On interquantile smoothness of censored quantile regression with induced smoothing

Quantile regression has emerged as a useful and effective tool in modeling survival data, especially for cases where noises demonstrate heterogeneity. Despite recent advancements, non-smooth components involved in censored quantile regression estimators may often yield numerically unstable results, which, in turn, lead to potentially self-contradicting conclusions. We propose an estimating equation-based approach to obtain consistent estimators of the regression coefficients of interest via the induced smoothing technique to circumvent the difficulty. Our proposed estimator can be shown to be asymptotically equivalent to its original unsmoothed version, whose consistency and asymptotic normality can be readily established. Extensions to handle functional covariate data and recurrent event data are also discussed. To alleviate the heavy computational burden of bootstrap-based variance estimation, we also propose an efficient resampling procedure that reduces the computational time considerably. Our numerical studies demonstrate that our proposed estimator provides substantially smoother model parameter estimates across different quantile levels and can achieve better statistical efficiency compared to a plain estimator under various finite-sample settings. The proposed method is also illustrated via four survival datasets, including the HMO (health maintenance organizations) HIV (human immunodeficiency virus) data, the primary biliary cirrhosis (PBC) data, and so forth.

Event history analysis of dynamic networks

Statistical analysis on networks has received growing attention due to demand from various emerging applications. In dynamic networks, one of the key interests is to model the event history of time-stamped interactions among nodes. We model dynamic directed networks via multivariate counting processes. A pseudo partial likelihood approach is exploited to capture the network dependence structure. Asymptotic results are established. Numerical experiments are performed to demonstrate the effectiveness of our proposal.

Censored quantile regression model with time-varying covariates under length-biased sampling

Quantile regression is a flexible and effective tool for modeling survival data and its relationship with important covariates, which often vary over time. Informative right censoring of data from the prevalent cohort within the population often results in length-biased observations. We propose an estimating equation-based approach to obtain consistent estimators of the regression coefficients of interest based on length-biased observations with time-dependent covariates. In addition, inspired by Zeng and Lin 2008, we also develop a more numerically stable procedure for variance estimation. Large sample properties including consistency and asymptotic normality of the proposed estimator are established. Numerical studies presented demonstrate convincing performance of the proposed estimator under various settings. The application of the proposed method is demonstrated using the Oscar dataset.

A validated composite model to predict risk of curve progression in adolescent idiopathic scoliosis

BACKGROUND

In adolescent idiopathic scoliosis (AIS), the continuous search for effective prognostication of significant curve progression at the initial clinical consultation to inform decision for timely treatment and to avoid unnecessary overtreatment remains a big challenge as evidence of the multifactorial etiopathogenic nature is increasingly reported. This study aimed to formulate a composite model composed of clinical parameters and circulating markers in the prediction of curve progression.

METHOD

This is a two-phase study consisting of an exploration cohort (120 AIS, mean Cobb angle of 25°± 8.5 at their first clinical visit) and a validation cohort (51 AIS, mean Cobb angle of 23° ± 5.0° at the first visit). Patients with AIS were followed-up for a minimum of six years to formulate a composite model for prediction. At the first visit, clinical parameters were collected from routine clinical practice, and circulating markers were assayed from blood.

FINDING

We constructed the composite predictive model for curve progression to severe Cobb angle > 40° with a high HR of 27.9 (95% CI of 6.55 to 119.16). The area under curve of the composite model is higher than that of individual parameters used in current clinical practice. The model was validated by an independent cohort and achieved a sensitivity of 72.7% and a specificity of 90%.

INTERPRETATION

This is the first study proposing and validating a prognostic composite model consisting of clinical and circulating parameters which could quantitatively evaluate the probability of curve progression to a severe curvature in AIS at the initial consultation. Further validation in clinic will facilitate application of composite model in assisting objective clinical decision.

