This was a randomized controlled trial, which compared an intervention group that received a Web-based coaching program using EHRs with a control group that did not receive the intervention after discharge from hospital. All participants received usual care before discharge, including medication guidance, lifestyle education, and regular reviews in order to help them manage their disease. Data were collected before the program and at 1, 3, 6, and 12 months after the program. Figure 1 illustrates the Consolidated Standards of Reporting Trials (CONSORT) flow diagram.

The trial was developed by the research team, which included a clinical nurse, a head nurse, a community nurse, two respiratory physicians, and two nursing students (data collector). The clinical nurse taught disease-related information in the predischarge phase. The community nurse was responsible for the postdischarge follow-up using EHRs.

The EHRs were divided into two separate fields (Figures 2 and 3). There were two user spaces, one dedicated to the participants in the study (the front end) and the other one dedicated to the administrators of the platform (the back end). The patient-accessible personal health records were activated by the administrator logging into the system and inputting a username and code. Each participant received a unique identifier code generated by the system. The system was constructed to allow the input of the patient’s demographic information, record of admission, discharge, and community information. The system build included a statistical analysis function so that a curve graph could be produced of data variation that visually informed medical staff and patients of the trajectory of the disease. Once registered, the individual participant record was retrieved by entering the unique participant identifier. When the appropriate record was retrieved from the system, the administrator could add new information to the record. After the administrator assigned the username and code for the particular participant, the participant could access information about their disease and also health education content entered by the administrator, which now appeared in the participant field. The participant could connect with the community administrator through the EHR's system. This communication is similar to email function. Medical practitioners and nurses could write suggestions to the participants in the system and, when the patients logged in, they could view the message. Participants could also ask questions directed to the medical team using the same messaging function.

The implementation of the Web-based coaching program was based on sharing information within EHRs. All relevant data were recorded in the EHRs to allow the participants and administrators to refer to particular information at their convenience by Internet. There were two main information sets in these particular EHRs. One was related to health education and provided a resource for participants that included information about COPD and pulmonary rehabilitation instructions. The information related to COPD consisted of the cause of the disease, development, acute exacerbation, prognosis, medication information (name, route, dosage, and adverse reactions), oxygen therapy, and diet. Pulmonary rehabilitation instructions consisted of abdominal contraction and lip breathing, respiratory muscle exercise, aerobic exercise, and the importance of smoking cessation. The second information set was the participant’s particular management based on the progress of the disease. Tailored contents that match a user’s preference are more useful to motivate behavioral change [28].

The EHRs could store all the patients’ health information from hospital to community care. In this particular project, the individual also had control of their own records and managed their files. The records could be packaged and sent electronically. Because the participant had control of the records, they could supply their own health information to their health practitioners. Participants could read all their information and discuss it with the administrators by Internet within the EHRs.

The research team telephoned the participants every two weeks and made home follow-up visits at 1, 3, 6, and 12 months after discharge to promote adherence to the provided intervention methods and to collect data. The participants could telephone the community administrators when they had problems.

Approximately 198 patients with COPD were asked to participate in the study from September 2008 to November 2014. A convenience sample of 130 participants agreed to participate and were recruited from two tertiary hospitals in Tianjin, China. The inclusion criteria were as follows: (1) a medically confirmed diagnosis of COPD based on the Chinese Medical Association Diagnostic Criteria, including percentage forced expiratory volume for 1 second (FEV₁%)≤80% and forced expiratory volume for 1 second divided by forced vital capacity (FEV₁/FVC)≤70%; (2) clear consciousness, able to speak Mandarin, and able to communicate; (3) discharged to a home where Internet and computer have been installed; and (4) able to be reached by telephone postdischarge. The exclusion criteria included the presence of comorbidities, such as allergic rhinitis, myocardial infarction, severe heart failure, and malignant tumor; living outside Tianjin; and no access to a computer and Internet at home.

The patients who consented to participate were assigned to the intervention or control group using a computer-generated randomized table. Permission to carry out the study was granted by Tianjin Medical University Ethics Committee. The study was conducted in accordance with the Declaration of Helsinki. All participants completed an informed consent form before entering the study.

Patients who were definitively diagnosed with COPD were contacted the day after their admission to hospital. The questionnaires and clinical variables were collected the day before the patient’s discharge from the hospital and at 1, 3, 6, and 12 months after discharge.

