Search

In survival analysis with competing risks, the transformation model allows different functions between the outcome and explanatory variables. However, the model's prediction accuracy and the interpretation of parameters may be sensitive to the choice of link function. We review the practical implications of different link functions for regression of the absolute risk (or cumulative incidence) of an event. Specifically, we consider models in which the regression coefficients β have the following interpretation: The probability of dying from cause D during the next t years changes with a factor exp(β) for a one unit change of the corresponding predictor variable, given fixed values for the other predictor variables. The models have a direct interpretation for the predictive ability of the risk factors. We propose some tools to justify the models in comparison with traditional approaches that combine a series of cause-specific Cox regression models or use the Fine-Gray model. We illustrate the methods with the use of bone marrow transplant data. [ABSTRACT FROM AUTHOR]

BACKGROUND: Patients with alcoholic cirrhosis are at higher risk for hepatocellular carcinoma (HCC). The role of HCC surveillance for these patients is undefined. OBJECTIVE: To provide population-based estimates of HCC incidence and comparisons of HCC-related mortality and total mortality among patients with alcoholic cirrhosis as a basis for assessing the role of HCC surveillance. DESIGN: Nationwide, registry-based, historical cohort study. SETTING: Denmark. PATIENTS: All Danish citizens with a first-time hospital diagnosis of alcoholic cirrhosis from 1993 to 2005. MEASUREMENTS: Hepatocellular carcinoma incidence and mortality starting 1 year after diagnosis of alcoholic cirrhosis through 2009; ratio of HCC-related mortality to total mortality. RESULTS: Among 8482 patients, 169 developed HCC. A total of 5734 patients died, 151 of whom had developed HCC. Five-year cumulative HCC risk was 1.0% (95% CI, 0.8% to 1.3%), and 5-year cumulative mortality was 43.7% (CI, 42.6% to 44.7%). Only 1.8% of all deaths were HCC-related. In sensitivity analyses that included all possible HCC diagnoses and a subpopulation of patients who were followed by hepatologists, the highest 5-year HCC risk was 1.9% (CI, 0.8% to 3.9%). These patients did not have higher mortality than patients in the nationwide cohort. LIMITATION: Cirrhosis and HCC diagnoses were made by hospital physicians without uniform clinical criteria, and use of registry data precluded detailed information on clinical care of patients, including HCC surveillance. CONCLUSION: Danish patients with alcoholic cirrhosis have a low risk for HCC, and HCC contributes little to their high mortality. On the basis of these data, HCC surveillance would be expected to have a minimal effect on mortality and is unlikely to be cost-effective. PRIMARY FUNDING SOURCE: None. [ABSTRACT FROM AUTHOR]

We review sources of bias which can affect non-randomized cohort studies of non-specific effects of vaccines on child mortality. Using examples from the literature on non-specific effects, we describe different sources of selection and information bias, and, where possible, outline analysis strategies to mitigate or eliminate such biases. [ABSTRACT FROM AUTHOR]

