Preventing “tipping points” in high comorbidity patients: A lifeline from health coaches – rationale, design and methods Mary E. Charlson a, Ilana Mittleman b, Rosio Ramos a, Andrea Cassells b, T.J. Lin b, Alice Eggleston c, Martin T. Wells d, James Hollenberg a, Paul Pirraglia a,f, Ginger Winston g, Jonathan N. Tobin b,e,* a Department of Medicine, Weill Cornell Medical College, 1300 York Ave, New York, NY 10021, USA b Clinical Directors Network (CDN), 5 West 37th Street, 10th Floor, New York, NY 10018, USA c AllianceChicago, 225 W Illinois Street, Suite 500, Chicago, IL 60654, USA d Cornell University Department of Statistics and Data Science, Comstock Hall, 1198, 129 Garden Ave, Ithaca, NY 14853, USA e The Rockefeller University Center for Clinical and Translational Science, 1230 York Avenue, New York, NY 10065, USA f UMass Chan – Baystate Regional Campus, Baystate Health Regional Campus, 759 Chestnut Street, Springfield, MA 01199, USA g George Washington University- School of Medicine and Health Sciences, 2300 Street NW, Washington, DC 20052, USA A R T I C L E I N F O Keywords: Chronic conditions Health coaches Cluster randomized controlled trial Federally qualified health centers Practice-based Research Networks (PBRNs) A B S T R A C T Background: This paper describes an innovative cluster randomized controlled trial design to evaluate the comparative effectiveness of two approaches to preventing significant destabilization, leading to unplanned hospitalization and increased disability for patients with high comorbidity, that is, multiple chronic diseases defined by an enhanced Charlson Comorbidity Index ≥4. Methods: A total of 1974 patients were randomized in four waves at each of the sixteen Federally Qualified Health Centers (FQHCs) in four health systems —two in New York and two in Chicago. The two interventions compared 1) Patient-Centered Medical Home (PCMH) as implemented by the FQHCs (usual care control); or 2) PCMH plus a coaching intervention delivered by Health Coaches (experimental) helping patients identify life goals to encourage self-management enhanced by a positive affect/self-affirmation strategy. The two primary patient-centered clinical outcomes are 1) Unplanned hospitalizations; and 2) Within-patient changes in quality of life and disability, as measured by the World Health Organization Disability Assessment Scale 2 (WHODAS 2.0). The hypotheses are: 1) intervention patients will have a 5 % relative reduction in unplanned hospitalizations as compared to control patients; and 2) reduced disability measured by WHODAS2.0; 3) destabilization or ‘tipping points’ leading to hospitalization will be more often triggered by psychosocial issues than by medical Issues. Conclusion: This cluster RCT has the potential to transform the care for patients with high comorbidity by helping motivate patients to engage in self-management and to successfully navigate the barriers, challenges, and stresses leading to destabilization, hospitalization, and increased disability. ClinicalTrials.gov registration number: NCT04176510 1. Introduction Improving outcomes for patients with multiple chronic diseases is a national priority for the Patient-Centered Outcomes Research Institute (PCORI), the Centers for Medicare and Medicaid Services, and the Department of Health and Human Services [1–4]. Patients with multiple chronic diseases have the greatest risk of hospitalization, especially repeated, unplanned hospitalizations [5]. Yet these patients are either excluded or ignored in clinical trials because they are heterogeneous and have worse outcomes, confounding results [4,6–9]. Most trials of com- plex care coordination have not focused on high-comorbidity patients [10,11]. As a result, almost none of the evidence-based guidelines address managing these complex patients [12,13] and disease-specific guidelines cannot be readily combined to manage them [3]. Thus, * Corresponding author at: Clinical Directors Network (CDN), 5 West 37th Street, 10th Floor, New York, NY 10018, USA. E-mail address: JNTobin@CDNetwork.org (J.N. Tobin). Contents lists available at ScienceDirect Contemporary Clinical Trials journal homepage: www.elsevier.com/locate/conclintrial https://doi.org/10.1016/j.cct.2025.107865 Received 17 May 2024; Received in revised form 14 February 2025; Accepted 22 February 2025 Contemporary Clinical Trials 152 (2025) 107865 Available online 28 February 2025 1551-7144/© 2025 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by- nc-nd/4.0/ ). http://ClinicalTrials.gov mailto:JNTobin@CDNetwork.org www.sciencedirect.com/science/journal/15517144 https://www.elsevier.com/locate/conclintrial https://doi.org/10.1016/j.cct.2025.107865 https://doi.org/10.1016/j.cct.2025.107865 http://crossmark.crossref.org/dialog/?doi=10.1016/j.cct.2025.107865&domain=pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ http://creativecommons.org/licenses/by-nc-nd/4.0/ there is almost no evidence about optimal strategies for managing pa- tients with multiple chronic diseases to prevent such hospitalizations. While uncontrolled chronic illness can lead to hospitalization, [14,15] a significant research gap exists in recognizing that destabili- zation and hospitalization are often precipitated by psychosocial and other socioeconomic stressors instead of purely medical issues [16,17]. Thus, while most interventions intensify the management of medical issues, [3] this cRCT will assess whether Health Coaches, by helping patients establish explicit life goals, learn self-management, and address urgent and unmet socioeconomic and psychosocial needs, will prevent destabilization, unplanned hospitalization, and increased disability. The objective of “Tipping Points”, a cluster randomized controlled trial focused on patients with multiple chronic diseases defined as an Enhanced Charlson Comorbidity Index (CCI) ≥4, is to evaluate the comparative effectiveness of two approaches to preventing significant destabilization that leads to unplanned hospitalization and increased disability. Established primary care patients of 16 Federally Qualified Health Centers (FQHCs), from four Health Systems: Erie Family Health Centers and Friend Health in Chicago, Family Health Centers at NYU, and Community Health Network in New York. Each system had four participating FQHCs. Each FQHC was randomized in four waves to either 1) the Patient-Centered Medical Home (PCMH) as implemented by their FQHCs (usual care, control) or 2) PCMH plus a structured Health Coach intervention that employed a standardized positive affect/ self-affirmation intervention to help motivate patients to learn self- management skills by setting life goals (intervention). There are two primary patient-centered clinical outcomes: 1) Un- planned hospitalizations, as determined by independent blinded re- viewers, which included all hospital admissions except those for a planned procedure [18]; and 2) patient-specific changes in disability, as measured by the World Health Organization Disability Assessment Scale 2.0 (WHODAS) [19]. Our hypotheses were that: 1) PCMH patients with health coaches will have a 5 % absolute reduction compared to PCMH patients in terms of unplanned hospitalizations. 2) Reducing hospitalization will result in reduced disability as assessed by WHODAS 2.0. 3) Destabilization or ‘Tipping Points’ leading to hospitalization or emergency department visits will be more often triggered by psycho- social issues – family, community, and environmental – than by medical Issues. 1.1. Rationale For the most part, interventions focused on patients with multiple chronic diseases have not succeeded in improving outcomes and reducing hospitalizations [20–22]. Many well-executed programs did not improve outcomes or reduce utilization, in part because they did not target the right patients. To succeed, interventions must accurately identify the small proportion of patients with the worst outcomes and/or highest utilization. If not targeted to a small percent of highest-risk patients, the interventions will not reduce hospitalizations or improve outcomes since the larger number of lower-risk patients will get the most attention [23,24]. In this cRCT, high-risk patients with comorbidity ≥4 have a higher risk for hospitalizations, repeated hospitalization, and readmissions [25]. Before study implementation, our Tipping Points’ Clinicians and Patient Stakeholder Advisory Committee, comprised of patients, clini- cians, and health systems leadership from participating FQHCs and community-based organizations, provided critical perspectives about patients’ needs. 