Left Truncation Research Articles

Semiparametric copula method for semi-competing risks data subject to interval censoring and left truncation: Application to disability in elderly.

We aim to evaluate the marginal effects of covariates on time-to-disability in the elderly under the semi-competing risks framework, as death dependently censors disability, not vice versa. It becomes particularly challenging when time-to-disability is subject to interval censoring due to intermittent assessments. A left truncation issue arises when the age time scale is applied. We develop a flexible two-parameter copula-based semiparametric transformation model for semi-competing risks data subject to interval censoring and left truncation. The two-parameter copula quantifies both upper and lower tail dependence between two margins. The semiparametric transformation models incorporate proportional hazards and proportional odds models in both margins. We propose a two-step sieve maximum likelihood estimation procedure and study the sieve estimators' asymptotic properties. Simulations show that the proposed method corrects biases in the marginal method. We demonstrate the proposed method in a large-scale Chinese Longitudinal Healthy Longevity Study and provide new insights into preventing disability in the elderly. The proposed method could be applied to the general semi-competing risks data with intermittently assessed disease status.

The challenges of time for studies on the population effects of the COVID-19 pandemic on perinatal outcomes.

The challenges of time for studies on the population effects of the COVID-19 pandemic on perinatal outcomes.

Least squares support vector regression for complex censored data

Least squares support vector regression (LS-SVR) is a robust machine learning algorithm for small sample data. Its solution is derived from solving a set of linear equations, making the calculation process straightforward. In order to overcome the difficulties of the regression estimations when the responses are subject to interval censoring or left truncation and right censoring, two LS-SVR methods are proposed. For interval-censored data, one can easily estimate the regression functions by combining the imputation techniques and LS-SVR for right-censored data. For left-truncated and right-censored data, a weight is used to reduce the effects of truncation and censoring on the LS-SVR procedure. Simulation results show that the proposed methods can reduce regression error and yield high accuracy and stability.

Incorporating delayed entry into the joint frailty model for recurrent events and a terminal event.

In studies of recurrent events, joint modeling approaches are often needed to allow for potential dependent censoring by a terminal event such as death. Joint frailty models for recurrent events and death with an additional dependence parameter have been studied for cases in which individuals are observed from the start of the event processes. However, samples are often selected at a later time, which results in delayed entry so that only individuals who have not yet experienced the terminal event will be included. In joint frailty models such left truncation has effects on the frailty distribution that need to be accounted for in both the recurrence process and the terminal event process, if the two are associated. We demonstrate, in a comprehensive simulation study, the effects that not adjusting for late entry can have and derive the correctly adjusted marginal likelihood, which can be expressed as a ratio of two integrals over the frailty distribution. We extend the estimation method of Liu and Huang (Stat Med 27:2665-2683, 2008. https://doi.org/10.1002/sim.3077 ) to include potential left truncation. Numerical integration is performed by Gaussian quadrature, the baseline intensities are specified as piecewise constant functions, potential covariates are assumed to have multiplicative effects on the intensities. We apply the method to estimate age-specific intensities of recurrent urinary tract infections and mortality in an older population.

Characterizing quantile-varying covariate effects under the accelerated failure time model.

An important task in survival analysis is choosing a structure for the relationship between covariates of interest and the time-to-event outcome. For example, the accelerated failure time (AFT) model structures each covariate effect as a constant multiplicative shift in the outcome distribution across all survival quantiles. Though parsimonious, this structure cannot detect or capture effects that differ across quantiles of the distribution, a limitation that is analogous to only permitting proportional hazards in the Cox model. To address this, we propose a general framework for quantile-varying multiplicative effects under the AFT model. Specifically, we embed flexible regression structures within the AFT model and derive a novel formula for interpretable effects on the quantile scale. A regression standardization scheme based on the g-formula is proposed to enable the estimation of both covariate-conditional and marginal effects for an exposure of interest. We implement a user-friendly Bayesian approach for the estimation and quantification of uncertainty while accounting for left truncation and complex censoring. We emphasize the intuitive interpretation of this model through numerical and graphical tools and illustrate its performance through simulation and application to a study of Alzheimer's disease and dementia.

