This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
Many patients with chronic obstructive pulmonary disease (COPD) suffer from exacerbations, a worsening of their respiratory symptoms that warrants medical treatment. Exacerbations are often poorly recognized or managed by patients, leading to increased disease burden and health care costs.
This study aimed to examine the effects of a smart mobile health (mHealth) tool that supports COPD patients in the self-management of exacerbations by providing predictions of early exacerbation onset and timely treatment advice without the interference of health care professionals.
In a multicenter, 2-arm randomized controlled trial with 12-months follow-up, patients with COPD used the smart mHealth tool (intervention group) or a paper action plan (control group) when they experienced worsening of respiratory symptoms. For our primary outcome exacerbation-free time, expressed as weeks without exacerbation, we used an automated telephone questionnaire system to measure weekly respiratory symptoms and treatment actions. Secondary outcomes were health status, self-efficacy, self-management behavior, health care utilization, and usability. For our analyses, we used negative binomial regression, multilevel logistic regression, and generalized estimating equation regression models.
Of the 87 patients with COPD recruited from primary and secondary care centers, 43 were randomized to the intervention group. We found no statistically significant differences between the intervention group and the control group in exacerbation-free weeks (mean 30.6, SD 13.3 vs mean 28.0, SD 14.8 weeks, respectively; rate ratio 1.21; 95% CI 0.77-1.91) or in health status, self-efficacy, self-management behavior, and health care utilization. Patients using the mHealth tool valued it as a more supportive tool than patients using the paper action plan. Patients considered the usability of the mHealth tool as good.
This study did not show beneficial effects of a smart mHealth tool on exacerbation-free time, health status, self-efficacy, self-management behavior, and health care utilization in patients with COPD compared with the use of a paper action plan. Participants were positive about the supportive function and the usability of the mHealth tool. mHealth may be a valuable alternative for COPD patients who prefer a digital tool instead of a paper action plan.
ClinicalTrials.gov NCT02553096; https://clinicaltrials.gov/ct2/show/NCT02553096.
Exacerbations in chronic obstructive pulmonary disease (COPD) are acute events of transient worsening of the respiratory condition. Exacerbations considerably affect patients’ health status [
Self-management strategies, such as the use of a written exacerbation action plan, have been shown to improve exacerbation outcomes, that is, decrease exacerbation duration [
Telemonitoring, in which patients record and send information on symptoms or physiological measurements to a supervising clinician, may be an alternative approach to self-management strategies to reduce the impact of COPD exacerbations. Beneficial effects have been reported on the number of hospital admissions [
In this study, we examined the clinical effectiveness of the mHealth tool. We hypothesized that in patients with COPD, the use of the tool would lead to more weeks without exacerbations; improvement in health status, self-efficacy, and self-management behavior; and a reduction in health care utilization compared with the use of a paper exacerbation action plan. We also evaluated how patients valued the tool’s supportive function and usability.
This study was a multicenter, parallel, 2-arm, randomized controlled trial with a follow-up of 12 months per patient. After signing informed consent, patients with COPD recruited from general practices and outpatient clinics followed a 20-min self-management educational group session primarily addressing early recognition and prompt treatment of exacerbations. Subsequently, patients were randomized to either (1) usual care according to current COPD guidelines, that is, exacerbation self-management support through the use of a paper exacerbation action plan (control group) or (2) exacerbation self-management support through the use of the mHealth tool (intervention group). Participants in the control group were provided with a written action plan if they did not have one. Participants in the intervention group were instructed not to use their action plan if they had one before participation. This study has been registered at ClinicalTrials.gov (Identifier: NCT02553096) and has been approved by the medical ethics review board, region Arnhem-Nijmegen, the Netherlands (file 2014-1270).
