Background: Depression and hypertension are common, costly, and destructive conditions among the rapidly aging population of China. The two disorders commonly coexist and are poorly recognized and inadequately treated, especially in rural areas. Methods: The Chinese Older Adult Collaborations in Health (COACH) Study is a cluster randomized controlled trial (RCT) designed to test the hypotheses that the COACH intervention, designed to manage comorbid depression and hypertension in older adult, rural Chinese primary care patients, will result in better treatment adherence and greater improvement in depressive symptoms and blood pressure control, and better quality of life, than enhanced Care-as- Usual (eCAU). Based on chronic disease management and collaborative care principles, the COACH model integrates the care provided by the older person’s primary care provider (PCP) with that delivered by an Aging Worker (AW) from the village’s Aging Association, supervised by a psychiatrist consultant. One hundred sixty villages, each of which is served by one PCP, will be randomly selected from two counties in Zhejiang Province and assigned to deliver eCAU or the COACH intervention. Approximately 2400 older adult residents from the selected villages who have both clinically significant depressive symptoms and a diagnosis of hypertension will be recruited into the study, randomized by the villages in which they live and receive primary care. After giving informed consent, they will undergo a baseline research evaluation; receive treatment for 12 months with the approach to which their village was assigned; and be re-evaluated at 3, 6, 9, and 12 months after entry. Depression and HTN control are the primary outcomes. Treatment received, health care utilization, and cost data will be obtained from the subjects’ electronic medical records (EMR) and used to assess adherence to care recommendations and, in a preliminary manner, to establish cost and cost effectiveness of the intervention. Discussion: The COACH intervention is designed to serve as a model for primary care-based management of common mental disorders that occur in tandem with common chronic conditions of later life. It leverages existing resources in rural settings, integrates social interventions with the medical model, and is consistent with the cultural context of rural life. Trial registration: ClinicalTrials.gov ID: NCT01938963; First posted: September 10, 2013. Keywords: Depression, Hypertension, Older adults, Rural China, Collaborative care * Correspondence: Yeates_conwell@urmc.rochester.edu Department of Psychiatry, University of Rochester Medical Center, 300 Crittenden Boulevard, Rochester, NY 14642, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Chen et al. BMC Geriatrics (2018) 18:124 Page 2 of 9 Background effectiveness of integrated depression care management in China’s population of older adults is growing rapidly. Esti- rural China, where access to care is more challenged, none mates are that by 2015 there will be over 140 million has addressed depression comorbid with other common people in China over age 65 years—equivalent to 44% of chronic conditions such as HTN, and none has comple- the total U.S. population (U.S. Census). The prevalence of mented medication management with a psychosocial treat- affective illness among older adults is significant. Almost ment component, which is often a preferred option of 6% of community-dwelling Chinese over age 60 have a de- older adult Chinese . pressive disorder [1, 2]. Among those with chronic med- Our objective with this study is to address these short- ical conditions, the prevalence of depression is higher, comings of previous work. Using a cluster randomized con- about 13% . Depression is associated with functional trolled design, the Chinese Older Adult Collaborations in impairment, greater use and higher costs of healthcare, Health (COACH) Study will compare the COACH inter- and increased risks of suicide and all-cause mortality . vention to enhanced Care-as-Usual (eCAU) for the treat- Hypertension (HTN) is one of the most common chronic ment of comorbid depression and HTN in older residents conditions in China, estimated to have affected 57% of rural of rural Chinese villages. COACH integrates the care pro- village residents over age 60 . HTN is a major cause of vided by a PCP with that delivered by an Aging Worker strokes and ischemic heart disease , which is the only (AW, a staff person of either the Women’s Federation or disorder to top depression in the WHO’s list of greatest Aging Association of the village), and supervised by a expected causes of global illness burden by 2020 . psychiatrist consultant (Psychiatrist). The study’s five aims Depression and HTN commonly coexist [3, 8, 9]. A are designed to test the hypotheses that, relative to subjects study in Shanghai shows that over 78% of medical inpa- who receive eCAU, those in the COACH intervention will tients with HTN had clinically significant depressive symp- show (Aim 1a) better adherence to antidepressant treat- toms . The combination of HTN and depression ment recommendations and (Aim 1b) greater improve- complicates treatment; for example, depression may affect ments in their depressive disorders over 12 months of patients’ hypertension medication adherence. However, involvement in the study; and further that relative to sub- there is low recognition of the need to treat depression in jects who receive eCAU, those in the COACH intervention China [10, 11], and mental health care in general is difficult will show (Aim 2a) better adherence to antihypertensive to access, especially in rural areas. Primary care providers treatment recommendations and (Aim 2b) greater im- (PCP) in rural villages of China are charged with managing provements in blood pressure (BP) control. We will exam- their patients’ chronic conditions, including HTN, but they ine the temporal associations of change in depression and receive little or no training in diagnosis or treatment of BP control, hypothesizing that (Aim 3a) improvements in mental illnesses and are not permitted to prescribe treatment adherence precede symptomatic improvement in antidepressants. depression and hypertension, (Aim 3b) improvements in Approaches to management of