Biomedical subjects

Emmanuel Lesaffre

Publications and source records attributed to Emmanuel Lesaffre.

At least 19 recordsLinked to original sources

PubMed2007

Latent-variable models for longitudinal data with bivariate ordinal outcomes.

We use the concept of latent variables to derive the joint distribution of bivariate ordinal outcomes, and then extend the model to allow for longitudinal data. Specifically, we relate the observed ordinal outcomes using threshold values to a bivariate latent variable, which is then modelled as a linear mixed model. Random effects terms are used to tie all together repeated observations from the same subject. The cross-sectional association between the two outcomes is modelled through the correlation coefficient of the bivariate latent variable, conditional on random effects. Assuming conditional independence given random effects, the marginal likelihood, under the missing data at random assumption, is approximated using an adaptive Gaussian quadrature for numerical integration. The model provides fixed effects parameters that are subject-specific, but retain the population-averaged interpretation when properly scaled. This is particularly well suited for the situation in which population comparisons and individual level contrasts are of equal importance. Data from a psychiatric trial, the Fluvoxamine (an antidepressant drug) study, are used to illustrate the methodology.

Antidepressive Agents, Second-Generation↗

PubMed2006

Power and sample size calculations for discrete bounded outcome scores.

We consider power and sample size calculations for randomized trials with a bounded outcome score (BOS) as primary response adjusted for a priori chosen covariates. We define BOS to be a random variable restricted to a finite interval. Typically, a BOS has a J- or U-shaped distribution hindering traditional parametric methods of analysis. When no adjustment for covariates is needed, a non-parametric test could be chosen. However, there is still a problem with calculating the power since the common location-shift alternative does not hold in general for a BOS. In this paper, we consider a parametric approach and assume that the observed BOS is a coarsened version of a true BOS, which has a logit-normal distribution in each treatment group allowing correction for covariates. A two-step procedure is used to calculate the power. Firstly, the power function is defined conditionally on the covariate values. Secondly, the marginal power is obtained by averaging the conditional power with respect to an assumed distribution for the covariates using Monte Carlo integration. A simulation study evaluates the performance of our method which is also applied to the ECASS-1 stroke study.

Computer Simulation↗

PubMed2006

Stroke rehabilitation in Europe: what do physiotherapists and occupational therapists actually do?

BACKGROUND AND PURPOSE: Physiotherapy (PT) and occupational therapy (OT) are key components of stroke rehabilitation. Little is known about their content. This study aimed to define and compare the content of PT and OT for stroke patients between 4 European rehabilitation centers. METHODS: In each center, 15 individual PT and 15 OT sessions of patients fitting predetermined criteria were videotaped. The content was recorded using a list comprising 12 therapeutic categories. A generalized estimating equation model was fitted to the relative frequency of each category resulting in odds ratios. RESULTS: Comparison of PT and OT between centers revealed significant differences for only 2 of the 12 categories: ambulatory exercises and selective movements. Comparison of the 2 therapeutic disciplines on the pooled data of the 4 centers revealed that ambulatory exercises, transfers, exercises, and balance in standing and lying occurred significantly more often in PT sessions. Activities of daily living, domestic activities, leisure activities, and sensory, perceptual training, and cognition occurred significantly more often in OT sessions. CONCLUSIONS: This study revealed that the content of each therapeutic discipline was consistent between the 4 centers. PT and OT proved to be distinct professions with clear demarcation of roles.

Aged↗

PubMed2006

The logistic transform for bounded outcome scores.

The logistic transformation, originally suggested by Johnson (1949), is applied to analyze responses that are restricted to a finite interval (e.g. (0,1)), so-called bounded outcome scores. Bounded outcome scores often have a non-standard distribution, e.g. J- or U-shaped, precluding classical parametric statistical approaches for analysis. Applying the logistic transformation on a normally distributed random variable, gives rise to a logit-normal (LN) distribution. This distribution can take a variety of shapes on (0,1). Further, the model can be extended to correct for (baseline) covariates. Therefore, the method could be useful for comparative clinical trials. Bounded outcomes can be found in many research areas, e.g. drug compliance research, quality-of-life studies, and pain (and pain relief) studies using visual analog scores, but all these scores can attain the boundary values 0 or 1. A natural extension of the above approach is therefore to assume a latent score on 0,1) having a LN distribution. Two cases are considered: (a) the bounded outcome score is a proportion where the true probabilities have a LN distribution on (0,1) and (b) the bounded outcome score on [0,1] is a coarsened version of a latent score with a LN distribution on (0,1). We also allow the variance (on the transformed scale) to depend on treatment. The usefulness of our approach for comparative clinical trials will be assessed in this paper. It turns out to be important to distinguish the case of equal and unequal variances. For a bounded outcome score of the second type and with equal variances, our approach comes close to ordinal probit (OP) regression. However, ignoring the inequality of variances can lead to highly biased parameter estimates. A simulation study compares the performance of our approach with the two-sample Wilcoxon test and with OP regression. Finally, the different methods are illustrated on two data sets.

