Semiparametric Methods for Semi-competing Risks Problem with Censoring and Truncation

Studies of chronic life-threatening diseases often involve both mortality and morbidity. In observational studies, the data may also be subject to administrative left truncation and right censoring. Since mortality and morbidity may be correlated and mortality may censor morbidity, the Lynden-Bell estimator for left truncated and right censored data may be biased for estimating the marginal survival function of the non-terminal event. We propose a semiparametric estimator for this survival function based on a joint model for the two time-to-event variables, which utilizes the gamma frailty specification in the region of the observable data. Firstly, we develop a novel estimator for the gamma frailty parameter under left truncation. Using this estimator, we then derive a closed form estimator for the marginal distribution of the non-terminal event. The large sample properties of the estimators are established via asymptotic theory. The methodology performs well with moderate sample sizes, both in simulations and in an analysis of data from a diabetes registry.

‘Modeling and Analysis of Competing Risks Data’

there has been much research on analyzing various forms of competing risks data. Nevertheless, there are several occasions in survival studies, where the existing models and methodologies are inadequate for the analysis competing risks data. ldentifiabilty problem and various types of and censoring induce more complications in the analysis of competing risks data than in classical survival analysis. Parametric models are not adequate for the analysis of competing risks data since the assumptions about the underlying lifetime distributions may not hold well. Motivated by this, in the present study. we develop some new inference procedures, which are completely distribution free for the analysis of competing risks data.

Longitudinal outcomes of high-risk human papillomavirus (HPV) infections as competing-risks events following cervical HPV test at baseline visit in the *NIS-LAMS** cohort

Background: The complex natural history of human papillomavirus (HPV) infections following a single HPV test can be modeled as competing-risks events (i.e., no-, transient- or persistent infection) in a longitudinal setting. The covariates associated with these compet ng events have not been previously assessed using competing-risks regression models. Objectives: To gain further insights in the outcomes of cervical HPV infections, we used univariate- and multivariate competing-risks regression models to assess the covariaies associated with these competing events. Study Design and Methods: Covariates associated with three competing outcomes (no-, transient- or persistent HR-HPV infection) were analysed in a sub-cohort of 1,865 women prospectively followed-up in the NIS (n = 3,187) and LAMS Study (n = 12,114). Results: In multivariate competing-risks models (with two other outcomes as competing events), permanently HR-HPV negative outcome was significantly predicted only by the clearance of ASCUS+Pap during FU, while three independent covariates predicted transient HR-HPV infections: i) number of recent (< 12 months) sexual partners (risk increased), ii) previous Pap screening history (protective), and history of previous CIN (increased risk). The two most powerful predictors of persistent HR-HPV infections were persistent ASCUS+Pap (risk increased), and previous Pap screening history (protective). In pair-wise comparisons, number of recent sexual partners and previous CIN history increase the probability of transient HR-HPV infection against the HR-HPV negative competing event, while previous Pap screening history is protective. Persistent ASCUS+Pap during FU and no previous Pap screening history are significantly associated with the persistent HR-HPV outcome (compared both with i) always negative, and ii) transient events), whereas multiparity is protective. Conclusions: Different covariates are associated with the three main outcomes of cervical HPV infections. The most significant covariates of each competing events are probably distinct enough to enable constructing of a risk-profile for each main outcome.

Covariates of high-risk human papillomavirus (HPV) infections are distinct for incident CIN1, CIN2 and CIN3 as disclosed by competing-risks regression models

Background: In addition to the oncogenic human papillomavirus (HPV), several cofactors are needed in cervical carcinogenesis, but whether the HPV covariates associated with incident i) CIN1 are different from those of incident ii) CIN2 and iii) CIN3 needs further assessment. Objectives: To gain further insights into the true biological differences between CIN1, CIN2 and CIN3, we assessed HPV covariates associated with incident CIN1, CIN2, and CIN3. Study Design and Methods: HPV covariates associated with progression to CIN1, CIN2 and CIN3 were analysed in the combined cohort of the NIS (n = 3,187) and LAMS study (n = 12,114), using competing-risks regression models (in panel data) for baseline HR-HPV-positive women (n = 1,105), who represent a sub-cohort of all 1,865 women prospectively followed-up in these two studies. Results: Altogether, 90 (4.8%), 39 (2.1%) and 14 (1.4%) cases progressed to CIN1, CIN2, and CIN3, respectively. Among these baseline HR-HPV-positive women, the risk profiles of incident GIN I, CIN2 and CIN3 were unique in that completely different HPV covariates were associated with progression to CIN1, CIN2 and CIN3, irrespective which categories (non-progression, CIN1, CIN2, CIN3 or all) were used as competing-risks events in univariate and multivariate models. Conclusions: These data confirm our previous analysis based on multinomial regression models implicating that distinct covariates of HR-HPV are associated with progression to CIN1, CIN2 and CIN3. This emphasises true biological differences between the three grades of GIN, which revisits the concept of combining CIN2 with CIN3 or with CIN1 in histological classification or used as a common end-point, e.g., in HPV vaccine trials.

