Implementation and quantitative evaluation of chronic disease self-management programme in Shanghai, China: randomized controlled trial.

Implementation and quantitative evaluation of chronic disease self-management programme in Shanghai, China: randomized controlled trial Fu Dongbo,1 Fu Hua,1 Patrick McGowan,2 Shen Yi-e,1 Zhu Lizhen,3 Yang Huiqin,4 Mao Jianguo,4 Zhu Shitai,3 Ding Yongming,1 Wei Zhihua5 Objective To evaluate the effectiveness of the Shanghai Chronic Disease Self-Management Program (CDSMP). Methods A randomized controlled trial with six-month follow-up compared patients who received treatment with those who did not receive treatment (waiting-list controls) in five urban communities in Shanghai, China. Participants in the treatment group received education from a lay-led CDSMP course and one copy of a help book immediately; those in the control group received the same education and book six months later. Findings In total, 954 volunteer patients with a medical record that confirmed a diagnosis of hypertension, heart disease, chronic lung disease, arthritis, stroke, or diabetes who lived in communities were assigned randomly to treatment (n = 526) and control (n = 428) groups. Overall, 430 (81.7%) and 349 (81.5%) patients in the treatment and control groups completed the six-month study. Patients who received treatment had significant improvements in weekly minutes of aerobic exercise, practice of cognitive symptom management, self-efficacy to manage own symptoms, and self-efficacy to manage own disease in general compared with controls. They also had significant improvements in eight indices of health status and, on average, fewer hospitalizations. Conclusion When implemented in Shanghai, the CDSMP was acceptable culturally to Chinese patients. The programme improved participants’ health behaviour, self-efficacy, and health status and reduced the number of hospitalizations six months after the course. The locally based delivery model was integrated into the routine of community government organizations and community health services. Chinese lay leaders taught the CDSMP courses as successfully as professionals. Keywords Chronic disease/therapy; Self care; Community health services; Hypertension/therapy; Coronary disease/therapy; Heart failure, Congestive/therapy; Lung diseases, Obstructive/therapy; Arthritis/therapy; Cerebrovascular accident/therapy; Diabetes mellitus/therapy; Cultural characteristics; Outcome assessment (Health care); Program evaluation; Randomized controlled trials; China (source: MeSH, NLM). Mots cle´s Maladie chronique/the´rapeutique; Auto-traitement; Service public sante´; Hypertension arte´rielle/the´rapeutique; Vaisseaux coronaires, Maladies/the´rapeutique; Insuffisance cardiaque/the´rapeutique; Bronchopneumopathies obstructives/the´rapeutique; Arthrite/the´rapeutique; Accident vasculaire ce´re´bral/the´rapeutique; Diabe`te/the´rapeutique; Moeurs; Evaluation re´sultats (Sante´); Evaluation programme; Essai clinique randomise´; Chine (source: MeSH, INSERM). Palabras clave Enfermedad cro´nica/terapia; Autocuidado; Servicios de salud comunitaria/Hipertensio´n/terapia; Coronariopatı´a/ terapia; Insuficiencia cardı´aca congestiva/terapia; Neumopatı´as obstructivas/terapia; Artritis/terapia; Accidente cerebrovascular/ terapia; Diabetes mellitus/terapia; Caracterı´sticas culturales; Evaluacio´n de resultado (Atencio´n de salud); Evaluacio´n de programas; Ensayos controlados aleatorios; China (fuente: DeCS, BIREME ). Bulletin of the World Health Organization 2003;81:174-182. Voir page 181 le re´sume´ en franc¸ais. En la pa´gina 182 figura un resumen en espan˜ol. Introduction Chronic diseases — mainly heart disease, stroke, cancer, and lung disease — are becoming leading causes of disability and premature death in China. Noncommunicable diseases accounted for 81% of total deaths in China in 1996 (1). Chronic disease is also a major health care cost for China. In 1994, direct treatment costs related to chronic disease were more than four hundred billion Yuan (2). In China, Shanghai has the heaviest burden of noncommunic- able diseases, because it has the largest population and the largest ageing population. Prevention and management of chronic disease is an urgent primary health problem to be addressed in Shanghai. The inability of current major health care systems to deal with chronic conditions calls for prompt policy action and new approaches to people with chronic conditions. Self-management for people with chronic disease is now widely recognized as a necessary part of treatment. 1 School of Public Health, Fudan University, 138 Yi Xue Yuan Road, Shanghai, 200032, People’s Republic of China. Correspondence should be addressed to Dr Fu Hua (email: hfu@shmu.edu.cn). 2 Institute of Health Promotion Research, University of British Columbia, Vancouver, Canada. 3 Xinle Community Health Center, Shanghai, China. 4 Pudong Meiyuan Community Health Center, Shanghai, China. 5 Caojiadu Community Health Center, Shanghai, China. Ref. No. 02-0197 Research 174 Bulletin of the World Health Organization 2003, 81 (3) Chronic disease self-management has been shown in the United States, Canada, and the UnitedKingdom to be useful in maintaining and improving patients’ health behaviour and health status, while lowering health care utilization through improved self-management skills, ‘‘self-efficacy’’, and better communication between patients and health providers (3–6). The community-based generic self-management programme developed by Lorig et al. at Stanford University Patient Education Research Center — the Chronic Disease Self- Management Program (CDSMP) — is the most widely accepted self-management patient education programme worldwide, because it is designed to meet the needs of patients who have more than one chronic condition and is taught by trained lay