Generalized single index modeling of longitudinal data with multiple binary responses.

In health and clinical research, medical indices (eg, BMI) are commonly used for monitoring and/or predicting health outcomes of interest. While single-index modeling can be used to construct such indices, methods to use single-index models for analyzing longitudinal data with multiple correlated binary responses are underdeveloped, although there are abundant applications with such data (eg, prediction of multiple medical conditions based on longitudinally observed disease risk factors). This article aims to fill the gap by proposing a generalized single-index model that can incorporate multiple single indices and mixed effects for describing observed longitudinal data of multiple binary responses. Compared to the existing methods focusing on constructing marginal models for each response, the proposed method can make use of the correlation information in the observed data about different responses when estimating different single indices for predicting response variables. Estimation of the proposed model is achieved by using a local linear kernel smoothing procedure, together with methods designed specifically for estimating single-index models and traditional methods for estimating generalized linear mixed models. Numerical studies show that the proposed method is effective in various cases considered. It is also demonstrated using a dataset from the English Longitudinal Study of Aging project.

Comparing two hazard curves when there is a treatment time-lag effect.

In cancer and other medical studies, time-to-event (eg, death) data are common. One major task to analyze time-to-event (or survival) data is usually to compare two medical interventions (eg, a treatment and a control) regarding their effect on patients' hazard to have the event in concern. In such cases, we need to compare two hazard curves of the two related patient groups. In practice, a medical treatment often has a time-lag effect, that is, the treatment effect can only be observed after a time period since the treatment is applied. In such cases, the two hazard curves would be similar in an initial time period, and the traditional testing procedures, such as the log-rank test, would be ineffective in detecting the treatment effect because the similarity between the two hazard curves in the initial time period would attenuate the difference between the two hazard curves that is reflected in the related testing statistics. In this paper, we suggest a new method for comparing two hazard curves when there is a potential treatment time-lag effect based on a weighted log-rank test with a flexible weighting scheme. The new method is shown to be more effective than some representative existing methods in various cases when a treatment time-lag effect is present.

Multiblock partial least squares and rank aggregation: Applications to detection of bacteriophages associated with antimicrobial resistance in the presence of potential confounding factors.

Urban environments, characterized by bustling mass transit systems and high population density, host a complex web of microorganisms that impact microbial interactions. These urban microbiomes, influenced by diverse demographics and constant human movement, are vital for understanding microbial dynamics. We explore urban metagenomics, utilizing an extensive dataset from the Metagenomics & Metadesign of Subways & Urban Biomes (MetaSUB) consortium, and investigate antimicrobial resistance (AMR) patterns. In this pioneering research, we delve into the role of bacteriophages, or "phages"-viruses that prey on bacteria and can facilitate the exchange of antibiotic resistance genes (ARGs) through mechanisms like horizontal gene transfer (HGT). Despite their potential significance, existing literature lacks a consensus on their significance in ARG dissemination. We argue that they are an important consideration. We uncover that environmental variables, such as those on climate, demographics, and landscape, can obscure phage-resistome relationships. We adjust for these potential confounders and clarify these relationships across specific and overall antibiotic classes with precision, identifying several key phages. Leveraging machine learning tools and validating findings through clinical literature, we uncover novel associations, adding valuable insights to our comprehension of AMR development.

Multivariate single index modeling of longitudinal data with multiple responses.

In medical studies, composite indices and/or scores are routinely used for predicting medical conditions of patients. These indices are usually developed from observed data of certain disease risk factors, and it has been demonstrated in the literature that single index models can provide a powerful tool for this purpose. In practice, the observed data of disease risk factors are often longitudinal in the sense that they are collected at multiple time points for individual patients, and there are often multiple aspects of a patient's medical condition that are of our concern. However, most existing single-index models are developed for cases with independent data and a single response variable, which are inappropriate for the problem just described in which within-subject observations are usually correlated and there are multiple mutually correlated response variables involved. This paper aims to fill this methodological gap by developing a single index model for analyzing longitudinal data with multiple responses. Both theoretical and numerical justifications show that the proposed new method provides an effective solution to the related research problem. It is also demonstrated using a dataset from the English Longitudinal Study of Aging.

Effective disease surveillance by using covariate information.