A Prospective Cohort Study of Fecal miR-223 and miR-451a as Noninvasive and Specific Biomarkers for Diagnosis of Necrotizing Enterocolitis in Preterm Infants

OBJECTIVE

The objective of this study is to evaluate the usefulness of fecal microRNA (miR)-223 and miR-451a, as novel noninvasive biomarkers for early diagnosis of necrotizing enterocolitis (NEC) in preterm infants.

METHODS

Among the top-listed target miRNAs in our previous differential microarray analysis, miR-223 and miR-451a were quantified in a pilot validation case-controlled study (NEC vs. non-NEC/nonsepsis infants; n = 6 in each group). A definitive prospective cohort study (n = 218) further assessed their clinical usefulness as noninvasive and specific diagnostic biomarkers. Fecal calprotectin was quantified in parallel for comparison.

RESULTS

Of 43 proven NEC cases in the cohort study, 24 (55.8%) had fecal samples recovered within the first 3 days of clinical presentation. Fecal miRNA-223 (10.5 fold), miR-451a (4.5 fold), and calprotectin (2.1 fold) concentrations were significant higher in NEC compared with the non-NEC group (p < 0.009). Accepting a minimum sensitivity of 0.75, the positive predictive values (PPVs) ranged between 0.19 and 0.20. Combining fecal biomarkers and CRP (Day 1) could marginally increase the PPVs (0.31-0.34) but adversely lowered the sensitivity (0.54-0.63).

CONCLUSIONS

Although fecal miRNA biomarkers and calprotectin concentrations were significantly higher in the NEC group, the considerable overlapping of concentrations between groups and low recovery of stool specimens within 72 h of clinical presentation rendered fecal noninvasive tests of limited clinical value in guiding diagnosis of NEC during the acute phase. A further study is underway to evaluate their roles in surveillance for predicting high-risk premature infants developing NEC.

Plasma miR-1290 Is a Novel and Specific Biomarker for Early Diagnosis of Necrotizing Enterocolitis-Biomarker Discovery with Prospective Cohort Evaluation

OBJECTIVE

To discover specific circulating microRNA (miRNA) biomarkers for the early differentiation of necrotizing enterocolitis (NEC) from neonatal sepsis and inflammatory conditions.

STUDY DESIGN

The study comprised 3 distinct phases: differential microarray analysis to compare plasma miRNA expression profiles of NEC vs sepsis and non-NEC/nonsepsis cases, a case-control study to quantify dysregulated miRNAs as potential specific biomarkers of NEC, and a prospective cohort study to assess the diagnostic usefulness of the best miRNA biomarker(s).

RESULTS

A distinct miRNA expression profile was observed in the NEC compared with the sepsis and non-NEC/nonsepsis groups. miR-1290, miR-1246, and miR-375 were discovered to be specific biomarkers of NEC in the case-control study. In the cohort study (n = 301), plasma miR-1290 (day 0; >220 copies/µL) provided the greatest diagnostic usefulness for identifying both mild medical and severe surgical NEC cases. Of 20 infants with miR-1290 >650 copies/µL, 15 were diagnosed with NEC. Incorporating C-reactive protein (day 1; >15.8 mg/L) for cases with intermediate levels (220-650 copies/µL) in a 2-stage algorithm further optimized the diagnostic profile with a sensitivity of 0.83, a specificity of 0.96, a positive predictive value of 0.75, and a negative predictive value of 0.98. Importantly, 7 of 36 infants with NEC (19.4%) could be diagnosed 7.8-32.2 hours earlier (median, 13.3 hours) using miR-1290.

CONCLUSIONS

Plasma miR-1290 is a novel and specific biomarker that can effectively differentiate NEC cases from neonatal sepsis. miR-1290 facilitates neonatologists to confidently and timely reach a decision for early transfer of sick infants with NEC from community-based hospitals to tertiary surgical centers.

