Comparison of 7-day and repeated 24-h recall of type 2 diabetes.

PURPOSE: Patient reporting of type 2 diabetes symptoms in a questionnaire with a 7-day recall period was expected to be different from symptom reports using a 7-day diary with repeated 24-h recall based on cognitive theory of memory processes and prior literature. This study compared these two types of recall in patients diagnosed with type 2 diabetes (T2D). METHODS: One hundred and forty adults with T2D completed a daily diary for 7 days containing 9 T2D-related symptom and impact items. On day 7, patients completed the same items with a 7-day recall period. We examined the concordance of 7-day recall with summary descriptors of the daily reports and compared the scores and the discriminant ability of 7-day recall and mean of daily reports. RESULTS: Seven-day recall was most concordant with the mean of daily reports. The average difference in scores was small (range 0.22-0.77 on 11-point scale) and less than 0.5 standard deviations. For some items, the difference was positively associated with the variation in daily reports. The discriminant ability was comparable. CONCLUSIONS: In this study population, a questionnaire with 7-day recall provided information consistent with a daily diary measure of the average week-long experience of T2D symptoms and impacts.

Effects of patient management guidelines on physician practice patterns: the Community Hospital Oncology Program experience.

The Community Hospital Oncology Program (CHOP), funded under contracts by the National Cancer Institute (NCI) from 1981 to 1984, was designed to be a model for delivery of the most up-to-date cancer care in the community setting. Site specific patient management guidelines (PMGs) were developed by physicians who saw a majority of cancer patients in each community and represent a consensus of the most current information on pretreatment evaluation and management for each cancer site. There was the potential for PMGs to have a strong effect on physician practice patterns. A patterns of care study (POC) was conducted in the 17 CHOPs to determine the influence of guidelines on practice patterns. The practice elements examined in this report are clinical staging for breast and small-cell lung cancer, medical oncology consultation for breast cancer patients with positive lymph nodes, and radiation therapy consultation for rectal and small-cell lung cancer. These elements were in all CHOP guidelines. Except for physicians most active in the CHOP, the data provide no evidence of diffusion of guideline principles to the majority of practicing physicians. Although all guidelines contained discussion of the importance of staging before definitive treatment for breast cancer, only 33% of the 1,922 charts examined had a stage recorded. In small-cell lung cancer, 67% of the 388 charts examined had stage recorded. The years in practice of the primary physician had an inverse relationship to practice patterns as defined by the guidelines. Physician specialty was also a determinant of practice patterns for small-cell lung cancer. However, participation in guideline development and their dissemination did not significantly influence the patterns of care elements examined in this study.

Early detection and control of cancer in clinical practice.

As part of the Community Cancer Care Evaluation, a random-sample survey of practicing physicians in 12 geographic areas was conducted in 1985 to provide information about physician practice patterns with reference to cancer detection, control, and treatment. All respondents were asked whether they routinely performed comprehensive physical examinations, breast palpations, mammography, rectal examinations, chest roentgenography, and stool guaiac examinations on normal healthy patients older than 50 years. Responses were examined in terms of American Cancer Society and National Cancer Institute (Bethesda, Md) recommendations. Conformity with recommendations was dependent on the geographic area, the specific procedure, and the specialty of the physician. Across all procedures, frequency of performance varied with years since graduation from medical school, with more recent graduates more likely to conform to recommended standards.

Diabetes in American Indians of Washington, Oregon, and Idaho.

The prevalence of diabetes among 29,000 American Indians living on or near 10 reservations in the Pacific Northwest was established from clinical data sources. The age- and sex-adjusted prevalence rate in 1987 was three times higher [95% confidence interval (CI) 2.8-3.1] than the comparable United States rate in 1980. The age-adjusted prevalence for women was 1.3 times that for men (95% CI 1.2-1.4). The prevalence varied by culture area. Reservation communities whose principal tribe came from the Great Basin culture area had 3.6 times more diabetes than found in the U.S. as a whole. Reservation communities whose principal tribe was from the Plateau culture had 3.0 times more diabetes than found in the U.S., and those from the Northwest Coast culture had 1.9 times more diabetes than found in the U.S. The reasons for these variations are unknown and warrant investigation.