Anticholesteremic Agents↗

PubMed2006

Adverse outcomes in Belgian acute hospitals: retrospective analysis of the national hospital discharge dataset.

OBJECTIVE: The prevalence and variability of adverse outcome rates in Belgian acute hospitals is examined by using the national hospital discharge database. DESIGN: setting, and participants. Retrospective analysis based on administrative data of all Belgian acute hospitals, covering the full medical (n = 1 024 743) and surgical (n = 633 027) in-patients population for the year 2000. MAIN OUTCOME MEASURES: For 11 adverse outcomes and failure-to-rescue, the rates and variability among hospitals were studied. The all patient refined diagnostic-related groups (APR-DRG) method was used for risk adjustment. RESULTS: The prevalence of adverse outcomes was 7.12% in the medical and 6.32% in the surgical group. Rates ranged from 6.25 (deep venous thrombosis) to 32.3 (urinary tract infection) outcomes per 1000 discharges in the medical group and from 3.39 (deep venous thrombosis) to 17.6 (urinary tract infection) outcomes per 1000 discharges in the surgical group. The failure-to-rescue rate was 240 and 211 per 1000 discharges, respectively. Except for pressure ulcers and hospital-acquired sepsis, the prevalence of adverse outcomes was significantly higher (P = 0.001) in the medical group. All adverse outcome rates varied substantially among the hospitals surveyed. CONCLUSIONS: This study identifies the occurrence of adverse outcomes in a national population. It adds information to the growing body of knowledge in predominantly Anglo-Saxon countries about adverse outcomes. Striking variation exists in the risk-adjusted adverse outcome rates across Belgian acute hospitals, revealing a large potential for quality gains that encourage further action.

Adult↗

PubMed2006

Effect of intervention through a pharmaceutical care program on patient adherence with prescribed once-daily atorvastatin.

AIMS: Correct execution of prescribed dosing regimen(s) is essential for patients to benefit from lipid lowering treatments. The objective of this study was to estimate the effect of a pharmaceutical care program on the adherence of once-daily atorvastatin treatment in patients with elevated cholesterol levels. METHODS: In both linguistic regions of Belgium, two districts were randomized between usual care and a supportive intervention program. Eligible patients included hyperlipemic subjects taking atorvastatin for at least 3 months. 'Adherence' was defined as the proportion of days during which the electronic device record showed that the patient had taken the daily dose. 'Persistence' quantifies how long the treatment is executed. RESULTS: A total of 392 patients from 35 pharmacies were included in the intent-to-treat (ITT) analysis of the data (194 patients received intervention and 198 in the control group). The intervention resulted in a 6.5% increase in post-baseline adherence (p < 0.001) mainly driven by a 13% increase in persistence at 1 year (p = 0.002). CONCLUSION: A supportive pharmaceutical care program consisting of review by patient and pharmacist of each patient's electronically compiled dosing history, plus educational reminders, improves patient adherence to the once-daily atorvastatin dosing regimen, mainly by extending persistence.

Atorvastatin↗

PubMed2006

Estimating disease prevalence in a Bayesian framework using probabilistic constraints.