Efficient estimation for the proportional hazards model with left‐truncated and interval‐censored data

SummaryInterval‐censored data often arise in prospective studies involving periodical follow‐up for monitoring the failure event occurrence. In addition to censoring, left truncation also occurs if only participants who have not experienced the failure event are enrolled in the study, which clearly induces the selection bias and makes the analysis more complicated. This work provides an efficient maximum likelihood estimation approach that appropriately adjusts the biased sampling for the proportional hazards model with left‐truncated and interval‐censored data. A flexible and stable expectation–maximisation algorithm via a two‐stage data augmentation is developed to maximise the intractable likelihood function. The asymptotic properties of the proposed estimators are established with the empirical process theory. The numerical results obtained from extensive simulations suggest that the proposed method performs satisfactorily and has some prominent advantages over the competing methods. An application to a colon cancer dataset also demonstrates the usefulness of the proposed method.

New Probability Distributions in Astrophysics: XI. Left Truncation for the Topp-Leone Distribution

New Probability Distributions in Astrophysics: XI. Left Truncation for the Topp-Leone Distribution

SYSTEMATIC EXCLUSION AT STUDY COMMENCEMENT MASKS EARLIER MENOPAUSE FOR BLACK WOMEN IN THE STUDY OF WOMEN’S HEALTH ACROSS THE NATION (SWAN)

Abstract Assumptions regarding “normative” aging, including average age at onset of disease, are often based on White populations. However, evidence suggests that Black and Hispanic populations may experience “weathering” or accelerated health declines compared to Whites due to the cumulative impact of social and economic marginalization. If “weathering” leads to a differential probability of inclusion into a cohort study, it will likely misinform understanding of aging in minoritized populations. Using SWAN, a longitudinal multi-ethnic cohort of midlife women, and its cross-sectional screening study, we quantified the extent of potential selection bias at study commencement and re-estimated racial/ethnic differences in age at menopause, the main outcome of SWAN, with adjustment for various forms of potential selection bias. Left truncation was corrected for using inverse probability weighting and right censoring using multiple imputation. Two selection mechanisms were identified, eligibility and participation. Black and Hispanic women had the lowest probability of eligibility stemming from a high prevalence of surgical menopause. Their eligibility rates decreased with increasing age faster than in White women. Correcting for selection biases showed that uncorrected analyses overestimated of the median age of menopause in Black and Hispanic women, thereby underestimating racial/ethnic disparities. After adjustment, Black women had earlier natural and surgical menopause (average 1.2 years) versus White women. Overall, this study found that failure to account for different forms of selection can lead to mis-estimation of racial/ethnic disparities in health and aging. Selection bias is particularly acute at study commencement, and particularly affects minoritized populations.

AGE AT CARDIO-METABOLIC DISEASE ONSET IN A COHORT OF MIDLIFE WOMEN: SYSTEMATIC EXCLUSION MISESTIMATES THE MAGNITUDE OF RACIAL DISPARITIES

Abstract Cohort studies of aging recruit participants at an age before most of the population experiences the study outcome, to document its natural history. The age of study commencement is usually based on “normative” aging among Whites. However, “weathering” can cause accelerated health declines in minoritized populations compared to Whites due to cumulative experience of multiple forms of marginalization. Thus, considering if weathering among minoritized individuals could affect selection into cohort studies is necessary to effectively estimate and understand racial/ethnic disparities in aging and disease. Using the Study of Women’s Health Across the Nation (SWAN), a multi-ethnic longitudinal cohort, and its cross-sectional screening survey, we examine the effects of selection on the racial/ethnic differences in the age of onset of 4 cardio-metabolic outcomes (hypertension, isolated systolic hypertension, insulin resistance and diabetes). Selection at study commencement (left truncation and left censoring) had greater effects on outcomes with earlier age at onset (hypertension) and right censoring had greater effects on outcomes with later onsets (metabolic). Full adjustment led to an average 20-year decrease in predicted median age of onset for all groups across the 4 outcomes and tended to decrease the predicted disparity in age at onset. However, significantly earlier onset of each outcome for Black and Hispanic women compared to Whites remained. Not considering the full extent of selection bias in cohort studies can misinform our understanding of aging and disease, especially for minoritized populations who have higher prevalence of these leading causes of morbidity and mortality earlier in life.