leaders (3). The CDSMP is a community-based patient self-management education course. Several assump- tions make this programme suitable for different kinds of patients (7): (1) people with chronic diseases have similar concerns and problems; (2) people with chronic conditions can learn to take responsibility for the day-to-day management of their disease(s) and the physical and emotional problems caused by their disease(s); (3) lay people with chronic conditions, when given a detailed leader’s manual, can teach the CDSMP as effectively, if not more effectively, than health professionals; and (4) the process or way the CDSMP is taught is as important, if not more important, than the subject matter taught. The content and process of teaching the CDSMP course are based on self-efficacy theory (8, 9). Self-efficacy refers to people’s beliefs in their abilities to perform specific behaviours, which is a key factor to behaviour change and health functioning (9). The courses incorporate strategies suggested by Bandura to enhance self-efficacy (10), including weekly action planning and feedback, modelling of behaviours and problem-solving by participants for one another, reinterpretation of symptoms, several different management techniques, group problem-solving, and individual decision- making. The leaders of the courses act more as facilitators than lecturers. The process is documented in a detailed protocol — the Chronic disease self-management leader’s manual (11). Topics covered include: exercise; use of cognitive symptom manage- ment techniques; nutrition; fatigue and sleep management; use of community resources; use of medications; dealing with the emotions of fear, anger, and depression; communication with others, including health professionals; problem-solving; and decision-making. The content of the course was published as Living a healthy life with chronic conditions (12), which was used as a textbook for course participants. To date, no research has tested whether CDSMP would be culturally acceptable for Chinese people and could benefit participants in China, or whether Chinese lay people could be trained to teach CDSMP courses as capably as professionals. This study examined the course’s feasibility, benefits, and cultural sensitivity in Shanghai. The study aimed to develop a Chinese culturally acceptable CDSMP, develop locally based sustainable CDSMP delivery models, and evaluate the effectiveness (changes in self-management behaviours, self-efficacy, health status, and health service utilization) of CDSMP for patients with at least one chronic condition through a randomized controlled trial in Shanghai. It also aimed to investigate whether trained lay leaders could teach CDSMP courses in China as effectively as professionals. Methods Patients Eligibility criteria were men and women aged 520 years, at least one of hypertension, heart disease (coronary heart disease or congestive heart disease), chronic lung disease (asthma, chronic bronchitis, or emphysema), arthritis, stroke (com- pleted cerebrovascular accident with neurological handicap and normal mentation), or diabetes confirmed by medical record, and community dwelling. Patients with other condi- tions as well as at least one of those described above were also included. Patients with compromised mentation, cancer patients who had received chemotherapy or radiation within the past year, patients for whom problems could be expected with compliance or follow-up, patients who had participated in another trial or study during the past 30 days, stroke patients with severe physical disability (that would interfere with them taking the CDSMP course), and patients aged <20 years were excluded. The community health centre and general practi- tioners were not informed of the participants’ study status (treatment or control). Recruitment and randomization From June 1999 to February 2000, participants were recruited through public service announcements in the mass media, posters at community senior centres (activity centres for the community-dwelling elderly people), referrals from flyers left in community clinics, and interpersonal persuasion. The programme was run at multiple community sites in five communities in Shanghai, China, to ensure that it was easily accessible to patients. Randomization was conducted at each site. After the patient’s eligibility and willingness to participate in the study were confirmed, each participant completed a baseline questionnaire. After informed consent was obtained, participants at each site were randomized into ‘‘treatment groups’’ (spring course) and ‘‘control groups’’ (autumn course), according to a random-number table with a randomization ratio designed to yield no fewer than ten and no more than 15 participants in a group. Neither the participants nor the data collectors were blind to the treatment assignment. As a result, 526 participants were placed in treatment groups and 428 in waiting-list control groups. Treatment group members received the CDSMP course immediately. Control group members received the CDSMP course six months after the treatment groups. Fig. 1 shows the trial profile. Shanghai Chronic Disease Self-Management Program The parts of Chronic disease self-management leader’s manual and Living a healthy life with chronic conditions culturally unsuitable for the Chinese population were deleted or replaced. For example, because Chinese patients normally feel uncomfortable dis- cussing issues of ‘‘death’’ and no ‘‘advance directives for health care’’ in China, the topic ‘‘making your wishes known: advance directives for health care’’ was deleted.We replaced some parts of ‘‘exercise for flexibility and strength’’ with physical activities that the Chinese people are familiar with and would feel comfortable doing, such as shadowboxing (Taiji) and eighteen exercises (a