Biomedical subjects

Theo Stijnen

Publications and source records attributed to Theo Stijnen.

At least 19 recordsLinked to original sources

PubMed2007

A new method to determine mean adult height from incomplete follow-up data.

BACKGROUND/AIM: In long-term growth studies with adult height (AH) as outcome, reporting is often required while data are incomplete because some participants have not yet reached AH whereas others might be lost to follow-up. Current practice is to analyze only participants who did reach AH, which can easily give biased results. We introduce a new method into the area of growth research. METHODS: We used the data of patients from a registration database and a growth study. The new method uses growth data in time intervals. The percentage of children still growing and the mean growth at each interval are used to determine mean AH. RESULTS: With the new method, estimated mean AHs had smaller bias and standard error than with commonly used methods. The method is not hampered by a correlation between AH and age at reaching AH, unlike methods merely using patients who have reached AH. CONCLUSION: In contrast to commonly used methods, the new method provides valid results on mean AH when complete actual measurements of AH are not (yet) available, provided that drop-out, if any, is not related to (disappointing) growth. As it also uses observed data of children with incomplete follow-up, the method employs the data more effectively.

Adolescent↗

PubMed2006

Prediction of adult height in growth-hormone-treated children with growth hormone deficiency.

CONTEXT: Several studies have searched for factors that significantly influence adult height (AH) of children with GH deficiency (GHD) who have been treated with biosynthetic GH, but a prediction model for AH has not yet been presented. OBJECTIVE: Our objective was to develop models for prediction of AH, using information available at the start of GH treatment or after 1 yr of treatment. DESIGN AND SETTING: For this retrospective study, data were collected from the National Registry of Growth Hormone Treatment in Children, which contained data of Dutch children treated with GH. PATIENTS/INTERVENTION: Patients included males born before 1985 and females born before 1987 with either diagnosis of GHD (syndromes, tumors, and other diseases were excluded) or a maximal GH response during provocation tests of less than 11 ng/ml, treated with biosynthetic GH for at least 1 yr. To be able to use the complete group of 342 children for the development of the models, multiple imputation was used for missing values. MAIN OUTCOME MEASURE: We assessed AH sd scores (SDS). RESULTS: Each prediction model contained both target height SDS and current height SDS. The change in height SDS during the first year proved an important predictor for AH. In all models, addition of GH dose was not significant. The percent explained variance, after correction for overfitting, ranged from 37% (prepubertal children, prediction at start) to 60% (pubertal children, prediction after 1 yr). CONCLUSION: The presented prediction models give accurate predictions of AH for children with GHD at start and after 1 yr of GH treatment. They are useful tools in the treatment of these children.

Adolescent↗

PubMed2006

Annual influenza vaccination in community-dwelling elderly individuals and the risk of lower respiratory tract infections or pneumonia.

BACKGROUND: Influenza vaccination has been associated with a reduction in the number of hospitalizations for respiratory conditions in elderly persons over the period from 1996 to 2002. Little is known, however, about the effect of influenza vaccination on the whole range of severity of respiratory tract infections. METHODS: We investigated the effect of annual influenza vaccination on the occurrence of lower respiratory tract infections (LRTIs) in community-dwelling elderly individuals. From 1996 to 2002, we performed a population-based cohort study, using the computerized Integrated Primary Care Information database in the Netherlands, of community-dwelling subjects who were 65 years or older on January 1 of the year of study entry. For each year, the individual cumulative exposure to influenza vaccination since study entry was computed. We compared the risk of LRTI after a first vaccination or revaccination with the risk for no vaccination using a time-varying multivariate Cox proportional hazard model, adjusted for age, sex, smoking, and underlying disease. RESULTS: In the study population of 26 071 subjects, 3412 developed LRTIs during follow-up. During the influenza epidemic periods, a first vaccination did not reduce risk for LRTI. In the total population, the hazard ratio following a first vaccination was 0.86 (95% confidence interval [CI], 0.71 to 1.05); in the population without or with comorbidity, these ratios were 0.90 (95% CI, 0.56 to 1.45) and 0.83 (95% CI, 0.66 to 1.04), respectively. During epidemic periods, revaccination reduced risk of LRTI by 33% (95% CI, 8% to 52%) in individuals without comorbidity. In individuals with comorbidity, the risk reduction of 5% was nonsignificant (95% CI, -10% to 18%). CONCLUSION: In this study, annual influenza revaccination was associated with a reduction in LRTI in community-dwelling elderly individuals.