Patients were recruited between June 2015 and July 2016 at the pulmonary outpatient clinics of 3 Dutch hospitals and 9 general practices in the city of Nijmegen and the surroundings in the Netherlands. All participating centers delivered care according to the current Dutch COPD guideline [
We used a computer-generated 2-block randomization procedure, stratifying for the health care center. All patients from the participating centers who met the inclusion criteria received a questionnaire from their health care professional with questions related to exacerbations in the previous 12 months. Patients who responded and had experienced 2 or more symptom-based exacerbations in the previous 12 months (see inclusion criteria) were invited by the research team to participate in this study. The allocation order was determined by the order in which eligible patients responded to our invitation to participate (kept by the research assistant). Participants were assigned to one of the groups after signing informed consent during the group meeting by the researcher (LB). Owing to the type of intervention, patients and health care professionals could not be blinded for group assignment. In addition, the research team could not be blinded as it was responsible for the personalization and technical support of the mHealth tool. The study statistician (RA) who was responsible for analyzing the data was blinded for study assignment of the participants until the analyses had been finished.
Before randomization and after signing the written informed consent, all participants received a 20-min educational session based on the Dutch version of the Living Well with COPD self-management program provided by the nurse in groups of 4 to 10 participants to establish a homogeneous baseline in exacerbation self-management knowledge [
Participants in the intervention group were instructed to visit the nurse within 2 weeks after allocation for instructions on the use of the mHealth tool. The tool consisted of a mobile phone (provided by the research team), a pulse oximeter (CMS50D, Contec Medical Systems,), a spirometer (PiKo-1 monitor, nSpire), and a forehead thermometer (FTN, Medisana AG). Patients answered 12 yes-or-no questions concerning changes in symptoms, physical limitations, and emotions using the touch screen on the mobile phone complemented by measurements with the pulse oximeter, spirometer, and forehead thermometer (see
Before the trial started, participants in the intervention group were instructed to use the system daily for 2 weeks to get familiarized with the app, mobile phone, spirometer, pulse oximeter, and forehead thermometer. Data were sent to a secured Web-based interface and were monitored by the research team to make sure participants practiced sufficiently. After this 2-week run-in period, the nurses evaluated patients’ use of the system, including the physiological measurements. Reference values for each patient’s FEV1 and peripheral oxygen saturation were set. Then, the 12-month follow-up period started. Patients were instructed to use the tool every time they experienced or had any doubts about any change in symptoms or disease burden.
Participants in the control group visited the nurse within 2 weeks of allocation for instructions on the use of a paper exacerbation action plan. When patients did not already possess a written action plan at that moment, the nurses provided the action plan of the Living Well with COPD program [
At the 3-month follow-up, patients in both the intervention and control groups were invited by their nurse to evaluate their self-management of COPD exacerbations. In the intervention group, only the nurses received the patients’ entries in the mHealth tool from the research team to enable tailoring of feedback on self-management behavior. In the control group, the nurses evaluated the use of the paper action plan. Patients in both the groups did not receive any feedback on self-management behavior before or after this nurse contact. All patients in both the intervention and control groups continued to have complete access to their health care professionals during the follow-up.
Our primary outcome was the difference in the number of exacerbation-free weeks between the intervention and control groups. An exacerbation-free week was defined as a week in which there had not been episodes of 2 or more consecutive days with worsening of 2 major symptoms (ie, dyspnea, sputum purulence, and sputum amount) or 1 major and 1 or more minor symptoms (ie, colds, wheeze, sore throat, and cough) [
The secondary outcomes included the following:
Exacerbation-related outcomes, that is, the number of unscheduled health care contacts, the number of exacerbations treated with antibiotics and/or prednisolone, and the number of exacerbation-related hospital admissions, all retrieved from patients’ medical records, and the number of symptom-based exacerbations as assessed with TEXAS.
Exacerbation-related self-management behavior, measured with TEXAS or the online questionnaire, and defined as taking 1 or more of the following 3 actions during symptom-based exacerbations: (1) contacting the health care professional, (2) starting a course of prednisolone and/or antibiotics, or (3) maximizing bronchodilator use. We also assessed the time between the date of exacerbation onset and the date of 1 of these 3 actions, defining actions taken within 3 days of exacerbation onset as adherence to the instructions.
Exacerbation-related self-efficacy, measured with an exacerbation-related self-efficacy scale containing 5 questions. This questionnaire was created for the purpose of this study as, to our knowledge, no questionnaire existed that measured exacerbation-related self-efficacy. Reliability analyses showed a Cronbach alpha of .69 at baseline and .81 at follow-up.