Data were analyzed using SPSS for Windows version 19.0 software (SPSS Inc, Chicago, IL, USA), with the level of statistical significance set at P<.05. We used descriptive statistics to analyze the participants’ characteristics. Repeated variant analysis was performed to test the data from the two groups collected predischarge and at 1, 3, 6, and12 months after discharge. Differences between the two groups were evaluated by t tests.

Expecting a difference between the groups at follow-up equivalent to an effect size of approximately 0.8, with a power of .80 and alpha at P=.05, the number of participants needed in each group was estimated to be 33.

During the intervention, seven of 62 participants (11%) in the study group dropped out (three changed addresses, two had cerebral infarction, one had a fracture, and one declined ongoing participation). Three of 68 participants (4%) in the control group were lost to follow-up (one had a cerebral infarction, one had myocardial infarction, and one declined ongoing participation). Therefore, 120 participants completed the study. The number of participants in the intervention group was 55 and in the control group was 65 (Figure 1).

The mean age of the 120 participants (47.5%, 58/120 men) was 70.6 (8.0) years. The severity of disease was evaluated as Global Initiative for Chronic Obstructive Lung Disease (GOLD) II (23.3%, 28/120), GOLD III (47.5%, 57/120), GOLD IV (29.2%, 35/120). There were no statistical differences between the intervention and control group in the main sociodemographic characteristics at baseline (Table 1).

The variation of SGRQ between the intervention and control group is described in Table 4.

There was a tendency to change in overall impact (P<.001), symptoms (P<.001), activities (P=.03), and psychosocial impact (P=.047) in a year. The intervention method and time showed an interaction effect in each variable. The variation of variables is shown in Table 5.

For overall impact, there were statistically significant differences in the intervention and control groups (P<.001). From 1 month to 12 months, the intervention group had higher values than the control group (P=.01, P<.001, P<.001, and P<.001, respectively).

There were statistically significant differences in symptoms in the intervention and control groups (P<.001). From 3 months to 12 months, the intervention group had lower values than the control group (P=.02, P<.001, and P<.001, respectively).

For activities, there were significant differences in the intervention group (P<.001). However, there were no significant differences in the control group. From 1 month to 12 months, the intervention group had lower values than control group (P=.002, P=.001, P<.001, and P<.001, respectively).

There were statistically significant differences in impact in the intervention (P=.02) and control groups (P<.001). From 1 month to 12 months, the intervention group had lower values than the control group (P=.03, P<.001, P<.001, and P<.001, respectively).

The variation of the MMRC and 6MWT for the intervention and control groups are described in Table 6. There is a tendency to change in MMRC (P<.001) and 6MWT (P<.001) in a year. The intervention method and time had an interaction effect in MMRC. However, there was no significant difference in the intervention time in the 6MWT. The variation of MMRC and 6MWT are shown in Table 7.

For the MMRC, there were statistically significant differences in the intervention (P=.01) and control groups (P=.001). From 1 month to 12 months, the intervention group had lower values than the control group (P<.001).

The 6MWT was only tested at baseline, 6 months, and 12 months when the participants were required to be reexamined in the outpatient service. There were significant differences between the intervention (P<.001) and control groups (P<.001). From 6 months to 12 months, the intervention group had higher values than the control group (P=.002 and P<.001, respectively).

This study examined the effects of a Web-based coaching program using EHRs for patients with COPD. The intervention group had a statistically significant better pulmonary function measured by FVC% and PEF, physical capacity measured by 6MWT, health-related quality of life measured by SGRQ, and lower dyspnea degree measured by MMRC during the 1-year follow-up.

The EHRs are databases that health care providers use to record patient-related information to track care [32,33]. The effects of EHRs have been validated effectively in a pharmacogenetic study and also in a study of asthma exacerbation among school-aged children [34,35]. However, one qualitative study found that participants’ low technological issues resulted in low satisfaction with the technology [36]. Because of several factors associated with high or low compliance with Web-based interventions [37,38], we recommend the use of a combination of Internet and non-Internet engagement to enhance follow-up of all participants.