Age Factors↗

PubMed2006

Vessel wall calcifications at multi-detector row CT angiography in patients with peripheral arterial disease: effect on clinical utility and clinical predictors.

PURPOSE: To evaluate retrospectively the effect of vessel wall calcifications on the clinical utility of multi-detector row computed tomographic (CT) angiography performed in patients with peripheral arterial disease and to identify clinical predictors for the presence of vessel wall calcifications. MATERIALS AND METHODS: The study was approved by the hospital institutional review board, and informed consent was obtained from all patients. For this study the authors included patients from two randomized controlled trials that measured the costs and effects of diagnostic imaging in patients with peripheral arterial disease. All patients underwent CT angiography and were followed up for 6 months. Clinical utility was measured on the basis of therapeutic confidence (rated on a 10-point scale) in the results of initial CT angiography and the need for additional vascular imaging. Univariable and multivariable logistic and linear regression analysis and the area under the receiver operating characteristic curve were used to evaluate the effect of vessel wall calcifications on the clinical utility of CT angiography and the use of patient characteristics to predict the number of calcified segments at CT angiography. RESULTS: A total of 145 patients were included (mean age, 64 years; 70% men). The authors found that the number of calcified segments was a significant predictor of the need for additional imaging (P = .001) and of the confidence scores (P < .001). The number of calcified segments discriminated between patients who required additional imaging after CT angiography and those who did not (area under the receiver operating characteristic curve, 0.66; 95% confidence interval: 0.54, 0.77). Age, diabetes mellitus, and cardiac disease were significant predictors of the number of calcified segments in both the univariable and multivariable analyses (P < .05). CONCLUSION: Vessel wall calcifications decrease the clinical utility of CT angiography in patients with peripheral arterial disease. Diabetes mellitus, cardiac disease, and elderly age (older than 84 years) are independently predictive for the presence of vessel wall calcifications.

Angiography, Digital Subtraction↗

PubMed2006

Channeling and prevalence of cardiovascular contraindications in users of cyclooxygenase 2 selective nonsteroidal antiinflammatory drugs.

OBJECTIVE: To assess use and channeling of cyclooxygenase 2 selective inhibitors (coxibs) over time and to estimate the percentage of coxib users with cardiovascular contraindications. METHODS: The study population comprised all coxib and nonselective nonsteroidal antiinflammatory drug (NSAID) users in the Integrated Primary Care Information project between January 2000 and December 2004. The prevalence of risk factors for NSAID-related upper gastrointestinal ulcer complications, cardiovascular disease, and cerebrovascular disease at the start of treatment was compared between users of coxibs and users of nonselective NSAIDs. RESULTS: The study population included 72,841 nonselective NSAID users and 10,739 coxib users. The prevalence of risk factors for NSAID-related gastrointestinal complications was higher in coxib users than nonselective NSAID users (odds ratio [OR] 1.18, 95% confidence interval [95% CI] 1.10-1.26). Similarly, the prevalence of prior cardiovascular disease was higher in coxib users than in nonselective NSAID users (OR 1.35, 95% CI 1.28-1.43). Channeling of coxibs to patients with NSAID-related gastrointestinal risk factors declined after 2001 but increased again in 2004, whereas the channeling of coxibs to patients with cardiovascular disease remained constant. Less than 15% of all coxib users had history of ischemic coronary or cerebrovascular disease. Among coxib users with increased risk for NSAID-related gastrointestinal disorders, 27% had history of ischemic coronary or cerebrovascular disease. CONCLUSION: This study demonstrates that coxibs were preferentially prescribed to patients with risk factors for NSAID-related gastrointestinal disorders and/or cardiovascular diseases. Only one-quarter of coxib users with increased risk for NSAID-related gastrointestinal complications had cardiovascular conditions compatible with recent European safety contraindications for coxibs.

