Multimorbidity and evidence generation.

Most people with a chronic disease actually have more than one, a condition known as multimorbidity. Despite this, the evidence base to prevent adverse disease outcomes has taken a disease-specific approach. Drawing on a conference, Improving Guidelines for Multimorbid Patients, the goal of this paper is to identify challenges to the generation of evidence to support the care of people with multimorbidity and to make recommendations for improvement. We identified three broad categories of challenges: 1) challenges to defining and measuring multimorbidity; 2) challenges related to the effects of multimorbidity on study design, implementation and analysis; and 3) challenges inherent in studying heterogeneity of treatment effects in patients with differing comorbid conditions. We propose a set of recommendations for consideration by investigators and others (reviewers, editors, funding agencies, policymaking organizations) involved in the creation of evidence for this common type of person that address each of these challenges. The recommendations reflect a general approach that emphasizes broader inclusion (recruitment and retention) of patients with multimorbidity, coupled with more rigorous efforts to measure comorbidity and comorbidity burden and the influence of multimorbidity on outcomes and the effects of therapy. More rigorous examination of heterogeneity of treatment effects requires careful attention to prioritizing the most important comorbid-related questions, and also requires studies that provide greater statistical power than conventional trials have provided. Relatively modest changes in the orientation of current research along these lines can be helpful in pointing to and partially addressing selected knowledge gaps. However, producing a robust evidence base to support patient-centered decision making in complex individuals with multimorbidity, exposed to many different combinations of potentially interacting factors that can modify the risks and benefits of therapies, is likely to require a clinical research enterprise fundamentally restructured to be more fully integrated with routine clinical practice.

Assistive devices in context: cross-sectional association between challenges in the home environment and use of assistive devices for mobility.

PURPOSE: A primary goal for aging policy is to optimize independence in later life. We estimate the cross-sectional association between physical and social challenge in the home environment and use of assistive devices (AD) for mobility in the home, controlling for lower extremity physical performance (short physical performance battery [SPPB]) and other factors. DESIGN AND METHODS: Data are from the Women's Health and Aging Study I, a prospective study of the factors related to physical disability in a sample of moderately to severely disabled older women. We describe these associations in the baseline sample overall and also within subsets who do and do not have both a baseline and a 3-year follow-up observation. RESULTS: On average, physical challenge in the home environment is inversely associated with level of AD use (p < .05) in the overall sample, independent of SPPB, living alone, and other factors. We do not find a significant (p < .05) association between social challenge and the level of AD use in the overall sample. Findings by follow-up responder status were similar (with minor variability). IMPLICATIONS: Future cohorts who are better educated and more receptive to technology may confront challenges in the home environment that limit their ability to age in place. Our findings suggest that the physical challenges of the home are significantly related to AD use. Future analyses that explore the mechanisms of the home environment as a source of challenges to independent functioning could help in the design of future interventions for these cohorts as they age.

Benefits and harms of roflumilast in moderate to severe COPD.

BACKGROUND: Roflumilast, a phosphodiesterase 4 inhibitor, has been approved for the prevention of chronic obstructive pulmonary disease (COPD) exacerbations. It is unclear which patients will have a favourable benefit-harm balance with roflumilast. Our aim was to quantitatively assess the benefits and harms of roflumilast (500 µg/day) compared with placebo. METHODS: We used summary data released by the US Food and Drug Administration to estimate the treatment effects of roflumilast. Data from trials and observational studies were used to estimate the baseline risks for COPD exacerbations and gastrointestinal, neurological and psychiatric harms associated with roflumilast. Using simulation, we calculated the probability that roflumilast provides net benefit. We examined the impacts of different baseline risks for exacerbations and the severity of exacerbations, and varied weights (ie, relative importance) for outcomes and treated death as a competing risk in the analyses. RESULTS: The probability that roflumilast provides net benefit approximates 0% across different age categories of men and women with varying baseline risks for exacerbations. Using different weights for outcomes did not change the probability that roflumilast provides a net benefit. Only in the sensitivity analysis restricted to the prevention of severe exacerbations was there a probability of >50% that roflumilast provides a net benefit if the baseline risk of having at least one severe exacerbation per year exceeds 22%. CONCLUSIONS: Our results suggest that roflumilast only provides a net benefit to patients at a high risk of severe exacerbations. Guideline developers should consider different recommendations for patients with COPD at different baseline risks for exacerbations.

