The impact of competing risks on relative risks in occupational cohort studies

This paper develops a method for adjustment of competing causes of death in the calculation of relative risk. It has identified three factors determining the significance of competing risks: (1) the magnitude of the overall mortality risk of the study population; (2) differential risk (or the adjustment factor) for mortality other than cause of interest between two populations; and (3) age intervals used in the mortality calculation. Thus, the impact of competing risks is increased if the study cohort has a high mortality risk, if the mortality other than the cause of interest has a large differential risk or if the mortality calculation uses wide age intervals. Two examples from a refinery cohort and the US national population show that among certain age groups unadjusted for competing risks the relative risk is overestimated by 9%. The impact of competing risks in these two particular examples is relatively small. Furthermore, if relative risk is expressed in terms of the ratio of mortality rates, competing risks can be ignored.     

Breast cancer mortality in Russia and Ukraine 1963-2002: an age-period-cohort analysis

OBJECTIVE: To determine the reasons for the steady increase in breast cancer mortality in Russia and Ukraine. METHODS: Age-period-cohort analysis, supplemented by analysis of historical fertility trends. RESULTS: Mortality from breast cancer has risen steadily in both countries over the past 40 years, although faster in Russia than in Ukraine. There are strong birth cohort effects, which are consistent with known changes in fertility. Death rates were highest among those born in the first half of the 20th century, declining among those born after the 1950s. There has been a decline in mortality among younger women since the mid 1990s, which may reflect improvements in treatment. CONCLUSION: The increase in breast cancer mortality in Russia and Ukraine can largely be explained by known changes in fertility, while recent changes may reflect changes in treatment. Observed trends suggest that death rates from female breast cancer in the two countries are likely to stabilize or even decline in the future.     

A theory of health investment under competing mortality risks

In this paper, we present a theory of health investment when there are multiple causes of death. We analyze the optimal investment rules and the comparative statics. In particular, we examine the conditions under which a cause-specific intervention has a spillover effect. By spillover effect, we mean a price reduction in one cause-specific health investment would increase all other investments. We also study the strength of the spillover effect, which is measured in terms of cross price elasticities. We find that, while a cause-specific intervention might not be wasteful, the spillover effect would not be large either.     

Estimating and testing for center effects in competing risks

The problems of fitting Gaussian frailties proportional hazards models for the subdistribution of a competing risk and of testing for center effects are considered. In the analysis of competing risks data, Fine and Gray proposed a proportional hazards model for the subdistribution to directly assess the effects of covariates on the marginal failure probabilities of a given failure cause. Katsahianbiet al. extended their model to clustered time to event data, by including random center effects or frailties in the subdistribution hazard. We first introduce an alternate estimation procedure to the one proposed by Katsahian et al. This alternate estimation method is based on the penalized partial likelihood approach often used in fitting Gaussian frailty proportional hazards models in the standard survival analysis context, and has the advantage of using standard survival analysis software. Second, four hypothesis tests for the presence of center effects are given and compared via Monte-Carlo simulations. Statistical and numerical considerations lead us to formulate pragmatic guidelines as to which of the four tests is preferable. We also illustrate the proposed methodology with registry data from bone marrow transplantation for acute myeloid leukemia (AML).     

Testing treatment effects in the presence of competing risks

Competing risks are often encountered in clinical research. In the presence of multiple failure types, the time to the first failure of any type is typically used as an overall measure of the clinical impact for the patients. On the other hand, use of endpoints based on the type of failure directly related to the treatment mechanism of action allows one to focus on the aspect of the disease targeted by treatment. We review the methodology commonly used for testing failure specific treatment effects. Simulation results demonstrate that the cause-specific log-rank test is robust (in the sense of preserving the nominal level of the test) and has good power properties for testing for differences in the marginal latent failure-time distributions, whereas the use of a popular cumulative incidence based approach may be problematic for this aim.     