Variable selection for discrete survival model with frailty in presence of left truncation and right censoring: Studying association of environmental toxicants on time-to-pregnancy.

Understanding the association between mixtures of environmental toxicants and time-to-pregnancy (TTP) is an important scientific question as sufficient evidence has emerged about the impact of individual toxicants on reproductive health and that individuals are exposed to a whole host of toxicants rather than an individual toxicant. Assessing mixtures of chemical effects on TTP poses significant statistical challenges, namely (i) TTP being a discrete survival outcome, typically subject to left truncation and right censoring, (ii) chemical exposures being strongly correlated, (iii) appropriate transformation to account for some lipid-binding chemicals, (iv) non-linear effects of some chemicals, and (v) high percentage of concentration below the limit of detection (LOD) for some chemicals. We propose a discrete frailty modeling framework (named Discnet) that allows selection of correlated covariates while appropriately addressing the methodological issues mentioned above. Discnet is shown to have better and stable false negative and false positive rates compared to alternative methods in various simulation settings. We did a detailed analysis of the pre-conception endocrine disrupting chemicals and TTP from the LIFE study and found that older females, female exposure to cotinine (smoking), DDT conferred a delay in getting pregnant, which was consistent across various approaches to account for LOD as well as non-linear associations.

Registry-based cohort study of alpha-1 antitrypsin deficiency prevalence, incidence and mortality in Denmark 2000–2018

ObjectiveTo estimate the prevalence of diagnosed alpha-1 antitrypsin deficiency (dAATD) in Denmark as of 31 December 2018, and dAATD incidence and mortality from 1 January 2000 to 31 December 2018.Study...

Overall and Subgroup Results from the Third Interim Analysis of FIRE, a Real-World Study of Ibrutinib Treatment for CLL/SLL in France