Adolescent↗

PubMed2006

Complement factor H polymorphism, complement activators, and risk of age-related macular degeneration.

CONTEXT: The evidence that inflammation is an important pathway in age-related macular degeneration (AMD) is growing. Recent case-control studies demonstrated an association between the complement factor H (CFH) gene, a regulator of complement, and AMD. OBJECTIVES: To assess the associations between the CFH gene and AMD in the general population and to investigate the modifying effect of smoking, serum inflammatory markers, and genetic variation of C-reactive protein (CRP). DESIGN, SETTING, AND PARTICIPANTS: Population-based, prospective cohort study of individuals aged 55 years or older (enrollment between March 20, 1990, and July 31, 1993, and 3 follow-up examinations that were performed between September 1, 1993, and December 31, 2004) in Rotterdam, the Netherlands. The CFH Y402H polymorphism was determined in a total of 5681 individuals. Information on smoking, erythrocyte sedimentation rate, CRP serum levels, and haplotypes of the CRP gene were assessed at baseline. MAIN OUTCOME MEASURES: All severity stages of prevalent and incident AMD, graded according to the international classification and grading system for AMD. RESULTS: The frequency of CFH Y402H was 36.2% (4116/11,362 alleles). At baseline, there were 2062 persons (36.3%) with any type of AMD (prevalent cases), including 78 (1.4%) with late AMD (stage 4). During follow-up (mean, 8 years; median, 10 years), 1649 (35.5%) of 4642 participants progressed to a higher stage of AMD (incident cases), including 93 (5.6%) who developed late AMD. The odds ratio (OR) of AMD increased in an allele-dose manner with 2.00 (95% confidence interval [CI], 1.56-2.55) for stage 2 AMD, 4.58 (95% CI, 2.82-7.44) for stage 3 AMD, and 11.02 (95% CI, 6.82-11.81) for stage 4 (late, vision threatening) AMD for homozygous persons. Cumulative risks calculated by Kaplan-Meier analysis of late AMD by age 95 years were 48.3% for homozygotes, 42.6% for heterozygotes, and 21.9% for noncarriers. The population-attributable risk for CFH Y402H was 54.0%. Elevated erythrocyte sedimentation rates further increased the OR to 20.2 (95% CI, 9.5-43.0), elevated serum CRP levels to 27.7 (95% CI, 10.7-72.0), and smoking to 34.0 (95% CI, 13.0-88.6) for homozygotes compared with noncarriers without these determinants. The CRP haplotypes conferring high levels of CRP significantly increased the effect of CFH Y402H (P<.01). CONCLUSIONS: The CFH Y402H polymorphism may account for a substantial proportion of AMD in individuals similar to those in the Rotterdam Study and may confer particular risk in the presence of environmental and genetic stimulators of the complement cascade.

Aged↗

PubMed2006

Review: a gentle introduction to imputation of missing values.

In most situations, simple techniques for handling missing data (such as complete case analysis, overall mean imputation, and the missing-indicator method) produce biased results, whereas imputation techniques yield valid results without complicating the analysis once the imputations are carried out. Imputation techniques are based on the idea that any subject in a study sample can be replaced by a new randomly chosen subject from the same source population. Imputation of missing data on a variable is replacing that missing by a value that is drawn from an estimate of the distribution of this variable. In single imputation, only one estimate is used. In multiple imputation, various estimates are used, reflecting the uncertainty in the estimation of this distribution. Under the general conditions of so-called missing at random and missing completely at random, both single and multiple imputations result in unbiased estimates of study associations. But single imputation results in too small estimated standard errors, whereas multiple imputation results in correctly estimated standard errors and confidence intervals. In this article we explain why all this is the case, and use a simple simulation study to demonstrate our explanations. We also explain and illustrate why two frequently used methods to handle missing data, i.e., overall mean imputation and the missing-indicator method, almost always result in biased estimates.