Health status, measured with (1) the Nijmegen Clinical Screening Instrument (NCSI), which is a battery of instruments measuring 8 subdomains of health status—subjective symptoms, dyspnea emotions, fatigue, behavioral impairment, subjective impairment, general quality of life (QoL), health-related QoL, and satisfaction with relationships [
At the start and at 12 months, data were gathered on exacerbation history, self-efficacy, and health status. CCQ and EQ-5d were also completed at 3, 6, and 9 months of follow-up.
At 12 months, information on health care utilization, lung function, respiratory medication use, and comorbid conditions was extracted from the participants’ medical records. In addition, all participants were asked to evaluate the supportive function of either the mHealth tool or the paper action plan by using a paper survey including closed-ended questions regarding the use, difficulty in use, and intended future use of the mHealth tool or the paper action plan. Besides, 3 questions were asked related to clarity, suitability, and follow-up of the advice given by the mHealth tool or the paper action plan. All questions included answers on a 7-point rating scale from strongly disagree (score 1) to strongly agree (score 7). The survey also included 1 question about frequency of usage at times of symptom worsening, with answers on a 7-point rating scale varying from 1=never to 7=always. In addition, participants of the intervention group were asked to complete the System Usability Scale (SUS) [
Sample size calculation using analysis of variance showed that we needed 43 participants in each group for 80% power (alpha=.05, 2 sided) to detect an increase of 6 exacerbation-free weeks per year and anticipating a dropout rate of 20% (9/43). The calculation was based on a previous dataset [
We used the data recorded by the secured Web-based interface to assess the actual usage of the mHealth tool. We analyzed the answers to the paper evaluation survey to assess participants’ self-reported use of the mHealth tool and the paper action plan. Negative binomial regression analyses, controlling for follow-up time per participant, age and gender were used to analyze our primary outcome, that is, the number of exacerbation-free weeks, as well as the number of unscheduled health care contacts, self-reported exacerbations, exacerbations treated with antibiotics and/or prednisolone, and exacerbation-related hospital admissions.
To test the effect of the mHealth tool on the rate of symptom-based exacerbations and self-management behavior, we extracted exacerbation episodes from the TEXAS database. Each new episode was preceded by at least 2 exacerbation-free weeks or 2 weeks with missing data [
We used generalized estimating equation regression analyses to estimate the effect of the mHealth tool on changes between baseline and follow-up scores of the self-efficacy scale, NCSI, CCQ, and EQ-5d compared with the paper action plan. We analyzed the CCQ and EQ-5d with all 5 measurement time points. We used 2-tailed
Statistical significance was assumed at
Of the 87 patients included in the study, 43 were randomized to the intervention group. In addition, 45 patients were recruited from the hospitals and 42 from the general practices. Among them, 13% (11/87) dropped out of the study, 16% (7/43) in the intervention, and 9% (4/44) in the control group. A flowchart of the participants in the study is shown in
Mean duration of follow-up was 48.1 (SD 11.7) weeks, and 11 COPD-related hospital admissions (6 in the intervention group and 5 in the control group) were reported as serious adverse events to the medical ethics review board.
From the Web-based interface, it appeared that 38 of the 43 patients (88%) in the mHealth group used the app 727 times in total during follow-up. No data on usage was available for 5 patients. The range in frequency of usage was 1 to 250 times with a median of 7 (25%-75% interquartile range was 3-14). Results of the evaluation questionnaire showed that more patients reported to have used their mHealth tool often (scores 6 and 7 on the 7-point rating scale) compared with patients in the control group who reported to have used their paper action plan (44.4% vs 17.2%, respectively).
Flow diagram of the participants through the study. mHealth: mobile health.
Baseline and demographic characteristics of the study population (N=87) per treatment arm.
