Chronic obstructive pulmonary disease (COPD) affects over 250 million people worldwide [1] and almost 600,000 people in the Netherlands [2]. In 2020, it is expected to be the third leading cause of death worldwide [3]. COPD is a common disease characterized by persistent respiratory symptoms and airflow limitation due to airway and/or alveolar abnormalities [3]. The most common symptoms are dyspnea, chronic coughing, and sputum production [3-5]. An acute worsening of the symptoms is called an exacerbation [4,6]. Exacerbations lead to additional care [5] and often lead to hospital admission [7], with considerable costs involved [8].

Self-management interventions are also recognized to be important in reducing exacerbations [9] and hospital admissions [10,11], improving quality of life [9-11], and improving patients’ control over their health [9]. Self-management skills can be beneficial for patients with COPD to manage their disease on a daily basis [12], for example, for medication use, breathing techniques, physical activity, and symptom recognition [13]. Effing et al [12] defined these interventions for patients with COPD as structured, personalized, and often multi-component, with goals of motivating, engaging, and supporting patients to positively adapt their health behaviors. Relevant features for self-management interventions include smoking cessation, recognition and treatment of exacerbation, increasing physical activity, nutrition advice, and management of dyspnea [14].

Mobile apps are increasingly being used to provide patients with health and self-management interventions, for example, for remote monitoring of patients’ health status [15-17], self-report of symptoms or health status [16-18], education [16,19], and digital support or feedback [15,17,18]. This is often combined with feedback from a health care professional or automated via the app [17-19]. Multiple reviews have analyzed the effectiveness of self-management interventions supported by mobile apps for patients with COPD on hospital admissions [15,18], exacerbations [15,16], length of hospital stay [18], behavioral outcomes [15,19], health-related outcomes [15,19], and quality of life [15]. The use of smartphones can be feasible in providing patients with self-management interventions [20,21] and to improve behavioral change [21]. A recent review reported the effects of smartphone interventions on exacerbations and showed that these interventions may decrease exacerbations, compared with usual care [16]. However, the findings remain inconsistent [17] due to heterogeneity among interventions [9,16,17,19,22], target populations [9,22,23], outcomes [9,22,23], and small sample sizes [16]. Further research and analysis on relevant apps for apps to support patients with COPD is necessary [24], as evidence is limited [15].

Until now, much attention has been given to the effects on clinical health outcomes [11,25-27] and hospital services [11,28,29]. Self-management behavior is also found to be important in reducing hospital admissions [30]. Factors affecting use in daily clinical practice, such as patients’ satisfaction [31], technology acceptance [32,33], and health care professionals’ satisfaction [34], were examined to a lesser extent. It also remains unclear which patients benefit most from these digital interventions [35,36]. It is suggested that it may be beneficial for patients experiencing frequent exacerbations [37]; nevertheless, stable patients with COPD are often the target population [38]. Patients experiencing a hospital admission due to an exacerbation may require a different approach, as they often experience feelings of distress during this time [39]. Additional evidence on this specific subpopulation is still needed [36], especially in combination with mobile health (mHealth) solutions [16]. Health care professionals’ involvement is also essential for a successful self-management intervention in clinical practice [13].

Self-management interventions, which are increasingly supported by mobile apps in recent years, may improve disease management in patients with COPD and may decrease hospital admissions. However, not all patients qualify because of reasons such as socioeconomic status, internet access, and skills. Systematic, comprehensive information on implementation- and compliance-related aspects of mobile self-management apps is lacking. Additional evidence about the effectiveness of mobile self-management apps is needed, especially regarding factors affecting the use in clinical practice for high-use patients, such as those recently hospitalized due to an exacerbation.

The objective of this study is to evaluate the effects of a mobile health and self-management app (COPD app) in clinical practice for patients with COPD, after discharge from the hospital, on app use, self-management, expectations and experiences (technology acceptance), patients’ and nurses’ satisfaction, and hospital readmissions.