common Chinese exercise for improving joint flexibility and strength). In addition, we added some Chinese relaxation exercises and omitted some suggestions in ‘‘nearly 200 helpful hints’’, such as ‘‘going out by driving a new car’’, 175Bulletin of the World Health Organization 2003, 81 (3) Chronic disease self-management programme in China since such situations are not common in China. The materials then were translated and modified repeatedly after being pre- tested and reviewed by 20 patients with different common chronic diseases. The final Chinese versions of the two materials were written with common Chinese words and sentences. A community-based hypertension self-manage- ment group with ten volunteer patients with hypertension was established to pilot the content of the Shanghai CDSMP course and the process (the way in which the course was taught) (13). The experience and suggestions from the pilot study — for example that lay leaders should not be aged 575 years — contributed to the later implementation of Shanghai CDSMP. The Shanghai CDSMP included seven 2–2.5-hours sessions scheduled on consecutive weeks. The course was conducted in groups by two trained volunteer leaders who worked in pairs in community settings such as senior centres and community hospitals or in the patient’s house; then followed the Chronic disease self-management leader’s manual. One or both leaders were non-health professionals with a chronic disease themselves. Only 11 trained course leaders were health professionals in this study. Course participants took the CDMSP course and received a copy of the help book Living a healthy life with chronic conditions — the intervention group immediately and the control group six months later. The implementation of CDSMP in Shanghai According to participatory research principles (14), and considering the different local features of the five commu- 176 Bulletin of the World Health Organization 2003, 81 (3) Research nities, the Shanghai CDSMP developed two delivery models. The first was the ‘‘commonly participatory model’’, in which the community government, community health centre, and researchers worked together to design, implement, and evaluate the programme bymeans of a community programme advisory committee. The CDSMP was integrated into the routine of community government organizations and com- munity health services. In the second ‘‘professional dominated model’’, the community government organizations gave only passive support to the programme. The CDSMP was not integrated into their routines, but the community health centre contributed more to the programme than it did in the first model. Some community health care providers trained to be course leaders and taught part of the course; the researchers did the same. In terms of the process of CDSMP delivery, the scale and efficiency of community mobilization and programme advertisement in the commonly participatory model were higher than in the professional dominated model. Two communities used the first model and registered 588 course participants and 74 volunteer lay leaders. Three communities used the secondmodel and registered only 366 participants and 57 volunteer lay leaders. The total population in each of the five communities was nearly 50 000. The five communities were all located at the centre of Shanghai city. Sample size Previous similar research showed that the effect size of CDSMP in six months is small (3). Given that 0.20 represents a small effect size, power computations with Power Chart software (15) for a balanced design indicated that 395 parti- cipants in each group would be sufficient to achieve 80% power with two-tailed tests and a = 0.05. To assure 10– 15 patients were included in the treatment group, the overall ratio of treatment-to-control patients was not 1:1. Based on 80% power to detect a significant difference (P = 0.05, two- sided), with an assumed overall ratio of 5:4 and a loss to follow up of 20%, 534 patients were needed in the treatment group and 428 in the control group. Outcome measures and statistical analysis Outcomes were measured with a pre-tested self-administered Chinese version of the questionnnaire on chronic disease self- management studymeasures developed by Lorig et al. (7). This includes four primary classifications of outcome variables (health behaviours, health status, self-efficacy, and health service utilization), which enable a description of how well people are managing their chronic disease. The four categories of 20 outcome variables are: (1) self-management behaviour change score (exercise, practice of cognitive symptom management, and communication with doctor); (2) self- efficacy score (self-efficacy to manage disease in general and self-efficacy to manage symptoms); (3) health status (self-rated health, health distress, shortness of breath, pain, disability, depression, energy and fatigue, and social and role activity limitations); and (4) health service utilization (visits to physicians, visits to emergency departments, number of hospital stays, and nights spent in hospital). The questionnaire was completed by participants at baseline and after six months. To minimize acquiescent response, data collection was separated completely from the intervention and was per- formed by people who did not know the patients or their treatment status. The Mann-Whitney U-test was used to compare the baseline status of the treatment and control groups. Analysis of covariance was used to compare changes in the 20 outcome variables at six months between the treatment and control groups. The analysis controlled for the baseline value of the study variables that differed between the groups at baseline: age, sex, education, marital status, follow-up time, baseline number of minutes per week of stretching and strengthening exercise, cognitive symptom