Poverty and HIV seropositivity: the poor are more likely to be infected.

We analyzed demographic and behavioral risk factors for HIV seropositivity using data from 3601 clients of the main HIV counseling and testing clinic for high-risk people in Seattle, Washington, USA. Clients with lower income were found to be more likely to be HIV seropositive, before and after controlling for other demographic and risk factors with logistic regression. This result supports the hypothesis that the impoverished are at increased risk for HIV infection due to the physical and social circumstances in which their poverty places them. These may include poor access to risk-reduction information and less support for implementation of risk-reduction strategies.

Relationship between balance and abnormalities in cerebral magnetic resonance imaging in older adults.

BACKGROUND: Falling is a major cause of disability and morbidity among older adults. Because poor balance is a major reason for frequent falls, assessment of balance and its risk factors are important. In this study, we postulated that cerebral changes identified on magnetic resonance (MR) imaging are related to balance, and that older adults with balance problems would have significantly greater prevalence of such brain abnormalities than older adults without balance problems. DESIGN AND MEASUREMENTS: Several measures of balance were examined in more than 700 community-dwelling older men and women, blacks and whites. Balance measures included dynamic posturography, functional reach, Romberg and 1-foot stand tests, tandem stand, and 1-foot stand. Cerebral MR imaging assessments included ventricular size, sulcal widening, white matter disease, and ischemic infarctions. Cardiovascular disease and hypertension were determined and controlled for in the analyses. RESULTS: A summary of the balance measures was significantly related to each of the 4 MR imaging measures, with those with poorer balance having more disease. The strongest associations with balance were seen for white matter disease and ventricular size. All but the ischemic infarction variable remained significantly associated with balance after adjustments for sex, race, age, cardiovascular disease, and hypertension. CONCLUSION: Cerebral changes identified by MR imaging are associated with poorer balance among older adults.

Cluster analysis and patterns of findings on cranial magnetic resonance imaging of the elderly: the Cardiovascular Health Study.

OBJECTIVE: To characterize patterns of findings on cranial magnetic resonance imaging (MRI) of the elderly using a statistical technique called cluster analysis. SUBJECTS AND METHODS: The Cardiovascular Health Study is a population-based, longitudinal study of 5888 people 65 years and older. Of these, 3230 underwent cranial MRI scans, which were coded for presence of infarcts and grades for white matter, ventricles, and sulci. Cluster analysis separated participants into 5 clusters based solely on patterns of MRI findings. Participants comprising each cluster were contrasted with respect to cardiovascular risk factors and clinical manifestations. RESULTS: One cluster was low on all the MRI findings (normal) and another was high on all of them (complex infarcts). Another cluster had evidence for infarcts alone (simple infarcts), whereas the last 2 clusters lacked infarcts, one having enlarged ventricles and sulci (atrophy) and the other having prominent white matter changes and enlarged ventricles (leukoaraiosis). Factors that distinguished these clusters in a discriminant analysis were age, sex, several measures of hypertension, internal carotid artery wall thickness, smoking, and prevalent claudication before the MRI. The atrophy group had the highest percentage of men and the normal group had the lowest. Cognitive and motor performance also differed across clusters, with the atrophy cluster performing better than may have been expected. CONCLUSIONS: These MRI patterns identified participants with different vascular disease risk factors and clinical manifestations. Results of these exploratory analyses warrant consideration in other populations of elderly people. Such patterns may provide clues about the pathophysiology of structural brain changes in the elderly.

Neurologic outcome and blood glucose levels during out-of-hospital cardiopulmonary resuscitation.

We examined the interrelations of outcome, time elapsed during cardiopulmonary resuscitation (CPR), and blood glucose levels drawn from 83 patients with out-of-hospital cardiac arrest. Levels rose significantly during CPR. Although slope and intercept of regression lines differed for those dying in the field and those admitted, regression lines were similar for those who awoke and never awoke after admission. These results suggest that the previously reported association between poor neurologic recovery and high blood glucose level on admission after cardiac arrest is best explained by prolonged CPR, leading to both higher rise of blood glucose and worse neurologic outcome.

Optimal survey design for community intervention evaluations: cohort or cross-sectional?