The COPD app consisted of an 8-week health and self-management intervention, including the Lung Attack Action Plan, personalized medication overview, information about COPD, nutrition, physical activity, advantages of smoking cessation, weekly questionnaires monitored by nurses, and video consultation.

Pilot testing was used to receive feedback on a prototype of the COPD app. A total of 6 patients, admitted to a large teaching hospital (Rijnstate, Arnhem) for a COPD exacerbation, were provided with a tablet and access to the app. Patients received assignments such as Can you find and use the Lung Attack Action Plan, Can you find and open the questionnaire, and Can you find and read the information about nutrition. We also asked their opinion about the information (eg, if they missed information elements), frequency of notifications they would prefer, the readability, the frequency of new information, and their sociodemographic characteristics. Before starting the feasibility study, results from the pilot testing were used to improve the COPD app.

Patients were recruited from a large teaching hospital (Rijnstate, Arnhem). To be eligible, patients must be older than 18 years, diagnosed for COPD by a physician, admitted to the hospital for a COPD exacerbation (generally considered high-risk patients), have access to a smartphone or tablet, have a working internet connection, being able to use a smartphone or tablet, and be proficiency in Dutch language. Patients with cancer or (severe) cognitive or psychiatric conditions were excluded. At least one hospitalization for COPD exacerbation in the year preceding this study was also a criterion for accrual, but it only applied during the first month (of the inclusion period) because the number of eligible patients was too low.

Patients were informed about the study by a pulmonary nurse and the researcher during hospital admission. Patients received the study information letter and were asked to sign the informed form. They also received support to download apps. The Patient Journey App software (PJA version 4.0) [40] was used for the COPD app and Facetalk [41] for video consultation. The apps could be downloaded for free from the Google Play Store and the Apple App Store [41-43].

The COPD app provided patients with an 8-week self-management program. The app had 3 views: timeline, information page, and contact page (see Multimedia Appendix 1). The start date was the date of discharge of each patient. The timeline was classified in 8 weeks, and each week included the Lung Attack Action Plan, personalized (daily and extra) medication overview, information and education, and questionnaires. The first week also included a video of a pulmonologist explaining the purpose of the app and additional information about the functionalities of the COPD app. After 8 weeks (until 20 weeks), patients remained accessible to the information in the app, but the questionnaires, medication overview, video consultation, and Lung Attack Action Plan (including contact request) were no longer accessible.

Use was assessed using log data, extracted from the app software, after 8 and 20 weeks. Patients completed a baseline questionnaire during hospital admission, covering aspects of self-management (PIH), expectations with the COPD app, internet use, smartphone or tablet skills, and sociodemographics. After 8 weeks and 20 weeks, a questionnaire was sent on self-management, experiences with the app, and (overall) satisfaction. After 30 days, 8 weeks, and 20 weeks, the readmission rate was assessed, and data were extracted from the EMR. See Table 1 for an overview of the outcomes and measurement time points.

Data analysis was performed using IBM SPSS V22.0. Descriptive statistics were used to report the baseline characteristics, app use, expectations and experiences, satisfaction, and number of readmissions. Changes in self-management over time were analyzed using a linear mixed model. Using a linear mixed model allowed for the inclusion of cases with missing data. The relation between app use and self-management was analyzed using linear regression. Normally distributed variables were reported as mean and standard deviation, and non-normally distributed data were reported with medians and interquartile ranges (25th-75th percentiles).

The study was approved by the local ethical committee Commissie Mensgebonden Onderzoek Arnhem–Nijmegen.

A total of 6 patients participated in the pilot testing of a prototype of the COPD app: 3 men and 3 women. The age range was 58-78 years. A total of 4 patients used the internet (almost) every day and 2 patients (less than) 1 day per week. Moreover, 3 patients used a smartphone or tablet (almost) every day, 1 patient multiple days per week, and 2 patients never. Furthermore, 3 out of 6 patients perceived their smartphone or tablet skills not good or not bad, 1 bad, and 1 good. In addition, 3 (out of 6) patients did not miss information items in the COPD app.

