Biomedical subjects

Michal Abrahamowicz

Publications and source records attributed to Michal Abrahamowicz.

At least 37 records · Page 2Linked to original sources

PubMed2005

Modifiable risk factors associated with clearance of type-specific cervical human papillomavirus infections in a cohort of university students.

BACKGROUND: Previous findings regarding risk factors for human papillomavirus (HPV) persistence, other than viral determinants, identified from prospective cohort studies have been inconsistent in part because study designs have differed with respect to differing HPV detection methods and varying lengths of follow-up time. Therefore, the objectives of this study were to continue the search for epidemiologic risk factors of persistent cervical HPV infections and determine what behaviors differed between those women with transient HPV infections and those women who cannot clear their type-specific HPV infections. METHODS: Female university students (n = 621) in Montreal were followed for 24 months at 6-month intervals. At each visit, a cervical cell specimen was collected. HPV DNA was detected using the MY09/MY11 PCR protocol and 27 HPV genotypes were identified by the line blot assay (Roche Molecular Systems, Inc., Alameda, CA). Proportional hazards regression was used to estimate the crude and adjusted hazard ratios of clearing a type-specific high-risk (n = 222) or low-risk (n = 105) HPV infection over time according to specific baseline and time-dependent covariates. RESULTS: Daily consumption of vegetables seemed to increase the rate of HPV clearance independent of type. The use of tampons was associated with a reduced rate of high-risk HPV clearance, whereas regular condom use was associated with an increased rate of low-risk HPV clearance only. CONCLUSION: Some proactive measures can be taken to increase the rate of HPV clearance, and there may be some differences between the sets of predictors of low-risk and high-risk HPV clearance.

Adolescent↗

PubMed2005

Myocardial injury in critically ill patients: relation to increased cardiac troponin I and hospital mortality.

OBJECTIVE: To examine the relationship between myocardial injury, assessed by cardiac troponin I (cTnI) levels, and outcome in selected critically ill patients without acute coronary syndromes or cardiac dysfunction. DESIGN AND SETTING: Prospective, observational study in the emergency ICU of a university teaching hospital. POPULATION: Over a 6-month period, 217 consecutive patients admitted to the ICU were studied. METHODS AND RESULTS: cTnI assays were performed in all patients on admission to the ICU. The incidence of myocardial injury, defined by cTnI level > 0.1 ng/mL, was 32% (69 of 217 patients). Overall mortality was 27% (58 of 217 patients). Patients with myocardial injury had a mortality rate of 51%, compared with only 16% mortality for those without myocardial injury (p < 0.001). The hospital mortality rate was highest among older patients (71 +/- 14% vs 58.5 +/- 20%, p < 0.0001) and patients with higher simplified acute physiology scale (SAPS) II score (62 +/- 25% vs 37 +/- 17%, p < 0.0001). Mechanical ventilation was associated with higher in-hospital death (50% vs 31%, for patients who died in the hospital vs those who were discharged alive; p = 0.03). Elevated blood levels of cTnI were found to be independently associated with hospital mortality, regardless of the presence of SAPS II score and mechanical ventilation, in the logistic regression analysis (odds ratio, 2.09; 95% confidence interval, 1.06 to 4.11; p = 0.01). CONCLUSIONS: This study demonstrates the high frequency of myocardial injury (32%) in critically ill patients without acute coronary syndromes or cardiac dysfunction on admission to ICU. Myocardial injury is an independent determinant of hospital mortality. Assessment of myocardial injury on admission to ICU would make it possible to identify patients at increased risk of death.

Adult↗

PubMed2005

A comparison of prospective and retrospective evaluations of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index for systemic lupus erythematosus.

OBJECTIVE: To evaluate the comparability of prospective and retrospective evaluations of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SLICC/ACR DI). METHODS: Consecutive patients meeting ACR criteria for SLE were enrolled prospectively in our cohort. Prospective SLICC/ACR DI scores were collected on the 134 cohort members who were observed in the cohort between 1993-1999. The last available prospective SLICC/ACR DI scores were compared to scores that were retrospectively assigned (for the corresponding time point) from chart review by a research nurse blinded to the prospective values. Intra- and inter-observer agreement was assessed. Kappa coefficients with 95% confidence intervals (CI) were determined. RESULTS: The kappa correlation coefficient for agreement between prospective versus retrospective total damage scores was 0.68 (95% CI 0.54-0.81). Moderate to very good agreement was also observed with respect to the 12 individual organ systems itemized in this damage index. Substantial agreement was found between assessments done by different research nurses and for repeat assessments done by the same research nurse. CONCLUSION: These data suggest good agreement between prospective and retrospective evaluations of the SLICC/ACR DI scores.