Community intervention evaluations that measure changes over time may conduct repeated cross-sectional surveys, follow a cohort of residents over time, or (often) use both designs. Each survey design has implications for precision and cost. To explore these issues, we assume that two waves of surveys are conducted, and that the goal is to estimate change in behavior for people who reside in the community at both times. Cohort designs are shown to provide more accurate estimates (in the sense of lower mean squared error) than cross-sectional estimates if (1) there is strong correlation over time in an individual's behavior at time 0 and time 1, (2) relatively few subjects are lost to followup, (3) the bias is relatively small, and (4) the available sample size is not too large. Otherwise, a repeated cross-sectional design is more efficient. We developed methods for choosing between the two designs, and applied them to actual survey data. Owing to drop-outs and losses to followup, the cohort estimates were usually more biased than the cross-sectional estimates. The correlations over time for most of the variables studied were also high. In many instances the cohort estimate, although biased, is preferred to the relatively unbiased cross-sectional estimate because the mean squared error was smaller for the cohort than for the cross-sectional estimate. If these results are replicated in other data, they may result in guidelines for choosing a more efficient study design.

Predicting future years of healthy life for older adults.

Cost-effectiveness studies often need to compare the cost of a program to the lifetime benefits of the program, but estimates of lifetime benefits are not routinely available, especially for older adults. We used data from two large longitudinal studies of older adults (ages 65-100) to estimate transition probabilities from one health state to another, and used those probabilities to estimate the mean additional years of healthy life that an older adult of specified age, sex, and health status would experience. We found, for example, that 65-year-old women in excellent health can expect 16.8 years of healthy life in the future, compared to only 8.5 years for women in poor health. We also provide estimates of discounted years of healthy life and future life expectancy. These estimates may be used to extend the effective length of the study period in cost-effectiveness studies, to examine the impact of chronic diseases or risk factors on years of healthy life, or to investigate the relationship of years of life to years of healthy life. Several applications are described.

Do communities differ in health behaviors?

Communities differ in the prevalence of various health behaviors, but it is not known to what extent these differences are due to "different types" of people living in them. We used data from the evaluation of the Henry J Kaiser Family Foundation Community Health Promotion Grant Program to study individual-level and community-level variation in health behaviors for 15 communities. Our results show (1) there was significant variation among these communities in prevalences of smoking, consumption of alcohol and dietary fat, and use of seatbelts; (2) these differences persisted after control for demographic, health status, and other health behavioral characteristics of the people in the communities; (3) the community effect on a particular person's behavior, as represented by R2, was very small (less than 1%); and (4) the adjusted differences in prevalences among communities were potentially large (for example, a 7 percentage point difference in the probability of smoking). Unique features of communities may influence health behaviors. These findings affirm the potential importance of contextual effects on individual health behavior and thus support the theory that changing the community environment may offer effective ways to change individual health behavior.

Data analysis and sample size issues in evaluations of community-based health promotion and disease prevention programs: a mixed-model analysis of variance approach.

The growing interest in community-based approaches to health promotion and disease prevention (HP/DP) has been accompanied by a growing need to evaluate the effectiveness of such programs. Special issues that arise in these evaluation studies include (1) entire communities are assigned to intervention and control groups, (2) only a small number of communities can usually be studied, (3) the time course of changes in behavior and other outcomes is often of interest, and (4) surveys to measure such changes over time can be conducted with either repeated cross-sectional samples or with longitudinal samples. This paper shows how these issues can be addressed under a mixed-model analysis of variance approach. This approach serves to unify several ideas in the literature on evaluation of community studies, including use of time-series regression and the question of whether the individual or the community should be the unit of analysis. We also describe how the method can be used to estimate sample size requirements, statistical power, or minimum detectable program effect.

The evaluation of the Henry J. Kaiser Family Foundation's Community Health Promotion Grant Program: design.