The information was categorized per day in the prototype, meaning that a new information item was presented daily. During the assignments and observations, we found that it was not easy for patients to find information because the timeline was very long. A total of 4 (out of 6) patients preferred to receive all information items in 1 overview, ordered by information category (eg, nutrition). On the basis of the findings, we categorized the information per category (eg, nutrition, physical activity) instead of per day. To increase ease of use, the 8-week program was classified per week instead of per day. Patients’ opinion about the frequency of receiving a notification varied. Therefore, we decided to send a weekly reminder about the Lung Attack Action Plan and a reminder to fill out the weekly CCQ questionnaire.

Inclusion took place from November 19, 2018, to December 13, 2019. A total of 174 patients were assessed for eligibility. Moreover, 81 patients did not meet the inclusion criteria because they had no access to a smartphone or tablet (n=41), were not able to use a smartphone or tablet (n=19), no working internet connection (n=5), no proficiency in Dutch language (n=9), cancer, (severe) cognitive disability or psychiatric condition (n=7), or other reasons (n=24 eg, hospital admissions were too short, unclear diagnosis, or no reason was reported). In total, 28 patients declined to participate. Moreover, 2 patients signed the informed consent form, but they were excluded because the COPD app could not be installed on their smartphone or tablet. In total, 39 patients started the intervention. One patient died during the first 8 weeks, and 1 patient died before 20 weeks. Therefore, 39 patients were included in the analysis until 8 weeks, 38 patients were included in the analysis at week 8 and from week 8 to week 20, and 37 patients were included in the analysis at 20 weeks (Figure 1).

The baseline characteristics of the population included in the feasibility study are presented in Table 2.

The use of the COPD app, questionnaires, and consultations is described in more detail below and is presented in Table 3.

The COPD app was rated, on a scale of 1 to 10 (1: not satisfied at all to 10: very satisfied), with a 7.7 (SD 1.7) after 8 weeks and 7.0 (SD 2.4) after 20 weeks. Patients thought the app was easy to use and well-structured (26/28, 93%). Almost all patients reported that the Lung Attack Action Plan was easy to find (27/28, 96%) and easy to use (25/27, 93%), and more than half of the patients thought it actually helped them (18/27, 67%). The majority of patients also thought that the information was understandable (27/29, 93%), and all the patients (29/29, 100%) were satisfied with the information about nutrition. According to 33% (9/27) of patients, too much information was available in the COPD app. The majority of patients were satisfied with the video consultations (18/23, 78%) and thought it saved them time (19/29, 66%). See Multimedia Appendix 8 for more detailed information.

Knowledge and coping increased significantly over time (P=.04). However, there was no significant change in the recognition and management of symptoms (P=.14). See Multimedia Appendix 9.

No relation was not found between use of the app, the number of times the app was opened (mean page clicks during week 1-8), and the self-management elements knowledge and coping (P=.75) and recognition management and adherence (P=.92).

Patients’ expectations with the COPD app were relatively high. However, only 2 aspects improved over time. After using the app, more patients thought that it takes no effort to use it and that they had enough skills to use it. However, most aspects related to receiving support using the app decreased over time. See Multimedia Appendix 10 for more detailed information.

The use of the COPD app and monitoring of the weekly questionnaires were evaluated with 3 nurses. They rated the COPD app, on a scale of 1 to 10 (1: not satisfied at all to 10: very satisfied), on average with a 6.3 (SD 1.2) Most of them were satisfied with the app (2/3, 67%) and the information provided (2/3, 67%) and thought that better care was provided using the COPD app (2/3, 67%). However, use of the COPD app did not save time (3/3, 100%). They received a lot of questions from patients (3/3, 100%), and they mentioned that it took them a lot of time to explain it and answer questions (2/3, 67%). They also reported:

Only 1 nurse would recommend the COPD app to more patients. The nurses would not recommend it to their colleagues.