Studies sometimes estimate the prevalence of a disease from the results of one or more diagnostic tests that are applied to individuals of unknown disease status. This approach invariably means that, in the absence of a gold standard and without external constraints, more parameters must be estimated than the data permit. Two assumptions are regularly made in the literature, namely that the test characteristics (sensitivity and specificity) are constant over populations and the tests are conditionally independent given the true disease status. These assumptions have been criticized recently as being unrealistic. Nevertheless, to estimate the prevalence, some restrictions on the parameter estimates need to be imposed. We consider 2 types of restrictions: deterministic and probabilistic restrictions, the latter arising in a Bayesian framework when expert knowledge is available. Furthermore, we consider 2 possible parameterizations allowing incorporation of these restrictions. The first is an extension of the approach of Gardner et al and Dendukuri and Joseph to more than 2 diagnostic tests and assuming conditional dependence. We argue that this system of equations is difficult to combine with expert opinions. The second approach, based on conditional probabilities, looks more promising, and we develop this approach in a Bayesian context. To evaluate the combination of data with the (deterministic and probabilistic) constraints, we apply the recently developed Deviance Information Criterion and effective number of parameters estimated (pD) together with an appropriate Bayesian P value. We illustrate our approach using data collected in a study on the prevalence of porcine cysticercosis with verification from external data.

Animals↗

PubMed2006

A general method for dealing with misclassification in regression: the misclassification SIMEX.

We have developed a new general approach for handling misclassification in discrete covariates or responses in regression models. The simulation and extrapolation (SIMEX) method, which was originally designed for handling additive covariate measurement error, is applied to the case of misclassification. The statistical model for characterizing misclassification is given by the transition matrix Pi from the true to the observed variable. We exploit the relationship between the size of misclassification and bias in estimating the parameters of interest. Assuming that Pi is known or can be estimated from validation data, we simulate data with higher misclassification and extrapolate back to the case of no misclassification. We show that our method is quite general and applicable to models with misclassified response and/or misclassified discrete regressors. In the case of a binary response with misclassification, we compare our method to the approach of Neuhaus, and to the matrix method of Morrissey and Spiegelman in the case of a misclassified binary regressor. We apply our method to a study on caries with a misclassified longitudinal response.

Bias↗

PubMed2005

Use of time by stroke patients: a comparison of four European rehabilitation centers.

BACKGROUND AND PURPOSE: Differences exist between European countries in the proportion of patients who die or become dependent after stroke. The aim of the present study was to identify differences in the use of time by stroke patients in 4 rehabilitation centers in 4 European countries. METHODS: In each of the 4 centers, 60 randomly selected stroke patients were observed at 10-minute intervals using behavioral mapping. Observations took place on 30 weekdays selected at random, on equal numbers of morning, afternoon, and evening sessions. A logistic generalized estimating equation model with correction for differences in case mix and multiple testing was used for the analysis. RESULTS: Overall time available from different professions was the highest in the United Kingdom, but patients in the United Kingdom spent on average only 1 hour per day in therapy. This was significantly less than patients in Belgium and Germany, who spent approximately 2 hours, and patients in Switzerland who spent approximately 3 hours per day in therapy. In all centers, patients spent less than half their time in interactions and >72% of the time in nontherapeutic activities. CONCLUSIONS: Important differences in the use of time were established, which appeared dependent on management decisions rather than the number of staff available. Patients in the Swiss and German centers spent more time in therapy, possibly because of the structured organization of rehabilitation. Further studies will verify whether this has an effect on outcome.

Activities of Daily Living↗

PubMed2005

Factors influencing the survival of cryopreserved homografts. The second homograft performs as well as the first.

OBJECTIVE: To determine the life span of cryopreserved homografts implanted in the right ventricular outflow tract and the factors influencing it. METHODS: From 1989 through 2003, we reconstructed the pulmonary valve with 301 homografts in 272 patients (median age 13 years; range 4 days-69 years). Indications were tetralogy of Fallot (136), truncus (23), Rastelli repair (11), double outlet ventricle (13), endocarditis (5), and the Ross operation (84). Median follow-up was 5.7 years (range 0-14). We analyzed possible predictors of graft replacement by simple and multiple Cox regression. RESULTS: Actuarial survival was 96+/-1.2% at 1, 95+/-1.4% at 5, and 94+/-1.5% at 10 years follow-up. Three homografts were explanted because of endocarditis (excluded from the analysis). Freedom from explantation was 99.6+/-0.4% at 1, 94.5+/-1.7% at 5, and 81.8+/-4.1% at 10 years. Variables, significantly related to explantation in the univariate analysis, were younger age, small graft size, implantation in a non-anatomical position, the aortic donor homograft, a shorter aortic cross-clamp time and the implantation of a second homograft. In the multiple model, non-anatomical position (P=0.001), smaller graft size (P<0.0001) or younger age (on square root scale, P<0.0001) and clamp time (P=0.01) remain as independent risk factors. Immunological variables, like blood group incompatibility, implantation of a second homograft and short warm ischemic time were not significant. CONCLUSIONS: The life span of a cryopreserved homograft is determined by graft size (correlates with age) and the non-anatomic position (correlates with indication). In a specific patient, the second homograft performs as well as the first.