Risk factors for mortality in the nurses' health study: a competing risks analysis

Few studies have examined multiple risk factors for mortality or formally compared their associations across specific causes of death. The authors used competing risks survival analysis to evaluate associations of lifestyle and dietary factors with all-cause and cause-specific mortality among 50,112 participants in the Nurses' Health Study. There were 4,893 deaths between 1986 and 2004: 1,026 from cardiovascular disease, 931 from smoking-related cancers, 1,430 from cancers not related to smoking, and 1,506 from all other causes. Age, body mass index at age 18 years, weight change, height, current smoking and pack-years of smoking, glycemic load, cholesterol intake, systolic blood pressure and use of blood pressure medications, diabetes, parental myocardial infarction before age 60 years, and time since menopause were directly related to all-cause mortality, whereas there were inverse associations for physical activity and intakes of nuts, polyunsaturated fat, and cereal fiber. Moderate alcohol consumption was associated with decreased mortality. A model that incorporated differences in the associations of some risk factors with specific causes of death had a significantly better fit compared with a model in which all risk factors had common associations across all causes. In the future, this new model may be used to identify individuals at increased risk of mortality.     

Intragenerational mobility and mortality in Russia: Short and longer-term effects

This study uses the Russian Longitudinal Monitoring Survey to explore the relationship between mortality of men age 65 or younger and intragenerational mobility, measured objectively through household income and subjectively through social ranking. This relationship is considered in light of the social selection and social causation mechanisms developed in the literature as well as a proposed mechanism in which mobility itself is a consequential life event. The analysis spans the years 1994–2010, which covers the transitional period in Russia characterized by labor market restructuring and economic crisis as well as a later period of economic growth and recovery. Using Cox proportional hazard models, immediate and longer-term associations between mobility and mortality are estimated. Both subjective and objective downward mobility had an immediate positive association with mortality risk (increased by 44% and 24%, respectively). In contrast, upward mobility had a more pronounced effect over a longer-term horizon and lowered mortality risk by 17%. Controlling for destination status attenuated some associations, but findings were robust to the adjustment of selection-related factors such as alcohol consumption and health status in the year preceding mobility. Findings suggest that the negative relationship between upward mobility and mortality may be driven by social causation, whereas downward mobility may have an independent effect beyond selection or causation.     

A cohort study to estimate occupational mortality risks in Navarra

BACKGROUND: Few studies on occupational mortality have been conducted in Spain. The objective of this work was to analyse inequalities on global mortality and on mortality due to specific causes according to occupation in a historical cohort of males from the province of Navarra, Spain. METHODS: The base population for this historical cohort comprised all employed men over age 34 from Navarra in the 1986 population register. Age-standardised point estimates and confidence intervals for occupational-specific mortality risks were computed. RESULTS: There exist differences in mortality risks with respect to the overall risk of Navarra in certain occupational activities for several major causes of mortality. Some of the results corroborate previous findings in other works, such as the significant high risk that presents in leather, clothing workers and shoemakers when analysing kidney, bladder and other urinary malignant tumours, while others present a certain degree of novelty. CONCLUSION: This work contributes to filling the gap in the lack of works on occupational mortality in Spain. It also complements the information that other monitoring systems may provide on occupational health.     

Tutorial in biostatistics: competing risks and multi-state models

Standard survival data measure the time span from some time origin until the occurrence of one type of event. If several types of events occur, a model describing progression to each of these competing risks is needed. Multi-state models generalize competing risks models by also describing transitions to intermediate events. Methods to analyze such models have been developed over the last two decades. Fortunately, most of the analyzes can be performed within the standard statistical packages, but may require some extra effort with respect to data preparation and programming. This tutorial aims to review statistical methods for the analysis of competing risks and multi-state models. Although some conceptual issues are covered, the emphasis is on practical issues like data preparation, estimation of the effect of covariates, and estimation of cumulative incidence functions and state and transition probabilities. Examples of analysis with standard software are shown.     

Health lifestyles and political ideology in Belarus, Russia, and Ukraine

This paper examines the association of political ideology with health lifestyle practices and self-rated health in Belarus, Russia, and Ukraine. The political trajectory of post-Soviet societies has taken two divergent paths, either toward democracy or autocracy. The health trajectory has followed the same pattern with the more autocratic states continuing to experience a mortality crisis, while those former socialist countries that have embraced democracy and moved closer to the West have escaped this crisis. This paper investigates whether political ideology in three post-Soviet countries that are firmly (Belarus), increasingly (Russia), or recently (Ukraine) autocratic is related to health lifestyles and health self-ratings. Data were collected by face-to-face interviews (N = 8406) with a representative national sample of the adult population. The results show that respondents who are against restoring communism have healthier lifestyles and rate their health better than respondents who wish to see communism return.     

Primary liver cancer, other malignancies, and mortality risks following porphyria: a cohort study in Denmark and Sweden.

