causalCmprsk: An R package for nonparametric and Cox-based estimation of average treatment effects in competing risks data

BACKGROUND AND OBJECTIVE

Competing risks data arise in both observational and experimental clinical studies with time-to-event outcomes, when each patient might follow one of the multiple mutually exclusive competing paths. Ignoring competing risks in the analysis can result in biased conclusions. In addition, possible confounding bias of the treatment-outcome relationship has to be addressed, when estimating treatment effects from observational data. In order to provide tools for estimation of average treatment effects on time-to-event outcomes in the presence of competing risks, we developed the R package causalCmprsk. We illustrate the package functionality in the estimation of effects of a right heart catheterization procedure on discharge and in-hospital death from observational data.

METHODS

The causalCmprsk package implements an inverse probability weighting estimation approach, aiming to emulate baseline randomization and alleviate possible treatment selection bias. The package allows for different types of weights, representing different target populations. causalCmprsk builds on existing methods from survival analysis and adapts them to the causal analysis in non-parametric and semi-parametric frameworks.

RESULTS

The causalCmprsk package has two main functions: fit.cox assumes a semiparametric structural Cox proportional hazards model for the counterfactual cause-specific hazards, while fit.nonpar does not impose any structural assumptions. In both frameworks, causalCmprsk implements estimators of (i) absolute risks for each treatment arm, e.g., cumulative hazards or cumulative incidence functions, and (ii) relative treatment effects, e.g., hazard ratios, or restricted mean time differences. The latter treatment effect measure translates the treatment effect from probability into more intuitive time domain and allows the user to quantify, for example, by how many days or months the treatment accelerates the recovery or postpones illness or death.

CONCLUSIONS

The causalCmprsk package provides a convenient and useful tool for causal analysis of competing risks data. It allows the user to distinguish between different causes of the end of follow-up and provides several time-varying measures of treatment effects. The package is accompanied by a vignette that contains more details, examples and code, making the package accessible even for non-expert users.

Association of BCG Vaccine Treatment With Death and Dementia in Patients With Non-Muscle-Invasive Bladder Cancer

Importance

The BCG vaccine-used worldwide to prevent tuberculosis-confers multiple nonspecific beneficial effects, and intravesical BCG vaccine is currently the recommended treatment for non-muscle-invasive bladder cancer (NMIBC). Moreover, BCG vaccine has been hypothesized to reduce the risk of Alzheimer disease and related dementias (ADRD), but previous studies have been limited by sample size, study design, or analyses.

Objective

To evaluate whether intravesical BCG vaccine exposure is associated with a decreased incidence of ADRD in a cohort of patients with NMIBC while accounting for death as a competing event.

Design, Setting, and Participants

This cohort study was performed in patients aged 50 years or older initially diagnosed with NMIBC between May 28, 1987, and May 6, 2021, treated within the Mass General Brigham health care system. The study included a 15-year follow-up of individuals (BCG vaccine treated or controls) whose condition did not clinically progress to muscle-invasive cancer within 8 weeks and did not have an ADRD diagnosis within the first year after the NMIBC diagnosis. Data analysis was conducted from April 18, 2021, to March 28, 2023.

Main Outcomes and Measures

The main outcome was time to ADRD onset identified using diagnosis codes and medications. Cause-specific hazard ratios (HRs) were estimated using Cox proportional hazards regression after adjusting for confounders (age, sex, and Charlson Comorbidity Index) using inverse probability scores weighting.

Results

In this cohort study including 6467 individuals initially diagnosed with NMIBC between 1987 and 2021, 3388 patients underwent BCG vaccine treatment (mean [SD] age, 69.89 [9.28] years; 2605 [76.9%] men) and 3079 served as controls (mean [SD] age, 70.73 [10.00] years; 2176 [70.7%] men). Treatment with BCG vaccine was associated with a lower rate of ADRD (HR, 0.80; 95% CI, 0.69-0.99), with an even lower rate of ADRD in patients aged 70 years or older at the time of BCG vaccine treatment (HR, 0.74; 95% CI, 0.60-0.91). In competing risks analysis, BCG vaccine was associated with a lower risk of ADRD (5-year risk difference, -0.011; 95% CI, -0.019 to -0.003) and a decreased risk of death in patients without an earlier diagnosis of ADRD (5-year risk difference, -0.056; 95% CI, -0.075 to -0.037).