2. Design and methods Study Population and Setting: Eligible patients were recruited from participating FQHCs, which all had National Committee for Quality Assurance PCMH Level 3 Designation [26,27]. This cRCT was reviewed and approved by a central IRB (BRANY IRB #19–08-188) and registered with ClinicalTrials.gov (ID NCT04176510) [28]. Screening and Eligibility: The four Health Systems provided de- identified lists of adult patients for each of the four FQHCs with their ICD-10 codes so that the patients’ comorbidity could be calculated. Patients with comorbidity index ≥4 were re-identified by Health System so that their FQHCs could provide lists of potentially eligible patients to the Health Coaches assigned and credentialled at the specific site who verified eligibility with Electronic Health Records and with patients before enrollment. The specific comorbid conditions are listed in Ap- pendix A1. This enhanced index is designed to predict hospitalization and cost, distinct from the original Charlson comorbidity index, designed to predict mortality [29]. Patients with comorbidity ≥4 were re-identified so that their FQHCs could provide lists of potentially eligible patients to the Health Coaches assigned and credentialled at the specific site to verify eligibility with patients before enrollment. Exclusions: Patients with metastatic cancer, end-stage renal disease, post-transplant, or HIV positive were excluded because most patients with these conditions have specialized care in settings other than the FQHCs. Patients with dementia, severe mental illness, or drug/alcohol abuse were excluded because of the different care teams and strategies required. Patients who could not communicate in English or Spanish were also excluded. 2.1. Cluster randomization The study biostatistician, Dr. Martin Wells, completed randomiza- tion before RCT initiation. To achieve the sample size of 1920, the four health systems each recruited 480 patients in four waves. During each of the four waves of recruitment, approximately 30 patients were recruited at each FQHC, for a minimum of 120 per FQHC over four waves (Table 1). All patients were cluster-randomized and, therefore, each wave at each FQHC was assigned to either the experimental or control arm. Randomization was performed at the cluster level rather than at the patient level to prevent contamination. Health Coaches were blinded until the initiation of each wave. Health Coaches recruited both control and intervention patients. The Health Coach reached out to eligible patients by mail or telephone depending on the FQHC. While recruitment was initially in-person, in the spring of 2020, during the height of the COVID pandemic, in-person recruitment was suspended and resumed at different times according to each FQHC’s pandemic procedures. The Health Coach obtained written consent during the in-person visit and oral consent during remote enrollment. 2.2. Baseline measures The baseline assessment included demographics (age, gender, race, ethnicity, preferred language, residence, employment) and primary care clinician. It also included self-rated health [30] and PROMIS Item Bank v1.0 short-form questions on average pain and fatigue, pain interfer- ence, and fatigue impact [31]. The PHQ-8 [32], life events from the Social Readjustment Rating Scale [33], the 10-item Perceived Stress Scale [34], and International Physical Activity Questionnaire [35] were recorded. In addition, Social Determinants of Health were noted, using some sections of the PRAPARE scale such as housing situation, insur- ance, transportation, social service needs such as requesting resources for food, clothing, utilities, child care, elder care, medicine, or any health care needs [36]. The 36-item World Health Organization Disability Assessment Scale (WHODAS 2.0) was the primary patient-centered outcome, and assesses six major life domains [19]. Secondary outcomes included the Patient Activation Measure (PAM-13), a 13-item scale that assesses skill and confidence for self-management of chronic conditions [37]; the Health M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 2 http://ClinicalTrials.gov Education Impact Questionnaire (heiQ) evaluates eight domains, including positive and active engagement in life, health behavior, skill and technique acquisition, constructive attitudes and approaches, self- monitoring and insight, health services navigation, social integration/ support, and emotional well-being [38]. 2.3. Randomization groups Goal setting was completed in both the control and intervention arms. The goal-setting scripts differed for the intervention and control patients (see Appendix B.1 and B.2). Control patients were only asked about life goals and the confidence they have that they will achieve them, but the goals were not tied to self-management. Control Group (PCMH): Control patients were treated in accor- dance with the FQHCs’ specific implementation of PCMH. Intervention Group (PCMH plus Health Coach Multi-Component Intervention): The intervention patients received the Health Coach intervention in addition to PCMH services. The Health Coach interven- tion combined engagement in self-management to achieve life goals with positive affect and self-affirmation [39,40] (see Appendix C). • Setting life goals and self-management strategies: Health Coaches helped patients articulate their own life goals for the next year. Once the life goals were set, Health Coaches helped patients develop a self-management plan. • Behavioral contracting: Patients made a contract about what they will do, when they will do it, where they will do it, and for how long [41]. Patients rated their self-efficacy for adopting the self- management strategies, and coaches ensured that their self-efficacy was high enough for patient-selected strategies [42–44]. The goal is to build a repertoire of self-management skills [45]. • Positive affect and Self-affirmation intervention: Patients were taught a simple approach that combined a positive affect and self- affirmation intervention to promote greater engagement in self- management to achieve life goals. • Emotional and tangible support for life stresses: Patients were encouraged to reach out to the Health Coach if they need help, encounter challenges, or have had a hospital admission. The Health Coach can help to mobilize support from family, social, and com- munity services [46]. • Intervention follow-ups: Health Coaches called intervention pa- tients bi-weekly for the first three months and thereafter as needed. During the follow-up calls, intervention patients were reminded about their life goals and selected self-management activities, and were coached to overcome obstacles that they have encountered. Standard follow-up and close-out for all patients: All patients were followed up by phone at 6 months, 12 months, and 24 months. Patients in both groups were assessed identically using standardized instruments (See Appendix A2). To minimize both contamination and ascertainment bias, the follow- up assessments at 6, 12, and 24 months post-intervention were con- ducted by a different Health Coach who did not enroll the participant over the telephone. Interval psychosocial, socioeconomic, or medical events were documented at each follow-up. As well as two open-ended questions were included to assess interval health issues and life events [47]: 1) “Since we last spoke, how have you been feeling?” and. 2) “Since we last spoke, have you had any major life events or major difficulties?” Loss to follow-up: The patients’ home, cell, and work phone numbers, emails, and phone numbers of designated family or friends were also obtained at consent. If a patient was lost to follow-up, the Health Coach sent a letter or text to the patient or the patient’s desig- nated contacts. Health Coaches also reviewed the EHR to determine the future visit date to schedule the follow-up survey at their next visit. 