Search

For estimating the causal effect of treatment exposure on the occurrence of adverse events, inverse probability weights (IPW) can be used in marginal structural models to correct for time-dependent confounding. The R package ipw allows IPW estimation by modeling the relationship between the exposure and confounders via several regression models, among which is the Cox model. For right-censored data and time-dependent exposures such as treatment switches, the ipw package allows a single switch, assuming that patients are treated once and for all. However, to accommodate multiple switches, we extend this package by implementing a function that allows for multiple and intermittent exposure status in the estimation of IPW using a survival model. This extension allows for the whole exposure treatment trajectory in the estimation of IPW. The impact of the estimated weights on the estimated causal effect, with both methods, is assessed in a simulation study. Then, the function is illustrated on a real dataset from a nationwide prospective observational cohort including patients with inflammatory bowel disease. In this study, patients received one or multiple medications (thiopurines, methotrexate, and anti-TNF) over time. We used a Cox marginal structural model to assess the effect of thiopurines exposure on the cause-specific hazard for cancer incidence considering other treatments as confounding factors. To this end, we used our extended function which is available online in the Supporting Information., (© 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

Hospitalization rates over time of childhood cancer survivors (CCS) provide insight into the burden of unfavorable health conditions on CCS and health care resources. The objective of our study was to examine trends in hospitalizations of CCS and risk factors in comparison with the general population. We performed a medical record linkage study of a cohort of 1564 ≥five-year CCS with national registers. We obtained a random sample of the general population matched on year of birth, gender and calendar year per CCS retrieved. We quantified and compared hospitalization rates of CCS and reference persons from 1995 until 2005, and we analyzed risk factors for hospitalization within the CCS cohort with multivariable Poisson models. We retrieved hospitalization information from 1382 CCS and 25583 reference persons. The overall relative hospitalization rate (RHR) was 2.2 (95%CI:1.9-2.5) for CCS compared to reference persons. CCS with central nervous system and solid tumors had highest RHRs. Hospitalization rates in CCS were increased compared to reference persons up to at least 30 years after primary diagnosis, with highest rates 5-10 and 20-30 years after primary cancer. RHRs were highest for hospitalizations due to neoplasms (10.7; 95%CI:7.1-16.3) and endocrine/nutritional/metabolic disorders (7.3; 95%CI:4.6-11.7). Female gender (P<0.001), radiotherapy to head and/or neck (P<0.001) or thorax and/or abdomen (P = 0.03) and surgery (P = 0.01) were associated with higher hospitalization rates in CCS. In conclusion, CCS have increased hospitalization rates compared to the general population, up to at least 30 years after primary cancer treatment. These findings imply a high and long-term burden of unfavorable health conditions after childhood cancer on survivors and health care resources.

This paper presents an extension of the joint modeling strategy for the case of multiple longitudinal outcomes and repeated infections of different types over time, motivated by postkidney transplantation data. Our model comprises two parts linked by shared latent terms. On the one hand is a multivariate mixed linear model with random effects, where a low-rank thin-plate spline function is incorporated to collect the nonlinear behavior of the different profiles over time. On the other hand is an infection-specific Cox model, where the dependence between different types of infections and the related times of infection is through a random effect associated with each infection type to catch the within dependence and a shared frailty parameter to capture the dependence between infection types. We implemented the parameterization used in joint models which uses the fitted longitudinal measurements as time-dependent covariates in a relative risk model. Our proposed model was implemented in OpenBUGS using the MCMC approach., (© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

When estimating the probability of natural conception from observational data on couples with an unfulfilled child wish, the start of assisted reproductive therapy (ART) is a competing event that cannot be assumed to be independent of natural conception. In clinical practice, interest lies in the probability of natural conception in the absence of ART, as this probability determines the need for therapy. We thus want to estimate the marginal cumulative pregnancy distribution. Without assumptions on the dependence structure between the two competing events, this marginal distribution is not identifiable. We first use inverse probability of censoring weighting assuming that the factors influencing the choice to start ART are known. Then, we parameterize the event distributions for conception and for start of ART and use copulas to account for the dependency between both events. By using these two ways of correcting for the dependent risk of treatment, we obtain a plausible estimation region for the cumulative pregnancy curve and for the prognostic effect of tubal tests., (Copyright © 2014 John Wiley & Sons, Ltd.)

Markers are internal host factors that measure the current disease or recovery status of an individual. Individuals with more advanced disease progression are more likely to drop out, e.g. because they die. Marker data after dropout are missing. Such missingness is certainly not completely at random. A mixed effects model can be used if missingness of the marker data depends on measured marker values only (missing at random). If missingness is not at random, such models yield biased results. We describe various approaches that jointly model the marker development and dropout risk and may eliminate bias. One example of such a model is a random effects selection model. Based on a real data set with frequent follow-up, we compare results from a random effects model and a random effects selection model. Results are remarkably similar. In a simulation study, we investigate how the bias in the parameter estimates from a random effects model depends on the frequency of measurements and the time between the last measurement and the dropout or censoring time. Results from the simulation study confirm that the bias is small if follow-up is frequent.

