Research: Effects of Envol app on Symptoms, Overall Well-being and Energy levels

Introduction:

Background and Rationale

Chronic illness is a condition that lasts 1 year or more, requires ongoing medical attention and can prevent an individual from carrying out the simplest of tasks (Hodgkin, 2018). It is the leading driver of world death, and affects over 20% of the world's population (CDC Chronic Disease). For many patients, chronic or recurrent pain has a negative influence on their physical, social, and mental health and interferes with daily functioning, self-esteem and quality of life (Kristofferzon et al., 2018). This poses a great burden on society and highlights the importance of health research.

In wellness, there is never going to be a single approach that fits all, and interindividual approaches are crucial for addressing symptoms in diverse populations (Bland et al., 2017). Therefore, it is necessary to offer multidisciplinary therapeutic approaches that focus on, not just pain reduction, but enhancing daily functioning (Hodgkin, 2018). As such, frameworks such as interdisciplinary multimodal pain therapy (IMPT) have emerged as approaches to orientate goals towards assisting patients to live meaningful lives, despite their pain. These multidisciplinary frameworks draw from a biopsychosocial model, and frequently incorporate both pain physiology models and (cognitive) behavioural treatment principles (Kaiser et al., 2018).

The effectiveness of interdisciplinary, biopsychosocial models in facilitating patient wellbeing has been thoroughly demonstrated (Willis, 2016), but maintaining positive results on patients’ daily lives, over time, remains a key issue. Relapse is a problem that has been found across many health behaviour areas, not just in the realm of pain treatment (Elber s at al., 2020 ). As a result, several treatment programs have incorporated relapse prevention measures to help patients maintain their treatment achievements over time (Chadi et al., 2017; Nierenberg et al., 2016). Self-practice exercises (Shim et al., 2017) and encouraging patients to take notes throughout therapy via diaries (Nes et al., 2013) are examples of such tactics in the context of chronic pain and illness treatment.

Self-management support in particular is a popular method used by many clinicians dealing with patients with chronic illnesses (Gallant, 2003). This model commits to a patient-centered approach and increases the patients involvement in their health decision making. By helping patients to set goals and creating decimated follow up programs, clinicians can encourage chronic illness patients to take a more active role in managing their own health (Kralik et al., 2004). Nevertheless, many clinical studies neglect to fully describe the integration of these specific methods within the treatment program, as well as the underlying theoretical reasoning for how they might prevent relapse or persistent challenges in patient wellbeing (McDermott & While, 2013). This calls for further exploration into methods for effectively promoting long-term wellness.

Mobile Health

As the technological age evolves, more and more opportunities arise for the integration of technology to maximize health care. mHealth consists of self-management support applications delivered via mobile phones, and open new avenues for helping sustain and improve long-term treatment effects in individuals with chronic illness (Buck et al., 2021). These applications can offer psychological capabilities to patients by converting complicated clinical information into more digestible forms, providing medication reminders, offering behaviour regulation strategies, and offering a space for digital health communities to share treatment information and shared challenges (Marcolino et al., 2018). Furthermore, if used regularly, mHealth can promote reflective motivation, since patients can look back and evaluate how their health has progressed (Kampmeijer et al., 2016). In particular, research has found that mHealth can include regular self-management techniques in patients and lead to healthier habits (Smith et al., 2016).

However, despite mHealth’s potential and current popularity, the performance of mHealth apps on health related outcome metrics varies widely (Marcolino et al., 2018). This is due to the diverse ease of application use to achieve goals as well as level of engagement with different applications (Brown III et al., 2013). This current study seeks to assess the utility of the Envol app specifically in improving chronic illness symptoms. The hypothesis is that with more regular usage, Envol will provide greater benefits to patient’s health related outcomes and symptom severity.

Part 1: Health outcomes survey

Methodology

To assess the efficacy of Envol, individuals were recruited worldwide to use the app and take the SF-36 Health Survey to measure results after one month of usage. The SF-36 is a 36-item questionnaire that assesses health and wellbeing across several dimensions. It is one of the top scales used in medical research to evaluate quality of life, and has been used in similar studies in the past to test the effectiveness of health programs, including health mobile apps (Cajita et al., 2016). All respondents confirmed to have read the information sheet, understood that their participation is voluntary and that they are free to withdraw at any time without giving any reason. Also, all the respondents reported understanding that relevant sections of their medical notes and data collected during the study may be looked at by individuals from Envol and gave permission for this. Finally, all respondents confirmed to be 18 years or older, and agreed to take part in this study.