3. Outcomes 3.1. Primary outcomes The primary patient-centered outcomes are 1) unplanned hospitali- zations and 2) within-patient changes in WHODAS 2.0. Patient Report and Electronic Data Capture for Hospital Ad- missions and Emergency Department Visits Data on patient outcomes of unplanned hospitalizations and ED visits were gathered from both patient reports and electronic data sources and integrated into ClinvestiGator to alert Health Coaches of the need for follow-up. In New York City, hospitalization and ED visit data were obtained from daily hospitalization alerts through Healthix and BronxRHIO, which are NYC Health Information Exchanges and quar- terly from INSIGHT, a PCORI-funded CDRN [48]. In Chicago, admis- sion/discharge/transfer data were provided by Patient Ping/Bamboo Health, which is updated daily with hospitalization and ED visit infor- mation in patients’ EHRs. In addition, PCORnet CDRN alerts for patients in Chicago are sent quarterly through CAPriCORN [49]. Unplanned hospitalizations are hospitalizations for any reason other than elective procedures. The health coaches collected circum- stances surrounding hospitalization from patients. Two independent, blinded physician reviewers (P.P. and G.W.) classified all admissions as planned or unplanned. These physicians reviewed both the patient re- ported and EHR data about the hospitalization and the open-ended summaries of events leading to the hospitalization, including baseline diagnoses, while remaining blinded to intervention or control status. Table 1 Cluster Randomization/Assignment to Health Coach versus Usual Care (n = 64 clusters, 1920 patients). NYC PBRN (CDN) Chicago PBRN (AllianceChicago) N ¼ 960 N ¼ 960 NY1 NY2 Chicago1 Chicago2 Site 1 Site 2 Site 3 Site 4 Site 5 Site 6 Site 7 Site 8 Site 9 Site 10 Site 11 Site 12 Site 13 Site 14 Site 15 Site 16 Wave 1 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 Wave 2 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 Wave 3 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 Wave 4 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 Total N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 N = 120 M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 3 Patient-Reported Disability and Function. The 36-item World Health Organization Disability Assessment Scale (WHODAS 2.0) as- sesses six major life domains: cognition, mobility, self-care, getting along, life activities, and participation in society [19]. WHODAS 2.0 generates one global score and six domain-specific scores, validated against the SF-36 and other scales [50] and is reproducible and responsive to change [19,51]. 3.2. Secondary outcomes Emergency Department (ED) Visits included visits to emergency departments defined in PCORnet CDRNs within the Common Data Model in terms of Type and Provider. Within-patient changes in Patient Activation were evaluated with the Patient Activation Measure (PAM-13), a 13-item scale that assesses skill and confidence for self-management of chronic conditions [37]. Within-patient changes in the Health Education Impact Ques- tionnaire evaluates eight domains, including positive and active engagement [38]. 3.3. Health coach training Health Coaches, bilingual in English and Spanish, were drawn from the same communities as patients. In multiple studies, our experience has been that health coaches who are familiar with the patient’s com- munity and common challenges can create a trusted relationship with patients [52]. Health Coaches underwent a 14-day in-person training, followed by an additional 14 days of remote training, which used a combination of role-playing to impart effective communication and practical problem-solving skills. Training topics included motivational interviewing, life goal setting, behavioral contracts, problem-solving, and mapping local community resources. The training combined in- person and virtual training with many hours of role-playing and shad- owing experienced health coaches. A random sample of 7 % of all sessions was observed via real-time Zoom monitoring to evaluate adherence to the intervention protocol and the use of problem-solving. Treatment fidelity was evaluated using the Lichstein treatment fidelity model [53]. 3.4. Data management system ClinvestiGator provides Health Coaches, supervisors, and in- vestigators with a HIPAA-compliant data management system, creating a secure and reliable custom-tailored data capture system. Clinvesti- Gator also provides tools for monitoring recruitment and follow-up rates and readily exports well-curated data into analytic programs. 3.5. Effect sizes for the main patient-centered outcomes Unplanned hospitalization and the WHODAS 2.0 were the principal patient-centered outcomes. Determination of effective sample size (n = 1920), with 16 FQHCs and 4 waves (64 clusters), utilized an assigned power of 0.8, an alpha (2-tailed) = 0.05, an ICC = 0.05, 32 clusters per arm, and 30 participants per cluster. The goal was to detect clinically important differences in the two primary outcomes in patients with higher comorbidity: a 5 % decrease in unplanned hospital admissions and a 10-point change in the WHODAS 2.0 score [50]. The 5 % decrease in hospitalization was selected as a clinically important difference in the context of two populations. First, in a pop- ulation of working adults with a CCI ≥ 4, the 5 % absolute decline in the percent hospitalized over one year went from 14 % to 9 % [5]. Second, in a Medicaid-managed care population, this corresponds to an absolute decrease from 17 % to 12 % and a relative decrease of 29 % [5]. In both studies, a 5 % difference in hospitalization would be clinically signifi- cant. Both populations involved one-year data, and the hospitalizations over two years will likely be higher. In addition, 4 % of the working population and 7 % of the Medicaid population were hospitalized more than once, so the absolute reduction in total hospitalizations is likely to be greater. While we anticipate greater reductions, the study was pow- ered to detect a minimum absolute difference of 5 %. Similarly, a 10-point drop in WHODAS 2.0 score represented a clinically important difference including: (1) within-patient change in response to depression treatment [54]; (2) the difference between those who improved and those who did not improve after treatment for anx- iety [55]; and (3) half the difference between those who had a good recovery vs a poor recovery after surgery (21 vs 0) [56] and thus, rep- resents a clinically significant difference. 3.6. Sample size and power for secondary outcomes For Emergency Department Visits, we previously observed a baseline unplanned hospitalization rate of 17 %, so assuming rates of ED Visits greater than or equal to unplanned hospitalization rates, and given baseline ED Visit rates of 20 %, 25 %, and 30 %, we have sufficient statistical power (>80 %) with n = 1920 to detect differences as small as 4.9 %, 5.3 % and 5.7 %, respectively. For the Patient Activation Measure-13 (PAM-13), which runs from 1 to 100, with a sample size of 1920 and an expected baseline mean (±SD) = 54.18 ± 19.7, we can detect differences as small as 1.38, which is equivalent to 0.1 SD. When converted into quartiles, this sample size can detect a shift from Level 2 (lower levels of activation/self- management) to Level 3 (higher levels of activation/self-management [57]. For the Health Education Impact Questionnaire (heiQ), at base- line, we expect 33 % of all the participants to be highly engaged in self- management, and at follow-up, we can detect an intervention group difference in heiQ as small as 6.1 % with a sample size of 1920 [58]. 3.7. Analytic plan All main effects analyses will follow the intent-to-treat (ITT) princi- ple. Apart from attrition (expected to be <20 % at 2 years), we antici- pated very little missing data (see Appendix D). Models for each hypothesis account for levels of nesting in the analysis of our cluster design trial. We initially focus on multi-level modeling, using random effects for wave, FQHC network, and site. Fixed effects modeling also accounts for demographic, physical, psy- chological, and social assessments. We use a model selection strategy with objective information about theoretic measures to select between fixed and random effects approaches. In addition, we assess the impact of COVID-19 on our trial using a moderation analysis. We use indicator variables for time periods (pre/post or pre/during/post) to capture COVID-19 effects. The regression models use interaction terms with the intervention, and carry out a test for significance to assess the COVID-19 effect, as well as unequal wave duration. Modeling the COVID-19 moderation effect increases the explained variation in the regression model, reducing the standard error of the estimated intervention effect, and hence increasing the power relative to omitting the COVID-19 effect. Hypothesis 1 (Main effect): Unplanned hospitalization will be reduced by 5 % among Intervention PCMH patients with Health Coaches compared to PCMH patients. A multi-level regression model controls for the randomization group (PCMH alone vs. PCMH plus Health Coach), demographics, psychosocial and medical confounders, a COVID-19 wave disruption indicator, and Health Coach and FQHC site design random effect. Hypothesis 2 (Main effect): Reducing hospitalization will result in reduced disability (WHODAS). A multi-level regression analyses examines the six domains of the WHODAS 2.0 and the global score. The multi-level regression models controls for the randomization group, demographics, psychosocial and medical confounders, a COVID-19 wave disruption indicator, and Health Coach and FQHC site design