Identifying Patterns of Depression Comorbidities Using Association Rule Learning: Insights from Maryland Medicaid Data

Research on Multiple Chronic Conditions

This issue of the Annals, as well as the previous one, confronts the enormous public health challenges of multimorbidity. More than 1 in 4 Americans has multiple (2 or more) chronic conditions, including physical and behavioral health problems, accounting for an estimated two-thirds of total US health care spending.1 An individual’s risks for a variety of adverse health outcomes (eg, poor functional status, unnecessary hospitalizations, and adverse drug events) rise as the number of multiple chronic conditions increases.2 The Centers for Medicare and Medicaid Services (CMS) has just released even more detailed information with respect to its Medicare fee-for-service populations,3 exposing the exceptional complexity and sheer burden that multiple chronic conditions pose for patients, health facilities, payers, and clinicians. In its recently released chart book Chronic Conditions Among Medicare Beneficiaries,3 CMS describes detailed demographics and prevalence measures of multiple medical conditions in this population and the dramatic impact on service utilization and spending. Examples of key findings are that two-thirds (20.7 million beneficiaries) had at least 2 or more chronic conditions; about 50% of beneficiaries with stroke or heart failure had 5 or more additional chronic conditions; beneficiaries with 6 or more chronic conditions accounted for about one-half of Medicare spending on hospitalizations; more than one-quarter of beneficiaries with 6 or more chronic conditions had a hospital readmission within 30 days; and the 12% of beneficiaries with 6 or more chronic conditions accounted for 43% of Medicare spending. For health systems that have traditionally focused on research and treatment of single conditions, these tremendous challenges have forced many to escalate efforts to identify and implement solutions. How do we as a society bring a greater sense of order to this vexing challenge? As part of the response, the US Department of Health and Human Services (HHS), in conjunction with partner organizations and other stakeholders, used a deliberative process to develop Multiple Chronic Conditions—A Strategic Framework: Optimum Health and Quality of Life for Individuals with Multiple Chronic Conditions.2 The framework serves as a national-level road map to assist HHS programs and public and private stakeholders in ensuring a more coordinated and comprehensive approach to improving the health status of individuals with multiple chronic conditions.2,4 Released to stakeholders and the public in late 2010, the framework is organized into 4 major goal areas with subsets of objectives and action strategies for use by clinical practitioners, policy makers, researchers, and others. The framework’s goals encompass the interdependent domains of (1) strengthening the health care and public health systems; (2) empowering the individual to use self-care management; (3) equipping health care clinicians with tools, information, and other interventions; and (4) supporting targeted research about individuals with multiple chronic conditions and effective interventions. The 3 articles on multimorbidity included in this issue of the Annals,5–7 which focus on measurement, represent progress in better understanding the epidemiology of multiple chronic conditions, a key aspect of the HHS strategic framework’s fourth goal. The articles further illustrate that there are multiple operational definitions of multimorbidity, each with their respective strengths and weaknesses. Though each may be appropriate according to the outcome of interest, there also may be utility for some degree of standardization in characterizing this heterogeneous population. The definition for multiple chronic conditions used in the HHS strategic framework utilized an approach of simple counts of conditions (ie, 2 or more). The article by Huntley et al, consistent with this approach, finds that simple counts of diseases perform almost as well as complex measures in predicting most outcomes.5 Ultimately, however, from a policy perspective, what generally holds true irrespective of the selected measure is that as the magnitude of multimorbidity increases, patient outcomes decline and costs rise. Thus, targeting care management efforts on multimorbid populations should, in principle, accelerate the country’s progress toward the goals of delivery system reform. We invite readers to further their familiarity with the complex issues posed by multiple chronic conditions by reviewing the HHS strategic framework and related activities (http://www.hhs.gov/ash/initiatives/mcc).8 A current HHS collaborative initiative—led by CMS, the Centers for Disease Control and Prevention, and the Agency for Healthcare Research & Quality—is not only examining options for a conceptual framework to improve definition and measurement of priority chronic conditions, but also is conducting coordinated analyses of multiple data sets to improve descriptive epidemiological characterization of multiple chronic conditions. Increasing life expectancy and the aging of the population will only intensify the challenge for the future. The articles in this issue begin to address an enormous health system challenge that demands our urgent attention.