Characteristic | Mobile health tool intervention group (n=43) | Paper action plan control group (n=44) | |
Recruited in hospital, n (%) | 21 (49) | 24 (55) | |
Follow-up in weeks, mean (SD) | 48.3 (12.6) | 49.8 (10.9) | |
Age (years), mean (SD) | 69.3 (8.8) | 65.9 (8.9) | |
Male sex, n (%) | 25 (58) | 29 (66) | |
Postbronchodilator FEV1a (% predicted), mean (SD) | 53.0 (21.5) | 52.1 (19.8) | |
Medical Research Council dyspnea score, mean (SD) | 2.5 (1.2) | 2.6 (1.3) | |
Currently smoking, n (%) | 13 (30) | 11 (25) | |
Use of paper action plan prior to follow-up, n (%) | 11 (26) | 17 (39) | |
|
|||
|
Long-acting bronchodilators | 27 (63) | 26 (59) |
|
Short-acting bronchodilators | 30 (70) | 31 (71) |
|
Inhaled corticosteroids (ICS) | 7 (16) | 13 (30) |
|
Long-acting bronchodilators + ICS | 22 (51) | 24 (55) |
Low education level, n (%) | 19 (44) | 17 (39) | |
Diagnosis of COPD >5 years, n (%) | 29 (68) | 28 (64) | |
Currently working, n (%) | 6 (14) | 10 (23) | |
|
|||
|
Joint disorders | 13 (30) | 13 (30) |
|
Cardiac disorders | 12 (28) | 12 (27) |
|
Back pain | 8 (19) | 14 (32) |
|
Diabetes | 3 (7) | 3 (7) |
|
Depression and/or anxiety | 3 (7) | 2 (5) |
aFEV1: forced expiratory volume in 1 second.
Comparison of exacerbation-related outcomes between intervention group and control group (N=85).
Outcome | Mobile health tool intervention group (n=41), mean (SD) | Paper action plan control group (n=44), mean (SD) | Rate ratio (95% CI)a | |
Exacerbation-free weeksb | 30.6 (13.3) | 28.0 (14.8) | 1.21 (0.77-1.90) | .40 |
Unscheduled health care consultations because of respiratory complaintsc | 1.6 (1.7) | 1.6 (2.0) | 0.89 (0.50-1.60) | .70 |
Symptom-based exacerbationsb | 4.5 (2.3) | 4.3 (2.1) | 1.07 (0.65-1.75) | .80 |
Exacerbations treated with antibiotics and/or prednisolonec | 1.1 (1.5) | 1.0 (1.3) | 1.01 (0.53-1.93) | .97 |
Exacerbation-related hospital admissionsc | 0.15 (0.43) | 0.14 (0.41) | 1.25 (0.35-4.44) | .74 |
aCalculated using negative binomial regression analyses, relative to participants’ follow-up time, controlling for age and gender.
bData retrieved from weekly patient reports.
cData retrieved from patient medical files.
Patients in the intervention group did not differ statistically significantly from patients in the control group in the number of weeks without exacerbations (mean 30.6 weeks, SD 13.3 weeks vs mean 28.0 weeks, SD 14.8 weeks, respectively; rate ratio [RR] 1.21; 95% CI 0.77-1.91; see
A total of 377 symptom-based exacerbation episodes were identified.
We found no statistically significant difference in exacerbation-related self-efficacy between the intervention and control groups when comparing baseline scores with 12-month follow-up scores. In addition, there were no differences between the groups in changes between baseline and follow-up scores of the subscales of the NCSI, CCQ and EQ-5d (
A total of 58 (67%) participants returned an evaluation form, of which 28 were in the intervention group. The mHealth tool was rated as a more useful support tool than the paper action plan (
Self-reported self-management behavior during exacerbation onset (N=377 exacerbations).
Self-management action | Mobile health tool intervention group (n=187), n (%) | Paper action plan control group (n=190), n (%) | Odds ratio (95% CI)a | |
Contact health care professional | 61 (32.6) | 68 (35.8) | 0.94 (0.51-1.73) | .83 |
Start prednisolone and/or antibiotics | 64 (34.2) | 62 (32.6) | 1.16 (0.55-2.44) | .69 |
Increase bronchodilator use | 135 (72.2) | 135 (71.1) | 1.08 (0.56-2.06) | .82 |
aCalculated using multilevel logistic regression analyses, including participant as cluster variable, controlling for age and gender.
Self-reported self-management behavior within 3 days of exacerbation onset.
