Biomedical subjects

Lisa M Lix

Publications and source records attributed to Lisa M Lix.

7 recordsLinked to original sources

PubMed2005

Robust tests for the multivariate Behrens-Fisher problem.

Hotelling's T2 procedure is used to test the equality of means in two-group multivariate designs when covariances are homogeneous. A number of alternatives to T2, which are robust to covariance heterogeneity, have been proposed in the literature. However, all are sensitive to departures from multivariate normality. We demonstrate how to obtain multivariate tests that are robust to covariance heterogeneity and non-normality with estimators of location and scale based on trimming and Winsorizing. The performance of six alternatives to T2 was examined via Monte Carlo methods when characteristics of the research design, degree of covariance heterogeneity, and degree of non-normality were manipulated. We have recently developed a program written in the SAS/IML language that can be used to implement these robust multivariate tests. Recommendations are provided on the specific data-analytic conditions under which these tests should be adopted.

Analysis of Variance↗

PubMed2004

Fracture risk among First Nations people: a retrospective matched cohort study.

BACKGROUND: Canadian First Nations people have unique cultural, socioeconomic and health-related factors that may affect fracture rates. We sought to determine the overall and site-specific fracture rates of First Nations people compared with non-First Nations people. METHODS: We studied fracture rates among First Nations people aged 20 years and older (n = 32 692) using the Manitoba administrative health database (1987-1999). We used federal and provincial sources to identify ethnicity, and we randomly matched each First Nations person with 3 people of the same sex and year of birth who did not meet this definition of First Nations ethnicity (n = 98 076). We used a provincial database of hospital separations and physician billing claims to calculate standardized incidence ratios (SIRs) and 95% confidence intervals (CIs) for each fracture type based on a 5-year age strata. RESULTS: First Nations people had significantly higher rates of any fracture (age- and sex-adjusted SIR 2.23, 95% CI 2.18-2.29). Hip fractures (SIR 1.88, 95% CI 1.61-2.14), wrist fractures (SIR 3.01, 95% CI 2.63-3.42) and spine fractures (SIR 1.93, 95% CI 1.79-2.20) occurred predominantly in older people and women. In contrast, craniofacial fractures (SIR 5.07, 95% CI 4.74-5.42) were predominant in men and younger adults. INTERPRETATION: First Nations people are a previously unidentified group at high risk for fracture.

Adult↗

PubMed2004

Decrease in antibiotic use among children in the 1990s: not all antibiotics, not all children.

BACKGROUND: Decreases in antibiotic use were widely reported in the 1990s. This study was undertaken to determine trends in the use of antibiotics from fiscal year (FY) 1995 (April 1995 to March 1996) to FY 2001 in a complete population of Manitoba children. METHODS: Using Manitoba's health care databases, we determined annual population-based rates of antibiotic prescription among children by antibiotic class (narrow-spectrum and broader-spectrum antibiotics), age group, physician diagnosis (e.g., otitis media or bronchitis) and neighbourhood income in urban areas (derived from the 1996 census). Antibiotic prescription rates were generated within a generalized linear model framework with general estimating equations, and differences between FY 2001 and FY 1995 were tested. Differences in antibiotic use over time were compared across antibiotic classes, age groups, diagnoses and income neighbourhoods. RESULTS: The overall antibiotic prescription rate decreased by almost one-third, from 1.2 prescriptions per child in FY 1995 to 0.9 prescriptions in FY 2001. Total antibiotic use declined for all respiratory tract infections; decreases were greatest for the sulfonamides (decrease to less than one-third the FY 1995 rate) and narrow-spectrum macrolides (decrease to less than half the FY 1995 rate). In contrast, the FY 2001 rate for broader-spectrum macrolides was as much as 12.5 times the FY 1995 rate. Otitis media accounted for one-quarter of the use of the latter agents. Preschool children and low-income children received the greatest number of antibiotic prescriptions. Declines in antibiotic prescriptions were of a lesser magnitude for low-income children (for whom rates in FY 2001 were four-fifths the rates in FY 1995) than for higher-income children (for whom rates in FY 2001 were about two-thirds the rates in FY 1995). INTERPRETATION: Overall, antibiotic use declined over the late 1990s in this population of Canadian children, but the increasing use of broader-spectrum macrolides and higher rates of antibiotic use among preschool and low-income children may have implications for antibiotic resistance.

Adolescent↗

PubMed2004

Multivariate tests of means in independent groups designs. Effects of covariance heterogeneity and nonnormality.

Health evaluation research often employs multivariate designs in which data on several outcome variables are obtained for independent groups of subjects. This article examines statistical procedures for testing hypotheses of multivariate mean equality in two-group designs. The conventional test for multivariate means, Hotelling's T2, rests on certain assumptions about the distribution of the data and the population variances and covariances. When these assumptions are violated, which is often the case in applied health research, T2 will result in invalid conclusions about the null hypothesis. This article describes parametric procedures that are robust, or insensitive, to assumption violations. A numeric example illustrates the statistical concepts that are presented and a computer program to implement these robust solutions is introduced.

Algorithms↗

PubMed2003

A generally robust approach to hypothesis testing in independent and correlated groups designs.

Standard least squares analysis of variance methods suffer from poor power under arbitrarily small departures from normality and fail to control the probability of a Type I error when standard assumptions are violated. These problems are vastly reduced when using a robust measure of location; incorporating bootstrap methods can result in additional benefits. This paper illustrates the use of trimmed means with an approximate degrees of freedom heteroskedastic statistic for independent and correlated groups designs in order to achieve robustness to the biasing effects of nonnormality and variance heterogeneity. As well, we indicate when a boostrap methodology can be effectively employed to provide improved Type I error control. We also illustrate, with examples from the psychophysiological literature, the use of a new computer program to obtain numerical results for these solutions.

Algorithms↗

PubMed2002

Trends in health and healthcare utilization in Manitoba.

Trends in health status and healthcare utilization were examined for regions of Manitoba from 1985 to 2000. While the provincial premature mortality rate decreased, the difference between the northern and southern regions increased. Hospital admissions remained stable despite major bed closures and an aging population; a decrease in hospital days per capita was observed in all regions. Physician contact rates also remained constant despite a 40% increase in the number of seniors.

Adolescent↗

PubMedDate not supplied

Demographic risk factors for fracture in First Nations people.

BACKGROUND: Recently, First Nations people were shown to be at high fracture risk compared with the general population. However, factors contributing to this risk have not been examined. This analysis focusses on geographic area of residence, income level, and diabetes mellitus as possible explanatory variables since they have been implicated in the fracture rates observed in other populations. METHODS: A retrospective, population-based matched cohort study of fracture rates was performed using the Manitoba administrative health data (1987-1999). The First Nations cohort included all Registered First Nations adults (20 years or older) as indicated in either federal and/or provincial files (n = 32,692). Controls (up to three for each First Nations subject) were matched by year of birth, sex and geographic area of residence. After exclusion of unmatched subjects, analysis was based upon 31,557 First Nations subjects and 79,720 controls. RESULTS: Overall and site-specific fracture rates were significantly higher in the First Nations cohort. Income quintile, geographic area of residence, and diabetes were fracture determinants but the excess fracture risk of First Nations ethnicity persisted even after adjustment for these factors. CONCLUSION: First Nations people are at high risk for fracture but the causal factors contributing to this are unclear. Further research is needed to evaluate the importance of other potential explanatory variables.

Adult↗