ESTIMATION OF A MONOTONE DENSITY IN S-SAMPLE BIASED SAMPLING MODELS

We study the nonparametric estimation of a decreasing density function g 0 in a general s-sample biased sampling model with weight (or bias) functions wi for i = 1, …, s. The determination of the monotone maximum likelihood estimator ĝn and its asymptotic distribution, except for the case when s = 1, has been long missing in the literature due to certain non-standard structures of the likelihood function, such as non-separability and a lack of strictly positive second order derivatives of the negative of the log-likelihood function. The existence, uniqueness, self-characterization, consistency of ĝn and its asymptotic distribution at a fixed point are established in this article. To overcome the barriers caused by non-standard likelihood structures, for instance, we show the tightness of ĝn via a purely analytic argument instead of an intrinsic geometric one and propose an indirect approach to attain the n -rate of convergence of the linear functional ∫ wi ĝn.

Estimation and Inference of Quantile Regression for Survival Data Under Biased Sampling

Biased sampling occurs frequently in economics, epidemiology, and medical studies either by design or due to data collecting mechanism. Failing to take into account the sampling bias usually leads to incorrect inference. We propose a unified estimation procedure and a computationally fast resampling method to make statistical inference for quantile regression with survival data under general biased sampling schemes, including but not limited to the length-biased sampling, the case-cohort design, and variants thereof. We establish the uniform consistency and weak convergence of the proposed estimator as a process of the quantile level. We also investigate more efficient estimation using the generalized method of moments and derive the asymptotic normality. We further propose a new resampling method for inference, which differs from alternative procedures in that it does not require to repeatedly solve estimating equations. It is proved that the resampling method consistently estimates the asymptotic covariance matrix. The unified framework proposed in this article provides researchers and practitioners a convenient tool for analyzing data collected from various designs. Simulation studies and applications to real datasets are presented for illustration. Supplementary materials for this article are available online.

Regulation of Circulating Hematopoietic Stem/Progenitor Cells in Preterm Infants with Septicemia

Preterm infants are at high risk of developing severe sepsis. Circulating hematopoietic stem and progenitor cells (HSPCs; CD45⁺CD34⁺) have been suggested to play a vital role in the host immunological defense against invading pathogens. The objectives were to investigate the regulation of circulating HSPCs in preterm infants during infection episodes, and to assess the relationship of CD45⁺CD34⁺ cells with immunological mediators and differential leukocyte populations. First, we conducted a cross-sectional case-control study comparing these parameters among infected infants (n = 23), gestational and postnatal age-matched noninfected infants (n = 46), and "healthy" control (CTL) infants (n = 12). Second, we investigated the longitudinal change of CD45⁺CD34⁺ cell concentrations in infected infants before, during, and after an infection episode, and compared them with the other two groups. Our cross-sectional results showed that CD45⁺CD34⁺ cell count and percentage were significantly reduced in infected infants during systemic infection, compared with the noninfected or CTL infants. There were significant positive correlation between levels of CD45⁺CD34⁺ cells and lymphocytes or monocytes, and significant negative correlation between CD45⁺CD34⁺ cells and neutrophils or interleukin (IL)-6 in infected infants. Longitudinal analysis showed that changes of CD45⁺CD34⁺ cells at the onset of sepsis relative to levels 1 week prior and 1 week postsepsis in infected infants were significantly different from those changes in the corresponding time points for the other two groups. Our findings suggested that circulating HSPCs were dynamically regulated during septicemia and could play an important role in the defense mechanism, plausibly contributing to replenishment of leukocytes during sepsis in preterm infants.

Attitudes of Parents and Health Care Workers to Major Surgery for High-Risk Preterm Infants

OBJECTIVES

To assess preferences of health care workers (HCWs) and parents of term and preterm infants to adverse health outcomes, and how perceived surgical mortality influences decision-making.

STUDY DESIGN

A total of 536 participants (157 HCWs, 201 parents of term infants, and 178 parents of preterm infants) were recruited to take part in a structured interview. Preferences related to treatment of a critically ill preterm infant with necrotizing enterocolitis were measured by health state rank permutation analysis and pivotal risk analysis. Between-group and subgroup comparisons were performed.