depression [12–14]and depression will precede improvements in BP control, and comorbid medical conditions [14, 15] in primary care set- (Aim 3c) those temporal patterns are associated with the tings have been developed and well tested in Western na- intervention received. We hypothesize also that the tions. Based on the chronic disease management model COACH intervention results in greater improvements in , they typically include the use of algorithm-driven, health-related quality of life than eCAU (Aim 4). Finally evidence-based depression treatments administered by the (Aim 5), we will explore resource utilization and costs asso- PCP in collaboration with a depression-care specialist ciated with delivering the eCAU and the COACH interven- (nurse, psychologist, social worker) who is embedded in tions over 12 months, important information with which to the practice and supervised by a psychiatrist; regular symp- guide subsequent research on implementation and dissem- tom monitoring; and education and support of patients ination of the model if it is shown to be clinically effective. and families to ensure their active engagement in care. This integrated or collaborative depression care management Methods approach has been associated with improved clinical out- Setting comes as well as cost savings relative to care-as-usual This ongoing study is being conducted in rural villages (CAU), including in older adults [14, 17]. It has rarely been of Tonglu and Jiande counties, two of 55 counties in applied in developing nations like China, however. One ex- Zhejiang Province, China. With population size and ception is a randomized controlled trial of collaborative de- average household income at the mean level of all coun- pression care management among older adults in urban ties in Zhejiang Province, Tonglu and Jiande encompass primary care clinics conducted by our group, the results of 283 and 319 villages respectively, with average popula- which showed significantly greater reductions of depressive tions of 1400 and 1600. Approximately 20% of their pop- symptoms over 12 months when compared with patients ulations are age ≥ 60 years. They are served by a total of who received CAU . No RCTs have examined the 602 PCPs (one in each village) and two mental hospitals Chen et al. BMC Geriatrics (2018) 18:124 Page 3 of 9 (one for each county), which combined have 19 psychia- to conduct follow up research interviews. The subjects’ trists who serve 0.72 million village residents. Every involvement in the study ends after 12 months. village has either a Women’s Federation or an Aging Association that is staffed by local residents, one of Comparison condition: Enhanced Care-as-Usual (eCAU) whom is the designated AW. PCPs in village clinics have on average three years of med- ical education after completing high school. They receive systematic training in the diagnosis and treatment of com- Sample strategy and procedures mon chronic medical illnesses. For HTN management they As a guard against bias associated with “leakage” of the are expected to follow guidelines developed by the Center intervention across arms of the study, the unit of for Disease Control and Prevention in China, which is randomization is the village, which is equivalent to clinic- analogous to those developed by the U.S. Joint National level randomization. Of all villages in Tonglu and Jiande Committee on Prevention, Detection, Evaluation, and County 160 have been randomly selected by Prof. Hengjin Treatment of High Blood Pressure (the “JNC 7 Report” Dong to receive either the eCAU or COACH. No two vil- ). The guidelines address detection, management by lages assigned to different intervention arms (COACH or both pharmacologic and non-pharmacologic means, and eCAU) are geographically contiguous. Upon random se- guidance on when to refer to more specialized care. lection of each village, each PCP and AW are approached Village PCPs, however, receive little training in the diag- by the study coordinator and their county-level supervi- nosis or treatment of mental disorders and ordinarily are sors for agreement to participate. If any PCP or AW provided with no practice guidelines to help with its man- declines, or if their assignment places them immediately agement. When depression is suspected by PCPs, current adjacent to a village in which the other intervention is to practice involves suggesting to patients (or family mem- be offered, another village will be selected from the pool. bers) that they consult a psychiatrist at the County Mental Hospital for diagnosis and treatment. There is no direct Subjects referral/transfer mechanism between PCPs and mental A total of approximately 2400 subjects will be recruited to health specialists, and it is uncommon for patients to take participate in the study for up to 12 months. They will in- the initiative. Provincial laws prohibit the village PCP from clude registered residents who live independently in the se- initiating antidepressant treatment. However, if the patient lected villages in Tonglu or Jiande County. They must be does see the County Mental Hospital psychiatrist who ≥60 years and have both clinically significant depression, then begins antidepressant medication, the PCP may defined as a baseline score on the Patient Health choose to renew the prescription, typically in two-week Questionnaire-9 (PHQ-9)  of 10 or more, and a diag- increments. nosis of HTN in their medical record. Subjects must have We refer to CAU as “enhanced” (eCAU) because PCPs intact cognitive functioning (Six-Item Screener [SIS]  are told when study subjects screen positive for depres- score < 3) and be capable of participating in study inter- sion and are provided with copies of depression practice views with informed consent. Subjects found on baseline guidelines developed for the study. Because of con- assessment to have mania, psychosis, alcohol abuse or de- straints on their knowledge and practice, however, we do pendence active in past 6 months, or suicidal ideation with not anticipate that access to this information will sub- intent are excluded from the study with a recommendation stantially impact patient outcomes [23–27]. made to the PCP that they be referred for psychiatric assessment at the County Mental