Arthur Allignol

Competing Risks and Multistate Models with R

Data examples.- An informal introduction to hazard-based analyses.- Competing risks.- Multistate modelling of competing risks.- Nonparametric estimation.- Proportional hazards models.- Nonparametric hypothesis testing.- Further topics in competing risks.- Multistate models and their connection to competing risks.- Nonparametric estimation.- Proportional transition hazards models.- Time-dependent covariates and multistate models.- Further topics in multistate modeling.

A competing risks analysis should report results on all cause-specific hazards and cumulative incidence functions.

Competing risks endpoints are frequently encountered in hematopoietic stem cell transplantation where patients are exposed to relapse and treatment-related mortality. Both cause-specific hazards and direct models for the cumulative incidence functions have been used for analyzing such competing risks endpoints. For both approaches, the popular models are of a proportional hazards type. Such models have been used for studying prognostic factors in acute and chronic leukemias. We argue that a complete understanding of the event dynamics requires that both hazards and cumulative incidence be analyzed side by side, and that this is generally the most rigorous scientific approach to analyzing competing risks data. That is, understanding the effects of covariates on cause-specific hazards and cumulative incidence functions go hand in hand. A case study illustrates our proposal.

Teratogenicity of mycophenolate confirmed in a prospective study of the European Network of Teratology Information Services.

After maternal exposure to mycophenolate in pregnancy a high number of fetal losses and a specific pattern of birth defects consisting of microtia, cleft lip, and other anomalies have been reported. However, so far, prospective data on pregnancy outcome allowing quantitative risk assessment are missing. We report on 57 prospectively ascertained pregnancies after maternal therapy with mycophenolate (mycophenolate mofetil or mycophenolate sodium) identified by European Teratology Information Services (ETIS) through their risk consultation process. The outcome of these prospective pregnancies was as follows: 16 spontaneous abortions, 12 elective terminations of pregnancy (ETOP) (including two late terminations for multiple malformations consistent with mycophenolate embryopathy), and 29 liveborn infants. The probability of spontaneous abortion was about 45% (95% CI 29 to 66%) estimated using survival analysis technique. Six out of 29 live born infants had major congenital defects: Two with external auditory canal atresia (EACA) (with and without microtia), one with tracheo‐esophageal atresia, one with severe hydronephrosis, one with an atrial septal defect (ASD) and one with a myelomeningocele. Thus, at least four fetuses/infants of our prospective case series had a clinical phenotype consistent with mycophenolate embryopathy. Our results confirm a high incidence of major malformations (26%) after first trimester exposure to mycophenolate. Apart from exposure to mycophenololate, the underlying maternal disease and concomitant medication may also have contributed to the other poor pregnancy outcomes such as a high rate of spontaneous abortions, prematurity (62%), and low birth weight (31%).

Atypical antipsychotic drugs and pregnancy outcome: a prospective, cohort study.

Abstract Women of childbearing age are often affected with psychotic disorders, requiring the use of antipsychotic medication during pregnancy. In the present study, we prospectively followed the pregnancies of 561 women exposed to second-generation antipsychotic agents (SGAs; study cohort) and compared these to 284 pregnant women exposed to first-generation antipsychotic agents (FGAs; comparison cohort I) and to 1122 pregnant women using drugs known as not harmful to the unborn (comparison cohort II). Subjects were enrolled through the Institute’s consultation service. Major malformation rates of SGA exposed were higher compared to comparison cohort II (adjusted odds ratio, 2.17; 95% confidence interval, 1.20–3.91), possibly reflecting a detection bias concerning atrial and ventricular septal defects. Postnatal disorders occurred significantly more often in infants prenatally exposed to SGAs (15.6%) and FGAs (21.6%) compared to 4.2% of comparison cohort II. Cumulative incidences of elective terminations of pregnancy were significantly higher in both the study cohort (17%) and comparison cohort I (21%) compared to comparison cohort II (3%), whereas the rates of spontaneous abortions did not differ. The numbers of stillbirths and neonatal deaths were within the reference range. Preterm birth and low birth weight were more common in infants exposed to FGAs. To conclude, our findings did not reveal a major teratogenic risk for SGAs, making the better studied drugs of this group a treatment option during pregnancy. Because neonates exposed to SGAs or FGAs in the last gestational week are at higher risk of postnatal disorders, delivery should be planned in clinics with neonatal intensive care units.

A(H1N1)v2009: a controlled observational prospective cohort study on vaccine safety in pregnancy.

BACKGROUND A(H1N1)v2009 influenza vaccination of pregnant women was a challenge for health care providers, as little safety data were available. METHODS We prospectively followed the pregnancies of women who were vaccinated at any time during pregnancy or ≤ 4 weeks prior to conception and compared these outcomes to a control cohort matched by the estimated date of birth. Primary endpoints: rate of spontaneous abortion and major malformations. Secondary endpoints: preeclampsia, gestational age at birth, and birth weight. RESULTS Pregnancy outcome of 323 women immunized with adjuvanted or non-adjuvanted A(H1N1)v2009 influenza vaccines from 2009-09-28 to 2010-03-31 were compared to 1329 control subjects. The risk for spontaneous abortions (HR 0.89; 95% CI 0.36-2.19) and the rate of major malformations (all trimesters: OR 0.87; 95% CI 0.38-1.77; preconception and first trimester exposure: OR 0.79; 95% CI 0.13-2.64) did not vary between the two cohorts. Furthermore, there was no increase in preeclampsia, prematurity, and intrauterine growth retardation in the vaccinated cohort. CONCLUSION The results of our study do not indicate a risk for the pregnant woman and the developing embryo/fetus after H1N1 vaccination. We provide and apply methods novel in observational studies on pregnancy outcome, especially if a single dose exposure is investigated.