Self-management action | Mobile health tool intervention group | Paper action plan control group | Odds ratio (95% CI)a | |||
n | <3 days, n (%) | n | <3 days, n (%) | |||
Contact health care professional | 55 | 20 (36.4) | 63 | 17 (27.0) | 2.21 (0.78-6.23) | .13 |
Start prednisolone and/or antibiotics | 58 | 23 (39.7) | 57 | 23 (40.4) | 1.46 (0.48-4.42) | .50 |
Increase bronchodilator use | 122 | 87 (71.3) | 117 | 80 (68.4) | 1.15 (0.61-2.16) | .67 |
aCalculated using multilevel logistic regression analyses, including participant as cluster variable, controlling for age and gender.
Baseline and follow-up scores of exacerbation-related self-efficacy and measures of health status.
Outcome | Mobile health tool intervention group | Paper action plan control group | |||||
Baseline (n=43), mean (SD) | 12-month (n=35), mean (SD) | Baseline (n=44), mean (SD) | 12-month (n=41), mean (SD) | Beta (95% CI)a | |||
Exacerbation-related self-efficacyb | 2.91 (0.43) | 2.98 (0.41) | 2.84 (0.41) | 2.87 (0.52) | 0.03 (−0.17 to 0.22) | .91 | |
|
|||||||
|
NCSId QOLe | 12.90 (8.32) | 13.01 (8.27) | 19.08 (11.93) | 17.11 (12.14) | 2.53 (−1.28 to −6.33) | .19 |
|
NCSI HRQOLf | 4.19 (1.74) | 4.00 (1.61) | 4.68 (1.78) | 4.76 (1.93) | −0.16 (−0.89 to −0.57) | .66 |
|
NCSI relationship | 2.42 (0.93) | 2.58 (0.87) | 3.39 (1.74) | 3.24 (1.59) | 0.30 (−0.26 to −0.85) | .29 |
|
NCSI subjective impairment | 11.84 (5.80) | 11.20 (4.12) | 14.41 (6.57) | 13.22 (6.57) | 0.74 (−1.38 to −2.85) | .5 |
|
NCSI behavioral impairment | 22.11 (17.89) | 20.50 (15.58) | 19.11 (17.28) | 20.43 (21.76) | −1.77 (−7.20 to −3.66) | .52 |
|
NCSI subjective symptoms | 9.86 (4.88) | 9.23 (4.39) | 11.91 (4.86) | 11.05 (4.79) | 0.35 (−1.40 to −2.11) | .69 |
|
NCSI dyspnea emotions | 8.77 (2.42) | 8.71 (2.86) | 11.59 (3.92) | 10.73 (4.32) | 0.85 (−0.58 to −2.27) | .24 |
|
NCSI fatigue | 35.93 (10.96) | 35.23 (9.45) | 37.32 (10.17) | 37.73 (10.20) | −1.80 (−5.43 to −1.84) | .33 |
|
CCQg total | 2.06 (1.02) | 1.84 (0.77) | 2.31 (1.09) | 2.16 (1.05) | −0.06 (−0.38 to −0.26) | .7 |
|
CCQ symptoms | 2.41 (1.12) | 2.16 (0.80) | 2.60 (1.27) | 2.49 (1.24) | −0.22 (−0.67 to −0.23) | .34 |
|
CCQ functional status | 2.22 (1.38) | 2.03 (1.21) | 2.53 (1.36) | 2.41 (1.32) | 0.05 (−0.30 to −0.40) | .76 |
|
CCQ mental status | 1.03 (1.01) | 0.81 (0.79) | 1.30 (0.99) | 1.00 (1.05) | 0.09 (−0.34 to −0.53) | .68 |
|
EQ-5dh | 0.81 (0.15) | 0.79 (0.16) | 0.74 (0.20) | 0.77 (0.21) | −0.05 (−0.13 to −0.03) | .22 |
|
EQ VASi | 65.53 (17.37) | 70.94 (12.92) | 64.20 (15.35) | 62.63 (19.14) | 6.28 (−0.56 to −13.11) | .07 |
aBeta (B) indicates the difference between the intervention and control groups on differences between baseline and at 12 months.
bHigher score is positive.
cHigher score is negative.
dNCSI: Nijmegen Clinical Screening Instrument.
eQOL: quality of life.
fHRQOL: health-related quality of life.
gCCQ: Clinical Chronic Obstructive Pulmonary Disease Questionnaire.
hEQ-5d: EuroQol-5 dimensions.
iEQ VAS: EuroQol Visual Analogue Scale.