Adult↗

PubMed2004

Using generalized additive models to reduce residual confounding.

Traditionally, confounding by continuous variables is controlled by including a linear or categorical term in a regression model. Residual confounding occurs when the effect of the confounder on the outcome is mis-modelled. A continuous representation of a covariate was previously shown to result in a less biased estimate of the adjusted exposure effect than categorization provided the functional form of the covariate-outcome relationship is correctly specified. However, this is rarely known. In contrast to parametric regression, generalized additive models (GAM) fit a smooth dose-response curve to the data, without requiring a priori knowledge of the functional form. We used simulations to compare parametric multiple logistic regression vs its non-parametric GAM extension in their ability to control for a continuous confounder. We also investigated several issues related to the implementation of GAM in this context, including: (i) selecting the degrees of freedom; and (ii) alternative criteria for inclusion/exclusion of the potential confounder and for choosing between parametric and non-parametric representation of its effect. The impact of the shape and strength of the confounder-disease association, sample size, and the correlation between the confounder and exposure were investigated. Simulations showed that when the confounder has a non-linear association with the outcome, compared to a parametric representation, GAM modelling (i) reduced the mean squared error for the adjusted exposure effect; (ii) avoided inflation of the type I error for testing the exposure effect. When the true confounder-outcome relationship was linear, GAM performed as well as the parametric logistic regression. When modelling a continuous exposure non-parametrically, in the presence of a continuous confounder, our results suggest that assuming a linear effect of the confounder and focussing on the non-linearity of the exposure-outcome relationship leads to spurious findings of non-linearity: joint non-linear modelling is necessary. Overall, our results suggest that the use of GAM to reduce residual confounding offers several improvements over conventional parametric modelling.

Confounding Factors, Epidemiologic↗

PubMed2004

Updated risk factor values and the ability of the multivariable risk score to predict coronary heart disease.

Most existing coronary risk assessment methods are based on baseline data only. The authors compared the predictive ability of coronary multivariable risk scores based on updated versus baseline risk factors and investigated the optimal frequency of updating. Data from 16 biennial examinations of 4,962 subjects from the original Framingham Heart Study (1948-1978) were used. The predictive ability of three multivariable risk scores was evaluated through 10-fold cross-validation. The baseline-only multivariable risk score was computed using baseline values of coronary risk factors applied to a Cox model estimated from baseline data. The two other approaches relied on updated risk factors and included them in the models estimated from, respectively, baseline and updated data. All analyses were stratified by sex and age. For 30, 14, and 10 years of follow-up, the predictive ability of the baseline-only multivariable risk score was substantially poorer than that of the models using updated risk factors. Between the two latter models, the one estimated from updated data ensured better prediction than the one estimated from baseline data for 30 years of follow-up among younger subjects only. The results suggest that coronary risk assessment can be improved by utilizing updated risk factors and that the optimal frequency of updating may vary across subpopulations.

Adult↗

PubMed2004

Bias due to aggregation of individual covariates in the Cox regression model.

The impact of covariate aggregation, well studied in relation to linear regression, is less clear in the Cox model. In this paper, the authors use real-life epidemiologic data to illustrate how aggregating individual covariate values may lead to important underestimation of the exposure effect. The issue is then systematically assessed through simulations, with six alternative covariate representations. It is shown that aggregation of important predictors results in a systematic bias toward the null in the Cox model estimate of the exposure effect, even if exposure and predictors are not correlated. The underestimation bias increases with increasing strength of the covariate effect and decreasing censoring and, for a strong predictor and moderate censoring, may exceed 20%, with less than 80% coverage of the 95% confidence interval. However, covariate aggregation always induces smaller bias than covariate omission does, even if the two phenomena are shown to be related. The impact of covariate aggregation, but not omission, is independent of the covariate-exposure correlation. Simulations involving time-dependent aggregates demonstrate that bias results from failure of the baseline covariate mean to account for nonrandom changes over time in the risk sets and suggest a simple approach that may reduce the bias if individual data are available but have to be aggregated.