management practice, communication with medical doctor, and disability. The Mann-Whitney U-test also was used to detect differences in outcomes for those taught by professionals as opposed to those taught by lay leaders. Results Baseline characteristics Of 954 subjects who entered the study, 486 patients in the treatment groups and 418 in the control group had valid baseline data.Only 430 (81.7%) patients in the treatment group and 349 (81.5%) in the control group completed the six-month study. A comparison of baseline data showed that 779 parti- cipants who completed the six-month study had significantly worse (higher scores for) shortness of breath and disability than those who did not complete (P<0.05). Table 1 gives the demographics and disease characteristics of the participants Table 1. Patient characteristics. Values are numbers (percen- tages) unless otherwise specified Characteristic Treatment Control (n = 430) (n = 349) Mean+SDa age (years) 64.21+9.77 63.80+10.30 Age range 29.2–89.8 22.1–88.9 Sex Female 315 (73.3) 241 (69.1) Male 115 (26.7) 108 (30.9) Ethnicity Han 413 (96.0) 342 (98.0) Zhuang 9 (2.1) 5 (1.4) Miao 5 (1.2) 1 (0.3) Hui 3 (0.7) 1 (0.3) Mean+SD education (years) 9.48+3.89 9.88+3.88 Marital status Married 354 (82.3) 277 (79.4) Separated 8 (1.9) 4 (1.1) Widowed 60 (13.9) 60 (17.2) Single 6 (1.4) 6 (1.7) Divorced 2 (0.5) 2 (0.6) Mean number+SD of diseases 2.09+1.22 1.95+1.09 Disease prevalence Hypertension 223 (51.9) 208 (59.6) Heart disease 146 (34.0) 117 (33.5) Arthritis 137 (31.9) 101 (28.9) Lung disease 90 (20.9) 61 (17.5) Diabetes 75 (17.4) 46 (13.2) Cancer 15 (3.5) 8 (2.3) Other diseases 185 (3.0) 129 (40.0) Mean+SD follow-up time (months) 8.46+0.59 8.30+0.66b a SD = standard deviation. b F = 12.86, P<0.001. 177Bulletin of the World Health Organization 2003, 81 (3) Chronic disease self-management programme in China who completed the study. Only follow-up time was significantly different between those in the treatment and control groups (P<0.001). Table 2 gives baseline data for the treatment and control groups. Controls spent significantly less time on stretching and strengthening exercises per week and they had fewer cognitive symptom management practices and fewer communications with doctors but had significantly better disability scores than patients in the treatment group. Effectiveness of CDSMP intervention Table 3 shows the mean change in scores at six months for the treatment and control patients. Compared with controls, patients in the treatment group had significant improvements in two of the four self-management behaviours. Treatment subjects, on average, increased their duration of aerobic exercise to >25 minutes per week (P = 0.01) and scores in practice of cognitive symptom management by 0.37 (P = 0.005). The treatment group had significant improve- ments in both measures of self-efficacy: increases of 0.69 and 0.63 in self-efficacy tomanage symptoms and in self-efficacy to manage disease in general (both P = 0.001). Patients in the treatment group also had significant improvements in eight measures of health status (self-rated health, health distress, fatigue, shortness of breath, pain, disability, depression, and social and role activity limitations) (P<0.05). The treatment group had, on average, 0.12 fewer hospitalizations than the control group (P = 0.04). It was necessary to control for the baseline value of the study variables age, sex, education, marital status, follow-up time, baseline number of minutes per week spent on stretching and strengthening exercises, cognitive symptom management practice, communication with medical doctor, and disability to compare the mean change in scores for the treatment and control patients at six months. Among the control variables, analysis of covariance on six month post-test scores showed that each baseline study variable significantly influenced some six-month changes: age significantly influenced changes in self- rated health, fatigue, and disability (younger people had better changes in six months); education significantly influenced changes in self-rated health, fatigue and self-efficacy tomanage disease in general (patients with higher education levels had better improvements at six months); and disability significantly influenced changes in pain, shortness of breath, depression, and social and role activity limitation (patients with higher disability had worse improvements in those four variables in six months). Table 2. Baseline for treatment and control patients: self-management behaviour, health status, and health service utilization Variable Mean+SD P-valuea Treatment (n = 430) Control (n = 349) Self-management behaviour Stretching and strengthening exercise (minutes/week) 20.26+45.65 (n = 422)b 13.21+39.81 (n = 344) 0.002 Aerobic exercise (minutes/week) 179.54+130.50 (n = 421) 177.18+133.05 (n = 344) 0.27 Cognitive symptom managementc 1.19+0.98 (n = 407) 0.79+0.81 (n = 340) 0 Communication with medical doctorc 1.48+0.97 (n = 416) 1.24+1.07 (n = 341) 0 Self-efficacyd Managing symptoms 7.08+2.22 (n = 392) 7.04+2.22 (n = 336) 0.82 Managing disease in general 7.36+2.12 (n = 398) 7.23+2.21 (n = 333) 0.56 Health status Self-rated healthe 3.93+0.59 3.89+0.60 0.30 Energyc 2.56+0.89 (n = 409) 2.53+0.81 (n = 342) 0.83 Health distressf 0.90+0.91 (n = 411) 0.83+0.86 (n = 342) 0.31 Fatigueg 3.53+2.23 (n = 424) 3.29+2.10 (n = 346) 0.12 Shortness of breathg 2.18+1.87 (n = 424) 2.09+1.73 (n = 346) 0.71 Paing 2.36+2.07 (n = 424) 2.27+1.99 (n = 346) 0.60 Disabilityk 0.12+0.28 (n = 423) 0.07+0.22 (n = 342) 0.02 Illness intrusivenessh 25.48+14.36 (n = 415) 25.96+14.01 (n = 335) 0.54 Depressioni 7.68+4.25 (n = 403) 8.04+4.19 (n = 339) 0.15 Social/role activity limitationsj 0.75+0.81 (n = 415) 0.70+0.77 (n = 340) 0.36 Health care utilization (no. of times in previous six months) Physician visits 8.05+8.98 7.61+7.44 0.59 Emergency room visits 0.16+0.84 0.13+0.53 0.93 Hospital stays 0.13+0.38 0.11+0.32 0.53 Nights in hospital 1.57+8.18 0.87+3.99 0.58 a Mann-Whitney U-test was used to compare baseline variables between treatment and control groups (two-tailed P-values). b Several variables had data missing, which is shown in the table. c 0–5 increase = improvement. d 1–10 increase = improvement. e 1–5 decrease = improvement. f 0–5 decrease = improvement. g 0–10 decrease = improvement. h 13–91 decrease = improvement. i 0–30 decrease = improvement. j 0–4 decrease = improvement. k 0–3 decrease = improvement. 