Statistical analysis

Before proceeding to the main statistical analyses, the reliability of the scales was assessed by computing Cronbach alpha coefficients. In addition, before conducting the main analyses, the assumptions of the normality of distributions and the absence of outliers were tested. In the cases in which the assumptions were met, or in which the sample size was large enough for the Central limit theorem to be applicable (Kwak & Kim, 2017), paired sample t-tests were used to compare the scores obtained at the start and the end phase. When the sample size was too small for the Central limit theorem to be applicable, a non-parametric equivalent of the paired samples t-test (i.e., related samples Wilcoxon signed-rank test) was performed to compare the scores from the two phases. All the analyses were conducted using the SPSS, version 25 (IBM Corp, 2017).

In order to directly compare the scores obtained in the start and the end phase, the two datasets were merged using the respondents’ unique identification numbers. Respondents for whom the data were missing in either of the two phases were excluded from the final sample. The final sample consisted of 218 respondents.

Reliability analysis

Before proceeding to the main analysis, reliability was assessed using Chronbach alpha coefficient. Reliability coefficients were computed separately for each of the two phases (Table 1). All coefficients were deemed to be at least acceptable (Cronbach α > .7), and were comparable to those reported at the official website of the RAND corporation.

Table 1

Reliability: Chronbach alpha coefficients for eight health variables

Scale	Items	Alpha_start	Alpha_end
Physical functioning	10	.93	.93
Role functioning/physical	4	.86	.84
Role functioning/emotional	3	.79	.84
Energy/fatigue	4	.80	.89
Emotional well-being	5	.83	.85
Social functioning	2	.87	.86
Pain	2	.90	.92
General health	5	.79	.81

Outliers and normality of the distributions

The absolute Z values > 1.96 are considered as potential outliers, whereas the absolute values > 2.58 are considered probable outliers. Only the values > 3.29 are thought to be extreme outliers (Field, 2018, p.340). In the present sample, none of the values had a Z score equal to or higher than 3.29, so none of the values was treated as an outlier.

The central limit theorem postulates that, with large samples, sampling distributions are normal even if raw scores are not (Tabachnik & Fidell, 2013). That is, according to the central limit theorem, the shape of the data does not affect significance tests if sample is large enough, meaning that the statistical tests in this case are robust (Field, 2018, pp.331). As the sample size in our case was large enough, parametric tests were used.

Results on the use of Envol: Descriptive statistics

All respondents reported using the iOS operating system. In terms of the prior use of Envol, the majority of respondents (96.3%) reported not using the app in the past.

Over the course of this study, 97 respondents (45.5%) reported using Envol either several times a day or at least once a day. Additional 35.7% reported using the app 3-5 days a week. Similar distribution was observed for using Envol Toolbox including meditations, 3D sound journeys, visualization music, affirmations and breathing (Table 2).

Table 2

Use of Envol and Envol toolbox

	Frequency	Percent
Use of Envol
Several times a day	45	21.1
Once a day	52	24.4
3-5 days a week	76	35.7
1-2 days a week	29	13.6
Less often	11	5.2
Use of Envol toolbox
Several times a day	34	15.9
Once a day	56	26.2
3-5 days a week	69	32.2
1-2 days a week	38	17.8
Less often	17	7.9

In addition to being asked about the frequency of using Envol, respondents were asked about the reasons for doing so. Specifically, they were asked to rate the extent to which they agree with using Envol for the following reasons: to increase their energy, reduce the symptom severity, feel better and improve health. Of the four reasons, “using Envol to feel better” was endorsed by the highest percentage of respondents. Specifically, 84.1% of respondents agreed or strongly agreed to using the app for that particular reason. The second most frequent reason was to improve health, reported by 56.5% of the respondents. As for symptom severity and increased energy, 37.4% and 40.6% (for detailed review see Table 3) rated using Envol for these reasons, respectively (for detailed review see Table 3).

Table 3

Reasons for using Envol

Reason for using Envol	Frequency	Percent
To increase energy
Strongly agree	12	5.6
Agree	68	31.8
Neutral	113	52.8
Disagree	16	7.5
Strongly disagree	5	2.3
To reduce symptom- severity
Strongly agree	12	5.6
Agree	75	35.0
Neutral	100	46.7
Disagree	22	10.3
Strongly disagree	5	2.3
To feel better
Strongly agree	51	23.8
Agree	129	60.3
Neutral	31	14.5
Disagree	3	1.4
To improve health
Strongly agree	24	11.2
Agree	97	45.3
Neutral	81	37.9
Disagree	10	4.7
Strongly disagree	2	.9

Results of Envol on improving mental wellbeing:

To evaluate the impact of Envol on improving mental wellbeing, we asked users to fill out the SF-36, a self-report survey on quality of life. Based on these answers, we were able to assess improvement on the following key performance indicators (KPIs): users' overall mood, anxiety symptoms, depressive symptoms, energy levels, and productivity. Relevant performance indicators were combined by summing up weighted scores on individual survey items. The ratio between before and after measurements, minus one, was used to calculate subsequent improvement scores.