Assessing heterogeneity of treatment effect in a clinical trial with the proportional interactions model.

Understanding how individuals vary in their response to treatment is an important task of clinical research. For standard regression models, a proportional interactions model first described by Follmann and Proschan (1999) offers a powerful approach for identifying effect modification in a randomized clinical trial when multiple variables influence treatment response. In this paper, we present a framework for using the proportional interactions model in the context of a parallel-arm clinical trial with multiple prespecified candidate effect modifiers. To protect against model misspecification, we propose a selection strategy that considers all possible proportional interactions models. We develop a modified Bonferroni correction for multiple testing that accounts for the positive correlation among candidate models. We describe methods for constructing a confidence interval for the proportionality parameter. In simulation studies, we show that our modified Bonferroni adjustment controls familywise error and has greater power to detect proportional interactions compared with multiplcity-corrected subgroup analyses. We demonstrate our methodology by using the Studies of Left Ventricular Dysfunction Treatment trial, a placebo-controlled randomized clinical trial of the efficacy of enalapril to reduce the risk of death or hospitalization in chronic heart failure patients. An R package called anoint is available for implementing the proportional interactions methodology.

A framework for the analysis of heterogeneity of treatment effect in patient-centered outcomes research.

Individuals vary in their response to a treatment. Understanding this heterogeneity of treatment effect is critical for evaluating how well a treatment can be expected to work for an individual or a subgroup of individuals. An overemphasis on hypothesis testing has resulted in a dichotomy of all heterogeneity of treatment effect analyses into confirmatory (hypothesis testing) and exploratory (hypothesis finding) analyses. This limited view of heterogeneity of treatment effect is inadequate for creating evidence that is useful for informing patient-centered decisions. An expanded framework for heterogeneity of treatment effect assessment is proposed. It recognizes four distinct goals of heterogeneity of treatment effect analyses: hypothesis testing, hypothesis finding, reporting subgroup effects for meta-analysis, and individual-level prediction. Accordingly, two new types of heterogeneity of treatment effect analyses are proposed: descriptive and predictive. Descriptive heterogeneity of treatment effect analyses report treatment effects for prespecified subgroups in accordance with prospectively specified analytic strategy. They need not be powered to detect heterogeneity of treatment effect. They emphasize estimation and reporting of subgroup effects rather than hypothesis testing. Sampling properties (e.g., standard error) of descriptive analysis can be characterized, thus facilitating meta-analysis of subgroup effects. Predictive heterogeneity of treatment effect analyses estimate probabilities of beneficial and adverse responses of individuals to treatments and facilitates optimal treatment decisions for different types of individuals. Procedures are also suggested to improve reliability of heterogeneity of treatment effect assessment from observational studies. Heterogeneity of treatment effect analysis should be identified as confirmatory, descriptive, exploratory, or predictive analysis. Evidence should be interpreted in a manner consistent with the analytic goal.

Prevalence of diabetes treatment effect modifiers: the external validity of trials to older adults.

BACKGROUND AND AIMS: Potential treatment effect modifiers (TEMs) are specific diseases or conditions with a well-described mechanism for treatment effect modification. The prevalence of TEMs in older adults with type 2 diabetes mellitus (DM) is unknown. Objectives were to 1) determine the prevalence of pre-specified potential TEMs; 2) demonstrate the potential impact of TEMs in the older adult population using a simulated trial; 3) identify TEM combinations associated with number of hospitalizations to test construct validity. METHODS: Data are from the nationally-representative United States National Health and Examination Survey, 1999-2004: 8646 Civilian, non-institutionalized adults aged 45-64 or 65+ years, including 1443 with DM. TEMs were anemia, congestive heart failure, liver inflammation, polypharmacy, renal insufficiency, cognitive impairment, dizziness, frequent mental distress, mobility difficulty, and visual impairment. A trial was simulated to examine prevalence of potential TEM impact. The cross-sectional association between TEM patterns and number of hospitalizations was estimated to assess construct validity. RESULTS: The prevalence of TEMs was substantial such that 19.0% (95% CI 14.8-23.2) of middle-aged adults and 38.0% (95% CI 33.4-42.5) of older adults had any two. A simulated trial with modest levels of interaction suggested the prevalence of TEMs could nullify treatment benefit in 3.9-27.2% of older adults with DM. Compared to having DM alone, hospitalization rate was increased by several combinations of TEMs with substantial prevalence. CONCLUSIONS: We provide national benchmarks that can be used to evaluate TEM prevalence reported by clinical trials of DM, and correspondingly their external validity to older adults.