M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 4 random effect. Hypothesis 3: Destabilization ‘tipping points’ are more often triggered by psychosocial issues – family, community, and envi- ronmental – than by medical issues. A survival analysis uses a time-to- event regression model, such as a multi-level Cox-type regression model. We first define the set of destabilization events and the sets of the trigger psychosocial and medical issues. The time from the first psychosocial and medical issue until destabilization is recorded, and if no destabili- zation occurs, the time is censored. The event and censoring times are analyzed with multi-level Cox–type regression models; each of the re- gressions controls for the randomization group and possible psychoso- cial and medical confounders, as well as the cluster sample design. Once the hazard regression model is fit, we compare the hazard rates for the psychosocial and medical issue event times using a partial-likelihood ratio test. The same analysis is carried out on the subsequent psycho- social and medical issues if fewer than 30 % of the event times have been censored. The baseline individual social determinants of health are analyzed in relation to the outcomes of unplanned hospitalization and ER visits. The Cox regressions accounts for the multi-level cluster sample design using an appropriate frailty model. Secondary outcomes: A series of regression analyses examines ED visits, the eight domains of the heiQ, and the Patient Activation Mea- sure. Each regression model controls for the randomization group, possible psychosocial and medical confounders, and the cluster sample design. 3.8. Data and safety monitoring A Data and Safety Monitoring Board (DSMB) included Charles McCulloch PhD (Chair), Barbara J. Turner, MD, MSED, and Brent Egan, MD. The DSMB reviewed and approved the final protocol, which was uploaded to www.ClinicalTrials.gov before trial launch. The DSMB conducts yearly interim analyses. After 50 % of the participants completed the intervention, the DSMB determined whether the differ- ence between the two arms had reached statistical significance to allow for early trial termination. 3.9. COVID-19 related impact Originally, five months were allocated for each of the four waves. Recruitment was paused during Wave 1 because of FQHC closures due to the COVID-19 pandemic. When recruitment resumed, it was done remotely, and oral consent replaced written consent with the approval of the IRB. Follow-up continued for all previously consented and enrolled patients. Table 2 shows the duration of each wave. The COVID-19 effect is assessed using a moderation analysis. Indi- cator variables for time periods (pre/post or pre/during/post) are used to capture COVID-19 effects. The regression models use interaction terms with the intervention, and a test for significance is carried out to assess the COVID-19 effect. If the COVID-19 effects are not modeled via a regression approach, its effect would be unaccounted for and would lead to an increased standard error for the estimated intervention effect. Modeling the moderation effect increases explained variation in the regression model, reduces the standard error of the estimated inter- vention effect, and hence increases the power relative to omitting the COVID-19 effect. 4. Discussion High comorbidity patients, [59] often face an array of complex so- cioeconomic, cultural, environmental, social, and behavioral problems [27,60]. Patients often cannot prevent psychosocial events but can learn skills and gain support to cope while continuing self-management. Patients with high comorbidity face greater challenges in self- management and must believe that success in self-management will allow them to achieve their own life goals. Patients with a higher sense of purpose in life are more motivated to engage in and sustain self- management [61]. Social cognitive theory suggests that goal-setting for what a patient values is more important to increasing their motivation to initiate changes, while tangible support helps to sustain them [42,62]. Patients with higher self-efficacy, that is, more confidence in their ability to perform a specific behavior to achieve a given outcome, are more likely to engage in the behavior, to maintain the behavior, and to recover after setbacks [63]. Health Coaches help patients identify their life goals for the next year and the specific self-management activities that they need to do to reach their goals [42,44,64]. The goal-setting process is based upon what was used in the SCALE Small Changes and Lasting Effects RCT (NHLBI U01 HL 097843), which showed that a Health Coach can help high comor- bidity patients set realistic self-management goals that they can achieve [40,65,66]. Tipping Points also builds on three simultaneous RCTs in 756 pa- tients with chronic cardiopulmonary disease, which found that positive affect buffered against the potential adverse behavioral impact of negative psychosocial changes [67–71]. Positive affect interventions have had marked effects on people’s motivation, self-efficacy, and ability to cope with stressful life events [72]. Positive affect also in- creases an individual’s perception of the value of their ultimate goals [73], their belief that the immediate self-management goal can be achieved, and their expectation that they can reach the goal [74,75]. In addition, self-affirmation interventions enhanced individuals’ confi- dence in their ability to avoid unhealthy behaviors and to adopt positive ones [67,68]. The project is not designed to isolate the individual effects of the multicomponent intervention, such as setting life goals and self- management goals, positive affect self-affirmation, coaching for self- management, and emotional and tangible support, but to assess its aggregate impact on outcomes. The Health Coach intervention is deliberately not focused on a specific disease but rather on patients with multiple chronic diseases, defined by the comorbidity, a reproducible, valid method of defining multiple chronic diseases, which is key to identifying the right patients. This project evaluated an innovative approach that combines a positive affect and self-affirmation interven- tion to promote greater engagement in self-management to achieve life goals and to successfully navigate the barriers, challenges, and stresses, which are often tipping points, that could otherwise lead to destabili- zation, unplanned hospitalization, and increased disability. Disclaimer All statements in this report, including its findings and conclusions, are solely those of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute (PCORI), its Board of Governors or Methodology Committee. Funding acknowledgement This work was supported through a Patient-Centered Outcomes Research Institute (PCORI) Award (IHS-2017C3-8923 to Clinical Di- rectors Network) with additional infrastructure support provided by the Agency for Healthcare Research and Quality (AHRQ N2-PBRN Grant #1 P30-HS-021667 to Clinical Directors Network); INSIGHT (PCORI Award Table 2 Wave Lengths. Wave Dates Number of Months 1 November 2019–February 2021 16 2 February 2021–November 2021 9 3 November 2021–January 2023 15 4 January 2023–February 2024 13 M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 5 http://www.ClinicalTrials.gov #R-1306-03961 to Weill Cornell Medicine); and CAPriCORN (PCORI Award #CDRN-1306-04737 to Northwestern Medicine). CRediT authorship contribution statement Mary E. Charlson: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Conceptuali- zation. Ilana Mittleman: Writing – review & editing, Project adminis- tration. Rosio Ramos: Writing – review & editing, Writing – original draft, Supervision, Project administration. Andrea Cassells: Writing – review & editing, Supervision, Project administration. T.J. Lin: Vali- dation, Project administration. Alice Eggleston: Writing – review & editing, Project administration. Martin T. Wells: Methodology, Formal analysis, Data curation. James Hollenberg: Software, Data curation. Paul Pirraglia: Formal analysis. Ginger Winston: Formal analysis. Jonathan N. Tobin: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Conceptualization. Declaration of competing interest Mary E. Charlson, Cornell University has filed a patent on the methods and systems implementing the enhanced comorbidity index in the management of healthcare resources. Cornell owns the copyright on the index and MEC receives a portion of license revenue from Cornell University under Cornell’s Inventions and Related Property Rights Policy. Data availability De-identified data will be made available on request and execution of a data use agreement. Acknowledgements We wish to acknowledge the following FQHCs for their