Hospital. Intervention: COACH The COACH intervention applies chronic disease man- Procedure for recruitment, consent and retention agement principles to the treatment of both depression Each village clinic has a standardized electronic medical and HTN by a team consisting of the village PCP, AW, record system (EMR) supported by the Zhejiang Center for and a Psychiatrist. Disease Control and Prevention (CDC) in which each pa- tient is registered. PCPs use the EMR to identify all village Primary care providers (PCP) residents age ≥ 60 years who have a diagnosis of HTN. The background and qualifications of the PCP are as de- Trained in the administration of the PHQ-9, they then visit scribed above for eCAU. The PCP’s role and responsibilities each potential subject, either in their home or the village in COACH include screening and ongoing monitoring of clinic, and administer the PHQ-9. Those who screen posi- blood pressure and depressive symptoms with the PHQ-9, tive are invited to meet with the research assistant (RA) management of depression and HTN using standardized who introduces him or her to the study, obtains written in- practice guidelines in collaboration with other COACH formed consent, and conducts the baseline research assess- team members, and education of the patient, family and ment. RAs return to the village 3, 6, 9, and 12 months later community about these chronic conditions. Chen et al. BMC Geriatrics (2018) 18:124 Page 4 of 9 Aging Workers (AW) (c) social environment as risk and protective factors; (d) A staff member of either the village’sWomen Federation conducting psychosocial assessment and developing a care or Aging Association is designated as the AW for the plan; (e) providing psychoeducation to subjects and their COACH team. S/he typically has a middle-school educa- families; (f) techniques to reinforce treatment adherence tion, knows the residents well, and is well connected with and healthy lifestyles; (g) communication and problem- the village leadership. Although the AWs have no special solving skills with older adults and their families; (e) eth- training in social work, they receive in-service training ical standards including confidentiality. from the Bureau of Civil Affairs and have experience in Finally, PCPs, AWs and Psychiatrists come together addressing the villagers’ social needs. Their responsibilities for 1 day to learn how to collaborate in care manage- for the COACH intervention include baseline and ment. They are taught about the roles of the other team ongoing assessment of social stressors and supports, members and review procedures guiding their communi- fostering social connectedness, behavioral activation, cations and information transfer. Instruction techniques adherence support, facilitating communications with the include structured didactics, role-plays, and process/ PCP, and education of the patient, family and community problem solving sessions exploring potential barriers to about depression and HTN and their management. communication and collaboration. PCPs, AWs, and Psychiatrists in the COACH study arm Psychiatrist and PCPs in eCAU villages receive a small salary stipend Each COACH team is linked with one of the psychia- from the Departments of Health of Tonglu and Jiande trists based at the Tonglu and Jiande County Mental counties for their added effort in support of the study. Hospital. These two hospitals have a full array of in- patient (40 and 42 beds respectively) and outpatient ser- Procedures guiding team implementation of COACH vices. Psychiatrists have on average 5 years of medical With initiation of the study the PCP meets with each sub- training after high school followed by 2 years of specialty ject to review management of their BP. The Psychiatrist training. Their scope of practice includes psychiatric travels to the village and conducts the assessment to estab- diagnosis and management of patients of all ages with lish the depression diagnosis in the subject’shomeorthe mental illness from across the county. Their role in village clinic according to the subject’s preference. S/he COACH is to provide baseline diagnosis and initial pre- then consults with the PCP and, if medications are indi- scription of depression treatment, and to provide on- cated, initiates treatment with an antidepressant according going consultation to the team in its algorithm-driven to the study’s treatment guidelines. The AW visits the sub- management of the patient’s illness. ject in his or her home to conduct a systematic assessment of the person’s functional status, social supports, lifestyle, Training requirements and procedures medication use, nutritional and financial status. The AW Following randomization, PCPs and AWs of villages and PCP then meet, review the findings of the AWs in- assigned to implement COACH undergo training at the home assessment and the patient’s physical and mental County Mental Hospital along with the Psychiatrist con- health, and construct a care plan (problem list, approach sultant. Each member of the COACH team requires to each problem, responsible person, target date for com- training in their individual role, and all require training pletion) that addresses social, physical, and mental health in how to work together collaboratively. needs in a coordinated fashion. The intervention as outlined in the curriculum for the The PCP continues to meet with the patient at the inter- PCPs includes four major components: (a) depression man- vals designated by the HTN and depression practice guide- agement guidelines adapted from the Duke Somatic Treat- lines, monitoring progress with repeated PHQ-9 and BP ment Algorithm for Geriatric Depression (STAGED)  measures and adjusting treatment as indicated. The AW and applied in our prior study of depression care manage- continues to work with the subject to address identified ment in urban Hangzhou clinics ; (b) hypertension problems; educate the subject and family about their management using the guidelines adopted by the primary illnesses and support adherence to depression and hyper- care system in China ; (c) case management using the tension treatment regimens; encourage good health behav- toolkit adapted from the MacArthur Initiative on Depres- iors (diet, exercise, smoking