Impact of age and comorbidities on long-term survival of patients with high-risk prostate cancer treated with radical prostatectomy: a multi-institutional competing-risks analysis

Survival after surgical treatment using competing-risk analysis has been previously examined in patients with prostate cancer (PCa). However, the combined effect of age and comorbidities has not been assessed in patients with high-risk PCa who might have heterogeneous rates of competing mortality despite the presence of aggressive disease.

Interpreting and comparing risks in the presence of competing events

Competing events are common in medical research. Ignoring them in the statistical analysis can easily lead to flawed results and conclusions. This article uses a real dataset and a simple simulation to show how standard analysis fails and how such data should be analysed

Unemployment Duration: Competing and Defective Risks

This paper examines the determinants of unemployment duration in a competing risks framework with two destination states: inactivity and employment. The innovation is the recognition of defective risks. A polynomial hazard function is used to differentiate between two possible sources of infinite durations. The first is produced by a random process of unlucky draws, the second by workers rejecting a destination state. The evidence favors the mover-stayer model over the search model. Refinement of the former approach, using a more flexible baseline hazard function, produces a robust and more convincing explanation for positive and zero transition rates out of unemployment.

Multilevel competing risk models to evaluate the risk of nosocomial infection

Introduction Risk factor analyses for nosocomial infections (NIs) are complex. First, due to competing events for NI, the association between risk factors of NI as measured using hazard rates may not coincide with the association using cumulative probability (risk). Second, patients from the same intensive care unit (ICU) who share the same environmental exposure are likely to be more similar with regard to risk factors predisposing to a NI than patients from different ICUs. We aimed to develop an analytical approach to account for both features and to use it to evaluate associations between patient- and ICU-level characteristics with both rates of NI and competing risks and with the cumulative probability of infection. Methods We considered a multicenter database of 159 intensive care units containing 109,216 admissions (813,739 admission-days) from the Spanish HELICS-ENVIN ICU network. We analyzed the data using two models: an etiologic model (rate based) and a predictive model (risk based). In both models, random effects (shared frailties) were introduced to assess heterogeneity. Death and discharge without NI are treated as competing events for NI. Results There was a large heterogeneity across ICUs in NI hazard rates, which remained after accounting for multilevel risk factors, meaning that there are remaining unobserved ICU-specific factors that influence NI occurrence. Heterogeneity across ICUs in terms of cumulative probability of NI was even more pronounced. Several risk factors had markedly different associations in the rate-based and risk-based models. For some, the associations differed in magnitude. For example, high Acute Physiology and Chronic Health Evaluation II (APACHE II) scores were associated with modest increases in the rate of nosocomial bacteremia, but large increases in the risk. Others differed in sign, for example respiratory vs cardiovascular diagnostic categories were associated with a reduced rate of nosocomial bacteremia, but an increased risk. Conclusions A combination of competing risks and multilevel models is required to understand direct and indirect risk factors for NI and distinguish patient-level from ICU-level factors.

Regression models for bivariate survival data

Multivariate lifetime data arise in various forms including recurrent event data when individuals are followed to observe the sequence of occurrences of a certain type of event; correlated lifetime when an individual is followed for the occurrence of two or more types of events, or when distinct individuals have dependent event times. In most studies there are covariates such as treatments, group indicators, individual characteristics, or environmental conditions, whose relationship to lifetime is of interest. This leads to a consideration of regression models.The well known Cox proportional hazards model and its variations, using the marginal hazard functions employed for the analysis of multivariate survival data in literature are not sufficient to explain the complete dependence structure of pair of lifetimes on the covariate vector. Motivated by this, in Chapter 2, we introduced a bivariate proportional hazards model using vector hazard function of Johnson and Kotz (1975), in which the covariates under study have different effect on two components of the vector hazard function. The proposed model is useful in real life situations to study the dependence structure of pair of lifetimes on the covariate vector . The well known partial likelihood approach is used for the estimation of parameter vectors. We then introduced a bivariate proportional hazards model for gap times of recurrent events in Chapter 3. The model incorporates both marginal and joint dependence of the distribution of gap times on the covariate vector . In many fields of application, mean residual life function is considered superior concept than the hazard function. Motivated by this, in Chapter 4, we considered a new semi-parametric model, bivariate proportional mean residual life time model, to assess the relationship between mean residual life and covariates for gap time of recurrent events. The counting process approach is used for the inference procedures of the gap time of recurrent events. In many survival studies, the distribution of lifetime may depend on the distribution of censoring time. In Chapter 5, we introduced a proportional hazards model for duration times and developed inference procedures under dependent (informative) censoring. In Chapter 6, we introduced a bivariate proportional hazards model for competing risks data under right censoring. The asymptotic properties of the estimators of the parameters of different models developed in previous chapters, were studied. The proposed models were applied to various real life situations.