Bias↗

PubMed2004

A proportional hazards model with time-dependent covariates and time-varying effects for analysis of fetal and infant death.

Birth-weight- and gestational-age-specific perinatal mortality curves intersect when compared by race and maternal smoking. The authors propose a new measure to replace fetal and infant mortality and an analytic strategy to assess the effects of risk factors on this outcome. They used 1998 data for US Blacks and Whites. Age-specific post-last menstrual period (LMP) mortality rate was defined as the proportion of deaths (stillbirth, perinatal death, or infant death) at a given age post-LMP. The authors used extended Cox regression with time-varying covariates and hazard ratios to model the effects of race and smoking on post-LMP mortality. Perinatal mortality rates (conventional calculation) for Blacks and Whites showed the expected crossover. However, analyses of post-LMP mortality showed no crossover. For the Black-White comparison, a hazard ratio of 1.72 (95% confidence interval: 1.67, 1.77) was obtained. The hazard was higher for smokers than for nonsmokers, but the hazard ratio increased from 1.09 (95% confidence interval: 0.98, 1.22) at 22 weeks to 1.82 (95% confidence interval: 1.72, 1.92) at 40 weeks. The hazard ratio associated with birth was also time dependent: higher than 1 for preterm gestation and lower than 1 for term gestation. The increasing adverse effect of smoking with gestational age suggests an accumulating effect of smoking on mortality. Modeling post-LMP mortality eliminates the crossover paradox for race and maternal smoking in a single statistical model.

Birth Weight↗

PubMed2004

Mortality rates in elderly patients who take different angiotensin-converting enzyme inhibitors after acute myocardial infarction: a class effect?

BACKGROUND: Several randomized, controlled trials show that angiotensin-converting enzyme (ACE) inhibitors improve survival in patients who have had an acute myocardial infarction. However, existing data from trials do not address whether all ACE inhibitors benefit patients similarly. OBJECTIVE: To evaluate whether all ACE inhibitors are associated with similar mortality in patients 65 years of age or older who have had an acute myocardial infarction. DESIGN: Retrospective cohort study that used linked hospital discharge and prescription databases containing information on 18 453 patients 65 years of age or older who were admitted for an acute myocardial infarction between 1 April 1996 and 31 March 2000. SETTING: 109 hospitals in Quebec, Canada. PATIENTS: 7512 patients who filled a prescription for an ACE inhibitor within 30 days of discharge and who continued to receive the same drug for at least 1 year. MEASUREMENTS: The association between the specific drugs and clinical outcomes was measured by using Cox proportional hazards models, with adjustment for demographic, clinical, physician, and hospital variables and dosage categories, represented by time-dependent variables. RESULTS: Enalapril, fosinopril, captopril, quinapril, and lisinopril were associated with higher mortality than was ramipril; the adjusted hazard ratios and 95% CIs were 1.47 (95% CI, 1.14 to 1.89), 1.71 (CI, 1.29 to 2.25), 1.56 (CI, 1.13 to 2.15), 1.58 (CI, 1.10 to 2.82), and 1.28 (CI, 0.98 to 1.67), respectively. The adjusted hazard ratio associated with perindopril was 0.98 (CI, 0.60 to 1.60). LIMITATIONS: The administrative databases did not contain detailed clinical information, and unmeasured factors associated with a patient's risk for death may have influenced physicians' prescription choices. CONCLUSION: Survival benefits in the first year after acute myocardial infarction in patients 65 years of age or older seem to differ according to the specific ACE inhibitor prescribed. Ramipril was associated with lower mortality than most other ACE inhibitors.

Aged↗

PubMed2004

Longitudinal patterns of new Benzodiazepine use in the elderly.