Conclusions and Relevance

In this study, BCG vaccine was associated with a significantly lower rate and risk of ADRD in a cohort of patients with bladder cancer when accounting for death as a competing event. However, the risk differences varied with time.

The Protective Effect of Moderate Maternal Peanut Consumption on Peanut Sensitization and Allergy

BACKGROUND

The LEAP trial showed that the early introduction of peanuts in the diet of infants at risk for peanut allergies prevents peanut allergy. The effect of maternal consumption of peanuts on subsequent peanut sensitization or peanut allergy has not been studied to date.

OBJECTIVE

To determine whether maternal consumption of peanut protein while breastfeeding protects against peanut allergic outcomes, absent the consumption of peanuts in the infant.

METHODS

We performed an analysis on the data from the peanut avoidance arm of the LEAP study to discern the effects of maternal consumption of peanuts while pregnant and breastfeeding on an infant's peanut allergic outcomes.

RESULTS

Of the 303 infants in the avoidance group, 31 mothers consumed more than 5 g of peanut per week, 69 consumed less than 5 g of peanut per week and 181 did not consume peanut while breastfeeding. Peanut sensitization (p=0.027) and peanut allergy (p=0.073) occurred less frequently in infants whose mothers consumed a moderate amount of peanuts while breastfeeding when compared to those who either did not consume peanuts while breastfeeding or those who consumed a large amount of peanuts when breastfeeding. Ethnicity (OR 0.47, p = 0.046, 95% CI 0.22- 0.99), baseline peanut SPT stratum (OR 4.87, p = 0.00, 95% CI 2.13- 11.12), no maternal peanut consumption while breastfeeding (OR 3.25, p = 0.008, 95% CI 1.36- 7.77), and baseline SCORAD >40 (OR 2.78, p = 0.007, 95% CI 1.32- 5.85) were all significant contributors to peanut sensitization or allergy at 60 months of age.

CONCLUSION

Moderate consumption (<5 g/w) of peanuts while breastfeeding provides a significant protective effect against peanut sensitization and a noticeable but not statistically significant protective effect against peanut allergy later on in life in high-risk infants in the context of delayed peanut introduction.

Causal inference in medical records and complementary systems pharmacology for metformin drug repurposing towards dementia

Metformin, a diabetes drug with anti-aging cellular responses, has complex actions that may alter dementia onset. Mixed results are emerging from prior observational studies. To address this complexity, we deploy a causal inference approach accounting for the competing risk of death in emulated clinical trials using two distinct electronic health record systems. In intention-to-treat analyses, metformin use associates with lower hazard of all-cause mortality and lower cause-specific hazard of dementia onset, after accounting for prolonged survival, relative to sulfonylureas. In parallel systems pharmacology studies, the expression of two AD-related proteins, APOE and SPP1, was suppressed by pharmacologic concentrations of metformin in differentiated human neural cells, relative to a sulfonylurea. Together, our findings suggest that metformin might reduce the risk of dementia in diabetes patients through mechanisms beyond glycemic control, and that SPP1 is a candidate biomarker for metformin's action in the brain.

Nonparametric estimation of the survival distribution under covariate-induced dependent truncation

There is often delayed entry into observational studies, which results in left truncation. In the estimation of the distribution of time-to-event from left-truncated data, standard survival analysis methods require quasi-independence between the truncation time and event time. Incorrectly assuming quasi-independence may lead to biased estimation. We address the problem of estimation of the survival distribution when dependence between the event time and its left truncation time is induced by shared covariates. We introduce propensity scores for truncated data and propose two inverse probability weighting methods that adjust for both truncation and dependence, if all of the shared covariates are measured. The proposed methods additionally allow for right censoring. We evaluate the proposed methods in simulations, conduct sensitivity analyses, and provide guidelines for use in practice. We illustrate our approach in application to data from a central nervous system lymphoma study. The proposed methods are implemented in the R package, depLT.

How are adults who had Perthes’ disease functioning?

Aims

Perthes’ disease (PD) is a childhood hip disorder that can affect the quality of life in adulthood due to femoral head deformity and osteoarthritis. There is very little data on how PD patients function as adults, especially from the patients’ perspective. The purpose of this study was to collect treatment history, demographic details, the University of California, Los Angeles activity score (UCLA), the 36-Item Short Form survey (SF-36) score, and the Hip disability and Osteoarthritis Outcome score (HOOS) of adults who had PD using a web-based survey method and to compare their outcomes to the outcomes from an age- and sex-matched normative population.