Maternal exposure to ambient temperature and the risk of preterm birth and stillbirth in Brisbane, Australia

Almost 10% of all births are preterm and 2.2% are stillbirths globally. Recent research has suggested that environmental factors may be a contributory cause to these adverse birth outcomes. The authors examined the relationship between ambient temperature and preterm birth and stillbirth in Brisbane, Australia between 2005 and 2009 (n = 101,870). They used a Cox proportional hazard model with live birth and stillbirth as competing risks. They also examined if there were periods of the pregnancy where exposure to high temperatures had a greater effect. Exposure to higher ambient temperatures during pregnancy increased the risk of stillbirth. The hazard ratio for stillbirth was 0.3 at 12 °C relative to the reference temperature at 21 °C. The temperature effect was greatest for fetuses of less than 36 weeks of gestation. There was an association between higher temperature and shorter gestation, as the hazard ratio for live birth was 0.96 at 15 °C and 1.02 at 25 °C. This effect was greatest at later gestational ages. The results provide strong evidence of an association between increased temperature and increased risk of stillbirth and shorter gestations.

The influence of ambient temperature on birth outcomes in Brisbane, Australia

Lately, there has been increasing interest in the association between temperature and adverse birth outcomes including preterm birth (PTB) and stillbirth. PTB is a major predictor of many diseases later in life, and stillbirth is a devastating event for parents and families. The aim of this study was to assess the seasonal pattern of adverse birth outcomes, and to examine possible associations of maternal exposure to temperature with PTB and stillbirth. We also aimed to identify if there were any periods of the pregnancy where exposure to temperature was particularly harmful. A retrospective cohort study design was used and we retrieved individual birth records from the Queensland Health Perinatal Data Collection Unit for all singleton births (excluding twins and triplets) delivered in Brisbane between 1 July 2005 and 30 June 2009. We obtained weather data (including hourly relative humidity, minimum and maximum temperature) and air-pollution data (including PM10, SO2 and O3) from the Queensland Department of Environment and Resource Management. We used survival analyses with the time-dependent variables of temperature, humidity and air pollution, and the competing risks of stillbirth and live birth. To assess the monthly pattern of the birth outcomes, we fitted month of pregnancy as a time-dependent variable. We examined the seasonal pattern of the birth outcomes and the relationship between exposure to high or low temperatures and birth outcomes over the four lag weeks before birth. We further stratified by categorisation of PTB: extreme PTB (< 28 weeks of gestation), PTB (28–36 weeks of gestation), and term birth (≥ 37 weeks of gestation). Lastly, we examined the effect of temperature variation in each week of the pregnancy on birth outcomes. There was a bimodal seasonal pattern in gestation length. After adjusting for temperature, the seasonal pattern changed from bimodal, to only one peak in winter. The risk of stillbirth was statistically significant lower in March compared with January. After adjusting for temperature, the March trough was still statistically significant and there was a peak in risk (not statistically significant) in winter. There was an acute effect of temperature on gestational age and stillbirth with a shortened gestation for increasing temperature from 15 °C to 25 °C over the last four weeks before birth. For stillbirth, we found an increasing risk with increasing temperatures from 12 °C to approximately 20 °C, and no change in risk at temperatures above 20 °C. Certain periods of the pregnancy were more vulnerable to temperature variation. The risk of PTB (28–36 weeks of gestation) increased as temperatures increased above 21 °C. For stillbirth, the fetus was most vulnerable at less than 28 weeks of gestation, but there were also effects in 28–36 weeks of gestation. For fetuses of more than 37 weeks of gestation, increasing temperatures did not increase the risk of stillbirth. We did not find any adverse affects of cold temperature on birth outcomes in this cohort. My findings contribute to knowledge of the relationship between temperature and birth outcomes. In the context of climate change, this is particularly important. The results may have implications for public health policy and planning, as they indicate that pregnant women would decrease their risk of adverse birth outcomes by avoiding exposure to high temperatures and seeking cool environments during hot days.