In this study, we examined the clinical effects of a smart mHealth tool to support COPD patients in the detection and treatment of exacerbations without the interference of a health care professional. Our primary hypothesis that the use of mHealth would lead to more weeks without exacerbations than care as usual, that is, the use of a paper action plan, was not confirmed. In addition, we did not find differences in exacerbation frequency, health care utilization, or self-management behavior between patients who used the mHealth tool and patients who used the paper action plan. Furthermore, patients using the tool did not report higher exacerbation-related self-efficacy or better health status than patients using a paper action plan. Patients evaluated the usability of the mHealth tool as good and considered it as more supportive than the action plan.
So far, studies on the effects of electronic health (eHealth) in the management of COPD have focused on the use of telemonitoring apps [
In this study, we could not demonstrate positive effects on exacerbation-free weeks, exacerbation frequency, and health care utilization. Our primary outcome,
The major strength of this study is that we used a well-designed and validated mHealth tool to support self-management behavior [
However, this study also has limitations. Two important limitations may have led to the statistically nonsignificant results of our primary outcome exacerbation-free weeks. First, we were surprised that the mean exacerbation-free time and its standard deviation at the 12-month follow-up differed substantially from the study data on which our power calculation was based [
There are also other limitations. Of the 467 patients that responded to the study invitation and met the inclusion criteria, 283 were not willing to participate. This may have led to selection bias. We believe that the risk of contamination because of the individual randomization procedure is negligible, as both the mHealth tool and the paper action plan were used at home, outside the reach of the health care professional. Although in the last 4 months of the trial we had to switch to collecting exacerbation-related outcomes through a digital survey tool instead of the automated TEXAS telephone calls [
Although we were not able to demonstrate positive effects of our mHealth tool, we still believe that the use of eHealth tools, including machine learning techniques, better suits the goals of patient-centered care and self-management support than telemonitoring tools where a health care professional monitors the patient from a distance. Besides, in this study, the patients using the mHealth tool evaluated it as usable and more supportive than patients using a conventional supportive tool, that is, the written paper action plan. Future research should focus more on patients who are specifically interested in using digital tools in their daily life, as these patients may have greater benefit from them [
In this study, we examined the effectiveness of an mHealth tool designed to support COPD patients in their self-management of symptom worsening to reduce the impact of exacerbations. The app was not designed to replace the health care professional but to reduce patient delay. Patients evaluated the app’s usability as good and as more supportive than the paper action plan. Although this study did not show beneficial effects of the mHealth app compared with the use of a paper action plan, based on patient’s preference, it may be a valuable alternative to a paper action plan in the management of COPD.
Contents of the mobile health tool.
CONSORT-EHEALTH checklist (V 1.6.1).
Clinical Chronic Obstructive Pulmonary Disease Questionnaire
chronic obstructive pulmonary disease
electronic health
EuroQol-5 dimensions
forced expiratory volume in 1 second
mobile health
Nijmegen Clinical Screening Instrument
odds ratio
quality of life
rate ratio
System Usability Scale
Telephonic Exacerbation Assessment System
The authors would like to thank all patients who participated in this study for their commitment to the study as well as all team members, especially the nurses and physicians of the participating hospitals and general practices.
All authors contributed to the design of the study. LB, EB, and JV supervised the collection of the data. LB and RA performed the statistical analysis. LB led the writing of the paper, which was coled by EB, TS, and WA and assisted by all others. All authors had full access to all of the study, assisted in the interpretation of the data, and have seen and approved the final version of the paper.
None of the authors received any support from any company for the submitted work. LB, MvdH, RA, YH, JV, and WA have nothing to disclose. PL has a patent 20140206949 issued to Petrus Lucas, EB received personal fees from Boehringer Ingelheim and GlaxoSmithKline outside the submitted work, and TS received a grant from GlaxoSmithKline and personal fees from Boehringer Ingelheim outside the submitted work.