Effective surveillance of infectious diseases, cancers, and other deadly diseases is critically important for public health and safety of our society. Incidence data of such diseases are often collected spatially from different clinics and hospitals through a regional, national or global disease reporting system. In such a system, new batches of data keep being collected over time, and a decision needs to be made immediately after new data are collected regarding whether there is a disease outbreak at the current time point. This is the disease surveillance problem that will be focused in this article. There are some existing methods for solving this problem, most of which use the disease incidence data only. In practice, however, disease incidence is often associated with some covariates, including the air temperature, humidity, and other weather or environmental conditions. In this article, we develop a new methodology for disease surveillance which can make use of helpful covariate information to improve its effectiveness. A novelty of this new method is behind the property that only those covariate information that is associated with a true disease outbreak can help trigger a signal. The new method can accommodate seasonality, spatio-temporal data correlation, and nonparametric data distribution. These features make it feasible to use in many real applications.

Joint modeling of multivariate nonparametric longitudinal data and survival data: A local smoothing approach.

In many clinical studies, evaluating the association between longitudinal and survival outcomes is of primary concern. For analyzing data from such studies, joint modeling of longitudinal and survival data becomes an appealing approach. In some applications, there are multiple longitudinal outcomes whose longitudinal pattern is difficult to describe by a parametric form. For such applications, existing research on joint modeling is limited. In this article, we develop a novel joint modeling method to fill the gap. In the new method, a local polynomial mixed-effects model is used for describing the nonparametric longitudinal pattern of the multiple longitudinal outcomes. Two model estimation procedures, that is, the local EM algorithm and the local penalized quasi-likelihood estimation, are explored. Practical guidelines for choosing tuning parameters and for variable selection are provided. The new method is justified by some theoretical arguments and numerical studies.

Edge-Preserving Denoising of Image Sequences.

To monitor the Earth's surface, the satellite of the NASA Landsat program provides us image sequences of any region on the Earth constantly over time. These image sequences give us a unique resource to study the Earth's surface, changes of the Earth resource over time, and their implications in agriculture, geology, forestry, and more. Besides natural sciences, image sequences are also commonly used in functional magnetic resonance imaging (fMRI) of medical studies for understanding the functioning of brains and other organs. In practice, observed images almost always contain noise and other contaminations. For a reliable subsequent image analysis, it is important to remove such contaminations in advance. This paper focuses on image sequence denoising, which has not been well-discussed in the literature yet. To this end, an edge-preserving image denoising procedure is suggested. The suggested method is based on a jump-preserving local smoothing procedure, in which the bandwidths are chosen such that the possible spatio-temporal correlations in the observed image intensities are accommodated properly. Both theoretical arguments and numerical studies show that this method works well in the various cases considered.

Early detection of high disease activity in juvenile idiopathic arthritis by sequential monitoring of patients' health-related quality of life scores.

Juvenile idiopathic arthritis (JIA) is a chronic disease. During its "high disease activity (HDA)" stage, JIA can cause severe pain, and thus could seriously affect patients' physical and psychological health. Early detection of the HDA stage of JIA can reduce the damage of the disease by treating it at an early stage and alleviating the painful experience of the patients. So far, no effective cure of JIA has been found, and one major goal of disease management is to improve patients' quality of life. To this end, patients' health-related quality of life (HRQOL) scores are routinely collected over time from JIA patients. In this paper, we demonstrate that a new statistical methodology called dynamic screening system (DySS) is effective for early detection of the HDA stage of JIA. By this approach, a patient's HRQOL scores are monitored sequentially, and a signal is given by DySS once the longitudinal pattern of the scores is found to be significantly different from the pattern of patients with low disease activity. Dimension reduction of the observed HRQOL scores and the corresponding impact on the performance of DySS are also discussed.

Nonparametric estimation of the spatio-temporal covariance structure.

Spatio-temporal modeling is an active research problem with broad applications. In this problem, proper description and estimation of the data covariance structure plays an important role. In the literature, most available methods assume that the data covariance is stationary and follows a specific parametric form. In practice, however, such assumptions are hardly valid or difficult to verify. In this paper, we propose a new and flexible method for estimating the underlying covariance structure. Our proposed method does not require the covariance to be stationary or follow a parametric form. It can accommodate nonparametric space-time-varying mean structure of the observed data. Under some mild regularity conditions, it is shown that our estimated covariance structure converges to the true covariance structure. The proposed method is also justified numerically by a simulation study and an application to a hand, foot, and mouth disease data.

Errors-in-variables jump regression using local clustering.