participation: Community Health Network (Crown Heights, Harlem, Long Island City, Sutphin), Family Health Centers at NYU (Park Ridge, Park Slope, Sunset Park, Flatbush), Erie Family Health Centers (Foster, Waukegan, West Town, Division), Friend Health (Ashland, Pulaski, Cottage Grove, Western) We would like to acknowledge the project team members and stakeholders from the participating sites: Nicholas Martin, Dr. Taisha Benjamin, Ambrosia Elder, Dr. Kelly Horn, Michael Chapman, Dr. Rishi Dalal, Jodyann Wynter, and Dr. Andrea Ciano (CHN); Dr. Radhika Gore, Phil Hayward, Dr. Jorge Sastre, Claudya Verdiner, Dr. Ekaterina Olkhina, Anne Marie Sabella, Dr. Ramiro Jervis, and Dr. Sandeep Bhat (FHCs at NYU). Thank you to the Tipping Points research team, both current and former members: Mirta Milanes, Shelly Sital, Dr. Fred Rachman, Dr. Nivedita Mohanty, Dr. Eve Walter, Ariana Perdomo, Harpreet Kaur, Eilyn Candelo, Jacqueline Cruz, Maria Ruiz, Martha Muñoz, Victoria Sarita, Anisa Mian, Roxane Padilla, Andy Cruz, Jaileen Ocasio, Stepha- nie Vasquez, Andrea Chavarria, Azalia Garcia, Linda Humaidan, Joel Blanco-Aguirre, Nataly Aguirre, Edward Castillo, Annette Rueda, Nees Calderon, and Cynthia Mofunanya. Thank you to the Community & Patient Stakeholder Advisory Committee: (Dr. Taisha Benjamin, Dr. Sandeep Bhat, Edwin Ruiz, Dr. Deborah Midgley, Elizbeth Tumiel, Dr. Geetha Govindarajan, Kymmie Sanders, Susan McCauley, Guillerma M Martinez) the Data and Safety Monitoring Board Members including Charles McCulloch PhD (Chair), Barbara J. Turner, MD, MSED and Brent Egan, MD and to our data partners at INSIGHT (Dr. Mark Weiner, Alexandra LaMar, Catherine Rabin, Rosie Ferris, Kanta Hague, Dmitry Morozyuk, Joshua Gelber, Peter Morrisey), CAPriCORN (Shelly Sital, Andrea VanderLaan, Kyra VanDoren) Healthix (Todd Rogow, Tim Tirrell, and Tom Moore) and BronxRHIO (Kathryn Miller, Megha Khatri Arora, Ralph Figueroa and Jianwen Wu). Appendix A. Appendix A.1. Enhanced Charlson Comorbidity Index for Tipping Points Conditions and Weights Chronic Condition Comorbidity Weight Myocardial infarction 1 Congestive heart failure 1 Peripheral vascular disease or bypass 1 Cerebrovascular disease or transient ischemic disease 1 Hemiplegia 2 Pulmonary disease/asthma 1 Diabetes 1 Diabetes with end organ damage 2 Moderate to severe renal disease 2 Mild liver disease 2 Severe liver disease 3 Gastric or peptic ulcer 1 Cancer 2 Rheumatic or connective tissue disease 1 Hypertension 1 Skin ulcers/cellulitis 2 Depression 1 Warfarin 1 Inflammatory bowel disease 1 Sickle cell disease 3 Hemophilia 3 Muscular dystrophy 2 Cystic fibrosis 3 Tay Sachs 3 Cerebral palsy 2 Uncontrolled seizures 3 (continued on next page) M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 6 (continued ) Chronic Condition Comorbidity Weight EXCLUDE: ​ Dementia or Alzheimer’s 1 Dialysis 2 Developmental delay 2 Mental retardation 2 Down’s syndrome 3 Bipolar disease 3 Antipsychotics 3 Drug or alcohol addiction 3 Schizophrenia 3 Autism 3 Metastatic solid tumor 6 HIV or AIDS 6 Any transplant: Renal, heart, liver, bone marrow, or lung 6 A.2. Baseline & Follow-up Measures Outcome or variable Measure Primary outcomes Hospitalization/ED visit Patient reports, Electronic Admission, Discharge and Transfer (ADT) data from several sources, including: In NYC: INSIGHT CRN, Healthix, BronxRHIO In Chicago CAPriCORN CRN, Alliance, and Bamboo Health. Disability Assessment WHO Disability Assessment Schedule (WHODAS 2.0) Secondary outcomes & covariates Chronic conditions Enhanced Charlson Comorbidity Index (CCI) COVID-19 vaccinations COVID-19 vaccinations Pain, fatigue PROMIS Pain and fatigue Depression Patient Health Questionnaire (PHQ 8) Life events Social Readjustment Rating Scale Social support system Social Network Assessment Patient education & self-management Health Education Impact Questionnaire (heiQ) Physical activity International Physical Activity Questionnaire (IPAQ) Self-efficacy Patient Activation Measure (PAM) Stress Perceived Stress Scale (PSS) Emotions Trait Affect Scale Social Determinants of Health – individual level PRAPARE survey Social Determinants of Health – neighborhood level Social Deprivation Index (SDI), Area Deprivation Index (ADI). Appendix B. Appendix B.1. Goal-Setting and Contract for Control Arm 1. What are the important goals, [key life goal(s)] that you would like to achieve in the next year? _________________________________________________________________________ 2. Why is it important to you to be able to achieve your life goal(s) in the next year? _________________________________________________________________________ 3. What can you do to make sure you achieve your important goal(s). _________________________________________________________________________ On a scale of 1 to 10, how confident are you that you can do this? Let’s say 1 is not at all confident and 10 is completely confident. How confident are you that you can do this? _____ We will make a personal contract that reminds you of your goal. Participant will: (write in their responses) (Do what) _____________________________________________________ (When) _______________________________________________________ (How often)____________________________________________________ (For how long)___________________________________________________ I think this is an excellent step to help you better manage yourself and your health so you will be able to achieve your life goals. B.2. Goal Setting and Contract for Intervention Arm 1. What are the important goals, [key life goal(s)] that you would like to achieve in the next year? M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 7 _________________________________________________________________________ 2. Why is it important to you to be able to achieve your life goal(s) in the next year? _________________________________________________________________________ 3. What can you do to make sure you achieve your important goal(s) _________________________________________________________________________ 4. Do you believe that in order to achieve your life goal(s), you also have to manage your health conditions first? _________________________________________________________________________ 5. What are your challenges in terms of the management of your medical conditions? _________________________________________________________________________ 6. In thinking about how to do this, what are the challenges or obstacles? _________________________________________________________________________ 7. What would help you overcome these challenges? What extra support do you need to achieve the goal(s)? _______________________________________________________________________ 8. Briefly, so this is what you will do to better manage your health to reach your life goals: (prompt: Summarize) _________________________________________________________________________ On a scale of 1 to 10, how confident are you that you can do this? Let’s say 1 is not at all confident and 10 is completely confident. How confident are you that you can do this? _____ [PROMPT to HC: if 7 or less, please select another goal] We will make a personal contract that reminds you of your goal. Participant will: (write in their responses). (Do what) _____________________________________________________ (When) _______________________________________________________ (How often) ____________________________________________________ (For how long) ___________________________________________________ I think this is an excellent step to help you better manage yourself and your health so you will be able to achieve your life goals. Appendix C. Appendix C.1. Positive Affect/Self-Affirmation OK. As part of this study, I’m going to ask you to think of these proud and positive moments along with the ___________________________________________________ (selected goals). First, when you get up in the morning, think about the small things that you said make you feel good, like __________________________________________. Then as you go through your day, notice these and other small things that make you feel good and take a moment to enjoy them. Second, when you encounter some difficulties or are in a situation that makes it hard for you to keep being physically active, think about things you enjoy or proud moments in your life, like ______________________. Do you think you can do these things? Will you think of one of these when you start your day? And as you go through your day will you take time to enjoy them? Will you also think of one of these when you encounter some difficulties or are in a situation that makes it hard for you to keep being physically active? Appendix D. Missing data Assuming that missing data are either Missing Completely