIVO: Examinar la clasificación de la gravedad del deterioro del habla a lo largo del tiempo en una cohorte longitudinal de niños con parálisis cerebral (PC) entre las edades de 4 y 10 años. MÉTODO: Un total de 101 niños (58 varones, 43 mujeres) con PC participaron en este estudio longitudinal. La gravedad del habla se evaluó utilizando la Viking Speech Scale (VSS), una escala de calificación de clasificación de cuatro niveles, a los 4, 6, 8 y 10 años (edad 4 años: media = 52 meses [3 DE]; edad 6 años: media = 75 meses [2 DE]; edad 8 años: media = 100 meses [4 DE]; edad 10 años: media = 125 meses [5 DE]). Utilizamos la regresión logística ordinal de efectos mixtos bayesianos para modelar (1) la medida en que la severidad del habla cambió con el tiempo y (2) los patrones de cambio entre los grupos de edad y los niveles de clasificación de los grupos. RESULTADOS: Las calificaciones de VSS disminuyeron (la severidad del habla se volvió menos severa) entre los 4 y los 10 años de edad. Los niños que fueron clasificados por primera vez en los niveles I, II o III de VSS a los 4 años tenían una alta probabilidad de permanecer en el nivel I de VSS o mejorar al nivel I de VSS a los 10 años. Los niños que fueron clasificados por primera vez en el nivel IV de VSS a los 4 años tenían una alta probabilidad de permanecer en el nivel IV de VSS a los 10 años. INTERPRETACIÓN: El desempeño temprano del habla es altamente predictivo de las habilidades del habla en la niñez posterior. Los niños con cualquier nivel de discapacidad del habla a la edad de 4 años deben recibir terapia del habla. Aquellos con impedimentos del habla más severos deben ser introducidos a la comunicación aumentativa y alternativa tan pronto como sea posible.

Mudança longitudinal na classificação da fala entre 4 e 10 anos em crianças com paralisia cerebral OBJETIVO: Analisar a classificação da gravidade do comprometimento da fala ao longo do tempo em uma coorte longitudinal de crianças com paralisia cerebral (PC) entre 4 e 10 anos. MÉTODO: Um total de 101 crianças (58 meninos, 43 meninas) com PC participaram deste estudo longitudinal. A gravidade da fala foi avaliada usando a Viking Speech Scale (VSS), uma escala de classificação de quatro níveis, aos 4, 6, 8 e 10 anos (idade 4 anos: média = 52 meses [3 DP]; idade 6 anos: média = 75 meses [2 DP]; idade 8 anos: média = 100 meses [4 DP]; idade 10 anos: média = 125 meses [5 DP]). Usamos a regressão logística ordinal Bayesiana de efeitos mistos para modelar (1) a extensão em que a gravidade da fala mudou ao longo do tempo e (2) padrões de mudança entre as faixas etárias e os níveis de classificação do grupo. RESULTADOS: As classificações de VSS diminuíram (a gravidade da fala tornou-se menos grave) entre 4 e 10 anos de idade. As crianças que foram classificadas pela primeira vez nos níveis VSS I, II ou III aos 4 anos de idade tiveram uma alta probabilidade de permanecer ou melhorar para o nível VSS I em 10 anos. As crianças que foram classificadas pela primeira vez em VSS nível IV aos 4 anos tiveram alta probabilidade de permanecer no VSS nível IV aos 10 anos. INTERPRETAÇÃO: O desempenho precoce da fala é altamente preditivo de habilidades de fala na infância posteriormente. Crianças com qualquer nível de deficiência de fala aos 4 anos de idade devem receber terapia da fala. Aqueles com deficiências de fala mais graves devem ser introduzidos à comunicação aumentativa e alternativa o mais rápido possível.

Validation of remote atopic dermatitis severity assessment with the Eczema Area and Severity Ondex in children using caregiver-provided photos and videos.