Background: FIRE is a noninterventional, multicenter, observational study of patients with a confirmed diagnosis of chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL) treated with ibrutinib in France. Aim: The aim of the study is to collect real-world data to describe ibrutinib treatment patterns and assess its effectiveness and safety in CLL/SLL. Methods: This retrospective (ret) and prospective (pro) study started on May 12, 2016, and is planned to complete after ~6 years. Eligible patients were ≥ 18 years, with a confirmed diagnosis of CLL/SLL, and who initiated ibrutinib therapy on or after November 21, 2014, (date of ibrutinib commercialization) either in previously untreated patients with del17p and/or TP53 mutation, or in patients with relapsed or refractory disease R/R. Pro patients initiated ibrutinib treatment within 30 days of study start. Ret patients initiated ibrutinib treatment more than 30 days before inclusion in the study. Primary outcome measure was progression-free survival (PFS); secondary outcome measures included overall response rate (ORR), overall survival (OS), time to next treatment (TTNT), and safety. Ret-collected adverse events (AEs) included only those considered related to ibrutinib, whereas pro-collected AEs included all treatment-emergent AEs (TEAEs). Therefore, AEs are reported separately for each cohort. Total population effectiveness results were adjusted with left truncation. Patients were treated according to usual local clinical practice. Results: We report results from the third interim analysis for patients with previously untreated and R/R CLL/SLL, with a median follow-up of 47.2 months, and with a focus on subgroup analysis. The effectiveness and safety populations comprised 394 patients (pro, n = 200; ret, n = 194) and 400 patients (pro, n = 202; ret, n = 198), respectively. Median age at ibrutinib initiation was 72 years (range, 39.0-93.0); 66.2% were male; median time from diagnosis to ibrutinib initiation was 7 years (range, 0.0-35.0). Patients received 0 (15.0%), 1 (36.3%), 2 (28.2%), or ≥ 3 (20.6%) prior lines of therapy (LOT). At baseline, complete response (CR; 50.0%) or partial response (PR; 36.2%) were recorded with the last prior therapy. Most patients (87.4%) progressed following last prior therapy, and median time from progression to initiation of ibrutinib was 2.1 months (range, 0.0-100.1). Of reported patients (n = 399), massive or symptomatic lymphadenopathy and/or splenomegaly was the most common reason (42.6%) for initiating ibrutinib treatment. Median PFS was 47.5 months (48.5/51.6 months in pro/ret). LOT and age (≤ 75 vs > 75 years) were statistically significant predictive factors for PFS. In patients with 0, 1, 2, and ≥ 3 prior LOT, median PFS was: not estimable (NE), 53.9 months, 47.5 months, and 33.5 months, respectively (Figure). At 48 months’ follow-up, ORR was 91.4%. Median TTNT was 54.6 months (NE/54.8 months in pro/ret). Median OS was NE overall and in pro/ret cohorts. Main effectiveness outcomes according to prior LOT and age are presented in the Table. Serious TEAEs related to ibrutinib were reported in 25.8% and 28.2/23.2% in pro/ret cohorts, respectively. AEs led to dose interruption (37.6/29.8%), withdrawal (30.2/22.7%), or reduction (26.7/11.1%) in the pro/ret cohorts. TEAEs of interest in the pro/ret cohorts were infection (68.8/57.1%), diarrhea (27.2/15.7%), arthralgia/myalgia (25.2/13.1%), arrhythmia (15.8/13.6% [including atrial fibrillation {9.4%/7.1%}]), hypertension (14.9/14.6%), rash (10.4/8.1%), and major bleeding (7.9/1.5%). Ibrutinib was permanently discontinued in 234 patients (58.5%); reported reasons (n = 212) were mostly due to AEs (36.8%) and disease progression (31.1%). Conclusions: In this longer follow-up of the real-world FIRE study reflecting clinical practice in France, ibrutinib was shown to be an effective treatment for patients with CLL/SLL, and patients who received ibrutinib in earlier LOT achieved better PFS. The effectiveness and adverse event profiles are consistent with observations from clinical trials. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Effectiveness and Safety of Ibrutinib in Patients with Mantle Cell Lymphoma (MCL) in Belgian Routine Clinical Practice: 3-Year Follow-up