Background: Declining residual renal function, as indicated by the glomerular filtration rate (GFR), is associated with an increased mortality risk in patients with end-stage renal disease (ESRD) on dialysis., Methods: We monitored GFR and mortality in 1800 haemodialysis (HD) and peritoneal dialysis (PD) patients in 1996-2006. We used a marginal structural model to estimate the causal effects both of GFR when it was not completely lost and of the subsequent full loss of GFR on mortality, avoiding the drawbacks of standard regression models that include covariates to adjust for confounding. Instead, effect estimates were adjusted for possible baseline and time-varying confounders using inverse probability weighting., Results: We estimated a hazard ratio (HR) corresponding to the effect of the full loss of GFR on mortality, as compared to not having fully lost GFR, of 1.50 [95% confidence interval (CI) 1.09-2.07]. The HR corresponding to the effect of GFR when GFR is not (yet) fully lost on mortality was 0.97 (95% CI 0.92-1.02) (per mL/min/1.73 m(2)). We found no significant difference in the effect of GFR on mortality between patients starting on PD and HD., Conclusions: Preventing or delaying the full loss of GFR can improve survival in dialysis patients. This supports the importance that is given to the effect of treatment options for patients with ESRD on the rate of decline of the residual renal function.

Summary The standard estimator for the cause-specific cumulative incidence function in a competing risks setting with left truncated and/or right censored data can be written in two alternative forms. One is a weighted empirical cumulative distribution function and the other a product-limit estimator. This equivalence suggests an alternative view of the analysis of time-to-event data with left truncation and right censoring: individuals who are still at risk or experienced an earlier competing event receive weights from the censoring and truncation mechanisms. As a consequence, inference on the cumulative scale can be performed using weighted versions of standard procedures. This holds for estimation of the cause-specific cumulative incidence function as well as for estimation of the regression parameters in the Fine and Gray proportional subdistribution hazards model. We show that, with the appropriate filtration, a martingale property holds that allows deriving asymptotic results for the proportional subdistribution hazards model in the same way as for the standard Cox proportional hazards model. Estimation of the cause-specific cumulative incidence function and regression on the subdistribution hazard can be performed using standard software for survival analysis if the software allows for inclusion of time-dependent weights. We show the implementation in the R statistical package. The proportional subdistribution hazards model is used to investigate the effect of calendar period as a deterministic external time varying covariate, which can be seen as a special case of left truncation, on AIDS related and non-AIDS related cumulative mortality., (© 2010, The International Biometric Society.)

When comparing the causal effect of peritoneal dialysis (PD) and hemodialysis (HD) treatment on lowering mortality in renal patients, using observational data, it is necessary to adjust for different forms of confounding and informative censoring. Both the type of dialysis treatment that is started with and mortality are affected by baseline covariates. Longitudinal and baseline variables can affect both the probability of switching from one type of dialysis to the other, and mortality. Longitudinal and baseline variables can also affect the probability of receiving a kidney transplant, possibly causing informative censoring. Adjusting for longitudinal variables by including them as covariates in a regression model potentially causes bias, for instance by losing a possible indirect effect of dialysis on mortality via these longitudinal variables. Instead, we fitted a marginal structural model (MSM) to estimate the causal effect of dialysis type, adjusted for confounding and informative censoring. We used the MSM to compare the hazard of death as well as cumulative survival between the potential treatment trajectories "always PD" and "always HD" over time, conditional on age and diabetes mellitus status. We used inverse probability weighting (IPW) to fit the MSM.

Objective: Reducing viral load, highly active antiretroviral therapy has the potential to limit onwards transmission of HIV-1 and thus help contain epidemic spread. However, increases in risk behaviour and resurgent epidemics have been widely reported post-highly active antiretroviral therapy. The aim of this study was to quantify the impact that highly active antiretroviral therapy had on the epidemic., Design: We focus on the HIV-1 epidemic among men who have sex with men in the Netherlands, which has been well documented over the past 20 years within several long-standing national surveillance programs., Methods: We used a mathematical model including highly active antiretroviral therapy use and estimated the changes in risk behaviour and diagnosis rate needed to explain annual data on HIV and AIDS diagnoses., Results: We show that the reproduction number R(t), a measure of the state of the epidemic, declined early on from initial values above two and was maintained below one from 1985 to 2000. Since 1996, when highly active antiretroviral therapy became widely used, the risk behaviour rate has increased 66%, resulting in an increase of R(t) to 1.04 in the latest period 2000-2004 (95% confidence interval 0.98-1.09) near or just above the threshold for a self-sustaining epidemic. Hypothetical scenario analysis shows that the epidemiological benefits of highly active antiretroviral therapy and earlier diagnosis on incidence have been entirely offset by increases in the risk behaviour rate., Conclusion: We provide the first detailed quantitative analysis of the HIV epidemic in a well defined population and find a resurgent epidemic in the era of highly active antiretroviral therapy, most likely predominantly caused by increasing sexual risk behaviour.