Results:

92% improvement in overall mood across users
86% improvement in managing anxious symptoms across users
82% improvement in managing depressive symptoms across users
92% improvement in energy levels across users
71% improvement in productivity levels across users

Direct comparison between the scores obtained in two time points (original sample)

In order to directly compare the start and end scores on the eight variables of interest (1. Physical functioning, 2. Role functioning/physical, 3. Role functioning/emotional, 4. Energy/fatigue, 5. Emotional well-being, 6. Social functioning, 7. Pain, 8. General health perceptions), a series of paired-sample t-tests were computed.

Results showed that all the differences were statistically significant. Specifically, in the case of each of the eight health concepts measured with the SF36, the higher scores were obtained in the end phase, compared to the start phase (Table 4).

Table 4

Comparison of the eight health concepts between start and end phase (original sample)

Health concept	M_start	SD_start	M_end	SD_end	t	df	p
Physical functioning	70.28	27.29	74.40	26.02	-4.172	217	.000
Role functioning/physical	29.70	38.00	46.67	40.84	-6.580	217	.000
Role functioning/emotional	33.95	39.37	55.56	42.65	-6.680	215	.000
Energy/fatigue	30.86	19.77	44.61	23.44	-10.440	215	.000
Emotional well-being	51.46	19.35	63.85	18.50	-10.208	215	.000
Social functioning	44.33	26.35	58.62	28.84	-9.142	215	.000
Pain	51.79	25.58	65.39	25.81	-9.009	215	.000
General health	42.85	20.29	52.31	21.23	-10.624	213	.000

Note: To ensure that non-normality of some of the variables did not bias the results, the eight comparisons were repeated using a non-parametric equivalent to paired-samples t-test, namely the related-samples Wilcoxon signed-rank test. All the differences remained statistically significant.

It is worth noting that 34.1% of respondents reported using not only Envol, but other health / wellness / meditation apps in the past few weeks. Thus, the positive effects registered in this study might not be entirely attributable to Envol. However, when asked if they would recommend Envol to their friends and family, the majority of respondents (95.8%) responded positively.

Direct comparison between the scores obtained in two time points (subsample of Envol users who did not use other similar apps)

The first part of the analyses was performed using the whole sample. However, the original sample included not only the respondents who exclusively used Envol, but also those who, in addition to Envol, have used other similar apps. To get a more valid estimation of the efficiency of Envol in improving several health aspects, those who used other similar apps in the past couple of weeks were excluded from the sample, and the analyses are then repeated.

As was the case with the original sample, results showed statistically significant differences in terms of the eight health concepts. Specifically, the higher scores on each of the eight concepts were obtained in the end phase, compared to the start phase, with the higher scores indicating higher levels of health (Table 5).

Table 5

Comparison of the eight health concepts between the start and the end phase

Health concept	M_start	SD_start	M_end	SD_end	t	df	p
Physical functioning	69.82	27.43	73.90	26.34	-3.596	140	.000
Role functioning/physical	29.08	37.40	46.99	40.47	-5.299	140	.000
Role functioning/emotional	31.44	37.96	51.54	42.44	-4.876	140	.000
Energy/fatigue	29.36	19.37	44.50	23.48	-9.298	140	.000
Emotional well-being	50.13	19.73	64.06	18.37	-9.049	140	.000
Social functioning	44.77	24.69	60.55	28.16	-8.240	140	.000
Pain	50.98	24.96	66.40	26.01	-8.090	140	.000
General health	41.84	19.32	51.52	20.74	-8.871	140	.000

In addition to examining the differences concerning the health concepts, we examined the differences in symptom severity between the start and the end phase. Specifically, respondents were asked to rate the symptom severity at the moment of completing the survey, on average in the past few weeks prior to completing the survey, and at their worst in the past few weeks.

Results showed that the symptom severity was reduced in the end phase, compared to the start phase, regardless of the measure of symptom severity that was observed (right now, on average, or at worst) (Table 6).

Table 6

Comparison of the symptom severity between the start and the end phase

Symptom severity	M_start	SD_start	M_end	SD_end	t	df	p
right now	6.50	2.11	5.50	2.26	4.498	105	.000
on average	6.70	1.78	5.96	2.10	4.219	105	.000
at worst	8.45	1.93	7.72	2.01	3.477	105	.001

Note: To ensure that non-normality of some of the variables did not bias the results, the three comparisons were repeated using a non-parametric equivalent to paired-samples t-test, namely the related-samples Wilcoxon signed-rank test. All the differences remained statistically significant.