PURPOSE: To characterize longitudinal patterns of Benzodiazepine use in the elderly. METHODS: Prospective cohort of 78 367 community-dwelling Quebec residents aged 66 years or more who were new Benzodiazepine users, was followed for 5 years, 1989-1994. Data acquired from four population-based, provincial administrative databases were used to create time-dependent measures of change in dosage, switching or adding Benzodiazepines for 11 drugs listed in the provincial formulary. Subject-specific Spearman's rank correlation coefficients between dose and time were used to measure the tendency of increasing dose with consecutive periods of use. Multiple logistic regression and generalized estimating equations (GEE) models evaluated subject characteristics associated with increasing dose. RESULTS: The mean duration of uninterrupted Benzodiazepine use was 75.5 days. The mean daily dose was about half the recommended adult daily dose but 8.6% of subjects exceeded the recommended adult dose. Some of them (28.8%) switched medications at least once and 8.2% filled two or more prescriptions concurrently. For women, older age at date of first prescription was associated with increasing dose over time (odds ratio (OR) for 10 year age increase = 1.23, p < 0.001). CONCLUSION: Long periods of Benzodiazepine use are frequent among Quebec elderly. The evidence of increasing dose, particularly for older women, and long-duration of use has important implications for clinicians.

Age Factors↗

PubMed2004

Use of time-dependent measures to estimate benefits of beta-blockers after myocardial infarction.

PURPOSE: To estimate the reduction in all-cause mortality conferred by beta-blockers in a population-based cohort of elderly survivors of myocardial infarction during the year following hospital discharge. METHODS: A dynamic retrospective cohort was assembled from persons aged 66 years and over surviving myocardial infarction in Quebec between 1990 and 1993. Information on hospitalizations was linked to medication and physician claims, demographic characteristics and vital status. Subjects prescribed beta-blockers at hospital discharge had fewer comorbid medical conditions, less pre-existing cardiovascular disease and less severe infarcts. To control for these differences, analyzes were restricted to subjects receiving at least one beta-blocker and mortality was compared between periods with and without beta-blocker exposure using Cox proportional hazard models. RESULTS: Among 14,547 survivors of myocardial infarction, 41% were dispensed at least one beta-blocker. Among those subjects, the risk of dying during periods of beta-blocker use was reduced 40% (hazard ratio = 0.6; 95% CI: 0.5, 0.7). CONCLUSION: Confounding by indication threatens the validity of observational studies of intended effects of medications. For elderly survivors of myocardial infarction, the estimated benefit of beta-blockers from observational studies is greater than the estimate from randomized trials. Greater benefits do not seem to be an artifact arising from systematically prescribing beta-blockers to subjects with better prognosis. Reducing confounding by indication can enhance the validity of observational studies of medications and widen research applications of administrative health databases. While the actual benefits of medications are never truly known these studies can provide a credible range that brackets the truth.

Adrenergic beta-Antagonists↗

PubMed2004

Statistical measures were proposed for identifying longitudinal patterns of change in quantitative health indicators.

OBJECTIVE: To propose statistical measures to identify different longitudinal patterns of change in quantitative health indicators. METHODS: The authors propose several simple measures to discriminate between stable-unstable, increasing-decreasing, linear-nonlinear, monotonic-nonmonotonic patterns of change. They then suggest using factor analysis to select the subset of nonredundant measures, and cluster analysis, based on the selected measures, to identify subgroups of patients with similar longitudinal trajectories. The proposed approach is illustrated using data on osteoarthritis disability from a longitudinal study undertaken in Toronto, Ontario, in 1996-2001. Disability was measured at four points in time for 835 patients, using the Western Ontario and McMaster Universities Osteoarthritis (WOMAC) index. RESULTS: The proposed measures allowed the detection of individual patients with specific patterns of change and identification of four different groups of patients with WOMAC scores that are (1) regularly increasing, (2) regularly decreasing, (3) stable over time, or (4) highly unstable, with abrupt changes or short-term fluctuations. CONCLUSION: The proposed approach may be used to (1) screen even large databases to identify particular patterns of change; or (2) summarize different patterns of change by classifying patients into groups with similar trajectories. Once identified, the groups can be investigated to determine whether there are differences in patient characteristics or outcomes.

Aged↗

PubMed2004

Clinical, immunological and virological evolution in patients with CD4 T-cell count above 500/mm3: is there a benefit to treat with highly active antiretroviral therapy (HAART)?