at Random (MCAR) or Missing at Random (MAR), a maximum-likelihood multilevel modeling approach, in conjunction with the covariates to adjust for attrition bias (if necessary), yields intent-to-treat parameter estimates that are consistent. We have a macro that identifies each pattern of missing data that occurs for the items used to construct the scale; perform a separate regression analysis for each pattern predicting the scale score, for those with complete data, from the non-missing items; and use the regression equation to replace the missing scale score with the regression-based predicted score for people having that pattern of missing data. Missing data are only replaced with predicted scores when the regression analysis indicates that the predicted score is a sound estimate. Using this imputation approach, we expect that there will be almost no missing data. All patients will be accounted for, including any reasons for dropouts. M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 8 References [1] Multiple Chronic Conditions –A Strategic Framework: Optimum Health and Quality of Life for Individuals with Multiple Chronic Conditions, US Department of Health and Human Services, Washington DC, December 2010. Editor. [2] US Department of Health and Human Services, Inventory of Programs, Activities, and Initiatives Focused on Improving the Health of Individuals with Multiple Chronic Conditions (MCC). USDOHAHS. Services, Compiled by the HHS Interagency Workgroup on Multiple Chronic Conditions (MCC) under the direction Anand Parekh MD MPH, Deputy Assistant Secretary for Health (science and Medicine) Office of the Assistant Secretary for Health, September 2011. [3] Multiple Chronic Conditions Initiative: Private Sector Activities focused on improving the health of individuals with multiple chronic conditions: INNOVATIVE PROFILES, US Department of Health and Human Services, Washington DC, November 2012. Editor. [4] A.K. Parekh, et al., Managing multiple chronic conditions: a strategic framework for improving health outcomes and quality of life, Public Health Rep. 126 (4) (2011) 460–471. [5] M.E. Charlson, et al., Medicaid managed care: how to target efforts to reduce costs, BMC Health Serv. Res. 14 (2014) 461. [6] M. Fortin, et al., Randomized controlled trials: do they have external validity for patients with multiple comorbidities? Ann. Fam. Med. 4 (2) (2006) 104–108. [7] H.G. Van Spall, et al., Eligibility criteria of randomized controlled trials published in high-impact general medical journals: a systematic sampling review, Jama 297 (11) (2007) 1233–1240. [8] A.R. Jadad, et al., Consideration of multiple chronic diseases in randomized controlled trials, Jama 306 (24) (2011) 2670–2672. [9] C.M. Boyd, D. Vollenweider, M.A. Puhan, Informing evidence-based decision- making for patients with comorbidity: availability of necessary information in clinical trials for chronic diseases, PLoS One 7 (8) (2012) e41601. [10] R. Berry-Millett, T.S. Bodenheimer, Care management of patients with complex health care needs, Synth. Proj. Res. Synth. Rep. (19) (2009) 52372. [11] F.J. Huyse, F.C. Stiefel, P. de Jonge, Identifiers, or “red flags,” of complexity and need for integrated care, Med. Clin. North Am. 90 (4) (2006) 703–712. [12] E.S. Fisher, Paying for performance–risks and recommendations, N. Engl. J. Med. 355 (18) (2006) 1845–1847. [13] M. Lugtenberg, et al., Current guidelines have limited applicability to patients with comorbid conditions: a systematic analysis of evidence-based guidelines, PLoS One 6 (10) (2011) e25987. [14] L.N. Gitlin, et al., A randomized trial of a multicomponent home intervention to reduce functional difficulties in older adults, J. Am. Geriatr. Soc. 54 (5) (2006) 809–816. [15] L.P. Fried, et al., Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care, J. Gerontol. A Biol. Sci. Med. Sci. 59 (3) (2004) 255–263. [16] J.R. Vest, et al., Determinants of preventable readmissions in the United States: a systematic review, Implement. Sci. 5 (2010) 88. [17] C. Martin, J. Sturmberg, Complex adaptive chronic care, J. Eval. Clin. Pract. 15 (3) (2009) 571–577. [18] AHRQ, Unplanned readmission: hospital-wide all-cause, unplanned readmission rate (HWR), NQMC National Quality Measures Clearinghouse, October 2015. Available from: https://www.qualitymeasures.ahrq.gov/summaries/summary/4 9197/unplanned-readmission-hospitalwide-allcause-unplanned-readmission-ra te-hwr. [19] Ustun, T.B., et al., Developing the World Health Organization Disability Assessment Schedule 2.0. Bull. World Health Organ.. 88(11): p. 815–23. [20] Boult, C., et al., The effect of guided care teams on the use of health services: results from a cluster-randomized controlled trial. Arch. Intern. Med.. 171(5): p. 460–6. [21] S.R. Counsell, et al., Geriatric Resources for Assessment and Care of Elders (GRACE): a new model of primary care for low-income seniors, J. Am. Geriatr. Soc. 54 (7) (2006) 1136–1141. [22] S.R. Counsell, et al., Cost analysis of the Geriatric Resources for Assessment and Care of Elders care management intervention, J. Am. Geriatr. Soc. 57 (8) (2009) 1420–1426. [23] D. Peikes, et al., Third Report to Congress on the Evaluation of the Medicare Coordinated Care Demonstration, in: MPR reference number 8756-430, Princeton N.J, Mathematica Policy Research, January 1 2008. [24] D. Peikes, et al., Effects of care coordination on hospitalization, quality of care, and health care expenditures among Medicare beneficiaries: 15 randomized trials, Jama 301 (6) (2009) 603–618. [25] M. Charlson, et al., The Charlson comorbidity index can be used prospectively to identify patients who will incur high future costs, PLoS One 9 (12) (2014) e112479. [26] P.A. Nutting, et al., Transforming physician practices to patient-centered medical homes: lessons from the national demonstration project, Health Aff (Millwood) 30 (3) (2011) 439–445. [27] E. Rich, et al., Coordinating Care for Adults with Complex Care Needs in the Patient-Centered Medical Home: Challenges and Solutions, (HHSA290100900019/ HHSA29032005T), Editor, Agency for Healthcare Research and Quality White Paper, Rockville MD, 2012. [28] ClinicalTrials.gov, Preventing Tipping Points in High Comorbidity Patients: A Lifeline From Health Coaches (Tipping Points), Available from: https://clinicaltria ls.gov/study/NCT04176510, 2023. [29] M.E. Charlson, et al., A new method of classifying prognostic comorbidity in longitudinal studies: development and validation, J. Chronic Dis. 40 (5) (1987) 373–383. [30] J.E. Ware, et al., SF-36 physical and mental health summary scales : a user’s manual, Health Institute, New England Medical Center. 1 v. : ill, Boston, Mass, 1994. [31] K.F. Cook, et al., PROMIS measures of pain, fatigue, negative affect, physical function, and social function demonstrated clinical validity across a range of chronic conditions, J. Clin. Epidemiol. 73 (2016) 89–102. [32] K. Kroenke, et al., The PHQ-8 as a measure of current depression in the general population, J. Affect. Disord. 114 (1–3) (2009) 163–173. [33] D. Wallace, et al., The social readjustment rating scale: updated and modernised, PLoS One 18 (12) (2023) e0295943. [34] S. Cohen, T. Kamarck, R. Mermelstein, A global measure of perceived stress, J. Health Soc. Behav. 24 (4) (1983) 385–396. [35] M. Hagstromer, P. Oja, M. Sjostrom, The International Physical Activity Questionnaire (IPAQ): a study of concurrent and construct validity, Public Health Nutr. 9 (6) (2006) 755–762. [36] National Association of Community Health Centers [Internet], The Protocol for Responding to and Assessing Patient Assets, Risks, and Experiences (PRAPARE) by National Association of Community Health Centers, Association of Asian Pacific Community Health Organizations, and Oregon Primary Care Association, Available from: http://www.nachc.org/prapare, 2013. [37] J.H. Hibbard, et al., Development and testing of a short form of the patient activation measure, Health Serv. Res. 40 (6 Pt 1) (2005) 1918–1930. [38] R.H. Osborne, G.R. Elsworth, K. Whitfield, The Health Education Impact Questionnaire (heiQ): an outcomes and evaluation measure for patient education and self-management interventions for people with chronic conditions, Patient Educ. Couns. 