Adolescent↗

PubMed2005

A Bayesian analysis of multivariate doubly-interval-censored dental data.

A Bayesian survival analysis is presented to examine the effect of fluoride-intake on the time to caries development of the permanent first molars in children between 7 and 12 years of age using a longitudinal study conducted in Flanders. Three problems needed to be addressed. Firstly, since the emergence time of a tooth and the time it experiences caries were recorded yearly, the time to caries is doubly interval censored. Secondly, due to the setup of the study, many emergence times were left-censored. Thirdly, events on teeth of the same child are dependent. Our Bayesian analysis is a modified version of the intensity model of Harkanen et al. (2000, Scandinavian Journal of Statistics 27, 577-588). To tackle the problem of the large number of left-censored observations a similar Finnish data set was introduced. Our analysis shows no convincing effect of fluoride-intake on caries development.

Bayes Theorem↗

PubMed2005

Multivariate survival analysis for the identification of factors associated with cavity formation in permanent first molars.

This investigation used a multivariate survival model, which allowed for dependent caries data with possible censoring, to analyse the impact of timing of tooth emergence, gender and several reported oral hygiene and dietary habits on the incidence of cavity formation in permanent first molars (PFM) in young children. A 6-yr prospective oral health screening project in Flanders, Belgium, provided clinical and questionnaire data on 4,468 children. The results revealed that PFM in children who reported frequent brushing, who had no visible caries experience in the primary dentition, and who presented without occlusal plaque on the PFM had the best survival estimates. Girls had a higher incidence than boys of cavity formation in mandibular molars. The multiple survival analysis confirmed the major impact of the caries status of the deciduous dentition and self-reported oral hygiene habits on the incidence of cavities in permanent first molars.

Belgium↗

PubMed2005

An overview of methods for interval-censored data with an emphasis on applications in dentistry.

Interval-censored time-to-event data occur in many medical areas, with dentistry or AIDS research being typical representatives. This article reviews methods for the analysis of such data, with an emphasis on the use of the accelerated failure time (AFT) model. A flexible AFT model (avoiding parametric assumptions on the distribution of the error term) is described in greater detail and is used to solve a typical dental question in a longitudinal oral health study.

Belgium↗

PubMed2004

Effect of fluvastatin on long-term outcome after coronary revascularization with stent implantation.

We assessed the impact of long-term fluvastatin treatment on adverse atherosclerotic cardiac events (cardiac death, myocardial infarction, and revascularization excluding repeat interventions due to restenosis in the first 6 months) in 847 patients (fluvastatin [n = 417] or placebo [n = 430]) with average cholesterol levels treated with stents in the Lescol Intervention Prevention Study (LIPS). During the 4-year follow-up period, fluvastatin significantly decreased total cholesterol and low-density lipoprotein cholesterol levels and decreased the risk of first adverse atherosclerotic cardiac events by 30% compared with placebo (95% confidence interval -49 to -3.4, p = 0.03).

Adolescent↗

PubMed2004

Dose-response relationships within the parotid gland after radiotherapy for head and neck cancer.