cessation) based on established sion in Primary Care [21, 29]; and (d) psychoeducation (PE) psychoeducation methods; and connect the patient to and communication skills. PCP training is administered others and to the community (e.g., engage in social groups, over 4 days. There are no restrictions on concomitant care visits with friends and family). These activities can be or treatment, through working with individuals (e.g., home visits with the The AW curriculum, administered over 3 days, includes: subject and family), small groups (e.g., cooking demonstra- (a) an overview of depression, HTN, and their relation- tions, Tai Chi class), and community events (e.g., commu- ship; (b) self-management of depression and hypertension; nal meals, games.) AW outreach to the subject continues Chen et al. BMC Geriatrics (2018) 18:124 Page 5 of 9 weekly for 2 months, then biweekly for 2 months, then alcohol misuse, which are exclusion criteria. If a ques- monthly. If the patient continues to require AW support, tion of suicide risk is raised, the study’s safety protocol is more frequent visits are conducted. implemented and a decision made whether the subject The PCP and AW are instructed to meet weekly to re- should be excluded from the study. view their shared caseload. Each patient is very briefly Our primary outcome variables are adherence to anti- discussed and the treatment plan updated as indicated. depressant and antihypertensive medication recommenda- The Psychiatrist is available by telephone to make add- tions, depression symptom severity, and BP control. itional suggestions regarding depression management, Adherence is measured both by self-report (a single ques- and may refer the patient to the County Mental Hospital tion inquiring whether the subject has missed any doses as needed to assure safety. of her/his medications in the past) and by construction of a Medication Possession Ratio (MPR) from prescriptions Study measures and procedures written and filled, based on the patient’s EMR. The MPR Study measures are administered in face-to-face interviews then is calculated as the ratio of the total days of medica- conducted in the subject’s home or the clinic by unblinded, tion supplied to total days in a period , with values trained research assistants at study entry and 3-, 6-, 9-, and ≥80% considered medication adherent . The MPR is 12-months later. Data will be transferred from hard copy an accepted metric for the evaluation of adherence using formsby duplicateentry to theelectronicdata baseand retrospective data, including in China . maintained in secure servers at the Zhejiang University. The measure of depressive symptom change will be the Reasons for dropout and any occurence that could consti- Hamilton Depression Rating Scale – 17 item version tute an adverse event will be documented by the research (HDRS), which is widely used, reliable and valid in Chin- interviewers, examined by the study investigators, and re- ese . Consistent with other geriatric depression studies ported as indicated to the study’s institutional review  and our own work , we define response to treat- boards. As well, they will be assessed in semi-annual meet- ment as a change in the HDRS score > 50% from baseline, ings of the study’s Data and Safety Monitoring Board and remission as a follow-up HDRS score lower than 7. (DSMB) that is composed of Chinese experts independent BP control is established at follow up using guidelines from the sponsor and competing interests. specified by the Hypertension Treatment Guidelines [21, The measures we obtain fall into three domains: socio-- 22]. BP is taken according to the standards developed by demographic and baseline characteristics, outcome vari- the Center for Disease Control and Prevention in China ables, and covariates that may independently influence (auscultation with cuff deflation method after seated adherence and treatment response. The timing of their quietly for 5 min in the proper position; no caffeine, ex- administration is depicted in Table 1. ercise or smoking in preceding 30 min; appropriately Socio-demographic characteristics include age, sex, sized cuff; average of three measurements is recorded.) education and literacy level, and marital status. During For analyses, uncontrolled HTN is defined as systolic the baseline interview we use the Mini-International BP ≥ 130 or diastolic BP ≥ 80 for patients with diabetes Neuropsychiatric Interview (MINI)  to assess for the mellitus, coronary heart disease or renal disease, and presence in the last 6 months of mania, psychosis, and BP ≥ 140 or diastolic BP ≥ 90 for all others [21, 22]. Table 1 Timeline of measures Measure AIM Baseline 1 months 3 months 6 months 9 months 12 months HDRS, PHQ-9 x x x x x x BP x xxx xx MPR x x WHOQOL-BREF x x x BMI x x CCI x x ADL, IADL x x x MOS-SSS-C, SNS x x x CSQ-8 x x PS x x Cost data (EMR) x x Notes: ADL activities of daily living, BMI body mass index, BP blood pressure, CCI Charlson Comorbidity Index, CSQ-8 Client Satisfaction Questionnaire 8-item, EMR electronic medical record, HDRS Hamilton Depression Rating Scale, IADL instrumental activities of daily living, MOS-SSS-C, Medical Outcomes Study Social Support Survey, MPR medication possession ratio, SNS Social Network Size, PS Perceived Stigma Scale, WHOQOL-BREF WHO Quality of Life-BREF Chen et al. BMC Geriatrics (2018) 18:124 Page 6 of 9 Quality of life is measured using the WHOQOL-BREF Additional assumptions on which selection of the esti- [37, 38], a 26 item scale that yields four domain scores - mated sample of 2400 from villages randomized by village physical, psychological, social and environment - each of to the COACH or eCAU conditions was based include an which we will examine separately in analyses. The attrition rate of 20% over 12 months, informed by our pre- WHOQOL-BREF is a widely used measure the Chinese vious studies in Hangzhou. Since the village-randomized version of which has shown very good psychometric prop- study is a 3-level nested longitudinal design, power de- erties [39–41]. pends on the intra-class correlation (ICC) among the pa- In exploratory analyses we will estimate the costs associ- tients within the PCP and serial correlation between ated with two components of the intervention: (a) the in- repeated