A framework for organizing and selecting quantitative approaches for benefit-harm assessment.

BACKGROUND: Several quantitative approaches for benefit-harm assessment of health care interventions exist but it is unclear how the approaches differ. Our aim was to review existing quantitative approaches for benefit-harm assessment and to develop an organizing framework that clarifies differences and aids selection of quantitative approaches for a particular benefit-harm assessment. METHODS: We performed a review of the literature to identify quantitative approaches for benefit-harm assessment. Our team, consisting of clinicians, epidemiologists, and statisticians, discussed the approaches and identified their key characteristics. We developed a framework that helps investigators select quantitative approaches for benefit-harm assessment that are appropriate for a particular decisionmaking context. RESULTS: Our framework for selecting quantitative approaches requires a concise definition of the treatment comparison and population of interest, identification of key benefit and harm outcomes, and determination of the need for a measure that puts all outcomes on a single scale (which we call a benefit and harm comparison metric). We identified 16 quantitative approaches for benefit-harm assessment. These approaches can be categorized into those that consider single or multiple key benefit and harm outcomes, and those that use a benefit-harm comparison metric or not. Most approaches use aggregate data and can be used in the context of single studies or systematic reviews. Although the majority of approaches provides a benefit and harm comparison metric, only four approaches provide measures of uncertainty around the benefit and harm comparison metric (such as a 95 percent confidence interval). None of the approaches considers the actual joint distribution of benefit and harm outcomes, but one approach considers competing risks when calculating profile-specific event rates. Nine approaches explicitly allow incorporating patient preferences. CONCLUSION: The choice of quantitative approaches depends on the specific question and goal of the benefit-harm assessment as well as on the nature and availability of data. In some situations, investigators may identify only one appropriate approach. In situations where the question and available data justify more than one approach, investigators may want to use multiple approaches and compare the consistency of results. When more evidence on relative advantages of approaches accumulates from such comparisons, it will be possible to make more specific recommendations on the choice of approaches.

Resting metabolic rate in old-old women with and without frailty: variability and estimation of energy requirements.

OBJECTIVES: To measure resting metabolic rate (RMR) in old-old adults living in the community and examine the association between measured RMR and frailty status and compare it with expected RMR generated by a predictive equation. DESIGN: Physiological substudy conducted as a home visit within an observational cohort study. SETTING: Baltimore City and County, Maryland. PARTICIPANTS: Seventy-seven women aged 83 to 93 enrolled in the Women's Health and Aging Study II. MEASUREMENTS: Resting metabolic rate with indirect calorimetry, frailty status, fat-free mass, ambient and body temperature, expected RMR according to the Mifflin-St. Jeor equation. RESULTS: Average RMR was 1,119 ± 205 kcal/d (range 595-1,560 kcal/d). Agreement between observed and expected RMR was biased and poor (between-subject coefficient of variation 38.0%, 95% confidence interval = 35.1-40.8). Variability of RMR was greater in frail individuals (heteroscedasticity F-test P = .02). Low and high RMR were associated with being frail (odds ratio 5.4, P = .04) and slower self-selected walking speed (P < .001) after adjustment for covariates. CONCLUSION: Equations to predict RMR that are not validated in old-old adults appear to correlate poorly with measured RMR. RMR is highly variable in old-old women, with deviations from the mean predicting clinical frailty. These exploratory findings suggest a pathway to clinical frailty through high or low RMR.

The prevalence of clinically-relevant comorbid conditions in patients with physician-diagnosed COPD: a cross-sectional study using data from NHANES 1999-2008.