Errors-in-variables (EIV) regression is widely used in econometric models. The statistical analysis becomes challenging when the regression function is discontinuous and the distribution of measurement error is unknown. In the literature, most existing jump regression methods either assume that there is no measurement error involved or require that jumps are explicitly detected before the regression function can be estimated. In some applications, however, the ultimate goal is to estimate the regression function and to preserve the jumps in the process of estimation. In this paper, we are concerned with reconstructing jump regression curve from data that involve measurement error. We propose a direct jump-preserving method that does not explicitly detect jumps. The challenge of restoring jump structure masked by measurement error is handled by local clustering. Theoretical analysis shows that the proposed curve estimator is statistically consistent. A numerical comparison with an existing jump regression method highlights its jump-preserving property. Finally, we demonstrate our method by an application to a health tax policy study in Australia.

Spatiotemporal incidence rate data analysis by nonparametric regression.

To monitor the incidence rates of cancers, AIDS, cardiovascular diseases, and other chronic or infectious diseases, some global, national, and regional reporting systems have been built to collect/provide population-based data about the disease incidence. Such databases usually report daily, monthly, or yearly disease incidence numbers at the city, county, state, or country level, and the disease incidence numbers collected at different places and different times are often correlated, with the ones closer in place or time being more correlated. The correlation reflects the impact of various confounding risk factors, such as weather, demographic factors, lifestyles, and other cultural and environmental factors. Because such impact is complicated and challenging to describe, the spatiotemporal (ST) correlation in the observed disease incidence data has complicated ST structure as well. Furthermore, the ST correlation is hidden in the observed data and cannot be observed directly. In the literature, there has been some discussion about ST data modeling. But, the existing methods either impose various restrictive assumptions on the ST correlation that are hard to justify, or ignore partially or entirely the ST correlation. This paper aims to develop a flexible and effective method for ST disease incidence data modeling, using nonparametric local smoothing methods. This method can properly accommodate the ST data correlation. Theoretical justifications and numerical studies show that it works well in practice.

Objective Preoperative Parameters Predict Difficult Pelvic Dissections and Clinical Outcomes.

BACKGROUND: Objective criteria to predict difficult pelvic dissection with prognostic significance are lacking. Previous studies have focused on predicting intraoperative conversion and not evaluated factors specific to pelvic surgery. We aimed to develop an objective, prognostic, preoperative assessment to predict difficult pelvic dissections and clinical outcomes. Such a model is much needed, may facilitate objective comparisons between rectal cancer centers, or may serve as a stratification variable in clinical trials. MATERIALS AND METHODS: Patients who underwent low anterior resection or abdominoperineal resection for rectal cancer within 10 cm of the anal verge (2009-2014) were retrospectively analyzed. Procedures were categorized into "routine" or "difficult" based on predefined criteria. All patients underwent 14 measurements on preoperative imaging. Outcomes were compared between the two groups. Stepwise multivariate logistic regression was used to develop the prediction model, which was validated in an independent data set. RESULTS: Of the 280 patients analyzed, 80 fulfilled the inclusion criteria. Baseline characteristics were similar except for more males having a "difficult" pelvis. "Difficult" patients were significantly more likely to have a narrower pelvis, smaller pelvic volumes, a longer pelvis, more curved sacrum, and more acute anorectal angle. Difficult cases correlated significantly with higher blood loss, hospital costs, longer operative time, and length of stay. A practical model to predict difficult pelvic dissections was created and included male gender, previous radiation, and length from promontory to pelvic floor > 130 mm. Model validation was performed in 40 patients from an independent data set. CONCLUSIONS: An objective, validated model that predicts a difficult pelvic dissection and associated worse clinical outcome is possible.

High Generic Drug Prices and Market Competition: A Retrospective Cohort Study.