178 Bulletin of the World Health Organization 2003, 81 (3) Research An intention-to-treat analysis was conducted with 893 subjects. Baseline data were used at six months for the 114 subjects lost to follow-up, although the four treatment and seven control patients who diedwere excluded. In this analysis, all probability values remained unchanged except for social and role activity limitations (P = 0.06), although the change scores were slightly lower for both treatment and control groups. For example, the change in aerobic exercise was 26.81 minutes per week for the treatment group compared with 2.30 minutes for the control group, changes in self-rated general health were –0.26 and –0.03 and fatigue were –0.32 and 0.07, and the numbers of hospital stays were –0.055 and 0.049. Effect size of CDSMP intervention on significantly changed outcome variables To evaluate treatment effectiveness, figures were prepared to show the effect on significantly changed outcomemeasureswith a standardized statistic: the effect size. This allowed the comparison of effectiveness across measures with different scales and was calculated using the formulae below (15), in which mt = the mean change in outcome variables among members of the treatment group, mc = the mean change in outcome variables among members of the control group; s = the common standard deviation (which can be estimated by Sp); Sp= the pooled standard deviation; st 2 and sc 2 = the variances for the treatment and control groups, respectively; and nt and nc = the sample sizes for the treatment and control groups. ES = mt – mc s Sp =H (nt – 1)st 2 + (nc – 1)sc 2 (nt – 1) + (nc – 1) Table 4 shows that CDSMP had relatively large effects on cognitive symptom management, self-rated health, self- efficacy, and disability, but it had relatively small effects on depression. Comparison of lay-taught and professional-taught CDSMP courses The Mann-Whitney U-test was used to compare six-month changes of 20 outcome variables between 373 treatment Table 3. Six-month changes for treatment and control subjects: self-management behaviors, health status, and health service utilization Variable Mean+SD change P-valuea Treatment (n = 430) Control (n = 349) Self-management behaviour Stretching and strengthening exercise (minutes/week) 3.54+64.10 (n = 411)b 4.44+55.13 (n = 321) 0.07 Aerobic exercise (minutes/week) 27.93+175.51 (n = 406) 2.68+136.51 (n = 319) 0.01 Cognitive symptom managementc 0.37+0.95 (n = 391) 0.04+0.76 (n = 305) 0.005 Communication with medical doctorc 0.04+1.24 (n = 396) 0.11+1.32 (n = 308) 0.89 Self-efficacy in self-managementd Managing symptoms 0.52+2.11 (n = 378) –0.17+2.67 (n = 301) 0.001 Managing disease in general 0.22+2.52 (n = 377) –0.41+2.87 (n = 297) 0.001 Health status Self-rated healthe –0.28+0.79 –0.03+0.72 0.001 Energyc 0.03+1.03 (n = 380) 0.06+1.01 (n = 294) 0.93 Health distressf –0.24+1.01 (n = 386) –0.01+1.12 (n = 296) 0.001 Fatigueg –0.35+2.70 (n = 411) 0.09+2.52 (n = 326) 0.03 Shortness of breathg 0.05+2.39 (n = 411) 0.38+2.18 (n = 326) 0.01 Paing –0.04+2.38 (n = 412) 0.34+2.31 (n = 326) 0.02 Disabilityk –0.07+0.28 (n = 412) 0.01+0.32 (n = 322) 0.005 Illness intrusivenessh 0.07+17.01 (n = 403) 1.05+17.71 (n = 305) 0.06 Depressioni –1.20+5.23 (n = 385) –0.66+5.17 (n = 308) 0.004 Social/role activity limitationsj –0.09+0.98 (n = 399) 0.06+0.99 (n = 308) 0.046 Health care utilization (no. in previous six months) Physician visits –1.01+9.41 –0.84+7.76 0.72 Emergency room visits –0.04+0.96 –0.03+0.72 0.44 Hospital stays –0.06+0.46 (n = 430) 0.06+0.92 (n = 348) 0.04 Nights in hospital –0.55+9.60 0.44+6.72 0.40 a Analysis of covariance on six-month post-test scores controlling for the baseline value of study variables, age, sex, education, marital status, follow-up time, and baseline number of minutes per week of stretching and strengthening exercises, cognitive symptom management practice, communication with medical doctor, and disability (two- tailed P-values). b Several variables had data missing, which is shown in the table. c 0–5 increase = improvement. d 1–10 increase = improvement. e 1–5 decrease = improvement. f 0–5 decrease = improvement. g 0–10 decrease = improvement. h 13–91 decrease = improvement. i 0–30 decrease = improvement. j 0–4 decrease = improvement. k 0–3 decrease = improvement. 179Bulletin of the World Health Organization 2003, 81 (3) Chronic disease self-management programme in China subjects taught by lay leaders and 57 treatment subjects taught by professionals to determine whether outcomes were different. Only changes in self-rated health and fatigue differed significantly between the two groups. Patients in the lay-taught group had significantly better improvements in fatigue than those in the professional-taught group (–0.45 change in six months compared with 0.27; Z= –2.01; P= 0.04), but those in the professional-taught group had better self-rated health scores than those in the lay-taught group (–0.49 comparedwith –0.25; Z = –2.18; P = 0.03). Preliminary analysis of Shanghai CDSMP cost versus saving According to the decrease in number of hospitalizations in the six-month treatment period and the parameters shown in Table 5, a preliminary comparison of programme costs and savings was