The nurses were less satisfied with monitoring the results of the questionnaires and rated this with a 5.3 (SD 0.58), on a scale of 1 to 10 (1: not satisfied at all to 10: very satisfied). Only 1 nurse thought that monitoring the results of the questionnaires fitted well in their work process. They commented:

The nurses were less satisfied with the video consultations and mentioned the following:

In total, 39 patients were included in the study. A total of 12 patients (12/39, 31%) were readmitted 22 times during the study period (20 weeks), of which 5 patients (5/39, 13%) were readmitted 1 time in the first 30 days. Within 8 weeks, 8 patients (8/39, 21%) were readmitted 11 times. In the total study period (until 20 weeks), there were 22 readmissions for 12 patients (12/39, 31%). The main reasons for readmissions was COPD exacerbations, and 1 time it was due to a patient’s home situation.

In the year preceding the study, from November 2017 to November 2018, 340 patients were admitted 478 times to the hospital. In total, 48 patients (48/340, 14.1%) were readmitted 77 times within 30 days. There were 103 readmissions within 8 weeks for 61 patients (61/340, 17.9%), and 74 patients (74/340, 21.8%) were readmitted 129 times within 20 weeks.

In this study, a mobile self-management app for high-risk patients with COPD was evaluated in daily clinical practice. The COPD app was opened most often in the first week (median 6.0; IQR 3.5-10.0), but its use decreased over time (median 2.0; IQR 1.0-3.5 in week 8). Information, especially on physical activity (24/39, 62%), was read most often during the first week. The self-management element knowledge and coping increased significantly over time (P=.04), but a relation with app use was not found (P=.75). No significant change was found in recognition and management of symptoms, adherence to treatment (P=.14), or in relation with app use (P=. 92). Patients rated the COPD app on average with a 7.7 (SD 1.7) and nurses with a 6.3 (SD 1.2). Preliminary evidence about readmission rate showed that 13% (5/39) of patients were readmitted within 30 days, 21% (8/39) within 8 weeks, and 31% (12/39) within 20 weeks compared with 14.1% (48/340), 17.9% (61/340), and 21.8% (74/340), respectively, in a preresearch cohort.

The use of mobile apps itself is not applicable to all patients [51,52]. In total, 37.4% (65/174) of all patients in our study had to be excluded because of lack of access to a mobile device or internet or skills to use it. This is in line with other findings of mHealth use in patients with COPD, in which only a minority owned a smartphone (23%) [53]. Technical issues and low compliance are recognized issues for digital interventions [54], and digital literacy among patients with COPD remains a challenge [52]. As a result of the pilot testing, the app we implemented was already simplified. However, digital literacy may still have been an issue during this study. Therefore, ease of use seems to be an essential element in digital interventions for this patient population [20,27]. A total of 16.1% (28/174) of those possibly qualifying declined to participate, among other things, because it was too much of a burden or effort at the time. Patients may have experienced high levels of distress after experiencing an exacerbation [55], and therefore, they may be less willing to engage in a self-management intervention [38]. Therefore, these interventions are not applicable to all patients who are recently discharged from the hospital [38], as they may still feel (too) sick and/or are not able to focus on the intervention [34]. This emphasizes the importance of timing [39] and tailoring [56] an intervention.

Until now, the effects of self-management interventions on patients recently discharged from the hospital were scarcely evaluated [38] in combination with mobile apps. The direct effects [57] of app supported self-management and health interventions, for example, technology acceptance, self-management, and patients’ and nurses’ satisfaction are relevant for use in clinical practice. We found that the app was especially used during the first week after discharge. The Lung Attack Action Plan (9/39, 23%) and request for contact using the contact page (3/39, 8%) were used to a limited extent. However, the majority (29/39, 74%) completed the weekly CCQ questionnaires during the whole intervention period and the HADS in week 8 (33/39, 85%). Patients received frequent reminders by email, in the app and sometimes from nurses, to complete the questionnaires. The use of the COPD app and the Lung Attack Action Plan was more optional, rather at patients’ own initiative. Receiving feedback can be important [56], and this may explain that the majority of patients completed the questionnaires, but that the use of the COPD app decreased over time. Low frequency of use can also be due to lack of self-management or technological skills [56].