To assess the clinical, immunological and virological evolution in HIV-1 infected patients with CD4 T-cell count above 500/mm3, a historical cohort of 202 untreated and 96 patients treated with HAART was longitudinally studied (median follow-up 36 months). Fourteen untreated and 2 treated patients experienced clinical progression (p = 0.09). The difference between baseline CD4 T-cell count and after 3 years, was -240/mm3 in the untreated group +19/mm3 in the HAART group (p < 10(-3)). A better immunological outcome was significantly associated with a HIV sexual contamination (p = 0.01), HAART (p = 0.01), high baseline CD4 T-cell count (p < 10(-3)) and low baseline HIV viral load (p = 0.01). In the HAART group, the incidence rate of antiretroviral modification due to tolerance difficulties was 0.23+/-0.36/patient year. A sustained undetectable HIV viral load was correlated with a low baseline HIV viral load (p = 0.003) and to be antiretroviral naive (p < 10(-3)). Thus, HAART provide a better immunological outcome in patients with high CD4 T-cell count. However, the CD4 decay slope after 3 years, the risk of therapeutic side-effects and the low risk of clinical progression do not support systematic treatment of those patients.

Adult↗

PubMed2004

Effect of amiodarone dose on the risk of permanent pacemaker insertion.

Bradyarrhythmia requiring permanent pacemaker insertion has been associated with amiodarone use but the effect of amiodarone dose has not been investigated. In order to determine the effect of amiodarone dose on the risk of requiring permanent pacemaker insertion, a cohort of 15,824 subjects with atrial fibrillation (AF) and prior myocardial infarction was established. This study included 1,340 subjects who received a first prescription of amiodarone at > 65 years of age. Cox regression with daily dose and cumulative dose (weighted for recency of exposure) represented by time dependent covariates was performed, adjusting for baseline risk factors and time dependent exposure to other cardiac medications. The incidence of pacemaker insertion was 2.2% per person-year during a mean of 1.8 +/- 1.5 years of follow-up, and 5.2% per person-year during the first 90 days of amiodarone exposure. Amiodarone daily doses > 200 mg were associated with an increased risk during the entire follow-up HR 2.0; 95% CI 1.0 to 4.1) as well as during the first 90 days (HR 3.1; 95% CI 1.1 to 9.0). Cumulative doses greater than the equivalent of continuous therapy with 200 mg per day were also associated with an increased risk (HR 2.8; 95% CI 1.4 to 5.5). Baseline conduction disorder or sinus node dysfunction was the only other significant predictor of pacemaker insertion. This study suggests that there is a dose dependent increased risk of permanent pacemaker insertion associated with amiodarone use that appears to be greatest during the initial months of treatment.

Age Factors↗

PubMed2004

Epidemiology of malaria in a hypoendemic Brazilian Amazon migrant population: a cohort study.

The present study describes aspects of the epidemiology of malaria in a migrant population living in a hypoendemic area in Brazil using an open cohort study design. Rural settlement residents in Leonislândia, Peixoto de Azevedo, Mato Grosso, Brazil were followed from September 1996 to April 1997. At baseline, an interview and malaria diagnoses were carried out and spleen size was measured. Incident cases were detected through follow-up visits and laboratory records. Cox regression was used to assess risk factors for time to malaria onset. Eighty percent (n = 414) of the study population (n = 521) contributed follow-up data. Overall, malaria prevalence during any study visit ranged from 0.3% to 5.4% and the malaria incidence rate (IR) was 4.49 (95% confidence interval = 3.66, 5.46) per 100 person-months. The IR of Plasmodium vivax malaria was approximately four times higher than the IR for P. falciparum malaria during follow-up. Among individuals who had had malaria during his or her lifetime, 14.03% reported hospitalization (median duration = 3 days) and 70.1% reported days of work lost (median duration = 4 days for P. falciparum malaria and 3 days for P. vivax malaria) related to the last malaria episode. No important risk factor was associated with the malaria IR. The fact that neither work-related factors nor age was associated with the risk of malaria indicates that indoor/peri-domiciliary transmission by the local vector is more important or as important as workplace-related transmission.

Adolescent↗

PubMed2003

Evaluation of Cox's model and logistic regression for matched case-control data with time-dependent covariates: a simulation study.