The most common chronic condition experienced by adults is multimorbidity, the coexistence of multiple chronic diseases or conditions. In patients with coronary disease, for example, it is the sole condition in only 17% of cases.1 Almost 3 in 4 individuals aged 65 years and older have multiple chronic conditions, as do 1 in 4 adults younger than 65 years who receive health care.2 Adults with multiple chronic conditions are the major users of health care services at all adult ages, and account for more than two-thirds of health care spending.2 Despite the predominance of multiple chronic conditions, however, reimbursement remains linked to discrete International Classification of Diseases diagnostic codes, none of which are for multimorbidity or multiple chronic conditions. Specialists are responsible for a single disease among the patient’s many. Quality measurement largely ignores the unintended consequences of applying the multiple interventions necessary to adhere to every applicable measure. Uncertain benefit and potential harm of numerous simultaneous treatments, worsening of a single disease by treatment of a coexisting one, and treatment burden arising from following several disease guidelines are the well-documented challenges of clinical decision making for patients with multiple chronic conditions.3,4 To ensure safe and effective care for adults with multiple chronic conditions, particularly the millions of baby boomers entering their years of declining health and increasing health service use, health care must shift its current focus on managing innumerable individual diseases. To align with the clinical reality of multimorbidity, care should evolve from a disease orientation to a patient goal orientation, focused on maximizing the health goals of individual patients with unique sets of risks, conditions, and priorities. Patient goal–oriented health care involves ascertaining a patient’s health outcome priorities and goals, identifying the diseases and other modifiable factors impeding these goals, calculating and communicating the likely effect of alternative treatments on these goals, and guiding shared decision making informed by this information.4

Multiple chronic conditions are imposing an increasing health and economic burden on the Chinese health system. While exposure to residential greenness has been shown to provide various health benefits, its relationship with multiple chronic conditions remains largely unexplored. This study aims to investigate this relationship using high-resolution satellite imagery and data from the 6th Health Services Survey (HSS) cohort in Shandong province. We linked health data from the HSS with 12-month average Normalised Difference Vegetation Index (NDVI) measurements based on reported residential geocodes. Multiple chronic condition status was defined as having two or more chronic conditions concurrently, according to the HSS's predefined disease classification. Generalised mixed regression models were utilised to assess both the likelihood and count of multiple chronic conditions in relation to greenspace exposure. Additionally, using the pre-defined disease classes, we also explored how greenspace influences multiple chronic conditions across various physiological systems and disease categories. A total of 28 489 individuals were included in this cross-sectional analysis. After adjusting for potential confounding factors, we found that exposure to greenspace was significantly associated with a reduced prevalence and count of chronic conditions. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) for were: Q2 (aOR = 0.74; 95% CI = 0.62, 0.88), Q3 (aOR = 0.69; 95% CI = 0.55, 0.86), and Q4 (aOR = 0.70; 95% CI = 0.56, 0.88), respectively, compared against the baseline Q1 quartile. Subgroup analyses revealed that higher residential greenspace exposure reduced risks of blood, endocrine, nutritional and metabolic chronic diseases. No clear associations were found for other chronic disease classes. Additionally, consistent results were observed across spatial and temporal sensitivity analyses. Our findings underscore the potential beneficial effects of residential greenness on multiple chronic conditions, with implications for urban planning, environmental policy, and community development.

The association between an individuals' demographic and health characteristics and the presence of multiple chronic conditions is not well known among older United States (US) adults. This study aimed to identify the prevalence and associations of having multiple chronic conditions among older US adults with self-reported pain. This retrospective, cross-sectional study used data from the 2017 Medical Expenditure Panel Survey. Study subjects were aged≥50 years and had self-reported pain in the past four weeks. The outcome variable was multiple (≥5) chronic conditions (vs. <5 chronic conditions). Hierarchical logistic regression models were used to identify significant associations between demographic and health characteristics and multiple chronic conditions with significance indicated at an a priori alpha level of 0.05. The complex survey design was accounted for when obtaining nationally-representative estimates. The weighted population was 57,074,842 US older adults with pain, of which, 66.1% had≥5 chronic conditions. In fully-adjusted analyses, significant associations of≥5 comorbid chronic conditions included: age 50-64 vs.≥65 years (adjusted odds ratio [AOR]=0.478, 95% confidence interval [CI]=0.391, 0.584); male vs. female gender (AOR=1.271, 95% CI=1.063, 1.519); white vs. other race (AOR=1.220, 95% CI=1.016, 1.465); Hispanic vs. non-Hispanic ethnicity (AOR=0.614, 95% CI=0.475, 0.793); employed vs. unemployed (AOR=0.591, 95% CI=0.476, 0.733); functional limitations vs. no functional limitations (AOR=1.862, 95% CI=1.510, 2.298); work limitations vs. no work limitations (AOR=1.588, 95% CI=1.275, 1.976); little/moderate vs. quite a bit/extreme pain (AOR=0.732, 95% CI=0.599, 0.893); and excellent/very good (AOR=0.375, 95% CI=0.294, 0.480) or good (AOR=0.661, 95% CI=0.540, 0.810) vs. fair/poor physical health. Approximately 38 million of the 57 million US older adults with pain in this study had≥5 chronic conditions in 2017. Several characteristics were associated with multiple chronic conditions, which may be important for health care professionals to consider when working with patients to manage their pain. This study was approved by The University of Arizona Institutional Review Board (2006721124, June 12, 2020).