performed. This did not take into account indirect costs. The savings in health care expenditure approximated 726.79Yuan per participant—nearly nine times the cost of the intervention. Discussion This study was an implementation and evaluation of a generic chronic disease self-management patient education pro- gramme for Chinese people with one or more chronic conditions in Shanghai. The content of the Shanghai CDSMP course was repeatedly modified, pre-tested, and piloted to guarantee that the course was culturally acceptable to course participants. The course was delivered to groups mainly by lay leaders who worked in pairs. This new health education format was liked by most participants, which made the implementa- tion of CDSMP very successful, especially with the locally based delivery model— the commonly participatorymodel— in which the community government organizations, commu- nity health centre, and researchers worked together to design, implement, and evaluate the programme through a community programme advisory committee. Programme implementation was integrated into the routine of community government organizations and community health services. Overall, the Shanghai CDSMP was successful in increasing self-management behaviours, maintaining and improving health status, and decreasing health service utilization (Table 3). Chronic disease self-management approaches are widely accepted and are being disseminated (16–21). A five-year study by Lorig et al. showed that CDSMP can improve health behaviours (exercise, cognitive symptom management practice, and communication with physicians), health status (self-reported health, health distress, fatigue, disability, and social and role activity limitations), and decrease the number and duration of hospitalizations (3). This has prompted lay-led CDSMP courses to be disseminated internationally (22–24). No previous research has examined the feasibility and cultural appropriateness of CDSMP with Chinese people. This study showed that CDSMP can be accepted by Chinese people and can lead to benefits for the participants after it is modified carefully and delivered according to a locally based model. This has significant implications for noncommunicable disease prevention and control in China. Many patients with chronic conditions lack and need help with basic self-management knowledge and skills. At the same time, chronic disease is causing escalating health expenditure as the population in twenty-first century China rapidly ages. Another interesting finding in the study was the comparison of the effectiveness between lay-taught and professional-taught groups. Professional-taught groups had better changes in self-rated health scores, but lay-taught groups had better changes in fatigue scores. Although it is impossible to draw definitive conclusions, this suggests that Chinese lay people could be trained to teach CDSMP courses as successfully as health professionals. The use of lay people as community health educators is very uncommon in China, so the result has strong implications for community health education practice in the future. In addition, the findings strengthen evidence from other studies that the use of lay instructors in community-based health promotion and chronic illness self-management programme is cost-effective (25–27). One key question concerns the generalizability of the findings. As in all similar studies, participants in this study were volunteers who were recruited mainly through community mass media and interpersonal channels. The participants self- Table 4. Effect size for chronic disease self-management programme in terms of significantly changed outcome variables Outcome variables Effect size Self-management behaviours Aerobic exercise (minutes/week) 0.16 Cognitive symptom management 0.38 Self-efficacy in self-management Managing symptoms 0.29 Managing disease in general 0.24 Health status Self-rated health –0.33 Health distress –0.22 Fatigue –0.17 Shortness of breath –0.14 Pain –0.16 Disability –0.27 Depression –0.10 Social and role activity limitations –0.15 Health service utilization (no. in past six months) Hospital stays –0.17 Table 5. Parameters for cost benefit analysis of chronic disease self-management programme Parameter Cost (Chinese Yuan) Cost per participanta Programme advertisement 5.00 Training leaders 32.16 Volunteer leader stipends 20.00 Course materials 23.50 Administrative costs 2.00 Total 82.66 Cost per hospitalizationb 6745.45 a Six training workshops were finished and 100 qualified volunteer leaders trained; assumes two leaders taught each course, that each leader taught only one course, and that ten participants formed one self-management group. b Calculated from data from 330 participants’ baseline surveys. 180 Bulletin of the World Health Organization 2003, 81 (3) Research selected to be in the study and may have been more motivated than most patients with chronic disease. Evidence shows that 47% of patients repeatedly offered an opportunity to participate in an education programme similar to CDSMP will choose to do so (3, 28). This suggests that research on dissemination of CDSMP in Shanghai in the future may need to look at the reasons why patients attend or refuse to attend courses and to explore effective interventions to motivate patients to attend. The participants in this study had a high mean level of education: the mean duration of education was nine years. The feasibility and effectiveness of CDSMP when conducted with patients with lower education levels, especially in rural communities, need further study. The interpretation of this study is limited by the fact that nearly 20% of participants did not complete the study and that data are missing on some outcome variables. The real sample size (n = 779) was used to calculate statistical power with two- tailed tests and a = 0.05. We found that 779 participants gave 79% power — 1% less than the targeted statistical power (80%). It is important to note that the participants, community health centres, and general practitioners did not know participants’ study status. No link existed between the CDSMP and the patient’s individual treatment plans. The CDSMP did not alter participants’ treatment. The benefits achieved were additional to those achieved by usual care. Integration of CDSMP with usual care presumably would give participants more self-management support fromhealth care providers and would further enhance the benefits of the programme. Conclusions The Chronic Disease Self-Management Program developed in Shanghai from the generic Chronic Disease Self- Management Program was designed specifically to meet the needs of patients with one or more different chronic conditions, including those with comorbid conditions. The results of this study show that, when adapted, CDSMP is culturally acceptable to Chinese people and that it is feasible in China when delivered according to a locally- based model and integrated into the routine of community government organizations and community health services. In Shanghai, CDSMP had benefits similar to those demonstrated in previous studies in improving self- management behaviours and in health status and decreasing hospitalization, with a potential for substantial savings in health care costs. Chinese lay leaders can teach CDSMP courses as effectively as health professionals. If our results are replicated in more studies in China, CDSMP should be recommended for integration into the usual care of patients with chronic disease. n Acknowledgements We thank the Hsu Foundation and the Institute of Health Promotion Research, University of British Columbia, Canada, who trained the authors in how to implement and evaluate CDSMP in Shanghai. We also thank the China Medical Board, who partly supported this research. Conflicts of interest: none declared. Re´sume´ Mise en œuvre et e´valuation quantitative du programme d’auto-prise en charge des maladies chroniques a` Shanghai (Chine) : essai controˆle´ randomise´ Objectif Evaluer l’efficacite´ du programme d’auto-prise en charge des maladies chroniques (CDSMP) a` Shanghai. Me´thodes Lors d’un essai controˆle´ randomise´ avec suivi de six mois, on a compare´ les patients ayant rec¸u un traitement avec des te´moins non traite´s (patients inscrits sur une liste d’attente) dans cinq communaute´s urbaines de Shanghai (Chine). Les participants du groupe traite´ ont imme´diatement suivi un cours organise´ par le CDSMP et dispense´ par des non-professionnels de la sante´ et rec¸u une brochure d’appui ; les participants du groupe te´moin ont rec¸u les meˆmes prestations six mois plus tard. Re´sultats Au total, 954 volontaires dont le dossier me´dical confirmait un diagnostic d’hypertension, de maladie cardiaque, d’affection pulmonaire chronique, d’arthrite, d’accident vasculaire ce´re´bral ou de diabe`te et re´sidant dans la communaute´ ont e´te´ re´partis par tirage au sort entre un groupe traite´ (n = 526) et un groupe te´moin (n = 428). Sur ce total, 430 patients du groupe traite´ (81,7 %) et 349 patients du groupe te´moin (81,5 %) ont participe´ a` l’e´tude jusqu’a` la fin, c’est-a`-dire pendant six mois. Par rapport aux te´moins, les patients traite´s ont pre´sente´ une ame´lioration significative en ce qui concerne la dure´e hebdoma- daire (en minutes) d’exercice ae´robie, la pratique de la prise en charge des symptoˆmes cognitifs, l’auto-efficacite´ de la prise en charge de leurs symptoˆmes et l’auto-efficacite´ de la prise en charge de leur maladie en ge´ne´ral. Ils pre´sentaient e´galement une ame´lioration significative de huit indicateurs de l’e´tat de sante´ et subissaient en moyenne moins d’hospitalisations. Conclusion Mis en œuvre a` Shangai, le CDSMP e´tait cultu- rellement acceptable pour les patients chinois. Six mois apre`s le cours, le programme avait ame´liore´ le comportement des participants en rapport avec la sante´, leur auto-efficacite´ et leur e´tat de sante´ et avait re´duit le nombre d’hospitalisations. Le mode`le applique´, conc¸u au niveau de la communaute´, e´tait inte´gre´ dans les activite´s de routine des organismes publics communau- taires et des services de sante´ de la communaute´. Les Chinois non professionnels de la sante´ qui ont dispense´ les cours du CDSMP l’ont fait aussi efficacement que des professionnels. 181Bulletin of the World Health Organization 2003, 81 (3) Chronic disease self-management programme in China Resumen Aplicacio´n y evaluacio´n cuantitativa del programa de automanejo de enfermedades cro´nicas en Shanghai (China): ensayo controlado aleatorizado Objetivo Evaluar la eficacia del Programa de Automanejo de Enfermedades Cro´nicas de Shanghai (PAECS). Me´todos En un ensayo controlado aleatorizado con seguimiento de seis meses se procedio´ a comparar a los pacientes que recibieron tratamiento con los que no lo recibieron (controles de la lista de espera) en cinco comunidades urbanas de Shanghai (China). Los participantes en el grupo de tratamiento recibieron de inmediato formacio´n mediante un cursillo de PAECS dirigido por un no profesional de la salud, ası´ como un ejemplar de un libro de ayuda; los del grupo control recibieron la misma formacio´n y el mismo libro seis meses despue´s. Resultados En total, 954 pacientes voluntarios con una historia clı´nica que confirmaba un diagno´stico de hipertensio´n, cardiopatı´a, enfermedad pulmonar cro´nica, artritis, ictus o diabetes y que vivı´an en comunidades fueron asignados aleatoriamente al grupo de tratamiento (n=526) o al grupo control (n=428). Globalmente, 430 (81,7%) y 349 (81,5%) pacientes de los grupos tratado y control, repectivamente, terminaron el estudio de seis meses. Los pacientes que recibieron tratamiento