Survival models for right censored breast cancer data: theory, application and comparison

Background: Censoring frequently occurs in disease data analysis, which is a key characteristic of time to failure modeling. Typically, time to failure studies are conducted through non-parametric and semi-parametric modelling techniques. Parametric models provide more efficient estimates, but are seldomly used, because of some of the limitations and assumptions which need to be fulfilled to apply them. The aim of this study is to illustrate the theoretical and application limitations and performance of different flexible and standard parametric models to evaluate the prognostic value for mortality risk of breast cancer after recurrence among women. Methods: This article describes the theoretical properties of flexible parametric models and compares their performances to standard parametric models, by studying mortality in women diagnosed with breast cancer. We describe how time to failure data may be analyzed with nonlinear flexible models. In this regard, we apply fractional polynomials, spline models, piecewise exponential models, and piecewise exponential additive mixed models. We also illustrate properties of standard parametric models. All analyses have been conducted with multiple covariates to identify significant predictors. Information criteria have been used to evaluate performances of models. Results: Fractional polynomial and spline-based generalized additive models work well in capturing local fluctuations. Parameter estimation with a piecewise exponential additive mixed model (PAMM) as an extension of the piecewise exponential modelling (PEM) approach automatically penalizes model complexity, which is very helpful to avoid over fitting. Conclusions: Flexible parametric time to failure models are more efficient than standard parametric time to failure models. By incorporating time dependent covariates, PAMM is a good approach to perform in-depth studies of predictors over different finite intervals of follow-up time. Until now, this approach is rarely used in time to failure right censored studies.

Rheumatoid Arthritis, as a Clinical Disease, but Not Rheumatoid Arthritis-associated Autoimmunity is Linked to Cardiovascular Events

Background: Rheumatoid Arthritis (RA) is characterized by increased cardiovascular (CV) mortality. CV events are particularly high in patients with RA-specific autoimmunity, including rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA), raising the question whether RA-specific autoimmunity itself is associated with CV events.Methods: New CV events (myocardial infarction, stroke or death by CV cause) were recorded in 20,625 subjects of the Electricité de France – Gaz de France (GAZEL) cohort. Self-reported RA cases in the GAZEL cohort were validated by phone interview on the basis of a specific questionnaire. In 1,618 subjects, in whom serum was available, RF and ACPA were measured. A piecewise exponential Poisson regression was used to analyze the association of CV events with presence of RA as well as RA-specific autoimmunity (without RA).Results: CV events in GAZEL were associated with age, male sex, smoking, hypertension, hyperlipidemia and diabetes mellitus (HR from 1.06 to 1.87, p < 0.05). Forty-two confirmed RA cases were identified. Confirmed RA was significantly associated with CV risk increase (HR of 3.03; 95% CI: 1.13-8.11, p=0.03) independently of conventional CV risk factors. One hundred seventy-eight subjects showed RF or ACPA positivity without presence of RA. CV events were not associated with ACPA positivity (HR: 1.52, 95% CI: 0.47-4.84, p= 0.48) or RF positivity (HR: 1.15, 95% CI: 0.55-2.40, p= 0.70) in the absence of RA.Conclusions: RA, as a clinical chronic inflammatory disease, but not mere positivity for RF or ACPA in the absence of clinical disease is associated with increased CV risk.

Apprenticeship non-completion in Germany: a money matter?

German establishments heavily rely on the apprenticeship system for skill supply. With one in four apprenticeship contracts ending before successful completion, it is in the interest of establishments and policy-makers to determine factors, which reduce non-completion. This paper investigates the role of apprenticeship wages and income prospects after completion for apprenticeship non-completion in Germany. For this purpose, this study identifies incidences of apprenticeship non-completion in a large sample of administrative data on employment biographies and estimates a piecewise exponential model of the non-completion hazard with shared frailties by occupations. The results suggest a robust and significant association with both apprenticeship wages and skilled worker wages. All else at means, apprenticeships which are paid 5% more than the mean apprenticeship wage, on average have a 0.8 percentage points higher estimated survival rate. In turn, an apprenticeship expected to lead to a skilled job that is paid 5% above average, has an estimated survival rate, which is 3.1 percentage points higher on average. These findings highlight the importance of income prospects for apprenticeship non-completion.