assessments within the patient. We set the serial cremental costs of adding COACH resources to eCAU correlation at 0.5, but varied the ICC over 0.05, 0.1 and 0. (programmatic costs), and (b) medical costs attributable to 2 to get a sense of the impact of the latter on power. Based the care of the subjects in each arm. Programmatic costs on two-sided type I error = 0.05, power = 0.8 and an attri- will include expenses associated with training, travel for tion rate of 20%, the detectable effect size ranged from 0. consultant psychiatrists between Tonglu/Jiande and the vil- 17 to 0.26 for a continuous outcome. Since experiences lages for the intake assessment, staff time for intervention with multi-level designs suggest that ICC is generally team meetings, and any information system costs. Medical smaller than 0.2, the study is sufficiently powered to detect costs include costs of medical treatment documented by small effect sizes between the two intervention groups. the subjects’ EMR and Zhejiang Province insurance data For the dichotomous outcomes, we also set the serial cor- that document each office visit, drug prescription, labora- relation at 0.5 and varied the percent of variance between tory test performed, or hospital stay for medical or mental patients within the PCP over 0.05, 0.1 and 0.2. Based on health reasons. As well, we will include patient out-of- two-sided type I error = 0.05, power = 0.8, base rate 0.5 pocket healthcare costs in the preceding 3-month interval (most conservative) and an attrition rate of 20%, the de- obtained by subject interview at each research assessment tectable between-group proportion ranged from 11 to point. Cost data will be converted to US dollars. 17%, which is well within the range of clinically meaning- Additional variables for use as covariates in analyses in- ful differences in the primary care setting. clude factors known to influence adherence, treatment re- sponse or both, include the Charlson Comorbidity Index Statistical analytic approach  adapted for use with ICD-9 codes  and supple- Descriptive statistics (counts and proportions for categor- mented by questions about common disabling conditions ical variables and means [± SD] for continuous outcomes) of late life including diabetes, high cholesterol, heart dis- will be used to depict the characteristics of the sample (e.g., ease, and stroke; and the Body Mass Index (BMI), which age, gender, physical and functional status). We will com- is associated with hypertension treatment response. As a pare baseline characteristics between groups using t-test measure of functioning independent of the WHOQOL- (for continuous variables) and chi-square (for discrete BREF physical domain score, we measure the subject’sim- variables), and examine associations between outcome pairment in basic (ADL) and instrumental activities of variables and patients’ characteristics. Characteristics daily living (IADLs)  and cognitive function with the significantly differentiating the two groups will be treated Six-Item Screener (SIS) [45, 46]. We measure social sup- as covariates when testing between-group differences using port with the Chinese version of the Medical Outcomes longitudinal models and structural equation models (SEM). Study Social Support Survey (MOS-SSS-C)  and social network size (total number of persons with whom the Aim 1 – Depression respondent has discussed important matters in the past We hypothesize that relative to subjects who receive eCAU, 6months) . Participants will be queried about side ef- those in the COACH intervention will show (a) better ad- fects of both their antidepressant and antihypertensive herence to antidepressant treatment recommendations and medications using the Antidepressant Side-Effect Check- (b) greater improvements in their depressive disorders over list  and the Side-Effects and Symptoms Distress 12 months of involvement in the study. We will model each Checklist  respectively. repeatedly assessed variable of adherence to depression treatment as well as improvement in depression using gen- Sample size estimate eralized linear mixed effect models (GLMM) and weighted Primary outcomes of the COACH Study on which sam- generalized estimating equations (WGEE), two common ple size estimates were based are (1) adherence to approaches for modeling treatment differences over time depression and HTN treatment (MPR and % of subjects . Both approaches provide valid inference in the pres- with 80% MPR); (2) depression symptom change (pro- ence of missing data if the missing value follows the missing portion with 50% reduction in HDRS and % with HDRS at random (MAR) assumption, a popular mechanism that < 10); and (3) HTN control (% with BP controlled). applies to most studies in practice . If estimates differ Chen et al. BMC Geriatrics (2018) 18:124 Page 7 of 9 between the two approaches, only WGEE results will be re- than eCAU. We will use the same approach as in Aim 1 ported, as it provides significantly more robust inference, to test the hypothesis. The longitudinal model will be especially in the presence of missing data [51, 52]. We will applied to each of the WHOQOL-BREF domain scores. perform intention-to-treat analyses using all subjects ran- domized to the treatment groups. Aim 5 – Cost For each outcome, time and intervention will be predic- Finally, we will determine resource utilization and costs tors, adjusting for covariates. We will also assess the poten- associated with delivering the eCAU and the COACH tial interaction between time and intervention. To avoid interventions over 12 months and, in a set of exploratory oversimplifying temporal patterns with linear trend, we will analyses, conduct an economic feasibility evaluation use piece-wise or even polynomial functions of time based using cost-effectiveness analysis (CEA) and cost-benefit on the assessment points. Linear contrasts will be used to analysis (CBA) methods. We consider them exploratory assess COACH vs. eCAU differences over the 12-month because the 12-month duration of the intervention is period as well as any sub-intervals within this period. relatively short for such analyses. For CEA we will calculate an incremental cost-effective- Aim 2 – Hypertension ness ratio (ICER) based on two outcomes: (a) reduction