Bias↗

PubMed2006

Imputation of missing values is superior to complete case analysis and the missing-indicator method in multivariable diagnostic research: a clinical example.

BACKGROUND AND OBJECTIVES: To illustrate the effects of different methods for handling missing data--complete case analysis, missing-indicator method, single imputation of unconditional and conditional mean, and multiple imputation (MI)--in the context of multivariable diagnostic research aiming to identify potential predictors (test results) that independently contribute to the prediction of disease presence or absence. METHODS: We used data from 398 subjects from a prospective study on the diagnosis of pulmonary embolism. Various diagnostic predictors or tests had (varying percentages of) missing values. Per method of handling these missing values, we fitted a diagnostic prediction model using multivariable logistic regression analysis. RESULTS: The receiver operating characteristic curve area for all diagnostic models was above 0.75. The predictors in the final models based on the complete case analysis, and after using the missing-indicator method, were very different compared to the other models. The models based on MI did not differ much from the models derived after using single conditional and unconditional mean imputation. CONCLUSION: In multivariable diagnostic research complete case analysis and the use of the missing-indicator method should be avoided, even when data are missing completely at random. MI methods are known to be superior to single imputation methods. For our example study, the single imputation methods performed equally well, but this was most likely because of the low overall number of missing values.

Adult↗

PubMed2006

Using the outcome for imputation of missing predictor values was preferred.

BACKGROUND AND OBJECTIVE: Epidemiologic studies commonly estimate associations between predictors (risk factors) and outcome. Most software automatically exclude subjects with missing values. This commonly causes bias because missing values seldom occur completely at random (MCAR) but rather selectively based on other (observed) variables, missing at random (MAR). Multiple imputation (MI) of missing predictor values using all observed information including outcome is advocated to deal with selective missing values. This seems a self-fulfilling prophecy. METHODS: We tested this hypothesis using data from a study on diagnosis of pulmonary embolism. We selected five predictors of pulmonary embolism without missing values. Their regression coefficients and standard errors (SEs) estimated from the original sample were considered as "true" values. We assigned missing values to these predictors--both MCAR and MAR--and repeated this 1,000 times using simulations. Per simulation we multiple imputed the missing values without and with the outcome, and compared the regression coefficients and SEs to the truth. RESULTS: Regression coefficients based on MI including outcome were close to the truth. MI without outcome yielded very biased--underestimated--coefficients. SEs and coverage of the 90% confidence intervals were not different between MI with and without outcome. Results were the same for MCAR and MAR. CONCLUSION: For all types of missing values, imputation of missing predictor values using the outcome is preferred over imputation without outcome and is no self-fulfilling prophecy.

Adult↗

PubMed2006

Validation of a calibrated prediction model for response to growth hormone treatment in an independent cohort.

BACKGROUND: Prediction models, e.g. for prediction of response to growth hormone treatment, need validation in appropriate independent cohorts, comparing predicted and observed outcomes. In a previous validation of a model for predicting the first-year response to growth hormone treatment in children with idiopathic growth hormone deficiency, overfitting was observed. We modified the prediction formula and now report validation of this modified model. PATIENTS AND METHODS: The modified and original prediction models were applied to a group of patients selected from Lilly's GeNeSIS database using the same inclusion and exclusion criteria as for the original model. For both prediction methods, observed first-year height velocity was plotted vs. predicted height velocity in a calibration plot. For a valid prediction, the regression line should correspond to the line of identity (observed outcome is equal to predicted outcome); the regression lines for each prediction model were tested for significant differences from this line of identity. RESULTS: The number of patients fulfilling the criteria was 226. The regression line in the calibration plot of the modified model was not significantly different from the line of identity (p = 0.43), in contrast to the original model (p < 0.001). For the modified model the mean (SD) prediction error was -0.11 (2.05) cm/year and for the original model 0.28 (2.11) cm/year. CONCLUSION: The modified prediction method, obtained after calibration of the original model, performs well in an independent patient sample and gives more accurate predictions than the original model.