BACKGROUND/OBJECTIVES: We sought to quantify the reliability and validity of remote atopic dermatitis (AD) severity assessment using the Eczema Area and Severity Index (EASI) applied to caregiver-provided photos (p-EASI) and videos (v-EASI). METHODS: Children (0-17 years) with a physician diagnosis of AD were recruited. Caregivers took photos and a video of their child's skin. A clinician scored in-person EASI on the same day, then p-EASI and v-EASI for each participant 10 days or more between ratings. Two additional clinicians scored p-EASI and v-EASI. Lin's concordance correlation coefficient (CCC) was employed to assess criterion validity using in-person EASI as the gold standard. Intraclass correlation coefficients (ICCs) were calculated to assess interrater reliability of p-EASI and v-EASI. RESULTS: Fifty racially and ethnically diverse children (age [mean ± SD]: 4.3 ± 4.4 years; 42% female) with a range of AD severity (EASI: 6.3 ± 6.4) and Fitzpatrick skin types (1-2: 9%; 3-4: 60%; 5-6: 31%) were enrolled and received in-person EASI assessment. Fifty had p-EASI and 49 had v-EASI by the same in-person rater, and by two additional raters. The CCC and ICC for p-EASI were 0.89, 95% CI [0.83, 0.95] and 0.81, 95% CI [0.71, 0.89], respectively. The CCC and ICC for v-EASI were 0.75, 95% CI [0.63, 0.88] and 0.69, 95% CI [0.51, 0.81], respectively. CONCLUSIONS: In this diverse population with a range of skin tones, p-EASI showed good criterion validity and good interrater reliability. v-EASI showed moderate to good criterion validity and moderate interrater reliability. Both may be reliable and valid options for remote AD severity assessment.

power swgee: GEE-based power calculations in stepped wedge cluster randomized trials.

Stepped wedge cluster randomized trials are increasingly being used to evaluate interventions in medical, public health, educational, and social science contexts. With the longitudinal and crossover nature of a SW-CRT, complex analysis techniques are often needed which makes appropriately powering SW-CRTs challenging. In this paper, we introduce a newly-developed SW-CRT power calculator, embedded within the power command in Stata. The power calculator assumes a marginal model (i.e., generalized estimating equations [GEE]) for the primary analysis of SW-CRTs, for which other currently available SW-CRT power calculators may not be suitable. The program accommodates complete cross-sectional and closed-cohort designs, and includes multilevel correlation structures appropriate for such designs. We discuss the methods and formulae underlying our SW-CRT calculator, and provide illustrative examples of the use of power swgee. We provide suggestions about the choice of parameters in power swgee, and conclude by discussing areas of future research which may improve the program.

Impact of unequal cluster sizes for GEE analyses of stepped wedge cluster randomized trials with binary outcomes.

The stepped wedge (SW) design is a type of unidirectional crossover design where cluster units switch from control to intervention condition at different prespecified time points. While a convention in study planning is to assume the cluster-period sizes are identical, SW cluster randomized trials (SW-CRTs) involving repeated cross-sectional designs frequently have unequal cluster-period sizes, which can impact the efficiency of the treatment effect estimator. In this paper, we provide a comprehensive investigation of the efficiency impact of unequal cluster sizes for generalized estimating equation analyses of SW-CRTs, with a focus on binary outcomes as in the Washington State Expedited Partner Therapy trial. Several major distinctions between our work and existing work include the following: (i) we consider multilevel correlation structures in marginal models with binary outcomes; (ii) we study the implications of both the between-cluster and within-cluster imbalances in sizes; and (iii) we provide a comparison between the independence working correlation versus the true working correlation and detail the consequences of ignoring correlation estimation in SW-CRTs with unequal cluster sizes. We conclude that the working independence assumption can lead to substantial efficiency loss and a large sample size regardless of cluster-period size variability in SW-CRTs, and recommend accounting for correlations in the analysis. To improve study planning, we additionally provide a computationally efficient search algorithm to estimate the sample size in SW-CRTs accounting for unequal cluster-period sizes, and conclude by illustrating the proposed approach in the context of the Washington State study.

Two-wave two-phase outcome-dependent sampling designs, with applications to longitudinal binary data.

Two-phase outcome-dependent sampling (ODS) designs are useful when resource constraints prohibit expensive exposure ascertainment on all study subjects. One class of ODS designs for longitudinal binary data stratifies subjects into three strata according to those who experience the event at none, some, or all follow-up times. For time-varying covariate effects, exclusively selecting subjects with response variation can yield highly efficient estimates. However, if interest lies in the association of a time-invariant covariate, or the joint associations of time-varying and time-invariant covariates with the outcome, then the optimal design is unknown. Therefore, we propose a class of two-wave two-phase ODS designs for longitudinal binary data. We split the second-phase sample selection into two waves, between which an interim design evaluation analysis is conducted. The interim design evaluation analysis uses first-wave data to conduct a simulation-based search for the optimal second-wave design that will improve the likelihood of study success. Although we focus on longitudinal binary response data, the proposed design is general and can be applied to other response distributions. We believe that the proposed designs can be useful in settings where (1) the expected second-phase sample size is fixed and one must tailor stratum-specific sampling probabilities to maximize estimation efficiency, or (2) relative sampling probabilities are fixed across sampling strata and one must tailor sample size to achieve a desired precision. We describe the class of designs, examine finite sampling operating characteristics, and apply the designs to an exemplar longitudinal cohort study, the Lung Health Study.

