Search

Objective: The suggested "two subjects per variable" (2SPV) rule of thumb in the Austin and Steyerberg article is a chance to bring out some long-established and quite intuitive sample size considerations for both simple and multiple linear regression., Study Design and Setting: This article distinguishes two of the major uses of regression models that imply very different sample size considerations, neither served well by the 2SPV rule. The first is etiological research, which contrasts mean Y levels at differing "exposure" (X) values and thus tends to focus on a single regression coefficient, possibly adjusted for confounders. The second research genre guides clinical practice. It addresses Y levels for individuals with different covariate patterns or "profiles." It focuses on the profile-specific (mean) Y levels themselves, estimating them via linear compounds of regression coefficients and covariates., Results and Conclusion: By drawing on long-established closed-form variance formulae that lie beneath the standard errors in multiple regression, and by rearranging them for heuristic purposes, one arrives at quite intuitive sample size considerations for both research genres., (Copyright © 2016 Elsevier Inc. All rights reserved.)

Background and Objective: Reference limits are widely used in anthropometry, the behavioral sciences, medicine, and clinical chemistry. They describe the distribution of a quantitative variable in a healthy population, and are often a smooth function of age or another determinant. Thus, instead of estimating reference limits separately for several age groups, it is more economical and parsimonious to use regression methods to estimate reference limits as a function of age. Although the variability of regression-based reference limits has been addressed previously, the available methods to determine the sample sizes needed to estimate them are neither transparent nor user-friendly., Methods: We propose a simple and intuitive formula using margins of error, to project the sample sizes required to achieve a given degree of precision, for different sampling strategies., Results: We present two examples for the calculation of the sample size required to estimate a specific reference limit using various age distributions., Conclusion: We provide a simple formula to calculate the sample size needed to estimate a specific reference limit to a specified degree of precision. The structure of the formula can easily accommodate different age-sampling strategies.

Background and Objective: Persons experiencing changes in their physical health may change their values and rerate the importance of basic elements of health-related quality of life (HRQL), a process known as response shift. Developing an estimator of HRQL that differentiates between objective change and response shift is essential for the interpretation of the results. The purpose of the present article was to contrast three methodologic approaches for evaluating response shift to develop a proposed set of HRQL measurement recommendations under circumstances where response shift is expected to occur., Methods: The three approaches compared were a structural equation modeling (SEM) technique, the then test, and an individualized approach. The data collection procedures for these methods were incorporated into a poststroke randomized controlled trial., Results: The SEM did not show a response shift, contrary to the results of the then test and the individualized approaches. We discuss factors that effect the selection of a methodologic approach including feasibility, subjects' memory and more advanced cognitive tasks, and whether response shift was evaluated at the group or individual level., Conclusion: The evaluation of response shift is an integral part of HRQL evaluations, and further comparisons between methodologic approaches are needed.

Objective: The interpretability of changes in perceived health status over time is threatened if people experience a response shift. This study assessed whether the recovery process following stroke altered individuals' perceptions of past health status and the impact that change in internal standards (response shift) had on ratings over time. We hypothesized that individuals with stroke would experience changes in internal standards, not experienced by the control group. Two other hypotheses related to objective criterion measures also were tested., Study Design and Setting: Individuals were recruited through a randomized trial of acute poststroke care. Health status was evaluated at baseline (within the first week poststroke), 6 and 24 weeks later using the EQ VAS. At 6 and 24 weeks, subjects were asked to retrospectively re-evaluate their health status for the preceding evaluation using the then test technique., Results: The pattern of mean scores was indicative of changes in internal standards among individuals with stroke but not for the control group. Memory had an impact on estimates of response shift. Hypotheses related to the objective criterion measures were not supported., Conclusion: The results suggest that there was a change in internal standards, and that measures of improvement in health status are different based on prospective as compared to retrospective ratings. Further understanding of the impact of recall on the assessment of response shift using the then test is needed to validate the use of this technique.

Background and Objective: We investigated the effect of nonsteroidal anti-inflammatory drug (NSAID) use on stage of colon cancer and presence of distant metastases at diagnosis., Methods: Cases of incident colon cancer diagnosed between 1981 and mid-1995 were selected from the Saskatchewan Cancer Registry. Cases with more extensive cancer at diagnosis were compared with cases with less extensive cancer (characterized using the Dukes system). NSAID exposure was assessed using the Saskatchewan Prescription Drug Plan database., Results: Compared with non-users in the same period, nonmetastatic cancer was found to be less extensive in those with low NSAID use (odds ratio [OR] 0.67, 95% confidence interval [CI] 0.53-0.87) and high NSAID use (OR 0.55, 95%CI 0.36-0.84) during the 6 to 10 year periods preceding diagnosis, regardless of the cut-off used to define "more extensive cancer." NSAID use in the 2 to 5 year periods was also found to protect (although not statistically significantly) against the occurrence of distant metastasis., Conclusion: Results are consistent with the hypothesized mechanism of the action of NSAIDs, but confounding by healthy lifestyle cannot be ruled out.