BACKGROUND: Treatment of chronic diseases such as chronic obstructive pulmonary disease (COPD) is complicated by the presence of comorbidities. The objective of this analysis was to estimate the prevalence of comorbidity in COPD using nationally-representative data. METHODS: This study draws from a multi-year analytic sample of 14,828 subjects aged 45+, including 995 with COPD, from the National Health and Nutrition Examination Survey (NHANES), 1999-2008. COPD was defined by self-reported physician diagnosis of chronic bronchitis or emphysema; patients who reported a diagnosis of asthma were excluded. Using population weights, we estimated the age-and-gender-stratified prevalence of 22 comorbid conditions that may influence COPD and its treatment. RESULTS: Subjects 45+ with physician-diagnosed COPD were more likely than subjects without physician-diagnosed COPD to have coexisting arthritis (54.6% vs. 36.9%), depression (20.6% vs. 12.5%), osteoporosis (16.9% vs. 8.5%), cancer (16.5% vs. 9.9%), coronary heart disease (12.7% vs. 6.1%), congestive heart failure (12.1% vs. 3.9%), and stroke (8.9% vs. 4.6%). Subjects with COPD were also more likely to report mobility difficulty (55.6% vs. 32.5%), use of >4 prescription medications (51.8% vs. 32.1), dizziness/balance problems (41.1% vs. 23.8%), urinary incontinence (34.9% vs. 27.3%), memory problems (18.5% vs. 8.8%), low glomerular filtration rate (16.2% vs. 10.5%), and visual impairment (14.0% vs. 9.6%). All reported comparisons have p < 0.05. CONCLUSIONS: Our study indicates that COPD management may need to take into account a complex spectrum of comorbidities. This work identifies which conditions are most common in a nationally-representative set of COPD patients (physician-diagnosed), a necessary step for setting research priorities and developing clinical practice guidelines that address COPD within the context of comorbidity.

Assessing the applicability of trial evidence to a target sample in the presence of heterogeneity of treatment effect.

PURPOSE: To propose methods for the quantitative assessment of the applicability of evidence from a trial to a target sample using individual data. METHODS: Demonstration was with a trial of drug therapy to prevent mortality and an accompanying registry of people with heart failure. Principal components analysis with biplots did not identify measurement discrepancies. Multiple imputation with chained equations addressed missing predictor values. A proportional hazards model with interaction term, including graphical interpretation and a multivariate interaction test, identified heterogeneity of treatment effect. An interval of homogeneity of treatment effect was the interval of the baseline risk of outcome in which no two treatment effects were statistically significantly different. Absolute risk reduction for individuals was estimated for both benefit and harm outcomes and presented in a bivariate treatment effects scatterplot. RESULTS: Overall, the trial evidence applied to most of the registry according to overlapping distributions of estimated benefit and harm. However, 52% of trial and 33% of registry participants were estimated to have net benefit, and 14% of trial and 36% of registry participants were estimated to have strong net harmful treatment effect, that is, the individual estimate of harm was more than twice the estimate of benefit. CONCLUSIONS: The proposed methods provide quantitative assessment of the applicability of trial evidence to a target sample. They combine the strengths of different study designs, namely, unbiased effects estimation from trials and representation in observational studies, while addressing the practical challenges of combining information, namely, measurement discrepancies and missing data.

Association between inflammatory-related disease burden and frailty: results from the Women's Health and Aging Studies (WHAS) I and II.

Frailty is associated with a pro-inflammatory state, which has been characterized by elevated levels of systemic inflammatory biomarkers, but has not been related to the number of co-existing chronic diseases associated with inflammation. We sought to determine the extent to which a higher number of inflammatory-related diseases is associated with frailty and to identify the most common disease patterns associated with being frail in older adults. We performed binomial regression analyses to assess whether a higher count of inflammatory-related diseases increases the probability of frailty using data from the WHAS I and II, companion cohorts composed of 70-79-year-old community-dwelling older women in Baltimore, Maryland (n=620). An increase of one inflammatory-related disease was associated log-linearly with frailty (Prevalence Ratio (PR)=2.28, 95% Confidence Interval (CI)=1.81-2.87). After adjusting for age, race, education, and smoking status, the probability of frailty remained significant (PR=1.97, 95%CI=1.52-2.55). In the frail population, chronic kidney disease (CKD) and depressive symptoms (Prevalence=22.9%, 95%CI=14.2-34.8%); CVD and depressive symptoms (21.7%, 95%CI=13.2-33.5%); CKD and anemia (18.7%, 95%CI=11.1-29.7%); cardiovascular disease (CVD), CKD, and pulmonary disease (10.7%, 95%CI=5.2-21.0%); CKD, anemia, and depressive symptoms (8.7%, 95%CI=3.9-18.2%); and CVD, anemia, pulmonary disease, and depressive symptoms (5.0%, 95%CI=1.6-14.4%) were among the most frequent disease combinations. Their prevalence percentages were significantly higher in the frail versus non-frail women. A higher inflammatory-related disease count, perhaps reflecting a greater pro-inflammatory burden, increases the likelihood of frailty. Shared mechanisms among specific disease combinations may further contribute to this risk.