Hospital-acquired infections—appropriate statistical treatment is urgently needed!

Research on hospital-acquired infections (HAIs) requires the highest methodological standards to minimize the risk of bias and to avoid misleading interpretation. There are two major issues related specifically to studies in this area, namely the timing of infection and the occurrence of so-called competing risks, which deserve special attention. Just as a patient who acquires a serious infection during hospital admission needs appropriate antibiotic treatment, data being collected in studies on hospital-acquired infections need appropriate statistical analysis. We illustrate the urgent need for appropriate statistical treatment of hospital-acquired infections with some examples from recently conducted studies.The considerations presented are relevant for investigations on risk factors for HAIs as well as for outcome studies.

Boosting for high-dimensional time-to-event data with competing risks

MOTIVATION For analyzing high-dimensional time-to-event data with competing risks, tailored modeling techniques are required that consider the event of interest and the competing events at the same time, while also dealing with censoring. For low-dimensional settings, proportional hazards models for the subdistribution hazard have been proposed, but an adaptation for high-dimensional settings is missing. In addition, tools for judging the prediction performance of fitted models have to be provided. RESULTS We propose a boosting approach for fitting proportional subdistribution hazards models for high-dimensional data, that can e.g. incorporate a large number of microarray features, while also taking clinical covariates into account. Prediction performance is evaluated using bootstrap.632+ estimates of prediction error curves, adapted for the competing risks setting. This is illustrated with bladder cancer microarray data, where simultaneous consideration of both, the event of interest and competing events, allows for judging the additional predictive power gained from incorporating microarray measurements. AVAILABILITY The proposed boosting approach is implemented in the R package CoxBoost and prediction error estimation in the package peperr, both available from CRAN.

Time-dependent study entries and exposures in cohort studies can easily be sources of different and avoidable types of bias

OBJECTIVES To display and discuss the reasons and consequences of length and time-dependent bias. They might occur in presence of a time-dependent study entry or a time-dependent exposure which might change from unexposed to exposed. STUDY DESIGN AND SETTING Recalling the popular study of Oscar nominees and using a real-data example from hospital epidemiology, we give innovative and easy-to-understand graphical presentations of how these biases corrupt the analyses via distorted time-at-risk. Cumulative hazard plots and Cox proportional hazards models were used. We are building bridges to medical disciplines such as critical care medicine, hepatology, pharmaco-epidemiology, transplantation medicine, neurology, gynecology and cardiology. RESULTS In presence of time-dependent bias, the hazard ratio (comparing exposed with unexposed) is artificially underestimated. The length bias leads to an artificial underestimation of the overall hazard. When both biases coexist it can lead to different directions of biased hazard ratios. CONCLUSION Since length and time-dependent bias might occur in several medical disciplines, we conclude that understanding and awareness are the best prevention of survival bias.

Multistate modelling to estimate the excess length of stay associated with meticillin-resistant Staphylococcus aureus colonisation and infection in surgical patients

Currently available evidence on the excess length of stay (LOS) associated with nosocomial infections is limited by methodology, including time-dependent bias. To determine the excess LOS associated with nosocomial meticillin-resistant Staphylococcus aureus (MRSA) infection and colonisation, 797 MRSA-colonised, 167 MRSA-infected and 13,640 MRSA-negative surgical patients were included in a multistate model. The occurrence of MRSA infection or colonisation was the time-dependent exposure, and discharge or death was the study endpoint. The excess LOS was extracted by computing the Aalen-Johansen estimator of the matrix of transition probabilities. Multivariate Cox regression analysis was used to assess the independent effect of MRSA on excess LOS. MRSA infection prolonged LOS by 14.5 [95% confidence interval (CI): 7.8, 21.3] days compared to uninfected patients, and by 5.9 (95% CI: 0.1, 11.7) days compared to patients only colonised by MRSA. The hazard of discharge was reduced by nosocomial MRSA infection both with respect to MRSA-free patients and MRSA carriers [adjusted hazard ratio (HR): 0.69; 95% CI: 0.59, 0.81; and HR: 0.79; 95% CI: 0.65, 0.95, respectively]. MRSA carriage alone did not decrease the hazard of discharge after adjustment for confounding (HR: 1.00; 95% CI: 0.93, 1.07). Multistate modelling is a promising statistical method to evaluate the health-economic impact of nosocomial antibiotic-resistant infections.

Pregnancy outcome after paternal exposure to azathioprine/6-mercaptopurine ☆

There are only few studies with conflicting results on pregnancy outcome after paternal exposure to azathioprine or 6-mercaptopurine. In our study, pregnancy outcome of 115 prospectively followed pregnancies after paternal exposure to azathioprine or 6-mercaptopurine is compared to a control group of 341 pregnancies. The rate of major malformations was not increased (3.0% in exposed versus 2.2% in the control group). There was no specific pattern of birth defects and no indication for chromosomal aberrations in the exposed group. We observed a higher rate of elective terminations of pregnancy in the exposed group and a non-significant increase of spontaneous abortions (cumulative incidence 19% versus 13%, respectively). Further prospective studies are required to address the question of a possibly increased risk for spontaneous abortion.