Methods

The English REDCap-based survey was made available on a PD study group website. The survey included childhood and adult PD history, UCLA, SF-36, and HOOS. Of the 1,182 participants who completed the survey, the 921 participants who did not have a total hip arthroplasty are the focus of this study. The mean age at survey was 38 years (SD 12) and the mean duration from age at PD onset to survey participation was 30.8 years (SD 12.6).

Results

In comparison to a normative population, the PD participants had significantly lower HOOS scores across all five scales (p < 0.001) for all age groups. Similarly, SF-36 scores of the participants were significantly lower (p < 0.001) for all scales except for age groups > 55 years. Overall, females, obese participants, those who reported no treatment in childhood, and those with age of onset > 11 years had significantly worse SF-36 and HOOS scores. Pairwise correlations showed a strong positive correlation within HOOS scales and between HOOS scales and SF-36 scales, indicating construct validity.

Conclusion

Adult PD participants had significantly worse pain, physical, mental, and social health than an age- and sex-matched normative cohort. The study reveals a significant burden of disease on the adult participants of the survey, especially females. Cite this article: Bone Joint J 2022;104-B(12):1304–1312.

What is the adult experience of Perthes' disease? : initial findings from an international web-based survey

AIMS

Perthes' disease is an uncommon hip disorder with limited data on the long-term outcomes in adulthood. We partnered with community-based foundations and utilized web-based survey methodology to develop the Adult Perthes Survey, which includes demographics, childhood and adult Perthes' disease history, the University of California Los Angeles (UCLA) Activity Scale item, Short Form-36, the Hip disability and Osteoarthritis Outcome Score, and a body pain diagram. Here we investigate the following questions: 1) what is the feasibility of obtaining > 1,000 survey responses from adults who had Perthes' disease using a web-based platform?; and 2) what are the baseline characteristics and demographic composition of our sample?

METHODS

The survey link was available publicly for 15 months and advertised among support groups. Of 1,505 participants who attempted the Adult Perthes survey, 1,182 completed it with a median timeframe of 11 minutes (IQR 8.633 to 14.72). Participants who dropped out were similar to those who completed the survey on several fixed variables. Participants represented 45 countries including the USA (n = 570; 48%), UK (n = 295; 25%), Australia (n = 133; 11%), and Canada (n = 46; 4%). Of the 1,182 respondents, 58% were female and the mean age was 39 years (SD 12.6).

RESULTS

Ages at onset of Perthes' disease were < six years (n = 512; 43%), six to seven years (n = 321; 27%), eight to 11 years (n = 261; 22%), and > 11 years (n = 76; 6%), similar to the known age distribution of Perthes' disease. During childhood, 40% (n = 476) of respondents had at least one surgery. Bracing, weightbearing restriction, and absence of any treatment varied significantly between USA and non-USA respondents (p < 0.001, p = 0.002, and p < 0.001, respectively). As adults, 22% (n = 261) had at least one total hip arthroplasty, and 30% (n = 347) had any type of surgery; both more commonly reported among women (p = 0.002).

CONCLUSION

While there are limitations due to self-sampling, our study shows the feasibility of obtaining a large set of patient-reported data from adults who had childhood Perthes' from multiple countries. Cite this article: Bone Jt Open 2022;3(5):404-414.

Concentration-response relationships between hourly particulate matter and ischemic events: A case-crossover analysis of effect modification by season and air-mass origin