BACKGROUND: Prices for some generic drugs have increased in recent years, adversely affecting patients who rely on them. OBJECTIVE: To determine the association between market competition levels and the change in generic drug prices in the United States. DESIGN: Retrospective cohort study. SETTING: Prescription claims from commercial health plans between 2008 and 2013. MEASUREMENTS: The 5.5 years of data were divided into 11 study periods of 6 months each. The Herfindahl-Hirschman Index (HHI)-calculated by summing the squares of individual manufacturers' market shares, with higher values indicating a less competitive market-and average drug prices were estimated for the generic drugs in each period. The HHI value estimated in the baseline period (first half of 2008) was modeled as a fixed covariate. Models estimated price changes over time by level of competition, adjusting for drug shortages, market size, and dosage forms. RESULTS: From 1.08 billion prescription claims, a cohort of 1120 generic drugs was identified. After adjustment, drugs with quadropoly (HHI value of 2500, indicating relatively high levels of competition), duopoly (HHI value of 5000), near-monopoly (HHI value of 8000), and monopoly (HHI value of 10 000) levels of baseline competition were associated with price changes of -31.7% (95% CI, -34.4% to -28.9%), -11.8% (CI, -18.6% to -4.4%), 20.1% (CI, 5.5% to 36.6%), and 47.4% (CI, 25.4% to 73.2%), respectively, over the study period. LIMITATION: Study findings may not be generalizable to drugs that became generic after 2008. CONCLUSION: Market competition levels were associated with a change in generic drug prices. Such measurements may be helpful in identifying older prescription drugs at higher risk for price change in the future. PRIMARY FUNDING SOURCE: None.

Oral Human Papillomavirus Infection: Differences in Prevalence Between Sexes and Concordance With Genital Human Papillomavirus Infection, NHANES 2011 to 2014.

BACKGROUND: The burden of human papillomavirus (HPV)-positive oropharyngeal squamous cell carcinoma (OPSCC) is disproportionately high among men, yet empirical evidence regarding the difference in prevalence of oral HPV infection between men and women is limited. Concordance of oral and genital HPV infection among men is unknown. OBJECTIVE: To determine the prevalence of oral HPV infection, as well as the concordance of oral and genital HPV infection, among U.S. men and women. DESIGN: Nationally representative survey. SETTING: Civilian noninstitutionalized population. PARTICIPANTS: Adults aged 18 to 69 years from NHANES (National Health and Nutrition Examination Survey), 2011 to 2014. MEASUREMENTS: Oral rinse, penile swab, and vaginal swab specimens were evaluated by polymerase chain reaction followed by type-specific hybridization. RESULTS: The overall prevalence of oral HPV infection was 11.5% (95% CI, 9.8% to 13.1%) in men and 3.2% (CI, 2.7% to 3.8%) in women (equating to 11 million men and 3.2 million women nationwide). High-risk oral HPV infection was more prevalent among men (7.3% [CI, 6.0% to 8.6%]) than women (1.4% [CI, 1.0% to 1.8%]). Oral HPV 16 was 6 times more common in men (1.8% [CI, 1.3% to 2.2%]) than women (0.3% [CI, 0.1% to 0.5%]) (1.7 million men vs. 0.27 million women). Among men and women who reported having same-sex partners, the prevalence of high-risk HPV infection was 12.7% (CI, 7.0% to 18.4%) and 3.6% (CI, 1.4% to 5.9%), respectively. Among men who reported having 2 or more same-sex oral sex partners, the prevalence of high-risk HPV infection was 22.2% (CI, 9.6% to 34.8%). Oral HPV prevalence among men with concurrent genital HPV infection was 4-fold greater (19.3%) than among those without it (4.4%). Men had 5.4% (CI, 5.1% to 5.8%) greater predicted probability of high-risk oral HPV infection than women. The predicted probability of high-risk oral HPV infection was greatest among black participants, those who smoked more than 20 cigarettes daily, current marijuana users, and those who reported 16 or more lifetime vaginal or oral sex partners. LIMITATION: Sexual behaviors were self-reported. CONCLUSION: Oral HPV infection is common among U.S. men. This study's findings provide several policy implications to guide future OPSCC prevention efforts to combat this disease. PRIMARY FUNDING SOURCE: National Cancer Institute.

Evaluation of the treatment time-lag effect for survival data.

Medical treatments often take a period of time to reveal their impact on subjects, which is the so-called time-lag effect in the literature. In the survival data analysis literature, most existing methods compare two treatments in the entire study period. In cases when there is a substantial time-lag effect, these methods would not be effective in detecting the difference between the two treatments, because the similarity between the treatments during the time-lag period would diminish their effectiveness. In this paper, we develop a novel modeling approach for estimating the time-lag period and for comparing the two treatments properly after the time-lag effect is accommodated. Theoretical arguments and numerical examples show that it is effective in practice.

Comparison of multiple hazard rate functions.

