Jane F. Pendergast

Modelling covariance structure in the analysis of repeated measures data.

The term ‘repeated measures’ refers to data with multiple observations on the same sampling unit. In most cases, the multiple observations are taken over time, but they could be over space. It is usually plausible to assume that observations on the same unit are correlated. Hence, statistical analysis of repeated measures data must address the issue of covariation between measures on the same unit. Until recently, analysis techniques available in computer software only offered the user limited and inadequate choices. One choice was to ignore covariance structure and make invalid assumptions. Another was to avoid the covariance structure issue by analysing transformed data or making adjustments to otherwise inadequate analyses. Ignoring covariance structure may result in erroneous inference, and avoiding it may result in inefficient inference. Recently available mixed model methodology permits the covariance structure to be incorporated into the statistical model. The MIXED procedure of the SAS® System provides a rich selection of covariance structures through the RANDOM and REPEATED statements. Modelling the covariance structure is a major hurdle in the use of PROC MIXED. However, once the covariance structure is modelled, inference about fixed effects proceeds essentially as when using PROC GLM. An example from the pharmaceutical industry is used to illustrate how to choose a covariance structure. The example also illustrates the effects of choice of covariance structure on tests and estimates of fixed effects. In many situations, estimates of linear combinations are invariant with respect to covariance structure, yet standard errors of the estimates may still depend on the covariance structure. Copyright

A Rodent Model of Spontaneous Stereotypy: Initial Characterization of Developmental, Environmental, and Neurobiological Factors

Stereotypies are patterns of motor behavior that are repetitive, excessive, topographically invariant, and that lack any obvious function or purpose. In humans, stereotyped behaviors are associated with psychiatric, neurological, and developmental disorders. In animals, stereotypy has been frequently associated with adverse environmental circumstances and often related to alterations in striatal dopamine. To assess the development of stereotyped behaviors and to test the hypothesis that these behaviors are associated with environmental restriction, deer mice were housed in either standard laboratory cages or larger, enriched cages, and the development of stereotypy was followed from weaning over a 17-week period. Standard-caged deer mice engaged in stereotyped behaviors at a higher rate and developed these behaviors more quickly when compared to animals in enriched caging. Additionally, enriched caging was associated with higher rates of patterned running, whereas jumping and backward somersaulting were typically observed in standard cages. In addition, there was a significant effect of litter, but no effect of sex or cage, on the time to develop stereotypy. No differences were found in the density of either striatal D1 or D2 dopamine receptors or the concentration of striatal dopamine or its metabolites as a function of rearing condition or as a function of whether the animals developed stereotypy. These results characterize the development of stereotypies in this species, demonstrate the importance of environmental conditions in the genesis of stereotypy, and suggest that alterations in striatal dopamine content or dopamine receptor density do not account for the expression of stereotyped behaviors in this model.

Dietary caffeine, fluid intake and urinary incontinence in older rural women.

Abstract: Forty-one women completed the first phase (self-monitoring) of the Behavioral Management for Continence (BMC) intervention, while working with a nurse during home visits to reduce involuntary urine loss as part of the parent study involving older, rural women living at home. A decrease in dietary caffeine intake and an increase in fluid intake were most frequently recommended. The relationship between a decrease in the amount of dietary caffeine consumed and fewer daytime episodes of involuntary urine loss approached significance –P = 0.0744 – whereas an increase in the average amount of fluid intake was significantly related to an increase in the average volume of urine voided –P = 0.0479 – and not to involuntary urine loss.