Social support is seen as a facilitator for use [32,52]. The majority of the patients (28/37, 76%) expected to receive enough help using the COPD app. However, only 57% (17/30) of the patients indicated that they had received enough help (Multimedia Appendix 10). Tailored education can also facilitate use [52], but in this COPD app, only the medication overview was really personalized. Although the information items were aimed at high-risk patients with COPD, the information was generic. This might have contributed to the decrease in use. Tailored interventions [56], support [30], and patient engagement during development and implementation [56,58] may be beneficial for improved use.

A positive effect was found on knowledge and coping, which may partly be explained by the selection criteria for this study, as patients with cognitive disability and lack of skills with a mobile device were excluded. In addition, the provision of timely information using a mobile device can positively influence knowledge [59]. Self-management can also be enhanced by involving patients’ partners, enhancing self-efficacy, and support from health care professionals [30]. Although positive results on hospital readmissions were found in previous studies [6,18], these findings were inconsistent [15,28,60], which could be due to high methodological heterogeneity [16,19]. In our study, no large difference was observed, possibly due to low numbers. It would be interesting to verify the element of selection bias in view of the large percentage of patients that were excluded from this population.

Patients were satisfied with the COPD app, user-friendliness, and information. However, nurses addressed some concerns, for example, the increased workload and (lack of) integration in the work process. It is common that the degree of satisfaction between patients and health care professionals can differ. In general, patients report more favorable outcomes because mobile interventions are often provided as an extra service in addition to their usual care. For that same reason, health care professionals are generally less satisfied, especially because they often see it as an increase in workload [61]. The nurses in our study addressed concerns about the monitoring of the results of the questionnaires because they experienced a lack of integration in their work processes. Often a common pattern with the introduction of new innovations, this intervention was an addition to their current activities. Another reason might be that nurses had to work with different information technology systems that were not connected to the EMR. Lack of interoperability can be a barrier [58] for use, and this might explain the lack of monitoring of the first phase of the study. This improved after they received the scores in person by email. Health care professionals’ adoption is essential to ensure success; therefore, they should be involved in the development and implementation process [56].

COPD management requires a multidisciplinary approach that is fragmented [24], and this approach is often not sufficiently supported by information technology [62]. Therefore, future research should focus on self-management interventions with a multidisciplinary approach tailored to individual patients recently discharged from the hospital. Pragmatic trials [63] can be used to determine, at a more rapid pace, which elements of self-management interventions are effective for which subgroups of patients with COPD recently discharged and which characteristics of mHealth solutions are adopted by both patients and health care professionals. Subsequently, a larger controlled study specifically involving this frail subgroup of patients should focus on the effects on clinical outcomes and hospital services use (eg, readmissions).

Due to accrual issues, especially related to device availability and internet access, the COPD app was evaluated in a small sample, so we could not reach the power originally calculated for this trial. In addition, nurses found it difficult to comply with the contacting rules, so there were inconsistencies in the follow-up monitoring using the CCQ questionnaires. Some patients were only contacted a limited number of times when they had a high score on the CCQ questionnaire. After approximately 20 patients, we decided to send nurses a notification by email with the scores, and they were asked to take up contact (if necessary). As a consequence of the team setting, only 3 nurses were involved in this study, and we have to be careful about the related outcomes. Preliminary evidence on readmission rates was provided based on an earlier cohort, but this was not a matched exercise. Therefore, definitive conclusions on this aspect cannot be drawn.

The integration and use of a mobile self-management app for recently discharged patients with COPD in clinical practice is affected by multiple factors and is only feasible for a relatively small number of patients after hospital discharge. Patients were very positive about the COPD app; however, its use decreased over time. The findings of this study showed a significant positive change in the self-management element knowledge and coping. Nurses expressed concerns about integration in their work processes and increased workload. Tailored interventions, patient support, and active adoption by professionals are important elements to ensure successful mHealth interventions. Therefore, future research on digital self-management interventions in clinical practice should focus on including more difficult subgroups of target populations, on a multidisciplinary approach, on technology-related aspects (such as acceptability), and on finetuning its adoption in clinical pathways.