Many robust tests have been proposed in the literature to compare two hazard rate functions, however, very few of them can be used in cases when there are multiple hazard rate functions to be compared. In this article, we propose an approach for detecting the difference among multiple hazard rate functions. Through a simulation study and a real-data application, we show that the new method is robust and powerful in many situations, compared with some commonly used tests.

Statistical monitoring of the hand, foot and mouth disease in China.

In a period starting around 2007, the Hand, Foot, and Mouth Disease (HFMD) became wide-spreading in China, and the Chinese public health was seriously threatened. To prevent the outbreak of infectious diseases like HFMD, effective disease surveillance systems would be especially helpful to give signals of disease outbreaks as early as possible. Statistical process control (SPC) charts provide a major statistical tool in industrial quality control for detecting product defectives in a timely manner. In recent years, SPC charts have been used for disease surveillance. However, disease surveillance data often have much more complicated structures, compared to the data collected from industrial production lines. Major challenges, including lack of in-control data, complex seasonal effects, and spatio-temporal correlations, make the surveillance data difficult to handle. In this article, we propose a three-step procedure for analyzing disease surveillance data, and our procedure is demonstrated using the HFMD data collected during 2008-2009 in China. Our method uses nonparametric longitudinal data and time series analysis methods to eliminate the possible impact of seasonality and temporal correlation before the disease incidence data are sequentially monitored by a SPC chart. At both national and provincial levels, our proposed method can effectively detect the increasing trend of disease incidence rate before the disease becomes wide-spreading.

Surveillance of cardiovascular diseases using a multivariate dynamic screening system.

In the SHARe Framingham Heart Study of the National Heart, Lung and Blood Institute, one major task is to monitor several health variables (e.g., blood pressure and cholesterol level) so that their irregular longitudinal pattern can be detected as soon as possible and some medical treatments applied in a timely manner to avoid some deadly cardiovascular diseases (e.g., stroke). To handle this kind of applications effectively, we propose a new statistical methodology called multivariate dynamic screening system (MDySS) in this paper. The MDySS method combines the major strengths of the multivariate longitudinal data analysis and the multivariate statistical process control, and it makes decisions about the longitudinal pattern of a subject by comparing it with other subjects cross sectionally and by sequentially monitoring it as well. Numerical studies show that MDySS works well in practice.

Model selection and diagnostics for joint modeling of survival and longitudinal data with crossing hazard rate functions.

Comparison of two hazard rate functions is important for evaluating treatment effect in studies concerning times to some important events. In practice, it may happen that the two hazard rate functions cross each other at one or more unknown time points, representing temporal changes of the treatment effect. Also, besides survival data, there could be longitudinal data available regarding some time-dependent covariates. When jointly modeling the survival and longitudinal data in such cases, model selection and model diagnostics are especially important to provide reliable statistical analysis of the data, which are lacking in the literature. In this paper, we discuss several criteria for assessing model fit that have been used for model selection and apply them to the joint modeling of survival and longitudinal data for comparing two crossing hazard rate functions. We also propose hypothesis testing and graphical methods for model diagnostics of the proposed joint modeling approach. Our proposed methods are illustrated by a simulation study and by a real-data example concerning two early breast cancer treatments.

Predicting high-risk periods for weight regain following initial weight loss.

OBJECTIVE: The aim of this study was to develop a predictive algorithm of "high-risk" periods for weight regain after weight loss. METHODS: Longitudinal mixed-effects models and random forest regression were used to select predictors and develop an algorithm to predict weight regain on a week-to-week basis, using weekly questionnaire and self-monitoring data (including daily e-scale data) collected over 40 weeks from 46 adults who lost ≥5% of baseline weight during an initial 12-week intervention (Study 1). The algorithm was evaluated in 22 adults who completed the same Study 1 intervention but lost <5% of baseline weight and in 30 adults recruited for a separate 30-week study (Study 2). RESULTS: The final algorithm retained the frequency of self-monitoring caloric intake and weight plus self-report ratings of hunger and the importance of weight-management goals compared with competing life demands. In the initial training data set, the algorithm predicted weight regain the following week with a sensitivity of 75.6% and a specificity of 45.8%; performance was similar (sensitivity: 81%-82%, specificity: 30%-33%) in testing data sets. CONCLUSIONS: Weight regain can be predicted on a proximal, week-to-week level. Future work should investigate the clinical utility of adaptive interventions for weight-loss maintenance and develop more sophisticated predictive models of weight regain.