Adverse Events Among the Elderly Receiving Chemotherapy for Advanced Non–Small-Cell Lung Cancer

PURPOSE To describe chemotherapy use and adverse events (AEs) for advanced-stage, non-small-cell lung cancer (NSCLC) in community practice, including descriptions according to variation by age. METHODS We interviewed patients with newly diagnosed, stages IIIB and IV NSCLC in the population-based cohort studied by the Cancer Care Outcomes Research and Surveillance Consortium, and we abstracted the patient medical records. AEs were medical events occurring during chemotherapy. Using logistic regression, we assessed the association between age and chemotherapy; with Poisson regression, we estimated event rate ratios and adjusted the analysis for age, sex, ethnicity, radiation therapy, stage, histology, and presence and grade of 27 comorbidities. RESULTS Of 1,371 patients, 58% (95% CI, 55% to 61%) received chemotherapy and 35% (95% CI, 32% to 38%) had AEs. After adjustment, 72% (95% CI, 65% to 79%) of those younger than 55 years and 47% (95% CI, 42% to 52%) of those age 75 years and older received chemotherapy. Platinum-based therapies were less common in the older-age groups. Pretreatment medical event rates were 18.6% for patients younger than 55 years and were only 9.2% for those age 75 years and older (adjusted rate ratio, 0.49; 95% CI, 0.26 to 0.91). In contrast, older adults were more likely to have AEs during chemotherapy. The adjusted rate ratios compared with age younger than 55 years were 1.70 for 65- to 74-year-olds (95% CI, 1.19 to 2.43) and 1.34 for those age 75 years and older (95% CI, 0.90 to 2.00). CONCLUSION Older patients who received chemotherapy had fewer pretherapy events than younger patients and were less likely to receive platinum-based regimens. Nevertheless, older patients had more adverse events during chemotherapy, independent of comorbidity. Potential implicit trade-offs between symptom management and treatment toxicity should be made explicit and additionally studied.

Representativeness of participants in the cancer care outcomes research and surveillance consortium relative to the surveillance, epidemiology, and end results program.

BackgroundThe research goals of the Cancer Care Outcomes Research and Surveillance (CanCORS) Consortium are to determine how characteristics and beliefs of patients, providers, and health care organizations influence the treatments and outcomes of individuals with newly diagnosed lung and colorectal cancers. As CanCORS results will inform national policy, it is important to know how they generalize to the United States population with these cancers. Research DesignThis study assessed the representativeness of the CanCORS cohort of 10,547 patients with lung cancer (LC) or colorectal cancer (CRC) enrolled between 2003 and 2005. We compared characteristics (sex, race, age, and disease stage) with the Surveillance, Epidemiology, and End Results (SEER) population of 234,464 patients with new onset of these cancers during the CanCORS recruitment period. ResultsThe CanCORS sample is well matched to the SEER Program for both cancers. In CanCORS, 41% LC/47% CRC were female versus 47% LC/49% CRC in SEER. African American, Hispanic, and Asian cases differed by no more than 5 percentage points between CanCORS and SEER. The SEER population is slightly older, with the percentage of patients older than 75 years 33.1% LC/37.3% CRC in SEER versus 26.9% LC/29.4% in CanCORS, and also has a slightly higher proportion of early stage patients. We also found that the CanCORS cohort was representative within specific SEER regions that map closely to CanCORS sites. ConclusionsThis study demonstrates that the CanCORS Consortium was successful in enrolling a demographically representative sample within the CanCORS regions.

Effects of an Education Program for Community Pharmacists on Detecting Drug-related Problems in Elderly Patients

Community pharmacists are in a position to assume increased responsibility for preventing and resolving drug-related problems in ambulatory patients. Such an expanded role is mandated under provisions of the Omnibus Budget Reconciliation Act of 1990. The need for pharmacist oversight of drug therapy may be most acute in elderly patients. This study reports on a program to teach community pharmacists a process of assessing drug therapy of elderly patients and intervening to correct problems. Community pharmacists (N = 102) were assigned to treatment and control conditions. Both groups targeted patients meeting criteria and enrolled them into the study. Treatment group pharmacists, who participated in a training program, also assessed the medication use of enrolled patients to identify and resolve medication-related problems. Patients (N = 762) were telephoned by researchers 1 month after enrollment for an interview. Comparisons between treatment and control group patients were made on reports of pharmacist activities, knowledge of regimens, compliance, and potential drug therapy problems, such as interactions and side effects. Treatment patients were more likely to report that pharmacists provided information and assessed for problems than were control patients. These differences were maintained on 3-month follow-up questionnaires. No differences were found on the odds that patients indicated misunderstanding of regimens, noncompliance, or potential therapeutic problems.