in We hypothesize that relative to subjects who receive the average HDRS score between the eCAU and COACH eCAU, those in the COACH intervention will show (a) subjects (ΔEffect = hypertension [eCAU] – hypertension better adherence to antihypertensive treatment recom- [Coach]); and (b) Increase in the proportion of patients mendations and (b) greater improvements in BP control. with controlled BP: ΔEffect = cBP(eCAU) – cBP(Coach). The same approach as Aim 1 will be used to examine The ICER will be interpreted as the incremental cost per the hypotheses in this Aim. unit improvement in health as measured by the outcome of choice (HDRS score or % of patients with BP control). Aim 3 – Temporal associations For CBA, the benefit of the intervention will be estimated We will examine the temporal associations of change in as the reduction in the average healthcare expenses depression and BP control, hypothesizing that (a) im- between the patients receiving eCAU and COACH provements in treatment adherence precede improve- (ΔEffect = ΔBenefit = MedCost[CAU] – MedCost[Coach]). ment in depression and hypertension, and (b) The results of cost-benefit analysis could be interpreted as improvements in depression will precede improvements the return on investment (ROI; the ratio of reduction in in BP control. The dynamic relationships between adher- medical costs to the incremental cost of the intervention, ences, improvement in depression and BP control and with ratio > 1 indicating a significant return) or as a net mediation of the intervention effect on BP control by monetary benefit (NMB = the difference between the improvement in depression will be examined using reduction in medical costs and the incremental cost of the SEM. We will first examine the causal, or meditational, intervention, NMB > 0 indicating a monetary benefit of the relationship between the COACH intervention, adher- intervention). ence to antidepressant (antihypertensive) medications and improvement in depression (BP control), using SEM Discussion with intervention as the predictor, adherence as the me- The COACH study uses a cluster randomized controlled diator and depression (BP control) as the outcome. To trial design to compare an integrated, primary care- ensure the temporal order that changes in adherence based depression care management approach to eCAU take place before improvements in depression (BP con- for the treatment of comorbid depression and HTN in trol), the depression (BP control) variable will lag by one older adult Chinese rural village residents. Building on a visit with respect to the adherence variable. We will con- large body of research that has established the effective- struct direct, indirect, and total effects, and perform ness of integrated care models for depression and other tests to see if adherence to antidepressant (antihyperten- common mental disorders in Western countries, the sive) medications mediates the effect of the intervention COACH study is one of very few to apply the model in on improvement in depression (BP control). a low/middle income country (LMIC) in which mental The same approach will be used to test the mediation health resources are scarce and the need is enormous hypothesis in (b), where depression is the mediator and and growing rapidly. With 150 million adults over age BP control is the outcome. 65 currently, China’s older adult population is projected to grow to over 400 billion by 2050, even as the relative Aim 4 – Health-related quality of life proportion of the working aged population decreases. In addition to its impact on depression and BP control, Rapid changes are underway in family structure with mi- we will ask whether the COACH intervention results in gration of younger and middle aged adults from rural to greater improvements in health-related quality of life urban centers for work, leaving their older adult parents Chen et al. BMC Geriatrics (2018) 18:124 Page 8 of 9 behind in rural villages where the prevalence of depres- item screener; STAGED: Somatic treatment algorithm for geriatric depression; VAA: Village aging associations; WHOQOL-BREF: WHO Quality of Life-BREF sion is high and treatment is rarely accessible. Indeed, this situation is common in the developing world. Acknowledgements Research is required on means by which to give primary We thank the village residents and health care providers who participated in the study, research staff members who travelled so far and collected the care clinics the tools with which to address the mental study data so well, and the US National institutes of Health for support in health needs of the aged population in rural areas conducting this research. efficiently and effectively. Funding Although our previous study of depression care man- Following rigorous peer review the COACH study was supported by Grant agement in urban Chinese primary care clinics produced Number 1R01MH100298 to Drs. Conwell and Chen from the U.S. National promising results for treatment of late life depression, it Institutes of Health (NIH) that established collaborations in late life mental health research between investigators at the University of Rochester, was limited in a number of respects that the COACH University of Michigan, University of Pennsylvania, Zhejiang University, and study will attempt to address. COACH is implemented government ministries in China (Zhejiang Center for Disease Control and in rural villages, a necessary innovation given that 70% Prevention [CDC] and Zhejiang Provincial Committee on Aging [CoA]). The content is solely the responsibility of the authors and does not necessarily of the older population lives in rural China. As well, represent the official views of the NIH. COACH incorporates a psychosocial dimension to care with the AW that was lacking in previous work, couples Authors’ contributions SC, YC, LL, HRB, WT, and HD contributed to the design of the study. YC, SC, depression with care management of HTN to test the LL, JX and WT authored the manuscript, and all authors read and approved model’s impacts on commonly comorbid chronic condi- the final version prior to submission. tions, and will contribute to understanding the role that Ethics approval and consent to participate adherence plays in chronic disease