66 (2) (2007) 192–201. [39] E.G. Phillips, et al., Innovative approaches to weight loss in a high-risk population: the small changes and lasting effects (SCALE) trial, Obesity (Silver Spring) 25 (5) (2017) 833–841. [40] E.G. Phillips-Caesar, et al., Small Changes and Lasting Effects (SCALE) Trial: the formation of a weight loss behavioral intervention using EVOLVE, Contemp. Clin. Trials 41 (2015) 118–128. [41] K. Lorig, et al., The components of action planning and their associations with behavior and health outcomes, Chronic Illn. 10 (1) (2014) 50–59, https://doi.org/ 10.1177/1742395313495572. [42] A. Bandura, Self-efficacy: toward a unifying theory of behavioral change, Psychol. Rev. 84 (2) (1977) 191–215. [43] A. Bandura, Social Learning Theory, Prentice Hall, Englewood Cliffs NJ, 1977. [44] S. Rollnick, M. Morgan, Motivational interviewing: increasing readiness to change, in: Psychotherapy and Substance buse: A practitioner’s Handbook, W. A, Editor, Guilford press, New York, 1995. [45] H. Holman, K. Lorig, Patients as partners in managing chronic disease. Partnership is a prerequisite for effective and efficient health care, Bmj 320 (7234) (2000) 526–527. [46] E. Craig, D. Eby, J. Whittingon, Care coordination model: better care at lower cost for poeple with multiple health and social needs, in: IHI Inoovation Series White Paper, Institute for Health Care Improvement, Cambridge MA, 2011. [47] M. Charlson, et al., Interval life events are an important determinant of heterogeneity in outcomes in a randomised trial: a novel, simple method of assessment, BMJ Open 14 (7) (2024) e074623. [48] R. Kaushal, et al., Changing the research landscape: the New York City Clinical Data Research Network, J. Am. Med. Inform. Assoc. 21 (4) (2014) 587–590. [49] A.N. Kho, et al., CAPriCORN: Chicago area patient-centered outcomes research network, J. Am. Med. Inform. Assoc. 21 (4) (2014) 607–611. [50] Garin, O., et al., Validation of the "World Health Organization Disability Assessment Schedule, WHODAS-2" in patients with chronic diseases. Health Qual. Life Outcomes. 8: 51. [51] Von Korff, M., et al., Functional outcomes of multi-condition collaborative care and successful ageing: results of randomised trial. Bmj. 343 p. d6612. [52] S.A. Lewin, et al., Lay health workers in primary and community health care, Cochrane Database Syst. Rev. 1 (2005) CD004015. [53] K. Lichstein, B. Riedel, R. Grieve, Fair tests of clinical trials: A treament implementation model, Adv. Behav. Res. Therapy 16 (1994) 1–29. [54] L.A. Chwastiak, M. Von Korff, Disability in depression and back pain: evaluation of the World Health Organization Disability Assessment Schedule (WHO DAS II) in a primary care setting, J. Clin. Epidemiol. 56 (6) (2003) 507–514. [55] S.J. Perini, T. Slade, G. Andrews, Generic effectiveness measures: sensitivity to symptom change in anxiety disorders, J. Affect. Disord. 90 (2–3) (2006) 123–130. [56] M.A. Shulman, et al., Measurement of disability-free survival after surgery, Anesthesiology 122 (3) (2015) 524–536. [57] I. Bos-Touwen, et al., Patient and disease characteristics associated with activation for self-management in patients with diabetes, chronic obstructive pulmonary disease, chronic heart failure and chronic renal disease: a cross-sectional survey study, PLoS One 10 (5) (2015) e0126400. [58] S. Nolte, et al., The extent and breadth of benefits from participating in chronic disease self-management courses: a national patient-reported outcomes survey, Patient Educ. Couns. 65 (3) (2007) 351–360. [59] D.B. Reuben, M.E. Tinetti, Goal-oriented patient care–an alternative health outcomes paradigm, N. Engl. J. Med. 366 (9) (2012) 777–779. [60] M.M. Safford, J.J. Allison, C.I. Kiefe, Patient complexity: more than comorbidity. The vector model of complexity, J. Gen. Intern. Med. 22 (Suppl. 3) (2007) 382–390. [61] Friedman, E.M. and C.D. Ryff, Living well with medical comorbidities: a biopsychosocial perspective. J. Gerontol B Psychol. Sci. Soc. Sci.. 67(5): 535–44. [62] E.A. Locke, Towards a theory of task motivation and incentives, Organ. Behav. Hum. Perform. 3 (1968) 157–189. M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 9 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0005 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0005 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0005 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0010 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0010 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0010 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0010 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0010 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0010 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0015 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0015 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0015 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0015 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0020 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0020 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0020 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0025 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0025 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0030 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0030 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0035 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0035 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0035 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0040 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0040 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0045 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0045 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0045 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0050 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0050 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0055 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0055 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0060 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0060 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0065 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0065 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0065 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0070 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0070 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0070 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0075 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0075 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0075 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0080 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0080 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0085 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0085 https://www.qualitymeasures.ahrq.gov/summaries/summary/49197/unplanned-readmission-hospitalwide-allcause-unplanned-readmission-rate-hwr https://www.qualitymeasures.ahrq.gov/summaries/summary/49197/unplanned-readmission-hospitalwide-allcause-unplanned-readmission-rate-hwr https://www.qualitymeasures.ahrq.gov/summaries/summary/49197/unplanned-readmission-hospitalwide-allcause-unplanned-readmission-rate-hwr http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0095 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0095 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0095 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0100 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0100 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0100 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0105 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0105 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0105 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0110 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0110 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0110 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0115 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0115 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0115 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0120 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0120 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0120 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0125 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0125 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0125 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0125 https://clinicaltrials.gov/study/NCT04176510 https://clinicaltrials.gov/study/NCT04176510 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0135 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0135 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0135 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0140 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0140 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0140 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0145 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0145 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0145 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0150 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0150 