RESULTS

HCWs rank adverse health states less favorably than parents of preterm infants, consistently ranking 2 of the most adverse health states worse than death. Pivotal risk values of HCWs for all health states were consistently the lowest of the 3 groups. High operative mortality was associated uniformly with reduction in pivotal risks for all groups both in favorable and adverse health states. Subgroup analyses revealed significant discrepancies in preferences between fathers and mothers as well as doctors and nurses. Regular religious practice was significantly associated with increased pivotal risks in parental subgroups.

CONCLUSIONS

As discrepancies in health state preferences existed between subgroups (ie, doctors vs nurses, mothers vs fathers) and perceived operative mortality consistently biased parental and HCW health state preferences, we recommend that HCWs should first identify differences regarding patient management before interviewing the parents together. HCWs should be aware of inadvertently biasing parents when discussing the risks and outcomes of surgery in conjunction with the overall long-term prognosis of the underlying condition.

Design and analysis of clinical trials in the presence of delayed treatment effect

In clinical trials with survival endpoint, it is common to observe an overlap between two Kaplan-Meier curves of treatment and control groups during the early stage of the trials, indicating a potential delayed treatment effect. Formulas have been derived for the asymptotic power of the log-rank test in the presence of delayed treatment effect and its accompanying sample size calculation. In this paper, we first reformulate the alternative hypothesis with the delayed treatment effect in a rescaled time domain, which can yield a simplified sample size formula for the log-rank test in this context. We further propose an intersection-union test to examine the efficacy of treatment with delayed effect and show it to be more powerful than the log-rank test. Simulation studies are conducted to demonstrate the proposed methods.

Accelerated failure time model under general biased sampling scheme

Right-censored time-to-event data are sometimes observed from a (sub)cohort of patients whose survival times can be subject to outcome-dependent sampling schemes. In this paper, we propose a unified estimation method for semiparametric accelerated failure time models under general biased estimating schemes. The proposed estimator of the regression covariates is developed upon a bias-offsetting weighting scheme and is proved to be consistent and asymptotically normally distributed. Large sample properties for the estimator are also derived. Using rank-based monotone estimating functions for the regression parameters, we find that the estimating equations can be easily solved via convex optimization. The methods are confirmed through simulations and illustrated by application to real datasets on various sampling schemes including length-bias sampling, the case-cohort design and its variants.

A Unified Approach to Semiparametric Transformation Models under General Biased Sampling Schemes

We propose a unified estimation method for semiparametric linear transformation models under general biased sampling schemes. The new estimator is obtained from a set of counting process-based unbiased estimating equations, developed through introducing a general weighting scheme that offsets the sampling bias. The usual asymptotic properties, including consistency and asymptotic normality, are established under suitable regularity conditions. A closed-form formula is derived for the limiting variance and the plug-in estimator is shown to be consistent. We demonstrate the unified approach through the special cases of left truncation, length-bias, the case-cohort design and variants thereof. Simulation studies and applications to real data sets are presented.

H5N1 influenza viruses isolated from geese in Southeastern China: evidence for genetic reassortment and interspecies transmission to ducks

The H5N1 viruses (H5N1/97) associated with the "bird-flu" incident in the Hong Kong SAR have not been isolated since the slaughter of poultry in December 1997 brought that outbreak to an end. Recent evidence points to this virus as having arisen through a reassortment of a number of precursor avian viruses and a virus related to Goose/Guangdong/1/96 (H5N1) (Gs/Gd/96) was the likely donor of the H5 hemagglutinin. We characterize the Goose/Guangdong/1/96-like viruses isolated from geese and ducks imported into Hong Kong in the year 2000. Antigenically and genetically, these recent H5N1 viruses fall into two groups, one mainly associated with geese, and the other, recently transmitted to ducks. Further, viruses isolated from a goose and a duck in December 2000 have acquired NS, PA, M, and PB2 genes from the aquatic avian influenza gene pool through reassortment. For pandemic preparedness, it is important to monitor whether these reassortant viruses have the capacity for interspecies transmission to terrestrial poultry or mammals.