Regularized Ordinal Regression and the ordinalNet R Package.

Regularization techniques such as the lasso (Tibshirani 1996) and elastic net (Zou and Hastie 2005) can be used to improve regression model coefficient estimation and prediction accuracy, as well as to perform variable selection. Ordinal regression models are widely used in applications where the use of regularization could be beneficial; however, these models are not included in many popular software packages for regularized regression. We propose a coordinate descent algorithm to fit a broad class of ordinal regression models with an elastic net penalty. Furthermore, we demonstrate that each model in this class generalizes to a more flexible form, that can be used to model either ordered or unordered categorical response data. We call this the elementwise link multinomial-ordinal (ELMO) class, and it includes widely used models such as multinomial logistic regression (which also has an ordinal form) and ordinal logistic regression (which also has an unordered multinomial form). We introduce an elastic net penalty class that applies to either model form, and additionally, this penalty can be used to shrink a non-ordinal model toward its ordinal counterpart. Finally, we introduce the R package ordinalNet, which implements the algorithm for this model class.

Two-Phase, Generalized Case-Control Designs for the Study of Quantitative Longitudinal Outcomes.

We propose a general class of 2-phase epidemiologic study designs for quantitative, longitudinal data that are useful when phase 1 longitudinal outcome and covariate data are available but data on the exposure (e.g., a biomarker) can only be collected on a subset of subjects during phase 2. To conduct a study using a design in the class, one first summarizes the longitudinal outcomes by fitting a simple linear regression of the response on a time-varying covariate for each subject. Sampling strata are defined by splitting the estimated regression intercept or slope distributions into distinct (low, medium, and high) regions. Stratified sampling is then conducted from strata defined by the intercepts, by the slopes, or from a mixture. In general, samples selected with extreme intercept values will yield low variances for associations of time-fixed exposures with the outcome and samples enriched with extreme slope values will yield low variances for associations of time-varying exposures with the outcome (including interactions with time-varying exposures). We describe ascertainment-corrected maximum likelihood and multiple-imputation estimation procedures that permit valid and efficient inferences. We embed all methodological developments within the framework of conducting a substudy that seeks to examine genetic associations with lung function among continuous smokers in the Lung Health Study (United States and Canada, 1986-1994).

A latent variable approach to potential outcomes for emergency department admission decisions.

In emergency departments (EDs), care providers continuously weigh admissions against continued monitoring and treatment often without knowing their condition and health needs. To understand the decision process and its causal effect on outcomes, an observational study must contend with unobserved/missing information and a lack of exchangeability between admitted and discharged patients. Our goal was to provide a general framework to evaluate admission decisions from electronic healthcare records (EHRs). We describe admission decisions as a decision-making process in which the patient's health needs is a binary latent variable. We estimate latent health needs from EHR with only partial knowledge of the decision process (ie, initial evaluation, admission decision, length of stay). Estimated latent health needs are then used to understand the admission decision and the decision's causal impact on outcomes. For the latter, we assume potential outcomes are stochastically independent from the admission decision conditional on latent health needs. As a case study, we apply our approach to over 150 000 patient encounters with the ED from the University of Michigan Health System collected from August 2012 through July 2015. We estimate that while admitting a patient with higher latent needs reduces the 30-day risk of revisiting the ED or later being admitted through the ED by over 79%, admitting a patient with lower latent needs actually increases these 30-day risks by 3.0% and 7.6%, respectively.

Extending the Case-Control Design to Longitudinal Data: Stratified Sampling Based on Repeated Binary Outcomes.

We detail study design options that generalize case-control sampling when longitudinal outcome data are already collected as part of a primary cohort study, but new exposure data must be retrospectively processed for a secondary analysis. Furthermore, we assume that cost will limit the size of the subsample that can be evaluated. We describe a novel class of stratified outcome-dependent sampling designs for longitudinal binary response data where distinct strata are created for subjects who never, sometimes, and always experienced the event of interest during longitudinal follow-up. Individual designs within this class are differentiated by the stratum-specific sampling probabilities. We show for parameters associated with time-varying exposures, subjects who experience the event/outcome at some but not at all of the follow-up times (i.e., those who exhibit response variation) are highly informative. If the time-varying exposure varies exclusively within individuals (i.e., intraclass correlation coefficient is 0), then sampling all subjects with response variability can yield highly precise parameter estimates even when compared with an analysis of the original cohort. The flexibility of the designs and analysis procedures also permits estimation of parameters that correspond to time-fixed covariates, and we show that with an imputation-based estimation procedure, baseline covariate associations can be estimated with very high precision irrespective of the design. We demonstrate features of the designs and analysis procedures via a plasmode simulation using data from the Lung Health Study.