Body Height↗

PubMed2006

Decrease in motor performance in children with cancer is independent of the cumulative dose of vincristine.

BACKGROUND: Impaired performance on motor tasks in children treated for acute lymphoblastic leukemia (ALL) after completion of treatment is often attributed to vincristine (VCR). Whether motor problems persist in other children who receive different cumulative doses of VCR is to the authors' knowledge not known. The objective of the current study was to determine the extent of motor problems in children with ALL, Wilms tumor (WT), B-cell non-Hodgkin lymphoma (B-NHL), and malignant mesenchymal tumors (MMT) and whether these motor problems were related to VCR dose. METHODS: In 128 children ages 4-12 years who completed treatment after at least 1 year, motor performance was measured using the Movement Assessment Battery for Children (m-ABC). RESULTS: The m-ABC scores of the total study group were significantly lower than those of the normal population (P < 0.001). There were no differences in scores noted between children with ALL, WT, B-NHL, or MMT. There also were no differences noted between those children with ALL who had received pulses of VCR and steroids during maintenance therapy and those who had not. All groups demonstrated large variability in scores. Scores were not found to be significantly different between those children who had received low (< 20 mg/m(2)), intermediate (20-40 mg/m(2)), or high (> 40 mg/m(2)) cumulative doses of VCR. Cumulative doses of corticosteroids and methotrexate did not affect scores, nor did age at diagnosis and time since the completion of therapy. CONCLUSIONS: Although motor performance was impaired in all patient groups, no correlation was found between motor performance and the cumulative dose of VCR or other drugs, age, and follow-up time. Future studies have to address several issues, including whether polymorphisms in drug metabolizing genes or drug target genes explain the significant variability noted in the long-term motor outcome of children with cancer.

Antineoplastic Agents, Phytogenic↗

PubMed2006

An insulin-like growth factor-I promoter polymorphism is associated with increased mortality in subjects with myocardial infarction in an elderly Caucasian population.

We investigated whether an insulin-like growth factor I (IGF-I) promoter polymorphism is associated with excess mortality in elderly subjects with myocardial infarction (MI). This association was assessed in 7,983 subjects of the Rotterdam Study during 14 years of follow-up. Among 345 subjects who developed a MI, the risk of mortality was 1.49 times higher in the variant carriers of the IGF-I promoter polymorphism than in the nonvariant carriers (95% confidence interval 1.10 to 2.10, p = 0.02). The risk of death increased with the number of variant alleles. Our study suggests that genetically determined low IGF-I activity is an important determinant of mortality in subjects with MI.

Aged↗

PubMed2006

Prevalence, incidence and lifetime risk of atrial fibrillation: the Rotterdam study.

AIMS: We aimed to investigate the prevalence and incidence of atrial fibrillation (AF) in a large European population-based study. METHODS AND RESULTS: The study is part of the Rotterdam study, a population-based prospective cohort study among subjects aged 55 years and above. The prevalence at baseline was assessed in 6808 participants. Incidence of AF was investigated during a mean follow-up period of 6.9 years in 6432 persons. We identified 376 prevalent and 437 incident cases. Overall prevalence was 5.5%, rising from 0.7% in the age group 55-59 years to 17.8% in those aged 85 years and above. The overall incidence rate was 9.9/1000 person-years. The incidence rate in the age group 55-59 years was 1.1/1000 person-years, rose to 20.7/1000 person-years in the age group 80-84 years and stabilized in those aged 85 years and above. Prevalence and incidence were higher in men than in women. The lifetime risk to develop AF at the age of 55 years was 23.8% in men and 22.2% in women. CONCLUSION: In this prospective study in a European population, the prevalence and incidence of AF increased with age and were higher in men than in women. The high lifetime risk to develop AF was similar to North American epidemiological data.