management. This protocol received ethics approval from the study’s funding body, the Successful dissemination of the COACH model to U.S. National Institutes of Health (NIH) Institutional Review Board (IRB). As other counties and provinces in China, and to other well, the study protocol was reviewed and approved by the Research Subjects Review Board (RSRB) at University of Rochester, the Health Sciences rural LMIC settings, will require that it is more effective REB at the Zhejiang University, the Health Sciences IRB at the University of than eCAU in reducing depressive symptoms, improving Michigan, and the Health Sciences IRB at the University of Pennsylvania. quality of life, and improving BP control. It will require Subjects all will provide written informed consent to participate in the study. that the delivery model be acceptable to multiple stake- Competing interests holders in diverse sociocultural contexts, including the The authors declare that they have no competing interests. patients and their families, PCPs and other team mem- bers, village leaders and those who administer the health Publisher’sNote and aging services systems. To that end, COACH is Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. designed to leverage indigenous village resources to contain costs and to meet the needs of village residents Author details and providers in a manner that is culturally congruent. Department of Psychology, Zhejiang University, Hangzhou, China. Department of Psychiatry, University of Rochester Medical Center, 300 Successful dissemination will also require that the model Crittenden Boulevard, Rochester, NY 14642, USA. School of Social Work, be affordable, for which this study will provide some pre- 4 University of Michigan, Ann Arbor, USA. Department of Global Biostatistics liminary evidence. Nonetheless, additional modifications and Data Science, Tulane University, New Orleans, USA. Department of Family Medicine, University of Pennsylvania, Philadelphia, USA. Faculty of will likely be needed, for example, by linking the COACH Public Health Research, Zhejiang University, Hangzhou, China. teams to psychiatric supports using telehealth technology, and telementoring using the Project ECHO model to Received: 24 April 2018 Accepted: 4 May 2018 build the teams’ expertise over time, reducing the need for ongoing support and improving cost efficiency. With References these additional steps, Zhejiang Provincial officials may 1. Ma X, Xiang YT, Li SR, Xiang YQ, Guo HL, Hou YZ, et al. Prevalence and sociodemographic correlates of depression in an elderly population living support further testing on a larger scale. with family members in Beijing, China. Psycholog Med. 2008;38(12):1723–30. 2. Chen R, Wei L, Hu Z, Qin X, Copeland JR, Hemingway H. Depression in older Abbreviations people in rural China. Arch Intern Med. 2005;165:2019–25. ADL: Activities of daily living; AW: Aging worker; BMI: Body mass index; 3. Xue J, Chen S, Bogner HR, Tang W, Li L, Conwell Y. The prevalence of BP: Blood pressure; CBA: Cost-benefit analysis; CDC: Center for disease depressive symptoms among older patients with hypertension in rural control and prevention; CEA: Cost-effectiveness analysis; COACH: Chinese China. Int J Geriatr Psychiatry. 2017;32:1411–7. Older Adult Collaborations in Health; CSQ-8: Client satisfaction questionnaire 4. Blazer DG. Depression in late life: review and commentary. J Gerontol Biol 8-item; CVD: Cardiovascular disease; DM: Diabetes mellitus; eCAU: Enhanced Med Sci. 2003;58:249–65. Care-as-Usual; EMR: Electronic medical record; HDRS: Hamilton depression 5. Prince MJ, Ebrahim S, Acosta D, Ferri CP, Guerra M, Huang Y, et al. rating scale; IADL: Instrumental activities of daily living; ICER: Incremental Hypertension prevalence, awareness, treatment and control among older cost-effectiveness ratio; LMICs: Low-and middle-income countries; MINI: Mini people in Latin America, India and China: a 10/66 cross-sectional International Neuropsychiatric Interview; MOS-SSS-C: Medical outcomes study population-based survey. J Hypertension. 2012;30:177–87. social support survey; MPR: Medication possession ratio; PCP: Primary care 6. Ezzati M, Lopez AD, Rodgers A, Murray CJL. Comparative quantification of provider; PHQ-9: Patient Health Questionnaire-9; RA: Research assistant; health risks: global and regional burden of disease attributable to selected RCT: Randomized controlled trial; SEM: Structural equation models; SIS: Six- major risk factors OMS, vol. 2; 2004. (xxiv,2234) Chen et al. BMC Geriatrics (2018) 18:124 Page 9 of 9 7. Murray CJL, Lopez AD. Alternative projections of mortality and disability by cause 33. Wong M, Jiang JY, Griffiths SM. Antihypertensive drug adherence among 1990–2020: global burden of disease study. Lancet. 1997;349(9064):1498–504. 6408 Chinese patients on angiotensin-converting enzyme inhibitors in 8. Meng L, Chen D, Yang Y, Zheng Y, Hui R. Depression increases the risk of Hong Kong: a cohort study. J Clin Pharmacol. 2010;50:598–605. hypertension incidence: a meta-analysis of prospective cohort studies. J 34. Zheng YP, Zhao JP, Phillips M, Liu JB, Cai MF, Sun SQ, et al. Validity and Hypertension. 2012;30:842–51. reliability of the Chinese Hamilton Depression Rating Scale. Br J Psychiatry. 9. McConnell S, Jacka FN, Williams LJ, Dodd S, Berk M. The relationship between 1988;152:660–4. depression and cardiovascular disease. Int J Psychiatry Clin Pract. 2005;9:157–67. 35. Lecrubier Y. How do you define remission? Acta Psychiatr Scand. 2002; 106(s415):7–11. 10. Zhang J, Ye M, Huang H, Li L, Yang A. Depression of chronic medical 36. Bogner HR, Cary MS, Bruce ML, Reynolds CF III, Mulsant B, Ten HT, et al. The inpatients in China. Arch Psychiatr Nurs. 2008;22:39–49. role of medical comorbidity in outcome of major depression in primary 11. Huang J. Recognition and intervention of depression symptoms in in- care: the PROSPECT study. Am J Geriatr Psychiatry. 2005;13:861–8. patients. Mod Nurs (China). 2004;10:295–6. 