Background: Belgian ibrutinib Real-world Data (BiRD) is a multicenter, non-interventional, observational, cohort study of adults with a confirmed diagnosis of chronic lymphocytic leukemia, MCL, or Waldenström's macroglobulinemia. Aim: To assess the effectiveness and safety of ibrutinib treatment in relapsed/refractory MCL in routine clinical practice. Methods: Eligible patients initiated reimbursed ibrutinib treatment upon or after commercial availability in Belgium (August 1, 2015); or participated in the Medical Need Program and switched to reimbursed ibrutinib with commercial availability. The primary outcome measures were progression-free survival (PFS) and overall response rate (ORR); secondary outcome measures included overall survival (OS), time to next treatment (TTNT), and safety. Data were collected both prospectively (pro) and retrospectively (ret). Ret collected adverse events (AEs) included only those considered related to ibrutinib, whereas pro collected AEs included all treatment-emergent AEs (TEAEs); therefore, TEAEs/AEs are reported separately for the pro/ret groups. Total population effectiveness results were adjusted with left truncation. We report results from the third interim analysis (March 31, 2020 cutoff) for patients with relapsed/refractory MCL, with a median follow-up of 35.8 months. Results: The effectiveness and safety populations included 76 patients (pro, n = 20; ret, n = 56). Median age at ibrutinib initiation was 74 years (range, 47-88); 76.3% were male; median time from diagnosis to ibrutinib initiation was 2.8 years (range, 0.1-17.3). Most patients received 1 (51.3%) or 2 (31.6%) prior lines of therapy, which was most frequently combination therapy (72.4%). Common therapies received in the most recent prior line of treatment were R-CHOP/R-miniCHOP (23.7%), R-DHAP/R-DHAX/R-DHAC (14.5%), alternating R-CHOP/R-DHAP (11.8%), bendamustine (10.5%), and R-bendamustine (10.5%). Complete response (CR; 58.2%) or partial response (PR; 17.9%) was the most common best response to most recent prior therapy. The majority of patients progressed (87.7%) following the most recent prior therapy, and median time from progression to initiation of ibrutinib was 0.72 months (range, 0-25.1). Lymphadenopathy and/or splenomegaly (56.6%) was the most common reason for initiating ibrutinib treatment. Ibrutinib was initiated at 560 mg/day in 85.5% of patients, and 37 patients (48.7%) had ≥ 1 dose modification. With ibrutinib treatment, median PFS was 18.6 months in the overall MCL population (Figure), and 21.5/26.8 months in the pro/ret groups. ORR was 93.4% (CR 39.5%, PR 53.9%). Median TTNT was 18.9 months (34.9/29.1 months in pro/ret) and median OS was 32.2 months (not estimable/43.2 months in pro/ret). Main effectiveness outcomes stratified according to age, prior line of therapy, and progression ≤ 24 and > 24 months from last prior therapy (for patients with 1 prior line of therapy) are listed in the Table. TEAEs/AEs and serious TEAEs/AEs were reported in 100.0/80.4% and 85.0/50.0% in the pro/ret groups, respectively. TEAEs/AEs led to dose reduction, interruption, or withdrawal in 30.0/8.9%, 45.0/30.4%, and 30.0/17.9% of patients in the pro/ret groups, respectively. TEAEs/AEs of interest in the pro/ret groups were major bleeding (10.0/3.6%), infection (55.0/58.9%), hypertension (10.0/3.6%), atrial fibrillation (15.0/12.5%), arthralgia/myalgia (10.0/7.1%), diarrhea (35.0/17.9%), and rash (10.0/8.9%). Ibrutinib was permanently discontinued in 58 patients (76.3%), most commonly due to disease progression (n = 29, 54.7% [5 patients missing reasons for discontinuation]). Conclusions: Ibrutinib is an effective treatment for patients with relapsed/refractory MCL in a representative real-world population across various patient subgroups; results are consistent with those from clinical trials (Dreyling et al. Hemasphere. 2022;6(5):e712). TEAEs/AEs were consistent with the known safety profile of ibrutinib. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Bias in the association between advanced maternal age and stillbirth using left truncated data

Restriction to analysis of births that survive past a specified gestational age (typically 20 weeks gestation) leads to biased exposure-outcome associations. This bias occurs when the cause of restriction (early pregnancy loss) is influenced by both the exposure and unmeasured factors that also affect the outcome. The aim of this study is to estimate the magnitude of bias resulting from left truncated data in the association between advanced maternal age and stillbirth. We simulated data for the causal pathway under a collider-stratification mechanism. Simulation parameters were based on an observed birth cohort from Western Australia and a range of plausible values for the prevalence of early pregnancy loss, unmeasured factor U and the odds ratios for the selection effects. Selection effects included the effects of maternal age on early pregnancy loss, U on early pregnancy loss, and U on stillbirth. We compared the simulation scenarios to the observed birth cohort that was truncated to pregnancies that survived beyond 20 gestational weeks. We found evidence of marginal downward bias, which was most prominent for women aged 40 + years. Overall, we conclude that the magnitude of bias due to left truncation is minimal in the association between advanced maternal age and stillbirth.