Age Distribution↗

PubMed2006

Identifying key parameters in cost-effectiveness analysis using value of information: a comparison of methods.

Decisions in health care must be made, despite uncertainty about benefits, risks, and costs. Value of information analysis is a theoretically sound method to estimate the expected value of future quantitative research pertaining to an uncertain decision. If the expected value of future research does not exceed the cost of research, additional research is not justified, and decisions should be based on current evidence, despite the uncertainty. To assess the importance of individual parameters relevant to a decision, different value of information methods have been suggested. The generally recommended method assumes that the expected value of perfect knowledge concerning a parameter is estimated as the reduction in expected opportunity loss. This method, however, results in biased expected values and incorrect importance ranking of parameters. The objective of this paper is to set out the correct methods to estimate the partial expected value of perfect information and to demonstrate why the generally recommended method is incorrect conceptually and mathematically.

Bias↗

PubMed2006

Optimal treatment of premature ovarian failure after treatment for Hodgkin's lymphoma is often withheld.

BACKGROUND: The purpose of this study is to determine: 1. the effect of treatment for Hodgkin's lymphoma on ovarian function, and 2. the interventions to relieve postmenopausal symptoms. METHODS: Seventy-seven consecutive patients treated between 1989 and 2003 in the Rotterdam region for Hodgkin's lymphoma stages I and II were approached for this study. A questionnaire consisting of 45 questions was carried out to evaluate premature menopausal symptoms, hormonal replacement therapy and use of contraception, menstrual cycle, and subsequent pregnancies. RESULTS: After informed consent 67 patients were willing to participate in the study and 66 patients filled in a questionnaire. After antitumor treatment 13 patients developed treatment-related premature ovarian failure, 35 patients had a spontaneous cycle, and 18 patients could not be classified as they used hormonal contraception. Women who developed treatment-related premature ovarian failure had a significantly higher mean age at the start of treatment for Hodgkin's lymphoma than women who remained premenopausal (p < 0.002). Only 6 of these 13 women (46%) received hormonal substitution. In all, 21 women conceived after antitumor treatment, and 28 children were born. All pregnancies were the result of spontaneous conception. CONCLUSIONS: The effect of antitumor treatment for Hodgkin's lymphoma on ovarian function is age dependent (odds ratio of 1.18 per year). There is a striking inconsistency regarding the management of ovarian protection before and during antitumor treatment. Premenopausal women who undergo therapy for Hodgkin's lymphoma should be offered hormonal substitution therapy after loss of ovarian function.

Adolescent↗

PubMed2006

Free dissociable insulin-like growth factor I (IGF-I), total IGF-I and their binding proteins in girls with Turner syndrome during long-term growth hormone treatment.