37. Harper A, Power M. Development of the World Health Organization 12. Katon W, Von Korff M, Lin E, Walker E, Simon GE, Bush T, et al. Collaborative WHOQOL-BREF quality of life assessment. Psychol Med. 1998;28:551–8. management to achieve treatment guidelines: impact on depression in 38. Skevington SM, Lotfy M, O'Connell KA. The World Health Organization's primary care. JAMA. 1995;273:1026–31. WHOQOL-BREF quality of life assessment: psychometric properties and 13. Gilbody S, Whitty P, Grimshaw J, Thomas R. Educational and organizational results of the international field trial. A report from the WHOQOL group. interventions to improve the management of depression in primary care: a Qual Life Res. 2004;13:299–310. systematic review. JAMA. 2003;289:3145–51. 39. Leung KF, Wong WW, Tay MSM, Chu MML, Ng SSW. Development and 14. Unützer J, Katon W, Callahan CM, Williams JW Jr, Hunkeler E, Harpole L, et validation of the interview version of the Hong Kong Chinese WHOQOL- al. Collaborative care management of late-life depression in the primary BREF. Qual Life Res. 2005;14:1413–9. care setting: a randomized controlled trial. JAMA. 2002;288:2836–45. 40. Xia P, Li N, Hau K-T, Liu C, Lu Y. Quality of life of Chinese urban community 15. Hirschfeld RM. The comorbidity of major depression and anxiety disorders: residents: a psychometric study of the mainland Chinese version of the recognition and management in primary care. Prim Care Companion J Clin WHOQOL-BREF. BMC Med Res Methodol. 2012;12:37. https://doi.org/10. Psychiatry. 2001;3(Suppl):244. 1186/1471-2288-12-37 16. Wagner EH, Austin BT, Von Korff M. Organizing care for patients with 41. Yang S-C, Kuo P-W, Su S, Wang J-D, Lin M-l. Development and chronic illness. Milbank Q. 1996;74:511–44. psychometric properties of the dialysis module of the WHOQOL-BREF 17. Katon W, Unützer J, Wells K, Jones L. Collaborative depression care: history, Taiwan version. J Formos Med Assoc. 2006;105:299–309. evolution and ways to enhance dissemination and sustainability. Gen Hosp 42. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying Psychiatry. 2010;32:456–64. prognostic comorbidity in longitudinal studies: development and validation. 18. Chen S, Conwell Y, He J, Lu N, Wu J. Depression care management for J Chronic Dis. 1987;40:373–83. adults older than 60 years in primary care clinics in urban China: a cluster- 43. Chu Y-T, Ng Y-Y, Wu S-C. Comparison of different comorbidity measures for randomised trial. Lancet Psychiatry. 2015;2:332–9. use with administrative data in predicting short-and long-term mortality. 19. Tsang HW, Cheung L, Lak DC. Qigong as a psychosocial intervention for BMC Health Serv Res. 2010;10:140. https://doi.org/10.1186/1472-6963-10-140 depressed elderly with chronic physical illnesses. Int J Geriatr Psychiatry. 44. Tong AYC, Man DWK. The validation of the Hong Kong Chinese version of 2002;17:1146–54. the Lawton instrumental activities of daily living scale for institutionalized 20. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression elderly persons. OTJR. 2002;22:132–42. severity measure. J Gen Intern Med. 2001;16:606–13. 45. Callahan CM, Unverzagt FW, Hui SL, Perkins AJ, Hendrie HC. Six-item 21. Liu LS. 2010 Chinese guidelines for the management of hypertension. screener to identify cognitive impairment among potential subjects for Zhonghua Xin Xue Guan Bing Za Zhi. 2011;39:579. clinical research. Med Care. 2002;40:771–81. 22. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL, et al. Seventh 46. Chen MR, Guo QH, Cao XY, Hong Z, Liu XH. A preliminary study of the Six- report of the joint national committee on prevention, detection, evaluation, and Item Screener in detecting cognitive impairment. Neurosci Bull. 2010;26(4): treatment of high blood pressure. Hypertension. 2003;42:1206–52. 317–21. eng 23. Krahn DD, Bartels SJ, Coakley E, Oslin DW, Chen H, McIntyre J, et al. PRISM-E: 47. Yu DSF, Lee DTF, Woo J. Psychometric testing of the Chinese version of the comparison of integrated care and enhanced specialty referral models in Medical Outcomes Study Social Support Survey (MOS-SSS-C). Res Nurs depression outcomes. Psychiatr Serv. 2006;57:946–53. Health. 2004;27:135–43. 24. Schulberg HC, Block MR, Madonia MJ, Scott CP, Lave JR, Rodriguez E, et al. 48. Ruan D, Freeman LC, Dai X, Pan Y, Zhang W. On the changing structure of The ‘usual care’ of major depression in primary care practice. Arch Fam social networks in urban China. Soc Networks. 1997;19:75–89. Med. 1997;6:334–9. 49. Uher R, Farmer A, Henigsberg N, Rietschel M, Mors O, Maier W, et al. 25. Katon W, von Korff M, Lin E, Bush T, Ormel J. Adequacy and duration of Adverse reactions to antidepressants. Br J Psychiatry. 2009;195:202–10. antidepressant treatment in primary care. Med Care. 1992;30:67–76. 50. Testa MA, Hollenberg NK, Anderson RB, Williams GH. Assessment of quality 26. Lin EH, Von KM, Katon W, Bush T, Simon GE, Walker E, et al. The role of the of life by patient and spouse during antihypertensive therapy with atenolol primary care physician in patients’ adherence to antidepressant therapy. and nifedipine gastrointestinal therapeutic system. Am J Hypertension. Med Care. 1995;33:67–74. 1991;4(Pt 1):363–73. 27. Callahan CM, Dittus RS, Tierney WM. Primary care physicians’ medical 51. Tang W, He H, Tu XM. Applied categorical and count data analysis CRC decision making for late-life depression. J Gen Intern Med. 1996;11:218–25. press; 2012. 28. Steffens DC, McQuoid DR, Krishnan KR. The Duke Somatic Treatment 52. Lu N, Tang W, He H, Yu Q, Crits-Christoph P, Zhang H, et al. On the impact Algorithm for geriatric Depression (STAGED) approach. Psychopharmacol of parametric assumptions and robust alternatives for longitudinal data Bull. 2002;36:58–68. analysis. Biom J. 2009;51:627–43. 29. Han C, Voils CI, Williams JW Jr. Uptake of web-based clinical resources from the MacArthur initiative on depression and primary care. Commun Ment Health J. 2013;49:166–71. 30. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al. The MINI-International Neuropsychiatric Interview (MINI): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry. 1998;59:22–33. 31. Halpern MT, Khan ZM, Schmier JK, Burnier M, Caro JJ, Cramer J, et al. Recommendations for evaluating compliance and persistence with hypertension therapy using retrospective data. Hypertension. 2006;47: 1039–48. 32. Mallion J-M, Baguet J-P, Siche J-P, Tremel F, De Gaudemaris R. Compliance, electronic monitoring and antihypertensive drugs. J Hypertension. 1998; 16(Suppl):S75–9.
– Springer Journals
Published: May 29, 2018