The Kaiser Family Foundation's Community Health Promotion Grant Program (CHPGP) provides funding and technical assistance in support of community-based efforts to prevent major health problems. The first phase of the program was implemented in 11 communities in the western United States. This paper describes the evaluation design of the CHPGP in the West, the methods of data collection, and the baseline comparability of intervention and control communities. Major features of the evaluation design include: (1) the randomization of qualified communities making application into funded and unfunded comparison groups; (2) a second set of matched control communities for some intervention sites; (3) data gathering through repeated surveys of community residents (probability samples of adults and adolescents) and institutions (health-related organizations and randomly sampled grocery stores and restaurants); and (4) the use of secondary data to monitor health events. Selected baseline data show that intervention and control communities differ in racial/ethnic composition, but relevant health behaviors and ratings of community activation for health promotion appear comparable.

Interpreting excess mortality in a prevention trial for older adults.

BACKGROUND: In a randomized trial evaluating preventive services for older adults excess mortality was observed in the treatment group. We examined four explanations: unbalance of baseline characteristics, unintended effects of the intervention, consequence of an autonomy intervention (including increased number of living wills in the treatment group), and chance. We focus here on the effects of the autonomy intervention. METHODS: Preparation of living wills in the treatment and control groups was compared both at baseline and follow-up. A linear predictor of mortality was used to identify participants at high risk of dying. Charts of these 200 participants were reviewed for evidence of serious medical events and resuscitation decisions. Rates of life-sustaining treatment were compared between treatment and controls using logistic regression. RESULTS: More living wills (65%) were noted for the treatment group than control group (47%) at follow-up. Thirty-six per cent of participants were identified as having a serious medical event; of these, participants in the treatment group were over twice as likely not to receive life-sustaining treatment. CONCLUSIONS: Advance directives contributed to excess deaths, indicating the success of the autonomy intervention.

Patterns of self-rated health in older adults before and after sentinel health events.

OBJECTIVES: To describe and compare patterns of change in self-rated health for older adults before death and before and after stroke, myocardial infarction, congestive heart failure, cardiac procedure, hospital admission for cancer, and hip fracture. DESIGN: "Event cohort," measuring time in months before and after the event. SETTING: Four U.S. communities. PARTICIPANTS: 5888 participants in the Cardiovascular Health Study (CHS), sampled from Medicare rolls and followed up to 8 years. Mean age at baseline was 73. MEASUREMENTS: Self-rated health, including a category for death, assessed at 6-month intervals, and ascertainment of events. METHODS: We examined the percentage that was healthy each month in the 5 years before death and in the 2 years before and after the other events, and compared the patterns to a "no event" group and to one another, using graphs and linear regression. RESULTS: For people who died, health status declined slowly until about 9 months before death, when it dropped steeply. Comparing persons equally far from death, health was unrelated to age, but men and whites were healthier than women and blacks. Health for other events declined before the event, dropped steeply at the event, showed some recovery, and then declined further after the event. About 65% to 80% of the subjects were healthy 2 years before their event, but only 35% to 65% were healthy two years afterwards. Patterns were similar although less extreme for the "no event" group. CONCLUSION: Visualizing trajectories of health helps us understand how serious health events changes health. Conclusions about change must be drawn with care because of a variety of possible biases. We have described the trajectories in detail. Work is now needed to explain, predict, and possibly prevent such changes in health.

Mortality, health services use, and health behavior in a cohort of well older adults.

OBJECTIVES: To construct an evidence-based Good Health Behavior Score and examine the relationship between aggregate health behaviors, mortality, and health services utilization in the last year of life in a cohort of well older adults. DESIGN: A prospective cohort. SETTING: A large health maintenance organization. PARTICIPANTS: 1867 older enrollees who responded to a health promotion survey. MEASUREMENTS: A baseline self-administered questionnaire was used to ascertain health behaviors in 1987-1988, and vital status was determined 48 months later. A Good Health Behavior Score was calculated, and a Cox proportional hazards model was used to compare high, middle, and low score groups regarding risk of death. For those who died, differences in amount and type of health services utilization in relation to the summary score were compared for the year before death. RESULTS: During the 4 years of follow-up, the mortality rate for the mid-level score group was 50% less, and in the highest score group was 70% less, than in the lowest score group. Among decedents, no significant differences were found between high and low Good Health Behavior score groups for inpatient and outpatient utilization, pharmacy use, or total cost during the last year of life. CONCLUSION: An easily developed and simple health behavior score can predict short term mortality quite strongly. Medical care costs in the last year of life were similar in individuals with higher and lower health behavior scores.