OBJECTIVE: To investigate the effect of GH treatment on free IGF-I levels in girls with Turner syndrome (TS) and to verify relationships between free IGF-I levels and total IGF-I, IGFBP-1, 2 and 3. Additionally, to analyse whether free IGF-I, total IGF-I, IGFBP-3 or its ratio were related to IGF-I bioactivity outcome parameters. DESIGN: Sixty-five girls with TS were randomly assigned to three different GH-dosage groups (1.3, 2.0 and 2.7 mg/m2/day). Mean duration of GH therapy was mean (SD) 8.7(2.0) years. Free IGF-I, total IGF-I and IGFBP-1, -2, -3 were determined at baseline, first, second, third and fifth year of GH therapy, before the start of oestrogen therapy, during the final year of GH treatment, 6 months after GH and 4.8(2.0) years after GH discontinuation. MAIN OUTCOME: During GH treatment, mean free IGF-I levels stayed < +2 standard deviation score (sds), whereas mean total IGF-I and IGF-I/IGFBP-3 ratio were > +2 sds. There were no differences in free IGF-I levels between the three GH groups, whereas total IGF-I and ratio levels were significantly higher in the highest GH group. The following variables contributed significantly to predicting the square root of free IGF-I levels: age, GH dose, oestrogen dose, IGFBP-1, IGFBP-3, body mass index and total IGF-I or IGF-I/IGFBP-3 ratio. However, the explaining variance did not exceed 55%. Several IGF-I bioactivity outcome parameters positively correlated with total IGF-I and IGF-I/IGFBP-3 ratio, whereas free IGF-I did not. CONCLUSIONS: During long-term GH therapy in girls with TS, mean free IGF-I levels stayed within the normal range, whereas mean total IGF-I and IGF-I/IGFBP-3 ratio exceeded the upper normal range. Although total IGF-I and the IGF-I/IGFBP-3 ratio did not accurately represent free IGF-I levels, they seemed to better represent the IGF-I bioactivity than the measured free IGF-I.

Analysis of Variance↗

PubMed2006

IGF-I gene promoter polymorphism is a predictor of survival after myocardial infarction in patients with type 2 diabetes.

OBJECTIVE: Previously we observed that non-carriers of the most common alleles of an IGF-I promoter polymorphism have low circulating IGF-I levels and an increased risk of developing myocardial infarction (MI), particularly in patients with type 2 diabetes. DESIGN: We investigated whether this IGF-I promoter polymorphism is associated with survival of type 2 diabetes in a Caucasian population aged 55 years and older. METHODS: The study was embedded in the Rotterdam Study, a prospective population-based cohort study. At baseline, 668 patients with type 2 diabetes were diagnosed, among which, 55 incident MI were ascertained during follow-up. For the present study, we used two genotype groups: non-variant carriers (homozygous for 192, 194, or 192/194 bp genotypes), and variant carriers. RESULTS: During a median follow-up of 8.8 years, 396 out of the 668 patients with type 2 diabetes (59.3%) died of various causes. The frequency of type 2 diabetes variant carrier and non-variant carriers was 28.7 and 71.3% respectively. The survival in patients with type 2 diabetes without an MI did not differ between the IGF-I genotype groups (hazard ratio (HR) = 0.8, 95% confidence interval (CI): 0.7-1.1, P = 0.1). In contrast, in those who developed an MI, variant carriers had a 2.4 times higher risk of mortality than non-variant carriers (95% CI: 1.2-4.8, P = 0.01). CONCLUSION: Our study suggests that genetically determined low IGF-I activity is an important determinant of survival in patients with type 2 diabetes who developed an MI. The IGF-I promoter polymorphism, therefore, may help to predict the future mortality risk in this group of patients.

Aged↗

PubMed2005

Interim analysis on survival data: its potential bias and how to repair it.

We consider interim analyses in clinical trials or observational studies with a time-to-event outcome variable where the survival curves are compared using the hazard ratio resulting from a proportional hazards (PH) model or tested with the logrank test or another two-sample test. We show and illustrate with an example that if the PH assumption is violated, the results of interim analyses can be heavily biased. This is due to the fact that the censoring pattern in interim analyses can be completely different from the final analysis. We argue that, when the PH assumption is violated, interim analyses are only sensible if a fixed time horizon for the final analysis is specified, and at the time of the interim analysis sufficient information is available over the whole time interval up to the horizon. We show how the bias can then be remedied by introducing in the estimation and testing procedures an appropriate weighting that reflects the weights to be expected in the final analysis. The consequences for design and analysis are discussed and some practical recommendations are given.

Bias↗