The EHRs have contributed to the health management of people with chronic illness. Provision of patient-centered advice including personal diet and alimentary advice and personal intervention methods based on gender, age, health, environment, employment, and medical service can make a difference. The EHRs in our study were a continuous, comprehensive, individualized, health information database. The system had a large capacity to store content, was low cost, occupied a small amount of computer memory, and enabled a convenient reference point related to COPD. Patients could navigate their own records at any time at home.

This study was consistent with other studies which demonstrated that comprehensive therapy methods, such as medication, oxygen, respiratory exercise, diet, and health education could not change the decline of FEV₁% altogether. However, our program could postpone the decreasing speed of FEV₁%. The FEV₁% of the control group began to decline at the end of 3 months and patient quality of life also declined as identified by SGRQ. The rate of decline in FEV₁% is representative of the natural history of COPD [39]. With the aggravation of airway obstruction, FEV₁% has a declining trend and an irreversible pathological process. Although the decline of FEV₁% could not change, a study by Garcia et al [40] demonstrated that respiratory muscle exercise could increase FVC%, which was a way to judge airway obstruction and had a relationship with trachea resistance and conformance of pulmonary tissue. Our study also showed that the Web-based EHR coaching program was correlated with an increase in FVC%. Our intervention taught patients how to make a complete expiration through a breathing technique that can increase FVC%.

The program improved the quality of life for patients with COPD compared to the control group, whose quality of life declined. The SGRQ, an index predicting quality of life with COPD, was an effective and sensitive score to evaluate quality of life with COPD. This result shows the opposite finding compared to a recent study on the long-term effects of an Internet-mediated pedometer-based walking program in the United States that only reported there were no significant differences between the intervention and control groups [41]. Therefore, the content in the program made a big impact on patient quality of life. Our study showed that although quality of life in the intervention group improved notably, the intervention could not reverse airway obstruction, only postpone the ingravescence rate of airway obstruction. Although the irreversibility of airway obstruction was closely related with quality of life, the SGRQ is not a good indicator of irreversibility, which is related to the mortality and frequency of acute exacerbation [42,43].

The intervention contributes to a decrease in the degree of dyspnea. This program includes Web-based EHRs containing considerable information regarding prevention, treatment, pulmonary rehabilitation, and disease variation, which could encourage patients to actively engage in healthy behaviors and increase adherence to medication, oxygen therapy, and respiratory exercise. Consequently, this may improve patient endurance by decreasing systematic oxidative stress and decrease the dyspnea level by decreasing oxygen consumption and physiological dead space [44].

In our study, the program was related to an extended 6-minute walking distance, which strengthened endurance and improved quality of life. Physical capacity refers to a person’s overall ability to function and “undertake the physically demanding activities of daily living” [45]. Relevant studies have tested that different types of interventions, such as telephone support, a mobile telephone program, computer-based program, or diary-recorded home-walking intervention, could increase physical capacity for patients with COPD compared to standard care [46-49]. However, one study showed that telephone mentoring for home-based walking demonstrated no benefit to the exercise capacity measured by the 6MWT [50]. Therefore, this Web-based coaching program using EHRs gives patients more confidence to control dyspnea and is recommended for use in the hospital, community, and home setting. During the study period, the researchers found that the 6MWT was not an instrument currently used in hospitals in China for patients with COPD and only the participants who were recruited in this study used this measure. We had hoped that respiratory physicians would use this instrument more regularly as a necessary examination that they must consider to assess patients’ adaptation to exercise.

Our study had several limitations. First, we were not able to evaluate selection bias because we could not differentiate between participants who enrolled versus those who declined participation. Second, many participants were elderly and unable to use EHRs by themselves, needing help from their family members. This condition may have limited the participants’ ability to navigate the EHRs when their families were not present. Third, we could not control the frequency of computer use by participants or coaching compliance, which could influence the effect of the intervention. Finally, our study setting was limited to Tianjin and did not include those with mild severity of COPD (GOLD I), limiting the generalizability of our results in all the patients with four disease classifications of COPD in China.

The study established a Web-based coaching program using EHRs, which included Internet and non-Internet as a valuable component of intervention for COPD management in China. The program contributed to improved quality of life for people with COPD through better respiratory function and physical capacity. It also promoted hospital and community nurses sharing patients’ medical information, and achieved dynamic management and follow-up analysis.