In the context of the choice of treatment for end-stage renal disease (ESRD), three approaches to value assessment were examined for their repeatability over time within subjects. If formal decision analyses are to be used to advise patients about treatment choice, then repeatable value assessment methods, an essential component of such analyses, are needed. The methods assessed were standard gamble (SG), time trade-off (TTO), and visual analogue (VA). Sixty-six nephrology clinic patients were interviewed on two occasions, 10 days apart, by one of two interviewers. An information session was conducted 1 week before the first interview. Subjects were taught about the treatments using an information package developed expressly for the study and a video produced by a pharmaceutical company for use in this decision context. Patients differed widely in the values provided for the various treatments of ESRD, with responses that ranged across the entire scale (0 to 100). The repeatability of the three methods was poor, with the coefficients of repeatability (95% range of differences from one occasion to the next) observed as +/- 27.4 for SG, +/- 38.4 for TTO, and +/- 36.5 for VA. When subsets defined by characteristics that may have improved the repeatability were analyzed, the magnitude of the error did not vary substantially. The poor repeatability of these methods raises questions about their use for decision analyses applied to the individual context.

It is well known that there is an excess of physical and psychological health problems among family caregivers of elderly persons with Alzheimer's disease and other dementias. The objective of this study was to determine whether the higher level of morbidity translates into a higher level of medical care utilization. Data from a previously completed longitudinal study of caregivers for elderly persons with dementia were merged with data on physician visits obtained from the computerized records of the Quebec Health Insurance Board. Utilization of physician care (adjusted for age, sex, number of chronic diseases, and depression) was no higher for family caregivers of elderly patients with Alzheimer's disease and other dementias than for comparable family members of older persons without dementia. The annual cost of physician care was almost identical among caregivers and noncaregivers. However, the pattern of utilization for the two groups was somewhat different: there was a significantly higher frequency of physician utilization among caregivers for services billed by psychiatrists and internal medicine specialists. In multivariate analysis, physician utilization was significantly associated with having more than one chronic condition and with increased age. Future studies should focus on determining whether caregivers neglect their own health care needs as a result of the exigencies of the caregiving role.

We examined the presence, magnitude, and consequences of systematic and random errors caused by terminal digit preference in the measurement of highest systolic blood pressure during prenatal visits in 28,841 non-referred pregnant women who delivered between 1 January 1982 and 31 March 1990. In the overall distribution of terminal digit readings, 78% were read to 0, 15% to even digits other than 0, 5% to 5, and only 2% to odd digits other than 5. This preference for 0's was consistent across the entire distribution of blood pressure and for a variety of maternal characteristics. The relative frequency of the cutoff value of 140 mmHg (i.e. the percentage of readings on 140 mmHg) within the range containing the value (i.e. 138-142 mmHg) was similar to the relative frequency of other multiples of 0. This was true whether the comparison was made in the overall study sample, or in a pre-selected low-risk subgroup or high-risk subgroup, indicating no systematic bias. On the other hand, a strong tendency to read blood pressure values to the nearest 0 had a marked effect on the classification of hypertension. Changing the definition of hypertension from > or = 140 mmHg to > 140 mmHg produced a reduction in prevalence of hypertension from 25.9 to 13.3% in the overall study sample, from 15.4 to 6.3% in the low-risk subgroup, and from 43.3 to 25.3% in the high-risk subgroup. Epidemiologic studies that compare prevalences of hypertension in different populations based on routine clinical measurement of blood pressure and a single cutoff point should assess the consequences of terminal digit preference in defining hypertension.(ABSTRACT TRUNCATED AT 250 WORDS)

Most elderly persons with dementia are cared for at home, usually by the spouse or an adult child. The objective of the present study was to determine whether there is an excess of psychological and physical health problems among family caregivers (CGs) of elderly persons with dementia. Data were obtained by interview from close family members of dementia patients (CGs), and from a comparison group made up of close family members of patients undergoing cataract surgery (non-caregivers, NCGs). CGs had significantly higher levels of depression and physical symptoms than NCGs. The association between caregiving and the health variables was stronger among subjects who were the patient's spouse than among those who were the patient's child. Furthermore, greater behavioral disturbance in the demented patient was associated with higher levels of morbidity in the CG. The results suggest that CGs might benefit from careful monitoring of their health status, and from greater access to specialized support services.