Objective To examine whether late and very-late first-contact schizophrenia carry a risk for later development of dementia. Methods By linkage of the psychiatric and the somatic nation-wide registers of all out- and in-patients with hospital contact in Denmark, we identified all patients with first ever contact during the period from January 1994 to December 2001 with one of the index main diagnoses: late (age >=40) and very-late first-contact (age >=60) schizophrenia. First contact osteoarthritis patients as well as data on the general population were used as controls. The first diagnosis of dementia for each individual at discharge or at out-patient contact was established. The probability of getting a dementia diagnosis is estimated using Poisson regression models with dementia as the outcome of interest. Results Twelve thousand six hundred and sixteen and 7,712 individuals were included in the late and very-late sample, respectively. Follow-up time was between 3.00 and 4.58 years. The rate ratio (RR) of developing dementia in late and very-late first-contact schizophrenia compared to osteoarthritis patients were 3.47 (95%CI: 2.19-5.50) and 3.15 (95%CI: 1.93-5.14), respectively. Compared to the general population the RR were 2.36 (95%CI: 1.54-3.62) and 2.21 (95%CI: 1.39-3.50), respectively. Conclusion schizophrenic patients with late- and very-late first-contact with the psychiatric hospital system are at two to three times higher risk of subsequently getting a diagnosis of dementia compared to patients with osteoarthritis and compared to the general population. Copyright © 2008 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Objective: To examine whether very late first-contact delusional disorder carries a risk for later development of dementia.Methods: By linkage of the psychiatric and the somatic nationwide registers of all out- and in-patients with hospital contact in Denmark, we included all 60þ patients with first ever from 1 January 1994 to 31 December 2001 with the index main diagnosis: delusional disorder. First contact osteoarthritis patients as well as the general population were used as controls. A total of 1437 patients with persistent delusional disorder and 7302 patients with osteoarthritis were included. Median follow-up time until first diagnosis of dementia at discharge was 1.87 and 4.40 years, respectively. The probability of getting a dementia diagnosis was estimated using Poissonregression models with dementia as the outcome of interest.Results: Patients with very late first-contact delusional disorder had an 8.14 (95% CI, 6.51; 10.19) times increased rate of subsequently developing dementia compared with very late first contact osteoarthritis patients. Compared with the general population the rate ratio was 5.49 (95% CI, 4.81; 6.26).Conclusion: Very late first-contact delusional disorder increases the risk of subsequently getting a diagnosis of dementia 5-8 times compared with osteoarthritis patients and the general population. [ABSTRACT FROM AUTHOR]

Background: Depression and anxiety is prevalent among patients with dementia but the extent to which these conditions are treated with antidepressants has not previously been investigated.Methods: Nationwide register-based data in Denmark were used to identify all patients diagnosed with dementia or osteoarthritis at hospital admission or at first outpatient contact during the period 1995-2000. Rates of subsequent purchase of antidepressants from pharmacies were then calculated. Further, the rate of antidepressant use for patients with dementia was compared with the rate in a gender-, age-, and calendar-matched sample of the general population.Results: In total, 24,137 patients with a main diagnosis of dementia and 100,378 patients with a main first diagnosis of osteoarthritis were incorporated in the study. A total of 43.2% of patients with dementia received antidepressants during follow-up compared to 16.0% of patients with osteoarthritis. Among patients with a diagnosis of dementia, the rate of subsequently purchasing antidepressants was 4.17 times higher (95% CI = 4.05-4.29) than that of patients with a first diagnosis of osteoarthritis, and 8.85 times higher (95% CI = 8.68-9.03) than that of a gender-, age- and calendar-matched sample of the general population. The rate was increased in all subgroups of patients regardless of gender, age, socio-economic group and time since diagnosis.Conclusions: The findings challenge the widely held contention that depression and anxiety in patients with dementia is underdiagnosed and undertreated in clinical practice. [ABSTRACT FROM AUTHOR]

The authors investigated the shape, sex- and age-dependency, and possible confounding of the association between birth weight and systolic blood pressure (SBP) in 197,954 adults from 20 Nordic cohorts (birth years 1910-1987), one of which included 166,249 Swedish male conscripts. Random-effects meta-regression analyses were performed on estimates obtained from age- and sex-stratified analyses within each of the cohorts. There was an inverse association between birth weight and SBP, irrespective of adjustment for concurrent body mass index. The association was linear for males, but for females with a birth weight greater than 4 kg, SBP increased with birth weight (p < 0.01). The association was stronger in the older age groups (p < 0.05), although this could have been a birth cohort effect. The association was stronger among females than among males (p = 0.005) when birth weight was less than or equal to 4 kg. The estimated effect of birth weight on SBP at age 50 years was -1.52 mmHg/kg (95% confidence interval: -2.27, -0.77) in men and -2.80 mmHg/kg (95% confidence interval: -3.85, -1.76) in women. Exclusion of the Swedish conscripts produced nearly identical results. This meta-analysis supports the evidence of an inverse birth weight-SBP association, regardless of adjustment for concurrent body size. It also reveals important heterogeneity in the shape and strength of the association by sex and age. [ABSTRACT FROM AUTHOR]