Incident preclinical mobility disability (PCMD) increases future risk of new difficulty walking and reduction in walking activity.

PURPOSE: This study examines whether and how report of a change in walking behavior, incident PCMD, predicts subsequent reduction in walking activity. MATERIALS AND METHODS: Data are from a prospective study of 436 community-dwelling women age 70-79 years. Outcome measures include subjective and objective measures of walking ability at 3 years. PRINCIPAL RESULTS: Incident PCMD is associated with the loss of walking abilities at 3-years, regardless of baseline physical impairment. Compared to women without, women with incident PCMD at 1.5 years after baseline were 2.7 (95%CI 1.4-7.2) times more likely to report that they no longer walk outdoors at least 8 blocks and 4.9 (1.9-13.1) times more likely to report new difficulty walking. Incident PCMD was also associated with declines in objective outcomes. Incident PCMD is an independent marker of subsequent decreased walking activity. MAJOR CONCLUSIONS: Incident PCMD appears to be a target for programs to prevent declines in walking activity in older adults.

Difficulty assisting with health care tasks among caregivers of multimorbid older adults.

BACKGROUND: Family caregivers provide assistance with health care tasks for many older adults with chronic illnesses. The difficulty they experience in providing this assistance, and related implications for their well-being, have not been well described. OBJECTIVE: The objectives of this study are: (1) to describe caregiver's health care task difficulty (HCTD), (2) determine the characteristics associated with HCTD, and (3) explore the association between HCTD and caregiver well-being. DESIGN: This is a cross-sectional study. PARTICIPANTS: Baseline sample of caregivers to older (aged 65+ years) multimorbid adults enrolled in an ongoing cluster-randomized controlled trial (N = 308). MAIN MEASURES: The HCTD scale (0-16) is comprised of questions measuring self-reported difficulty in assisting older adults with eight health care tasks, including taking medication, visiting health care providers, and managing medical bills. Caregivers were categorized using this scale into no, low, medium, and high HCTD groups. We used ordinal logistic regression and multivariate linear regression analyses to examine the relationships between HCTD, caregiver self-efficacy, caregiver strain (Caregiver Strain Index), and depression (Center for Epidemiological Studies Depression Scale), controlling for patient and caregiver socio-demographic and health factors. KEY RESULTS: Caregiver age and number of health care tasks performed were positively associated with increased HCTD. The quality of the caregiver's relationship with the patient, and self-efficacy were inversely associated with increased HCTD. A one-point increase in self-efficacy was associated with a significant lower odds of reporting high HCTD (OR, 0.64; 95% CI, 0.54, 0.77).Adjusted linear regression models indicated that high HCTD was independently associated with significantly greater caregiver strain (B, 2.7; 95% CI, 1.12, 4.29) and depression (B, 3.01; 95% CI, 1.06, 4.96). CONCLUSIONS: This study demonstrates that greater HCTD is associated with increased strain and depression among caregivers of multimorbid older adults. That caregiver self-efficacy was strongly associated with HCTD suggests health-system-based educational and empowering interventions might improve caregiver well-being.

Financial Strain Is Associated with Malnutrition Risk in Community-Dwelling Older Women.