Evaluation of impacts of vaginal implant transmitter use in moose

AbstractVaginal implant transmitters (VITs) are a popular tool in ungulate parturition studies and have been used in moose (Alces alces), but their potential impact on animal performance has not been thoroughly assessed. We looked at potential short‐term impacts of VIT use on maternal condition through assessments of neonate calf mass, early calf survival, and parturition date. We also assessed long‐term impacts of VIT use on future reproduction and the inflammatory response after prolonged VIT retention through qualifying discharge at VIT removal and measures of acute phase proteins haptoglobin and ceruloplasmin. We captured 54 neonates and assessed over 900 potential parturition events of 278 adult moose over multiple years and we did not detect any evidence of a negative impact of VIT use on neonate calf mass, parturition date, or future reproduction. We did find elevated haptoglobin and ceruloplasmin levels in non‐pregnant individuals that had unintentional VIT retention past parturition, but this was not universal across all cases. We found a potential negative impact of VIT use on early calf survival which was likely driven by biased sampling. A bias can exist (i.e., left truncation) if sampling is mixed between cows that had a VIT (and, therefore, a known parturition time) and those without a VIT that were sampled opportunistically. Future evaluations need to overcome this bias when assessing early calf survival. With that caveat, we did not detect clear negative impacts of VIT use on moose and recommend late winter to early spring VIT insertion in pregnant females to minimize retention time.

Multifractal and Long-Term Memory of Impervious Surface Spatial Patterns in a Coastal City in China

An understanding of the multifractal and long-range dependence of impervious surfaces (IS) spatiotemporal patterns is helpful for regional environmental assessment and urban planning. Linear spectral-mixture analysis has been applied to compute IS in the coastal city of Xiamen, China, based on Landsat TM/OLI/TIRS images, and then the long-term trends and multifractal characteristics of IS patterns have been investigated using two-dimensional multifractal detrended fluctuation analysis. The IS spatial distribution displayed similar positive long-range correlations in study areas during the 1994–2015 period. IS have a long-memory characteristic within a certain spatial range, with an increase in the value of a pixel likely to cause an increment in the value of its neighbors. The multifractality of the IS distribution increased in Xiamen City during 1994–2015, but was lower than those in Xiamen Island. The multifractal spectra in Xiamen City vary in shape between years, capturing its evolution from right truncation with a long left tail to left truncations and long right tails and then a symmetrical shape. The fractal structure for Xiamen Island exhibits similar patterns of long right tails and left truncations. Economic and political considerations coupled with natural geographic conditions dominate the long-range trends in the IS spatial patterns.

A Comparison of Methods Examining Time-to-Readmission in the First Year of Life.

Readmissions analyses typically calculate time-to-readmission relative to hospital discharge. For newborns, this definition can be challenging when comparing groups with disparate birth stays. We compare 2 approaches to calculate readmissions and examine 1 year readmissions for newborns with versus without neonatal opioid withdrawal syndrome (NOWS; mean length of stay = 17 vs 2 days). Using birth discharge data from the Pediatric Health Information System (PHIS), we compared crude and adjusted 1 year readmissions by NOWS diagnosis using Cox regression models predicting time-to-readmission from: (1) birth discharge; and (2) birth (day-of-life), with left truncation allowing for delayed entry into the at-risk period at birth discharge. We included N = 155 885 birth discharges (n = 1467 with NOWS). At 1 year, 10% of infants with NOWS versus 6% without had been readmitted. Readmission risk was highest within 1 week since discharge or birth for newborns without NOWS, whereas those with NOWS were at higher risk later into infancy. NOWS was associated with a higher adjusted hazard of 1 year readmissions since discharge (adjusted hazard ratio [aHR]=1.58; 95% CI: 1.20-2.08) and a higher adjusted hazard of 1 year readmissions since birth (aHR = 1.56; 95% CI: 1.21-2.03). Estimates vary by choice of index date, particularly at early time-points, converging later into infancy. Our findings underscore the importance of methodological decisions for newborn readmissions. Although results were similar at 1 year with nearly identical adjusted hazards, approaches differed substantially through the neonatal period.