experimentaron mejoras significativas en lo que respecta al nu´mero de minutos semanales dedicados a ejercicios de aero´bic, los resultados de las pruebas cognitivas y la eficacia del automanejo de la propia enfermedad en general en comparacio´n con los controles. Presentaron asimismo mejoras importantes en ocho indicadores del estado de salud y, como media, menos hospitalizaciones. Conclusio´n Los pacientes chinos abarcados por el PAECS lo consideraron culturalmente aceptable. Al cabo de seis meses el programa habı´a mejorado el comportamiento sanitario, la eficacia del control auto´nomo y el estado de salud de los participantes, y reducido el nu´mero de hospitalizaciones. El modelo asistencial local se integro´ en las actividades ordinarias de las organizaciones de gobierno de la comunidad y de los servicios de salud comunitarios. Los chinos no profesionales impartieron los cursillos del PAECS con la misma eficacia que los profesionales. References 1. Jia ES. [The trends of disease burdens both in China and in the other countries. Foreign Medicine.] [Branch of Health Economic] 1999;16:7-11. In Chinese. 2. Lei HCH. [The changing trend of health expenditure in China and policy intervention.] [Chinese Health Care Quality Management] 1999;6:23-6. In Chinese. 3. Lorig K, Sobel D, Stewart A, Brown BW, Bandura A, Ritter P, et al. Evidence suggesting that a chronic disease self-management program can improve health status while reducing hospitalization: a randomized trial. Medical Care 1999;37:5-14. 4. Holroyd Kenneth A, Creer Thomas L. Self-management of chronic disease: handbook of clinical interventions and research. Orlando (FL): Academic Press; 1986. 5. Lorig KR, Mazonson PD, Holman HR. Evidence suggesting that health education for self-management in patient with chronic arthritis has sustained health benefits while reducing health care costs. Arthritis and Rheumatism 1993;36:439-46. 6. Watson PB, Town GI, Holbrook N, Dwan C, Toop LJ, Drennan CJ. Evaluation of a self-management plan for chronic obstructive pulmonary disease. European Respiratory Journal 1997;10:1267-71. 7. Lorig K, Stewart A, Ritter P, Gonzalez V, Laurent D, Lynch J. Outcome measures for health education and other health care interventions. Thousand Oaks (CA): Sage Publications; 1996. 8. Sieving R, Resnick MD, Bearinger L, Remafedi G, Taylor BA, Harmon B. Cognitive and behavioral predictors of sexually transmitted disease risk behavior among sexually active adolescents. Archives of Pediatrics and Adolescent Medicine 1997;151:243-51. 9. O’Leary A. Self-efficacy and health. Behaviour Research and Therapy 1985;23:437-51. 10. Bandura A. Social foundations of thought and action. Englewood Cliffs (NJ): Prentice Hall; 1986. 11. Lorig K, Holman H, Sobel D, Laurent D, Gonzalez V, Minor M, et al. Living a healthy life with chronic conditions: self-management of heart disease, arthritis, stroke, diabetes, asthma, bronchitis, emphysema & others. Palo Alto (CA): Bull Publishing Company; 1994. 12. Lorig K, Gonzalez V, Laurent D. The chronic disease self-management course leader’s manual. Stanford (CA): Stanford University; 1997. 13. Fu DB, Shen YE, Fu H, Dai JM, Li Y, Wu CE, et al. [Pilot of Hypertension Self-Management Approach.] [Shanghai Preventive Medicine] 2000;12:555-6, 59. In Chinese. 14. Frankish JC, Green LW. Participatory research in health promotion. Institute of Health Promotion research report. Vancouver (BC): University of British Columbia; 1995. 15. Lipsey MW. Design sensitivity: statistical power for experimental research. Newbury Park (CA): Sage Publications; 1990. 16. Von Korff M, Gruman J, Schaefer J, Curry SJ, Wagner EH. Collaborative management of chronic illness. Annals of Internal Medicine 1997;127: 1097-102. 17. Mangan M. Diabetes self-management education programs in the Veterans Health Administration. Diabetes Educator 1997;23:687-92, 695. 18. Glasgow RE, Fisher EB, Anderson BJ, LaGreca A, Marrero D, Johnson SB, et al. Behavioral science in diabetes contributions and opportunities. Diabetes Care 1999;22:832-43. 19. Higgins DM, Liu T. Asthma self-management: do patient education programs always have an impact? Archives of Internal Medicine 1999;159:2422-8. 20. Gifford AL, Sengupta S. Self-management health education for chronic HIV infection. AIDS Care 1999;11:115-30. 21. Nodhturft V, Schneider JM, Hebert P, Bradham DD, Bryant M, Phillips M, et al. Chronic disease self-management: improving health outcomes. Nursing Clinics of North America 2000;35:507-18. 22. Magyary D, Brandt P. A school-based self-management program for youth with chronic health conditions and their parents. Canadian Journal of Nursing Research 1996;28:57-77. 23. Brown SA. Hanis CL. A community-based, culturally sensitive education and group-support intervention for Mexican Americans with NIDDM: a pilot study of efficacy. Diabetes Educator 1995;21:203-10. 24. Lorig K, Gonzalez VM, Ritter P. Community-based Spanish language arthritis education program: a randomized trial. Medical Care 1999;37:957-63. 25. Lorig K, Feigenbaum P, Regan C, Ung E, Chastain RL, Holman HR. A comparison of lay-taught and professional-taught arthritis self-management courses. Journal of Rheumatology 1986;13:763-7. 26. Brownell KD, Stunkard AJ, McKeon PE. Weight reduction at the work site: a promise partially fulfilled. American Journal of Psychiatry 1985;142:47-52. 27. Davis C, Leveille S, Favaro S, LoGerfo M. Benefits to volunteers in a community- based health promotion and chronic illness self-management program for the elderly. Journal of Gerontological Nursing 1998;24:16-23. 28. Lorig K, Holman H. Patient education: accepted only by the highly motivated or a generalizable treatment? Abstract presented at the fifty-sixth annual meeting of the American College of Rheumatology, 1992. 182 Bulletin of the World Health Organization 2003, 81 (3) Research