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0155 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0155 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0160 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0160 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0165 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0165 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0165 http://www.nachc.org/prapare http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0175 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0175 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0180 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0180 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0180 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0180 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0185 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0185 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0185 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0190 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0190 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0190 https://doi.org/10.1177/1742395313495572 https://doi.org/10.1177/1742395313495572 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0200 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0200 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0205 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0210 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0210 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0210 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0215 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0215 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0215 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0220 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0220 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0220 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0225 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0225 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0225 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0230 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0230 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0235 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0235 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0240 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0240 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0245 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0245 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0250 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0250 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0250 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0255 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0255 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0260 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0260 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0265 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0265 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0265 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0265 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0270 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0270 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0270 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0275 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0275 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0280 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0280 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0280 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0285 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0285 [63] S. Rollnick, W. Miller, What is motivational interviewing? Behav. Cogn. Psychother. 23 (1995) 325–334. [64] W. Katon, et al., Integrating depression and chronic disease care among patients with diabetes and/or coronary heart disease: the design of the TEAMcare study, Contemp. Clin. Trials 31 (4) (2010) 312–322. [65] J.D. Eldridge, et al., Environmental influences on small eating behavior change to promote weight loss among black and Hispanic populations, Appetite 96 (2016) 129–137. [66] J.M. Hippolyte, et al., Recruitment and retention techniques for developing faith- based research partnerships, New York city, 2009-2012, Prev. Chronic Dis. 10 (2013). [67] C.M. Steele, The psychology of self-affirmation: Sustaining the intergrity of the self, in: L. Berkowitz (Ed.), Advances in Experimental Social Psychology, Academic press, New York, 1988, pp. 261–302. [68] C.M. Steele, A threat in the air: how stereotypes shape intellectual identity and performance, Am. Psychol. 52 (1997) 613–629. [69] C.M. Steele, T. Liu, Dissonance processes and self-affirmation 45, 1983, pp. 5–19. [70] M.E. Charlson, et al., Randomized controlled trials of positive affect and self- affirmation to facilitate healthy behaviors in patients with cardiopulmonary diseases: rationale, trial design, and methods, Contemp. Clin. Trials 28 (6) (2007) 748–762. [71] S. Cohen, S.D. Pressman, Positive affect and health, Curr. Dir. Psychol. Sci. 15 (2006) 122–125. [72] S.D. Pressman, S. Cohen, Does positive affect influence health? Psychol. Bull. 131 (6) (2005) 925–971. [73] J.P. Newhouse, et al., Adjusting capitation rates using objective health measures and prior utilization, Health Care Financ. Rev. 10 (3) (1989) 41–54. [74] B. Kahn, A.M. Isen, The influence of positive affect on variety-seeking among safe, enjoyable products, J. Consum. Res. 20 (1993) 257–270. [75] A.M. Isen, J. Reeve, The influence of positive affect on intrinsic and extrinsic motivation: facilitating enjoyment of play, responsible work behavior, and self- control, Motiv. Emot. 29 (4) (2006) 295–323. M.E. Charlson et al. Contemporary Clinical Trials 152 (2025) 107865 10 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0290 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0290 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0295 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0295 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0295 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0300 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0300 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0300 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0305 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0305 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0305 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0310 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0310 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0310 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0315 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0315 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0320 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0325 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0325 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0325 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0325 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0330 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0330 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0335 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0335 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0340 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0340 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0345 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0345 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0350 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0350 http://refhub.elsevier.com/S1551-7144(25)00059-X/rf0350 Preventing “tipping points” in high comorbidity patients: A lifeline from health coaches – rationale, design and methods 1 Introduction 1.1 Rationale 2 Design and methods 2.1 Cluster randomization 2.2 Baseline measures 2.3 Randomization groups 3 Outcomes 3.1 Primary outcomes 3.2 Secondary outcomes 3.3 Health coach training 3.4 Data management system 3.5 Effect sizes for the main patient-centered outcomes 3.6 Sample size and power for secondary outcomes 3.7 Analytic plan 3.8 Data and safety monitoring 3.9 COVID-19 related impact 4 Discussion Disclaimer Funding acknowledgement CRediT authorship contribution statement Declaration of competing interest datalink3 Acknowledgements Appendix A Appendix A.1 Enhanced Charlson Comorbidity Index for Tipping Points Conditions and Weights A.2 Baseline & Follow-up Measures Appendix B Appendix B.1 Goal-Setting and Contract for Control Arm B.2 Goal Setting and Contract for Intervention Arm Appendix C Appendix C.1 Positive Affect/Self-Affirmation Appendix D Missing data References