The association between treatment with Selective serotonin reuptake inhibitors (SSRIs) and suicide in children and adolescents on the individual and ecological level were examined in a nationwide Danish pharmacoepidemiological register-linkage study including all persons aged 10-17 years treated with antidepressants during the period 1995-1999 (n=2,569) and a randomly selected control population (n=50,000). A tripartite approach was used. In Part 1, changes in youth suicide and use of antidepressants were examined. In Part 2, we made an assessment of youth suicide characteristics. In Part 3, we analysed the relative risk (RR) of suicide according to antidepressant treatment corrected for psychiatric hospital contact to minimize the problem of confounding by indication. The use of SSRIs among children and adolescents increased substantially during the study period, but the suicide rate remained stable (Part 1). Among 42 suicides nationally aged 10-17 years at death, none was treated with SSRIs within 2 weeks prior to suicide (Part 2). There was an increased rate of suicide associated with SSRIs (RR=4.47), however, not quite significant (95% CI: 0.95-20.96), when adjusted for severity of illness (Part 3). Conclusively, we were not able to identify an association between treatment with SSRIs and completed suicide among children and adolescents. [ABSTRACT FROM AUTHOR]

OBJECTIVE: The sparse knowledge of the aetiology of pelvic pain in pregnancy makes evidence based prevention a limited option. The aim of this study was to examine the relation between pelvic pain in pregnancy and physical and psychosocial working conditions. METHODS: This study used self reported data on working conditions for 1219 cases and 1539 controls, sampled as a nested case-control study within the Danish national birth cohort. Cases and controls were selected on the basis of self reported pelvic pain intensity, pain localisation, and pain impact on daily living activities. Exposure data were collected prospectively; early in pregnancy and before the onset of pelvic pain. Main outcome measures were odds ratios for pelvic pain in pregnancy as a function of physical and psychosocial working conditions. RESULTS: Pregnant women with fixed evening work and with rotating shifts (without night shift) had odds ratios for pelvic pain in pregnancy of 1.76 (95% confidence intervals 1.04 to 2.96) and 1.65 (1.22 to 2.24), respectively, compared with women with day work. Physically strenuous work was associated with an almost 50% increased risk of pelvic pain in pregnancy (1.47; 1.17 to 1.84). In women who were under high psychosocial strain at work odds ratio was 1.39 (1.12 to 1.74) compared with women with low job strain. CONCLUSION: Both physically and psychosocially demanding working conditions, measured by physically strenuous work, rotating shifts, and high job strain, are associated with an increased reporting of pelvic pain in pregnancy. [ABSTRACT FROM AUTHOR]

BACKGROUND: It has been suggested that the level of high-density lipoprotein cholesterol (HDL-C) in the blood can be used as a marker of recent alcohol intake. However, before using HDL-C as a predictor of alcoholism, the relation between alcohol intake and HDL-C in the entire range of consumption must be explored. Most studies model the relation between alcohol intake and HDL-C linearly, although a threshold effect is expected. The objective of this study was to evaluate the shape of the relation between intake of alcohol and HDL-C and to determine whether there are differential effects of beer, wine, and spirits on HDL-C and whether they remain after adjusting for total alcohol. METHOD: The relation between alcohol intake and HDL-C was investigated by means of generalized additive models using data from the Copenhagen City Heart Study. RESULTS: A nonlinear effect of alcohol improved the model fit significantly, and the nonlinearity of alcohol was highly significant in both men and women. The relation was concave: HDL-C was stable in men and women who drank more than approximately 35 and 20 drinks per week, respectively. We found a significant nonlinear term of wine on HDL-C in men after adjustment for total alcohol intake. CONCLUSIONS: There was a concave relation between alcohol intake and HDL-C, indicating a threshold effect of alcohol on HDL-C. The association between wine and HDL-C in men after adjusting for total alcohol intake may be due to residual lifestyle confounding. [ABSTRACT FROM AUTHOR]