About half of U.S. adults have at least one chronic health condition, and the prevalence of multiple (two or more) chronic conditions increased from 21.8% in 2001 to 25.5% in 2012. Chronic conditions profoundly affect quality of life, are leading causes of death and disability, and account for 86% of total health care spending. Arthritis is a common cause of disability, one of the most common chronic conditions, and is included in prevalent combinations of multiple chronic conditions. To determine the impact of having arthritis alone or as one of multiple chronic conditions on selected important life domains, CDC analyzed data from the 2013 National Health Interview Survey (NHIS). Having one or more chronic conditions was associated with significant and progressively higher prevalences of social participation restriction, serious psychological distress, and work limitations. Adults with arthritis as one of their multiple chronic conditions had higher prevalences of adverse outcomes on all three life domains compared with those with multiple chronic conditions but without arthritis. The high prevalence of arthritis, its common co-occurrence with other chronic conditions, and its significant adverse effect on life domains suggest the importance of considering arthritis in discussions addressing the effect of multiple chronic conditions and interventions needed to reduce that impact among researchers, health care providers, and policy makers.

Cross-sectional study. The purposes of this study were to assess (i) prevalence of self-reported multiple chronic conditions (MCC) in a population-based cohort of persons with traumatic spinal cord injury (TSCI) and (ii) the association between health-related behaviors and MCC. Population-based TSCI cohort. Participants included 716 adults with TSCI of at least 1-year duration who were identified through a population-based TSCI surveillance system. Standard questions from the Behavioral Risk Factor Surveillance System measured cigarette smoking, binge drinking, planned exercises, and 10 chronic health conditions (CHC), including diabetes, heart attack, angina (or coronary artery disease), stroke, cancer, asthma, kidney disease, arthritis, depressive disorder, chronic obstructive pulmonary disease. MCC was defined as having two or more CHCs in this study. Multivariate logistic regression models were used to assess the association between health-related behaviors and MCC. Almost half (45%) of the study sample had MCC. After controlling for demographic and injury characteristics, participants with smoking history of at least 100 cigarettes were 59% more likely to develop MCC, and those who had planned exercises at least three times a week were 36% less likely to have MCC. We found MCC prevalence was high among people with TSCI, and MCC was associated with cigarette smoking and planned exercise.

Health in 2020 and beyond: what do the numbers tell us?

Evolution of multiple chronic conditions (MCC) follows a complex stochastic process, influenced by several factors including the inter-relationship of existing conditions, and patient-level risk factors. Nearly 20% of citizens aged 18 years and older are burdened with two or more (multiple) chronic conditions (MCC). Treatment for people living with MCC currently accounts for an estimated 66% of the Nation's healthcare costs. However, it is still not known precisely how MCC emerge and accumulate among individuals or in the general population. This study investigates major patterns of MCC transitions in a diverse population of patients and identifies the risk factors affecting the transition process. A Latent regression Markov clustering (LRMCL) algorithm is proposed to identify major transitions of four MCC that include hypertension (HTN), depression, Post- Traumatic Stress Disorder (PTSD), and back pain. A cohort of 601,805 individuals randomly selected from the population of Iraq and Afghanistan war Veterans (IAVs) who received VA care during three or more years between 2002-2015, is used for training the proposed LRMCL algorithm. Two major clusters of MCC transition patterns with 78% and 22% probability of membership respectively were identified. The primary cluster demonstrated the possibility of improvement when the number of MCC is small and an increase in probability of MCC accumulation as the number of co- morbidities increased. The second cluster showed stability (no change) of MCC overtime as the major pattern. Age was the most significant risk factor associated with the most probable cluster for each IAV. These findings suggest that our proposed LRMCL algorithm can be used to describe and understand MCC transitions, which may ultimately allow healthcare systems to support optimal clinical decision- making. This method will be used to describe a broader range of MCC transitions in this and non-VA populations, and will add treatment information to see if models including treatments and MCC emergence can be used to support clinical decision-making in patient care.