This study examined the relationship between financial strain, or difficulty acquiring necessities, and malnutrition risk in a community dwelling sample of frail and nonfrail women aged 70-79 in the Women's Health and Aging Study (n = 679). Malnutrition risk was measured with a modified version of the Mini-Nutritional Assessment Short Form (MNA-SF) and defined as a score <11, financial strain was measured by (1) sufficiency of money on a monthly basis and (2) adequacy of income for food, and income was measured by ordinal categories. Mean (SD) modified MNA-SF score was 12.2 (1.80), and 14.7% of women had malnutrition risk. Women who usually did not have enough money to make ends meet had more than four-fold increased odds of malnutrition risk (OR = 4.54; 95% CI: 2.26, 9.14) compared to their counterparts who had some money left over each month. This was only slightly attenuated after control for income and education, (OR = 4.08; 95% CI: 1.95, 8.52) remaining robust. These results show an association between financial strain and malnutrition risk, independent of income, in older women. Self-reported financial strain may be preferable to income as a screener for malnutrition risk in older adults in clinical and research settings.

Frailty status and altered glucose-insulin dynamics.

BACKGROUND: We examined women in their 80s and 90s and evaluated the hypothesis that abnormalities in the dynamic response of glucose and insulin to a glucose load are associated with frailty status. METHODS: We performed a 75 g oral glucose tolerance test in 73 community-dwelling women aged 84-95 years without known diabetes enrolled in the Women's Health and Aging Study II. We examined the association of frailty status (nonfrail, prefrail, or frail) with oral glucose tolerance test glucose and insulin levels at 0, 30, 60, 120, and 180 minutes using multiple linear regression models. RESULTS: Using American Diabetes Association criteria, only 27% of older women had normal glucose status, 48% had prediabetes, and 25% had undiagnosed diabetes. Fasting glucose, fasting insulin, homeostasis model of assessment-insulin resistance, and Matsuda index were similar by frailty status, adjusting for age and body mass index. Conversely, mean oral glucose tolerance test glucose levels were higher at 60 minutes (44.6 ± 18.1 mg/dL higher) and 120 minutes (67.1 ± 23.5 mg/dL higher) and to a lesser extent at 180 minutes (44.3 ± 22.5 mg/dL higher) in frail versus nonfrail women as was integrated glucose area after adjustment. Mean 120-minute insulin level was also higher in frail versus nonfrail women (45.7 ± 22.4 µU/mL higher). Overall, glucose and insulin responses were more exaggerated and prolonged in frail versus nonfrail or prefrail women. CONCLUSIONS: Our data demonstrate dysregulation in response to glucose challenge as a component of physiologic vulnerability associated with frailty in old-old women. Future studies should examine the timing of abnormal glucose-insulin dynamics with respect to the pathogenesis of frailty.

Community aging in place, advancing better living for elders: a bio-behavioral-environmental intervention to improve function and health-related quality of life in disabled older adults.

OBJECTIVES: To determine effect size and acceptability of a multicomponent behavior and home repair intervention for low-income disabled older adults. DESIGN: Prospective randomized controlled pilot trial. SETTING: Participants' homes. PARTICIPANTS: Forty low-income older adults with difficulties in one or more activities of daily living (ADLs) or two or more instrumental activities of daily living (IADLs). INTERVENTION: The Community Aging in Place, Advancing Better Living for Elders (CAPABLE), coordinated occupational therapy, nursing, and handyman visits, was compared with attention-control visits. The intervention consisted of up to six visits with an occupational therapist, up to four visits with a nurse, and an average of $1,300 in handyman repairs and modifications. Each intervention participant received all components of the intervention clinically individualized to risk profile and goals. Each attention-control participant received the same number of visits as the intervention participants, involving sedentary activities of their choice. PRIMARY OUTCOME: difficulty in performing ADLs and IADLs. SECONDARY OUTCOMES: health-related quality of life and falls efficacy. RESULTS: Thirty-five of 40 adults (87%) completed the 6-month trial, and 93% and 100% of the control and intervention group, respectively, stated that the study benefited them. The intervention group improved on all outcomes. When comparing mean change in the intervention group with mean change in the control group from baseline to follow-up, the CAPABLE intervention had effect sizes of 0.63 for reducing difficulty in ADLs, 0.62 for reducing difficulty in IADLs, 0.89 for quality of life, and 0.55 for falls efficacy. CONCLUSION: The CAPABLE intervention was acceptable to participants and feasible to provide and showed promising results, suggesting that this multicomponent intervention to reduce disability should be evaluated in a larger trial.