This paper deals with the competing risks model as a special case of a multi-state model. The properties of the model are reviewed and contrasted to the so-called latent failure time approach. The relation between the competing risks model and right-censoring is discussed and regression analysis of the cumulative incidence function briefly reviewed. Two real data examples are presented and a guide to the practitioner is given. [ABSTRACT FROM AUTHOR]

An introduction to event history analysis via multi-state models is given. Examples include the two-state model for survival analysis, the competing risks and illness–death models, and models for bone marrow transplantation. Statistical model specification via transition intensities and likelihood inference is introduced. Consequences of observational patterns are discussed, and a real example concerning mortality and bleeding episodes in a liver cirrhosis trial is discussed. [ABSTRACT FROM AUTHOR]

Background: Mortality from smoking-related diseases in women is increasing worldwide. Studies comparing hazards associated with smoking in women and men based on a sufficient number of heavy smokers of both genders are lacking.Methods: We used pooled data from three prospective population studies conducted in Copenhagen to compare total and cause-specific mortality in relation to smoking habits. A total of 30917 subjects, 44% women, with initial examinations between 1964 and 1992 were followed until 1994 for date and cause of death.Results: During follow up, 2900 women and 5744 men died. Smoking characteristics differed considerably with gender, particularly in the older subjects. Overall mortality rates in smokers were approximately twice those in people who never smoked. Positive associations with smoking in both men and women were confirmed for all-cause mortality as well as mortality from respiratory disease, vascular disease, lung cancer, and other tobacco-related cancers. Despite large gender differences in age at smoking debut, total and cause-specific relative mortality in smokers was similar in men and women. After excluding non-inhalers, relative risks associated with smoking for respiratory and vascular disease were higher for women than men whereas there were no gender differences in smoking related risk of cancers.Conclusions: The relative risks suggest that women may be more sensitive than men to some of the deleterious effects of smoking. However, because of lower baseline mortality rates in women, rate differences may be similar and results should be interpreted with caution. [ABSTRACT FROM AUTHOR]

Background: Two recent much cited publications have raised the concern that risk associated with cigarette smoking has so far been underestimated. In this study we wish to determine whether excess all-cause mortality associated with smoking has increased during the last 20-30 years in a study population representative of the general Danish population and whether any such changes relate to changes in smoking behaviour.Methods: Pooled data from three prospective population studies conducted in Copenhagen with detailed information on smoking habits. A total of 31,194 subjects, 17,669 males and 13,525 females, initially examined between 1964 and 1992 with examinations repeated at intervals from 1-10 years, were followed until 1995 for all-cause mortality. Relative mortality risk in smokers versus never-smokers was calculated within periods of five calendar years and compared throughout the study period.Results: Male smokers' exposure did not change during the study period whereas female smokers' exposure to tobacco increased in terms of age at smoking onset, quantity smoked and depth of inhalation. During follow-up 5744 males and 2900 females died. In males, death rate ratios (comparing continuous smokers with never-smokers) did not change in the study period. In females, ratios increased from 1964-1978 to 1979-1994 by a factor of 1.3 (95% confidence interval 1.0-1.8).Conclusions: In agreement with the observed changes in smoking habits, excess mortality in male smokers did not increase whereas excess mortality in female smokers increased slightly. [ABSTRACT FROM AUTHOR]

Editorial. Comments on the development of the field of survival analysis as an independent statistical discipline. Introduction of the framework for counting processes, martingales and stochastic integrals; Emphasis of the multistate models on medical and epidemiological applications; Evaluation of the handling of competing risks in a disease course.