Most studies linking cardiovascular disease with particulate matter (PM) exposures have focused on total mass concentrations, regardless of their origin. However, the origin of an air mass is inherently linked to particle composition and possible toxicity. We examine how the concentration-response relation between hourly PM exposure and ischemic events is modified by air-mass origin and season. Using telemedicine data, we conducted a case-crossover study of 1855 confirmed ischemic cardiac events in Israel (2005-2013). Based on measurements at three fixed-sites in Tel Aviv and Haifa, ambient PM with diameter < 2.5 μm (PM_2.5) and 2.5-10 μm (PM_10-2.5) concentrations during the hours before event onset were compared with matched control periods using conditional logistic regression that allowed for non-linearity. We also examined effect modification of these associations based on the geographical origin of each air mass by season. Independent of the geographical origin of the air mass, we observed concentration-response curves that were supralinear. For example, the overall odds ratios (ORs) of ischemic events for an increase of 10-μg/m³ in the 2-h average of PM_10-2.5 were 1.08 (95% confidence interval (CI): 1.03-1.14) and 1.00 (0.99-1.01) at the median (17.8 μg/m³) and 95th percentile (82.3 μg/m³) values, respectively. Associations were strongest at low levels of PM_10-2.5 when air comes from central Europe in the summer (OR: 1.27; 95% CI: 1.06, 1.52). Our study demonstrates that hourly associations between PM_2.5 and PM_10-2.5 and ischemic cardiac events are supralinear during diverse pollution conditions in a single population that experiences a wide range of exposure levels.

Bevacizumab Reduces Permeability and Concurrent Temozolomide Delivery in a Subset of Patients with Recurrent Glioblastoma

PURPOSE

Targeting tumor blood vessels is an attractive therapy in glioblastoma (GBM), but the mechanism of action of these agents and how they modulate delivery of concomitant chemotherapy are not clear in humans. We sought to elucidate how bevacizumab modulates tumor vasculature and the impact those vascular changes have on drug delivery in patients with recurrent GBM.

EXPERIMENTAL DESIGN

Temozolomide was labeled with [11C], and serial PET-MRI scans were performed in patients with recurrent GBM treated with bevacizumab and daily temozolomide. PET-MRI scans were performed prior to the first bevacizumab dose, 1 day after the first dose, and prior to the third dose of bevacizumab. We calculated tumor volume, vascular permeability (K trans), perfusion (cerebral blood flow), and the standardized uptake values (SUV) of [11C] temozolomide within the tumor.

RESULTS

Twelve patients were enrolled, resulting in 23 evaluable scans. Within the entire contrast-enhancing tumor volume, both temozolomide uptake and vascular permeability decreased after initiation of bevacizumab in most patients, whereas change in perfusion was more variable. In subregions of the tumor where permeability was low and the blood-brain barrier not compromised, increased perfusion correlated with increased temozolomide uptake.

CONCLUSIONS

Bevacizumab led to a decrease in permeability and concomitant delivery of temozolomide. However, in subregions of the tumor where permeability was low, increased perfusion improved delivery of temozolomide, suggesting that perfusion may modulate the delivery of chemotherapy in certain settings. These results support exploring whether lower doses of bevacizumab improve perfusion and concomitant drug delivery.

Inverse probability weighting methods for Cox regression with right-truncated data

Right-truncated data arise when observations are ascertained retrospectively, and only subjects who experience the event of interest by the time of sampling are selected. Such a selection scheme, without adjustment, leads to biased estimation of covariate effects in the Cox proportional hazards model. The existing methods for fitting the Cox model to right-truncated data, which are based on the maximization of the likelihood or solving estimating equations with respect to both the baseline hazard function and the covariate effects, are numerically challenging. We consider two alternative simple methods based on inverse probability weighting (IPW) estimating equations, which allow consistent estimation of covariate effects under a positivity assumption and avoid estimation of baseline hazards. We discuss problems of identifiability and consistency that arise when positivity does not hold and show that although the partial tests for null effects based on these IPW methods can be used in some settings even in the absence of positivity, they are not valid in general. We propose adjusted estimating equations that incorporate the probability of observation when it is known from external sources, which results in consistent estimation. We compare the methods in simulations and apply them to the analyses of human immunodeficiency virus latency.

Nonidentifiability in the presence of factorization for truncated data

A time to event, [Formula: see text], is left-truncated by [Formula: see text] if [Formula: see text] can be observed only if [Formula: see text]. This often results in oversampling of large values of [Formula: see text], and necessitates adjustment of estimation procedures to avoid bias. Simple risk-set adjustments can be made to standard risk-set-based estimators to accommodate left truncation when [Formula: see text] and [Formula: see text] are quasi-independent. We derive a weaker factorization condition for the conditional distribution of [Formula: see text] given [Formula: see text] in the observable region that permits risk-set adjustment for estimation of the distribution of [Formula: see text], but not of the distribution of [Formula: see text]. Quasi-independence results when the analogous factorization condition for [Formula: see text] given [Formula: see text] holds also, in which case the distributions of [Formula: see text] and [Formula: see text] are easily estimated. While we can test for factorization, if the test does not reject, we cannot identify which factorization condition holds, or whether quasi-independence holds. Hence we require an unverifiable assumption in order to estimate the distribution of [Formula: see text] or [Formula: see text] based on truncated data. This contrasts with the common understanding that truncation is different from censoring in requiring no unverifiable assumptions for estimation. We illustrate these concepts through a simulation of left-truncated and right-censored data.