Effectiveness and Safety of Ibrutinib for Chronic Lymphocytic Leukemia in Routine Clinical Practice: 3-Year Follow-up of the Belgian Ibrutinib Real-World Data (BiRD) Study

The multicenter observational BiRD study investigated the real-world effectiveness and safety of ibrutinib in patients with chronic lymphocytic leukemia (CLL), mantle cell lymphoma (MCL) and Waldenström’s macroglobulinemia (WM) in Belgium. This interim analysis reports results for patients with CLL, with a median follow-up of 34 months. Overall, patients had predominantly relapsed/refractory disease (73%) and were elderly (median age 72 years) with high-risk features such as del17p and/or TP53 mutations (59%). Patients were included either prospectively or retrospectively, and the total patient population effectiveness results were adjusted with left truncation. In the effectiveness population (N = 221: prospective, n = 71; retrospective, n = 150), the overall response rate was 90.0%. Median progression-free survival was 38.3 months (prospective, not estimable; retrospective, 51.5 months) and median overall survival was not yet estimable in the total, prospective and retrospective groups. Treatment-emergent adverse events (TEAEs) for the prospective and retrospective groups are reported separately. Any-grade TEAEs of interest in the prospective/retrospective groups included infections (67.1%/60.1%), diarrhea (20.5%/10.5%), hypertension (16.4%/9.8%) and atrial fibrillation (12.3%/7.2%). Major bleeding was reported in 5.5%/3.3% of prospective/retrospective patients, with little difference observed between those receiving versus not receiving antithrombotic treatment. Discontinuations due to toxicity were reported in 10.5% of patients. Results from this interim analysis show treatment with ibrutinib to be effective and tolerable, with no new safety signals observed. Future analyses will report on longer-term follow-up.

A flexible parametric approach for analyzing arbitrarily censored data that are potentially subject to left truncation under the proportional hazards model.

The proportional hazards (PH) model is, arguably, the most popular model for the analysis of lifetime data arising from epidemiological studies, among many others. In such applications, analysts may be faced with censored outcomes and/or studies which institute enrollment criterion leading to left truncation. Censored outcomes arise when the event of interest is not observed but rather is known relevant to an observation time(s). Left truncated data occur in studies that exclude participants who have experienced the event prior to being enrolled in the study. If not accounted for, both of these features can lead to inaccurate inferences about the population under study. Thus, to overcome this challenge, herein we propose a novel unified PH model that can be used to accommodate both of these features. In particular, our approach can seamlessly analyze exactly observed failure times along with interval-censored observations, while aptly accounting for left truncation. To facilitate model fitting, an expectation-maximization algorithm is developed through the introduction of carefully structured latent random variables. To provide modeling flexibility, a monotone spline representation is used to approximate the cumulative baseline hazard function. The performance of our methodology is evaluated through a simulation study and is further illustrated through the analysis of two motivating data sets; one that involves child mortality in Nigeria and the other prostate cancer.

Comparison of patient characteristics and long-term mortality between transferred and non-transferred COVID-19 patients in Dutch intensive care units: A national cohort study.

BackgroundCOVID‐19 patients were often transferred to other intensive care units (ICUs) to prevent that ICUs would reach their maximum capacity. However, transferring ICU patients is not free of risk. We aim to compare the characteristics and outcomes of transferred versus non‐transferred COVID‐19 ICU patients in the Netherlands.MethodsWe included adult COVID‐19 patients admitted to Dutch ICUs between March 1, 2020 and July 1, 2021. We compared the patient characteristics and outcomes of non‐transferred and transferred patients and used a Directed Acyclic Graph to identify potential confounders in the relationship between transfer and mortality. We used these confounders in a Cox regression model with left truncation at the day of transfer to analyze the effect of transfers on mortality during the 180 days after ICU admission.ResultsWe included 10,209 patients: 7395 non‐transferred and 2814 (27.6%) transferred patients. In both groups, the median age was 64 years. Transferred patients were mostly ventilated at ICU admission (83.7% vs. 56.2%) and included a larger proportion of low‐risk patients (70.3% vs. 66.5% with mortality risk <30%). After adjusting for age, APACHE IV mortality probability, BMI, mechanical ventilation, and vasoactive medication use, the hazard of mortality during the first 180 days was similar for transferred patients compared to non‐transferred patients (HR [95% CI] = 0.99 [0.91–1.08]).ConclusionsTransferred COVID‐19 patients are more often mechanically ventilated and are less severely ill compared to non‐transferred patients. Furthermore, transferring critically ill COVID‐19 patients in the Netherlands is not associated with mortality during the first 180 days after ICU admission.