Case-control studies are typically analysed using the conventional logistic model, which does not directly account for changes in the covariate values over time. Yet, many exposures may vary over time. The most natural alternative to handle such exposures would be to use the Cox model with time-dependent covariates. However, its application to case-control data opens the question of how to manipulate the risk sets. Through a simulation study, we investigate how the accuracy of the estimates of Cox's model depends on the operational definition of risk sets and/or on some aspects of the time-varying exposure. We also assess the estimates obtained from conventional logistic regression. The lifetime experience of a hypothetical population is first generated, and a matched case-control study is then simulated from this population. We control the frequency, the age at initiation, and the total duration of exposure, as well as the strengths of their effects. All models considered include a fixed-in-time covariate and one or two time-dependent covariate(s): the indicator of current exposure and/or the exposure duration. Simulation results show that none of the models always performs well. The discrepancies between the odds ratios yielded by logistic regression and the 'true' hazard ratio depend on both the type of the covariate and the strength of its effect. In addition, it seems that logistic regression has difficulty separating the effects of inter-correlated time-dependent covariates. By contrast, each of the two versions of Cox's model systematically induces either a serious under-estimation or a moderate over-estimation bias. The magnitude of the latter bias is proportional to the true effect, suggesting that an improved manipulation of the risk sets may eliminate, or at least reduce, the bias.

Canada↗

PubMed2003

A relative survival regression model using B-spline functions to model non-proportional hazards.

Relative survival, a method for assessing prognostic factors for disease-specific mortality in unselected populations, is frequently used in population-based studies. However, most relative survival models assume that the effects of covariates on disease-specific mortality conform with the proportional hazards hypothesis, which may not hold in some long-term studies. To accommodate variation over time of a predictor's effect on disease-specific mortality, we developed a new relative survival regression model using B-splines to model the hazard ratio as a flexible function of time, without having to specify a particular functional form. Our method also allows for testing the hypotheses of hazards proportionality and no association on disease-specific hazard. Accuracy of estimation and inference were evaluated in simulations. The method is illustrated by an analysis of a population-based study of colon cancer.

Aged↗

PubMed2003

Measuring differences between patients' and physicians' health perceptions: the patient-physician discordance scale.

We report on the development and validation of an instrument to assess discordance between physicians and their patients on evaluations of health-related information: the Patient-Physician Discordance Scale (PPDS). The 10-item questionnaire is designed to be employed across chronic diseases and can be used in clinical practice and research. It measures the extent of patient-physician discordance on five aspects of the patient's health status and five aspects of the office visit. A prospective study with 200 outpatients with inflammatory bowel disease and their treating physicians revealed that the 10-item discordance scores had good construct validity and satisfactory convergent validity. Overall discordance and the three subscales, discordance on symptoms and treatment, well-being, and communication and satisfaction, identified by factor analysis, had acceptable internal consistency. Patient and physician ratings demonstrated moderate-to-high concurrent validity. Study limitations and directions for future research with PPDS are discussed.

Abdominal Pain↗

PubMed2003

Occurrence of cervical infection with multiple human papillomavirus types is associated with age and cytologic abnormalities.

BACKGROUND: Few aspects of the occurrence of infections with multiple HPV types have been described. Since the immunity conferred by vaccines is type-specific, the epidemiology of such coinfections needs to be addressed. GOAL: The goal of the study was to document the prevalence and incidence of infection with multiple HPV types and the distribution of HPV types in coinfections. STUDY DESIGN: In a prospective cohort of 2075 Brazilian women, cervical specimens were collected for cytology and HPV detection. Information on potential risk factors was obtained by interview. RESULTS: The prevalence of HPV coinfections was 3% among cytologically normal women, 10% among women with ASCUS, 23% among those with LSIL, and 7% among those with HSIL. The incidence rate of coinfection declined markedly with age (Ptrend<0.001). Some HPV types co-occurred less frequently than expected, namely, HPV 16 and 18 occurring with other oncogenic HPV types and HPV 6/11. CONCLUSION: We have observed that occurrence of HPV coinfection was dependent both on age and on the presence of cytologic abnormalities. These results may have implications for vaccine development and for public health decisions about vaccination programs.

Adolescent↗