Fine and Coarse Particulate Matter Exposures and Associations with Acute Cardiac Events among Participants in a Telemedicine Service: A Case-Crossover Study

BACKGROUND

Subclinical cardiovascular changes have been associated with ambient particulate matter (PM) exposures within hours. Although the U.S. Environmental Protection Agency continues to look for additional evidence of effects associated with sub-daily PM exposure, this information is still limited because most studies of clinical events have lacked data on the onset time of symptoms to assess rapid increased risk.

OBJECTIVE

Our objective was to investigate associations between sub-daily exposures to PM and acute cardiac events using telemedicine data.

METHODS

We conducted a case-crossover study among telemedicine participants [Formula: see text] of age who called a service center for cardiac-related symptoms and were transferred to a hospital in Tel Aviv and Haifa, Israel (2002-2013). Ambient [Formula: see text] and [Formula: see text] measured by monitors located in each city during the hours before the patient called with symptoms were compared with matched control periods. We investigated the sensitivity of these associations to more accurate symptom onset time and greater certainty of diagnosis.

RESULTS

We captured 12,661 calls from 7,617 subscribers experiencing ischemic (19%), arrhythmic (31%), or nonspecific (49%) cardiac events. PM concentrations were associated with small increases in the odds of cardiac events. For example, odds ratios for any cardiac event in association with a [Formula: see text] increase in 6-h and 24-h average [Formula: see text] were 1.008 [95% confidence interval (CI): 0.998, 1.018] and 1.006 (95% CI: 0.995, 1.018), respectively, and for [Formula: see text] were 1.003 (95% CI: 1.001, 1.006) and 1.003 (95% CI: 1.000, 1.007), respectively. Associations were stronger when using exposures matched to the call time rather than calendar date and for events with higher certainty of the diagnosis.

CONCLUSIONS

Our analysis of telemedicine data suggests that risks of cardiac events in telemedicine participants [Formula: see text] of age may increase within hours of PM exposures. https://doi.org/10.1289/EHP2596.

Nonparametric estimation in the illness-death model using prevalent data

We study nonparametric estimation of the illness-death model using left-truncated and right-censored data. The general aim is to estimate the multivariate distribution of a progressive multi-state process. Maximum likelihood estimation under censoring suffers from problems of uniqueness and consistency, so instead we review and extend methods that are based on inverse probability weighting. For univariate left-truncated and right-censored data, nonparametric maximum likelihood estimation can be considerably improved when exploiting knowledge on the truncation distribution. We aim to examine the gain in using such knowledge for inverse probability weighting estimators in the illness-death framework. Additionally, we compare the weights that use truncation variables with the weights that integrate them out, showing, by simulation, that the latter performs more stably and efficiently. We apply the methods to intensive care units data collected in a cross-sectional design, and discuss how the estimators can be easily modified to more general multi-state models.

Comparing estimation approaches for the illness-death model under left truncation and right censoring

Left-truncated data arise when lifetimes are observed only if they are larger than independent truncation times. For example, in a cross-sectional sampling, only individuals who live long enough to be present on the sampling day are observed. There are several ways to perform statistical inference under this setting. One can do the following: (i) use an unconditional approach, (ii) condition on the value of the truncation variable, or (iii) condition on all the history up to the time of truncation. The latter two approaches are equivalent when analyzing univariate survival outcomes but differ under the multi-state framework. In this paper, we consider the illness-death model and compare between the three estimation approaches in a parametric regression framework. We show that approach (ii) is more efficient than the standard approach (iii), although it requires more computational effort. Approach (i) is the most efficient approach, but it requires knowledge on the distribution of the truncation variable and hence is less robust. The methods are compared using a theoretical example and simulations and are applied to intensive care units data collected in a cross-sectional design, where the illness state corresponds to a bloodstream infection.