Chronic health conditions pose a significant burden on the health of Americans, and the number of people with chronic conditions is rapidly rising. In 2000, 125 million Americans had ≥1 chronic conditions. That number rose to an estimated 145 million in 2009, and will further rise to a projected 171 million persons in 2030.1 In addition, today some 26% of adults and 68% of Medicare beneficiaries have ≥2 chronic conditions, and two thirds of total healthcare spending is devoted to their care.1–3 To address this growing challenge, the US Department of Health and Human Services (HHS) created the Multiple Chronic Conditions (MCC) Interagency Workgroup, a department-wide effort to address the needs of people with MCC and the health systems that serve them. The Strategic Framework on Multiple Chronic Conditions,4 released by HHS in 2010 as a result of this effort, is designed to address the challenge of MCC across the spectrum of all population groups through 4 overarching goals: Fostering healthcare and public health system changes to improve the health of individuals with multiple chronic conditions. Maximizing the use of proven self-care management and other services by individuals with multiple chronic conditions. Providing better tools and information to healthcare, public health, and social services workers who deliver care to individuals with multiple chronic conditions. Facilitating research to fill knowledge gaps about, and interventions and systems to benefit, individuals with multiple chronic conditions. The papers in this supplement, resulting from 45 MCC research grant awards made by the Agency for Healthcare Research and Quality (AHRQ) that make up the AHRQ MCC Research Network, offer important findings and commentary addressing these goals. First, although newer, more coordinated models of care now exist, such as the patient-centered medical home and accountable care organizations (goal 1), patients with MCC may be less likely than patients without MCC to benefit from these innovations, and more likely to experience fragmented care. Second, although much is required of many MCC patients in managing their care (goal 2), complex treatment regimens can often prove burdensome for patients to follow. The researchers in the MCC Research Network studied how health systems and healthcare professionals can better partner with patients living with MCC to create patient-centered management plans. The articles in the supplement also propose and describe useful tools for the care of patients with MCC (goal 3), including a new scale to measure patients’ ability to follow treatment regimens they are asked to follow and new treatment guidelines for the joint management of several comorbid conditions. Finally, many of the articles contribute to mapping the landscape for future MCC research (goal 4), through a new conceptual model, description of methodological challenges in MCC research and possible remedies, and intriguing descriptive data that poses new questions for future researchers. HHS remains committed to leading the way on improving the care of patients with MCC. Important opportunities in this area include rapid analyses of secondary data to further develop clinical guidelines for patients with MCC, inclusion of patients with MCC in clinical trials, and development of effective behavioral interventions and useful quality measures for front-line providers caring for patients with MCC. The National Institutes of Health recently issued 4 important new funding opportunity announcements in these areas.5–8 However, HHS cannot meet the needs of MCC patients alone. We welcome the participation of public and private sector stakeholders who wish to take a coleadership role in expanding knowledge about the care of this population and rapidly translating research findings into practice. We hope you enjoy these pages and look forward to working with public and private partners to better guide, empower, and support patients with MCC and the healthcare professionals who serve them. We applaud the 45 AHRQ grantees comprising the AHRQ MCC Research Network for their ground breaking accomplishments to date. ACKNOWLEDGMENTS The authors would like to thank Therese Miller, DrPH, the Project Officer at AHRQ, and the guest editors Mary Tinetti, MD and Jayasree Basu, PhD for their intellectual contributions and for making this supplement possible. They would also like to thank the Medical Care staff, including the Deputy Editors, Amy Rosen, PhD, and Robert Weech-Maldonado, MBA, PhD, the Publisher, Druanne Martin, the Managing Editor, Karen Doyle, and the Abt Associates team, Lisa LeRoy, MBA, PhD, Melanie Wasserman, PhD, Meghan Woo, PhD, and Emma Oppenheim, for their technical contributions and logistical support.

Although 1-in-5 adults have multiple (≥ 2) chronic conditions, limited attention has been given to the association between multiple chronic conditions and employment. Cross-sectional data (2011 National Health Interview Survey) and multivariate regression analyses were used to examine the association among multiple chronic conditions, employment, and labor force outcomes for U.S. adults aged 18-64 years, controlling for covariates. Among U.S. adults aged 18-64 years (unweighted, n = 25,458), having multiple chronic conditions reduced employment probability by 11-29%. Some individual chronic conditions decreased employment probability. Among employed adults (unweighted, n = 16,096), having multiple chronic conditions increased the average number of work days missed due to injury/illness in the past year by 3-9 days. Multiple chronic conditions are a barrier to employment and increase the number of work days missed, placing affected individuals at a financial disadvantage. Researchers interested in examining consequences of multiple chronic conditions should give consideration to labor force outcomes.