Cancer incidence and mortality risks were evaluated in a combined cohort of patients who were hospitalized for porphyria in Denmark (1977-1989) and Sweden (1965-1983). Patients were identified by using population-based hospitalization registries. The unique individual identification numbers of 530 patients with porphyria cutanea tarda (PCT) and 296 with acute intermittent porphyria (AIP) were linked to the nationwide cancer and death registries. Among patients with both types of porphyria, the authors found small but significantly elevated risks of all cancers combined (PCT: standardized incidence ratio (SIR) = 1.7, 95% confidence interval (CI) 1.3-2.2; AIP: SIR = 1.8, 95% CI 1.1-2.8) due to pronounced excesses of primary liver cancer (PCT: SIR = 21.2, 95% CI 8.5-43.7; AIP: SIR = 70.4, 95% CI 22.7-164.3) and moderate increases in lung cancer (PCT: SIR = 2.9, 95% CI 1.5-5.2; AIP: SIR = 2.8, 95% CI 0.3-10.2). PCT patients had a significantly increased risk of mortality from liver cirrhosis (standardized mortality ratio (SMR) = 8.4, 95% CI 3.1-18.4) or chronic obstructive pulmonary disease (SMR = 3.1, 95% CI 1.1-6.7). The increased risk of primary liver cancer and the increased risk of mortality from cirrhosis of the liver are consistent with findings from previous clinical surveys, but the new observations of excess lung cancer and chronic obstructive pulmonary disease require confirmation.     

The effects of particulate and ozone pollution on mortality in Moscow,Russia

The objectives of this study were (1) to evaluate how acute mortality responds to changes in particulate and ozone (O₃) pollution levels, (2) to identify vulnerable population groups by age and cause of death, and (3) to address the problem of interaction between the effects of O₃ and particulate pollution. Time-series of daily mortality counts, air pollution, and air temperature were obtained for the city of Moscow during a 3-year period (2003–2005). To estimate the pollution-mortality relationships, we used a log-linear model that controlled for potential confounding by daily air temperature and longer term trends. The effects of 10 μg/m³ increases in daily average measures of particulate matter ≤10 μm in aerodynamic diameter (PM₁₀) and O₃ were, respectively, (1) a 0.33% [95% confidence interval (CI) 0.09–0.57] and 1.09% (95% CI 0.71–1.47) increase in all-cause non-accidental mortality in Moscow; (2) a 0.66% (0.30–1.02) and 1.61% (1.01–2.21) increase in mortality from ischemic heart disease; (3) a 0.48% (0.02–0.94) and 1.28% (0.54–2.02) increase in mortality from cerebrovascular diseases. In the age group >75 years, mortality increments were consistently higher, typically by factor of 1.2 – 1.5, depending upon the cause of death. PM₁₀-mortality relationships were significantly modified by O₃ levels. On the days with O₃ concentrations above the 90th percentile, PM₁₀ risk for all-cause mortality was threefold greater and PM₁₀ risk for cerebrovascular disease mortality was fourfold greater than the unadjusted risk estimate.     

Asthma mortality in Australia 1920-94: age, period, and cohort effects.

STUDY OBJECTIVE: To investigate asthma mortality during 1920-94 in Australia in order to assess the relative role of period and birth cohort effects. DESIGN: Asthma mortality (both sexes) was age standardised and examined for changes over time. The data were also examined for age, period, and cohort (APC) effects using Poisson regression modelling. SETTING: National Australian mortality data. PARTICIPANTS: Population (both sexes) aged 15-34 years, 1920-94. MAIN RESULTS: Age adjusted period rates indicate an increase in asthma mortality during the 1950's, and increases and subsequent falls (epidemics) during the mid 1960s and late 1980s. APC modelling suggested an increasing cohort effect (adjusted for both age and period) from the birth cohort 1950-54 onwards. Period effects (adjusted for age and cohort) are characterized by an increase in the 1950s (possibly due to changes in diagnostic labelling), minimal or no increases in the mid 1960s and late 1980s (where period peaks had been noted when data were adjusted for age only), and declines in mortality risk subsequent to the periods where age-period analysis had noted increases. Thus, in Australia, some of the mid 1960s epidemic in asthma deaths, and all of the late 1980s mortality increase, seem to be attributable to cohort effects. CONCLUSIONS: The increase in asthma mortality cohort effect is consistent with empirical evidence of recent increases in prevalence (and presumably incidence) of asthma in Australia, and suggests the need for more research into the underlying environmental aetiology of this condition.     