BACKGROUND AND PURPOSE: To determine the salivary function, after parotid-sparing radiotherapy (RT), of different regions within the parotid gland and to evaluate dose-function relationships within the parotid glands and between patients. PATIENTS AND METHODS: Sixteen head and neck cancer patients, irradiated between September 1999 and November 2000 using a conformal parotid-sparing technique, were included in this study. Before RT and 7 months after RT (range 6-10 months), a salivary gland scintigraphy was performed in all patients combined with a single photon emission computed tomography (SPECT). The salivary excretion fraction (SEF) was measured, after stimulation, in 8-12 transverse 5mm SPECT slices of each parotid. Loss of salivary excretion fraction (dSEF %) of these slices was calculated as the proportion of SEF after RT as compared to SEF before RT. Since the planning CT-scan and the SPECT-scintigraphy were performed in the same treatment position, the dose to a transverse slice within the parotid gland could be matched to the loss of salivary excretion fraction of that respective slice. A non-linear model was fitted to the dose-loss of function data and the dose resulting in 50% loss of salivary excretion fraction (D50) was calculated. RESULTS: Before RT, all but one patient presented with normal salivary excretion fractions (SEF) of both parotid glands. Within the same parotid gland, the SEF's of the different slices were almost equal. Seven months after RT, the reduction in SEF was statistically significant (P-value<0.0001). A significant difference in loss of salivary excretion fraction (dSEF) was also observed between both parotid glands (P<0.0001) as a result of the parotid-sparing technique. When plotting the dSEF of a slice versus the dose given to that slice, doses as low as 10-15 Gy could result in a serious loss of function (dSEF>50%). After fitting a non-linear model to these plots, the mean dose resulting in 50% loss of salivary excretion fraction (D50) 7 months after RT was 22.5 Gy. A large inter-patient variability was found in D50. CONCLUSIONS: Salivary SPECT is a useful tool for the evaluation of the salivary function of different slices within the parotid gland. Before irradiation, the different slices within one parotid gland act as functional sub-units contributing equally to the function of the entire gland. Seven months after an average dose of 22.5 Gy (D50) the functional sub-unit has lost 50% of its excretion fraction. The high inter-patient variability in D50 and the observation that low doses (10-15 Gy) can induce serious loss of function should prompt us in the clinic to reduce the dose to the parotids even lower than the threshold of 22.5 Gy.

Adult↗

PubMed2004

Smooth random effects distribution in a linear mixed model.

A linear mixed model with a smooth random effects density is proposed. A similar approach to P-spline smoothing of Eilers and Marx (1996, Statistical Science 11, 89-121) is applied to yield a more flexible estimate of the random effects density. Our approach differs from theirs in that the B-spline basis functions are replaced by approximating Gaussian densities. Fitting the model involves maximizing a penalized marginal likelihood. The best penalty parameters minimize Akaike's Information Criterion employing Gray's (1992, Journal of the American Statistical Association 87, 942-951) results. Although our method is applicable to any dimensions of the random effects structure, in this article the two-dimensional case is explored. Our methodology is conceptually simple, and it is relatively easy to fit in practice and is applied to the cholesterol data first analyzed by Zhang and Davidian (2001, Biometrics 57, 795-802). A simulation study shows that our approach yields almost unbiased estimates of the regression and the smoothing parameters in small sample settings. Consistency of the estimates is shown in a particular case.

Biometry↗

PubMed2003

A note on non-parametric ANCOVA for covariate adjustment in randomized clinical trials.

Koch et al. recently (1998) proposed two covariate-adjusted approaches for the comparison of continuous, ordinal and binary responses in a randomized clinical trial (Statist. Med. 1998; 17: 1863-1892). The first is a randomization approach while the second assumes that the study is a sample of a population. Here, we study the second approach and consider the simplest cases of two treatments with a continuous response and with a binary response. Koch's second approach will be compared with the classical ANCOVA for a continuous response. From this relationship we demonstrate that Koch's method cannot preserve the probability of the type I error. Simulations with continuous responses as well as with binary outcomes confirm the aforementioned theoretical result on the performance of Koch's method under the null hypothesis of no treatment effect. However, this poses only a problem for relatively small to moderate sample sizes. Further, as specified in the original paper of Koch et al., the first approach does preserve the type I error for any sample size, as the P-values can be reported in an exact manner (Statist. Med. 1998; 17: 1863-1892). Finally, we propose a correction factor for Koch's test statistic that better preserves the type I error.

Computer Simulation↗

PubMed2003

A comparison of group sequential methods for binary longitudinal data.

Interim analyses are conducted to allow for early termination of the trial, for ethical as well as economical reasons. Here we consider interim analyses in repeated measurements studies where the measurements are binary. Two methods for analysing this kind of data are compared according to their operating characteristics. A subject-specific approach based on the logistic random-effects model is compared with the population-averaged approach based on the generalized estimating equations. The comparison is illustrated with simulations using a randomized clinical trial for toenail fungal infection.

Antifungal Agents↗