Morbidity and Mortality Rates of Boer and Central Highland Goats at Ataye Crossbreeding Program Research Site: Non-parametric Survival Analysis and Piecewise Exponential Model

Non-parametric survival analysis and piecewise exponential model (PEM) was used to estimate prevalence and incidence of goat mortality, to identify major clinical causes of morbidity and mortality related disease, and to investigate animal and environmental related risk factors affecting goat mortality at Ataye boer goat breeding and evaluation research site. A total of 671 kids and 347 adult (yearling) age goats were used for the analysis of non-parametric survival and piecewise exponential model for survival, mortality incidence rate and causes of morbidity and mortality analysis. The mortality incidence rate of kids and adult goats were 0.638 and 0.302 per animal year respectively. The 25th, 50th and 75th percentile of survival time of kids were 5, 157 and 1,274 days respectively and of adult goats were 280, 828 and 1,557 days respectively. The present mortality rate is relatively larger than reports of boer cross breeding and evaluation research sites in Ethiopia as well as goat mortality prevalence abroad. Gastro-intestinal related diseases, pneumonia, weak kid, agalactia, mismothering and hear water (cowdriosis) were most important causes of mortality. Constant piecewise exponential regression analysis of risk factors indicates that breed, kid birth weight (BWT), doe post-partum weight (PPWT), birth type, birth year and precipitation variables were associated with (p-value < 0.05) kid mortality rate. Pure boer kids compared with CHG cross boer goat, are 2.505 times at higher probability of mortality (p ≤ 0.001). A 1 kilo gram increase of kid birth weight and dam PPWT reduces mortality probability by 32.5% (p-value ≤ 0.001) and 6.4% (p-value ≤ 0.001) respectively. Twin birth kids are1.512 times higher rate of mortality (p-value = 0.001) compared with single born kids. A one-millilitre increment of 15 days average precipitation significantly reduces kid mortality by 7.8% (p-value ≤ 0.001). Fleshing of does during early meeting to improve the post-partum weight of does and kids is also important to reduce both kid and doe mortality at and after kidding. Immunization of new introduced and kids to common endemic diseases in the area, extensive control of ticks to breakdown heart water transmission and use of proper comfortable housing to reduce stress of goats is recommended. Improving nutrition particularly during scarce grazing and browsing feed availability is important to improving the health and reducing mortality of goats in intensively managed goat farms.

The design of a Bayesian platform trial to prevent and eradicate inhibitors in patients with hemophilia

Among individuals with the rare congenital bleeding disorder hemophilia A, the major challenge is inhibitor formation, which is associated with significant morbidity and cost. Yet, as the optimal approach to prevent and eradicate inhibitors is not known, we are at equipoise. Because classic trial design is not practical in a rare disease setting, we designed 2 48-week randomized trials comparing ELOCTATE and emicizumab to prevent and eradicate inhibitors. To achieve statistical efficiency, we incorporated historic data (Bayesian priors) on inhibitor formation to allow preferential randomization to emicizumab, piecewise exponential survival models to determine mean and 95% confidence interval for inhibitor formation in each arm, and simulations to determine the best model design to optimize power. To achieve administrative efficiency, the trials will be performed with the same sites, staff, visit frequency, blood sampling, laboratories, and laboratory assays, with streamlined enrollment so patients developing inhibitors in the first trial may be enrolled on the second trial. The primary end point is the probability of inhibitor formation or inhibitor eradication, respectively. The design indicates early stopping rules for overwhelming evidence of superiority of the emicizumab arms. Simulations indicate that, with 66 subjects, the Prevention Trial will have 84% power to detect noninferiority of emicizumab to ELOCTATE with a margin of 10% if emicizumab is truly 10% superior to ELOCTATE; with 90 subjects, the Eradication Trial will have 80% power to detect 15% superiority of ELOCTATE immune tolerance induction with vs without emicizumab. Thus, a platform design provides statistical and administrative efficiency to conduct INHIBIT trials.