IntroductionUnderstanding longitudinal relationships among multiple chronic conditions, limitations in activities of daily living, and health-related quality of life is important for identifying potential opportunities for health promotion and disease prevention among older adults.MethodsThis study assessed longitudinal associations between multiple chronic conditions and limitations in activities of daily living on health-related quality of life among older adults (≥65 years) from 2004 through 2006, using data from the Medicare Health Outcomes Survey (N = 27,334).ResultsUsing a longitudinal path model, we found the numbers of chronic conditions at baseline and 2-year follow-up were independently associated with more limitations in activities of daily living at 2-year follow-up. In addition, more limitations in activities of daily living at 2-year follow-up were associated with worse health-related quality of life during the follow-up time period. The association between multiple chronic conditions and indices of health-related quality of life was mediated by changes in limitations in activities of daily living.ConclusionBoth baseline and new multiple chronic conditions led to worse health in terms of activities of daily living and health-related quality of life and should be considered important outcomes to intervene on for improved long-term health. In addition, public health practitioners should consider addressing classes of multiple chronic conditions by using interventions designed to reduce the emergence of multiple chronic conditions, such as physical activity, reductions in smoking rates, and improved and coordinated access to health care services.

In the United States, over 60% of adults aged 65 years or older have multiple chronic health conditions, with consequences that include reduced quality of life, increasingly complex but less person-centered treatment, and higher health care costs. A previous trial of ElderTree, an eHealth intervention for older adults, found socioemotional benefits for those with high rates of primary care use. This study tested the effectiveness of an ElderTree intervention designed specifically for older patients with multiple chronic conditions to determine whether combining it with primary care improved socioemotional and physical outcomes. In a nonblinded randomized controlled trial, 346 participants recruited from primary care clinics were assigned 1:1 to the ElderTree intervention or an attention control and were followed for 12 months. All participants were aged 65 years or older and had electronic health record diagnoses of at least three of 11 chronic conditions. Primary outcomes were mental and physical quality of life, psychological well-being (feelings of competence, connectedness, meaningfulness, and optimism), and loneliness. Tested mediators of the effects of the study arm (ElderTree vs active control) on changes in primary outcomes over time were 6-month changes in health coping, motivation, feelings of relatedness, depression, and anxiety. Tested moderators were sex, scheduled health care use, and number of chronic conditions. Data sources were surveys at baseline and 6 and 12 months comprising validated scales, and continuously collected ElderTree usage. At 12 months, 76.1% (134/176) of ElderTree participants were still using the intervention. There was a significant effect of ElderTree (vs control) on improvements over 12 months in mental quality of life (arm × timepoint interaction: b=0.76, 95% CI 0.14-1.37; P=.02; 12-month ∆d=0.15) but no such effect on the other primary outcomes of physical quality of life, psychological well-being, or loneliness. Sex moderated the effects of the study arm over time on mental quality of life (b=1.33, 95% CI 0.09-2.58; P=.04) and psychological well-being (b=1.13, 95% CI 0.13-2.12; P=.03), with stronger effects for women than men. The effect of the study arm on mental quality of life was mediated by 6-month improvements in relatedness (α=1.25, P=.04; b=0.31, P<.001). Analyses of secondary and exploratory outcomes showed minimal effects of ElderTree. Consistent with the previous iteration of ElderTree, the current iteration designed for older patients with multiple chronic conditions showed signs of improving socioemotional outcomes but no impact on physical outcomes. This may reflect the choice of chronic conditions for inclusion, which need not have impinged on patients' physical quality of life. Two ongoing trials are testing more specific versions of ElderTree targeting older patients coping with (1) chronic pain and (2) greater debilitation owing to at least 5 chronic conditions. ClinicalTrials.gov NCT03387735; https://clinicaltrials.gov/study/NCT03387735. RR2-10.2196/25175.

PSS17 Prevalence and Predictors of Multiple Chronic Conditions Among United States Older Adults with PAIN and Prescribed Opioids

PSS15 Prevalence and Predictors of Multiple Comorbid Chronic Conditions Among a Nationally-Representative Sample of United States Older Adults with Self-Reported PAIN