The complications of trauma and their associated costs in a level I trauma center.

OBJECTIVES: To estimate the expected costs for acute trauma care, to quantify the costs associated with the development of complications in injury victims, and to determine the deficit incurred by patients in whom complications develop. DESIGN: A retrospective, cohort design. SETTING: A referral trauma center. PATIENTS: A total of 12,088 patients admitted to a single regional trauma center during a period of 5 years. INTERVENTIONS: This is an observational study, and no interventions specific to this study are included in the design. MAIN OUTCOME MEASURES: (1) The expected costs for injury victims based on readily available clinical data. (2) The costs associated with the most important complications of trauma. (3) The effect of complications on inadequate reimbursement for trauma care. RESULTS: The expected costs were estimated using a linear model incorporating demographic variables and measures of injury severity. The expected costs averaged $14,567, and the observed costs averaged $15,032. Six complications were important predictors of cost. These included adult respiratory distress syndrome, acute kidney failure, sepsis, pneumonia, decubitus ulceration, and wound infections. For 1201 individuals with these complications, the predicted costs averaged $23,266 and the observed costs averaged $47,457. The mean excess costs for a single complication ranged from $6669 to $18,052. Multiple complications led to greater increases in excess cost, averaging $110,007 for the 62 patients with 3 or more complications. Costs exceeded reimbursement to a much greater degree in those in whom any of the 6 complications developed. CONCLUSION: Expected hospital costs can be estimated using admission clinical data. Each of 6 complications was associated with enormous increases in costs, indicating their importance as a cause of avoidable expenditures in injury victims and identifying situations in which reimbursement may not be adequate.

Assessment of community-level influences on individuals' attitudes about cigarette smoking, alcohol use, and consumption of dietary fat.

A basic premise of community-based health promotion is that enduring changes in health behaviors are facilitated by changing community norms or the standards of acceptable behavior in the community. We examine whether community-level influences on individuals' normative attitudes can be related to cigarette smoking, alcohol consumption, and dietary fat intake. We conducted a random-digit dialing survey of 8,849 adults in 15 communities in the western United States as part of the evaluation of the Henry J. Kaiser Family Foundation Community Health Promotion Grant Program. We found independent associations among attitudes for tobacco, alcohol, and diet and both individual-level and community-level characteristics. A significant community effect on attitudes remained even when we controlled for individual demographics and health behaviors, as well as for the overall prevalence of the target health behavior in the community. The specific community characteristics that account for this effect were not clear in our analyses, suggesting that influential characteristics vary from community to community.

Testing the null hypothesis in small area analysis.

The goal of small area analysis is often to demonstrate that hospital admission rates or procedure rates vary greatly among regions, suggesting the occurrence of unnecessary admissions or procedures in some regions. Recent articles have shown that such variation may be largely due to chance, even if no underlying differences exist among the small areas; thus, it is important to test if the observed variation is larger than expected by chance. In this article we discuss how the appropriate method for testing the null hypothesis depends on the distribution of the number of admissions at the person level. If it is not possible for an individual to have more than one admission for a given procedure, the appropriate test is a simple chi-square test. If multiple admissions are possible, a modified chi-square test can be used to account for the excess variability due to multiple admissions. Failure to make the correct modification to the chi-square test in this latter case can result in spurious results. This underscores the importance of collecting data on multiple admissions in order to estimate the distribution of the number of admissions at the individual-patient level.

Factors explaining the use of health care services by the elderly.

The Anderson model of health services utilization, which relates use of service to predisposing, enabling, and need factors, has not often been applied to an elderly population. In this study, the factors of the Andersen model were used prospectively to predict utilization for a population sample of 1,317 elderly persons. Taken alone, the NEED construct was the most important single predictor of use of physician services, hospitalizations, ambulatory care, and home care. PREDISPOSING factors were better predictors of the use of dental services. Some of the variables studied were not related to utilization in the direction that would have been predicted from previous studies on general populations. Multivariate analyses demonstrated that the three constructs should be applied simultaneously when predicting use of services. These findings can be applied to the specific task of planning services for older people.