Modelling competing risks in cancer studies

Competing risks arise commonly in the analysis of cancer studies. Most common are the competing risks of relapse and death in remission. These two risks are the primary reason that patients fail treatment. In most medical papers the effects of covariates on the three outcomes (relapse, death in remission and treatment failure) are model by distinct proportional hazards regression models. Since the hazards of relapse and death in remission must add to that of treatment failure, we argue that this model leads to internal inconsistencies. We argue that additive models for either the hazard rates or the cumulative incidence functions are more natural and that these models properly partition the effect of a covariate on treatment failure into its component parts. We illustrate the use and interpretation of additive models for the hazard rate or for the cumulative incidence function using data from a study of the efficacy of two preparative regimes for hematopoietic stem cell transplantation.     

Model selection in competing risks regression

In the analysis of time‐to‐event data, the problem of competing risks occurs when an individual may experience one, and only one, of m different types of events. The presence of competing risks complicates the analysis of time‐to‐event data, and standard survival analysis techniques such as Kaplan–Meier estimation, log‐rank test and Cox modeling are not always appropriate and should be applied with caution. Fine and Gray developed a method for regression analysis that models the hazard that corresponds to the cumulative incidence function. This model is becoming widely used by clinical researchers and is now available in all the major software environments. Although model selection methods for Cox proportional hazards models have been developed, few methods exist for competing risks data. We have developed stepwise regression procedures, both forward and backward, based on AIC, BIC, and BICcr (a newly proposed criteria that is a modified BIC for competing risks data subject to right censoring) as selection criteria for the Fine and Gray model. We evaluated the performance of these model selection procedures in a large simulation study and found them to perform well. We also applied our procedures to assess the importance of bone mineral density in predicting the absolute risk of hip fracture in the Women's Health Initiative–Observational Study, where mortality was the competing risk. We have implemented our method as a freely available R package called crrstep. Copyright © 2013 John Wiley & Sons, Ltd.     

Cancer risks among iron and steel workers in Anshan,China, part I: Proportional mortality ratio analysis

A standardized proportional mortality ratio (SPMR) study of 8,887 deaths during 1980–1989 among male workers in a large integrated iron-steel complex in Anshan, China, was conducted to provide clues to occupational risk factors. Accidents and cancer accounted for a higher proportion of deaths among the iron-steel workers than among the general male population (SPMR = 1.21; 95% CI = 1.12–1.31 and 1.14; 95% CI = 1.10–1.18, respectively). Among all workers, SPMRs were significantly elevated for stomach, lung, and colorectal cancers (SPMR = 1.37, 1.37, 1.38, respectively), but not other cancers. Risks of stomach cancer appeared to be highest among workers employed in jobs with exposure to iron and coal dust, whereas significant increases in colorectal cancer were seen for loading and other dusty jobs and for administrative and sedentary jobs without dust exposure. Risks of lung cancer appeared increased for a variety of jobs throughout the complex, especially those with probable high levels of exposure to polycyclic hydrocarbons and asbestos. Risk of esophageal cancer was significantly elevated for fire-resistant brick makers, and risk of nonmalignant respiratory disease was significantly elevated for those employed as furnace workers, foundry workers, and fire-resistant brick makers. (This article is a US Government work and, as such, is in the public domain in the United States of America.) © 1996 Wiley-Liss, Inc.     

Cancer mortality in a cohort of licensed herbicide applicators

Objectives In order to expand our knowledge on the possible long-term health effects of exposure to herbicides, we updated the follow-up of a cohort of 1,341 licensed herbicide applicators in the Netherlands. The earlier report indicated that there might be an increased risk for multiple myeloma in this group. Although that finding was statistically significant, the result was based on a small number of cases.Methods We expanded the follow-up from 1 January 1988 to 1 January 2001, which added 13 years to the follow-up. We now report on the causes of death of 196 exposed workers.Results Our findings indicate that licensed herbicide applicators were at an increased risk for skin cancer mortality [standardized mortality ratio (SMR)=357.4, 95% confidence interval (CI) 115.1–827.0]. It is not clear if this excess of skin cancer should be attributed to herbicide exposure or to excess exposure to sunlight.