Physician Process and Patient Outcome Measures for Diabetes Care: Relationships to Organizational Characteristics

Background:Optimal diabetes management relies on providers adhering to evidence-based practice guidelines in the processes of care delivery and patients adhering to self-management recommendations to maximize patient outcomes. Purpose:To explore: (1) the degree to which providers adhere to the guidelines; (2) the extent of glycemic, lipid, and blood pressure control in patients with diabetes; and (3) the roles of organizational and patient population characteristics in affecting both provider adherence and patient outcome measures for diabetes. Design:Secondary data analysis of provider adherence and patient outcome measures from chart reviews, along with surveys of facility quality managers. Sample:We sampled 109 Veterans Affairs medical centers (VAMCs). Results:Analyses indicated that provider adherence to diabetes guidelines (ie, hemoglobin A1c, foot, eye, renal, and lipid screens) and patient outcome measures (ie, glycemic, lipid, and hypertension control plus nonsmoking status) are comparable or better in VAMCs than reported elsewhere. VAMCs with higher levels of provider adherence to diabetes guidelines had distinguishing organizational characteristics, including more frequent feedback on diabetes quality of care, designation of a guideline champion, timely implementation of quality-of-care changes, and greater acceptance of guideline applicability. VAMCs with better patient outcome measures for diabetes had more effective communication between physicians and nurses, used educational programs and Grand Rounds presentations to implement the diabetes guidelines, and had an overall patient population that was older and with a smaller percentage of black patients. Conclusions:Healthcare organizations can adopt many of the identified organizational characteristics to enhance the delivery of care in their settings.

Thiazide diuretics and bone mineral content in postmenopausal women.

This retrospective study of 54 postmenopausal women taking thiazide diuretics found that bone mineral measurements and bone fracture prevalence did not differ significantly from those of matched control subjects. Matching on the variables of type of menopause (surgical/nonsurgical), years postmenopausal, duration of estrogen therapy (if any), daily intake of dietary and supplemental calcium and vitamin D, and Quetelet index was done without knowledge of the bone mineral measurements. Bone mass was recorded as the bone mineral content and bone density of the distal and midshaft radius. Only fractures associated with osteoporosis (hip, rib, vertebrae, and wrist) were recorded. This study suggests that thiazide diuretics do not provide protection against osteoporosis.

Delineation of nursing administration research priorities.

The purposes of this national study were to determine: (a) a definition of nursing administration research and (b) priority nursing administration research questions. Brainstorming was used initially with Council on Graduate Education for Administration in Nursing (CGEAN) members in 1984, to delineate concepts for definitions. Definitions and research questions were generated and rated in a three-round Delphi survey by 129 experts from nursing and health administration. According to the most widely supported definition, nursing administration research is concerned with establishing costs of nursing care, examining the relationships between nursing services and quality patient care, and viewing problems of nursing service delivery within the broader context of policy analysis and delivery of health services. The research question of most importance was: What are the cost-effective components of clinical nursing care that yield high patient satisfaction, decrease the number of complications, and shorten hospital stay for identified groups of patients? In a scenario exercise at the 1985 CGEAN meeting, additional priorities identified were vulnerable populations and equity.

Comparing mean ranks for repeated measures data

Rank tests are considered that compare t treatments in repeated measures designs. A statistic is given that contains as special cases several that have been proposed for this problem, including one that corresponds to the randomized block ANOVA statistic applied to the rank transformed data. Another statistic is proposed, having a null distribution holding under more general conditions, that is the rank transform of the Hotelling statistic for repeated measures. A statistic of this type is also given for data that are ordered categorical rather than fully rankedo Unlike the Friedman statistic, the statistics discussed in this article utilize a single ranking of the entire sample. Power calculations for an underlying normal distribution indicate that the rank transformed ANOVA test can be substantially more powerful than the Friedman test.