Direct comparison between the scores obtained in two time points (subsample of those who calculated their scores daily and did not use other apps)

When the analyses were conducted on the subsample of respondents who used Envol daily (and did not use any other similar app), significant differences were found on six out of eight health concepts. Specifically, differences were significant for all health concepts except for physical functioning and role functioning physical (Table 7). In all other cases, the health scores were higher in the end phase, suggesting improved health in the end, compared to the start phase. It is worth noting that, even though the difference did not reach statistical significance in the case of physical functioning and role functioning/physical, the average end scores were higher than the average start scores in these two cases as well. One should have this in mind, especially as the sample size was small, making it more difficult to detect the significant differences even if they truly exist.

Table 7

Daily users: Comparison of the eight health concepts between the start and the end phase

Health concept	M_start	SD_start	M_end	SD_end	t	df	p
Physical functioning	57.73	30.62	61.82	32.05	-1.882	21	.074
Role functioning/physical	29.55	41.29	38.64	39.89	-1.053	21	.304
Role functioning/emotional	33.33	37.09	54.55	43.09	-2.246	21	.036
Energy/fatigue	27.05	21.31	40.68	24.36	-4.081	21	.001
Emotional well-being	51.64	22.18	68.36	14.81	-4.194	21	.000
Social functioning	39.20	21.58	56.25	30.31	-3.071	21	.006
Pain	44.89	24.60	61.93	29.25	-2.887	21	.009
General health	40.91	19.19	50.45	18.45	-4.149	21	.000

Note: To ensure that the non-normality of some of the variables did not bias the results (especially as the sample was rather small), the eight comparisons were repeated using a non-parametric equivalent to paired-samples t-test (the related-samples Wilcoxon signed-rank test). The results of the two sets of tests were almost identical (the only exception was role functioning/emotional, which was significant using the t-test (p=.036) and marginally significant (p=.05) using its non-parametric equivalent.

When we tested the differences concerning symptom severity, significant differences were found for “right now” and “on average in the past few weeks” (Table 8). The difference was not significant for the symptom severity “at worst”. However, as in the case of health concepts, it is worth noting that even “at worst”, on average, the symptoms were less severe in the end phase, albeit not reaching statistical significance.

Table 8

Daily users: Comparison of the symptom severity between the start and the end phase

Symptom severity	M_start	SD_start	M_end	SD_end	T	df	p
right now	6.25	2.81	5.10	2.20	2.981	19	.008
on average	6.85	2.11	5.85	1.93	3.082	19	.006
at worst	8.90	1.83	8.10	2.13	1.902	19	.072

Note: To ensure that non-normality of some of the variables did not bias the results, the eight comparisons were repeated using a non-parametric equivalent to paired-samples t-test, namely the related-samples Wilcoxon signed-rank test. The results were comparable to those obtained using t-test.

Direct comparison between the scores obtained in two time points (subsample of those who calculated their scores daily and used other apps in addition to Envol)

When the analyses were conducted on the subsample of respondents who used Envol daily (and also used another similar app), significant differences were found on seven out of eight health concepts. Specifically, differences were significant for all health concepts except for physical functioning (Table 9). In all other cases, the health scores were higher in the end phase, suggesting improved health in the end, compared to the start phase. It is worth noting that, even though the difference did not reach statistical significance in the case of physical functioning, the average end scores were higher than the average start scores in these two cases as well.

Table 9

Daily users: Comparison of the eight health concepts between the start and the end phase

Health concept	M_start	SD_start	M_end	SD_end	t	df	p
Physical functioning	66.41	30.56	67.69	31.26	-.754	38	.456
Role functioning/physical	30.13	39.39	46.79	41.43	-2.894	38	.006
Role functioning/emotional	41.03	39.34	64.96	41.14	-3.260	38	.002
Energy/fatigue	31.15	21.44	44.10	26.20	-4.383	38	.000
Emotional well-being	55.18	19.69	68.31	16.65	-4.396	38	.000
Social functioning	43.59	26.57	59.62	31.99	-3.743	38	.001
Pain	50.64	27.65	65.38	29.45	-3.869	38	.000
General health	43.33	21.01	54.23	19.82	-5.309	38	.000

Note: The results of the related-samples Wilcoxon signed-rank test were comparable to those obtained using t-test.

When we tested the differences concerning symptom severity, all the differences (right now, on average in the past few weeks, and at worst in the past few weeks) were statistically significant (Table 10). The symptoms were less severe in the end phase, compared to the start phase.

Table 10

Daily users: Comparison of the symptom severity between the start and the end phase

Symptom severity	M_start	SD_start	M_end	SD_end	T	df	p
right now	6.00	2.72	5.12	2.09	3.163	32	.003
on average	6.64	2.00	5.70	1.98	3.459	32	.002
at worst	8.55	1.66	7.70	2.05	2.727	32	.010

Note: The results of the related-samples Wilcoxon signed-rank test were comparable to those obtained using t-test.

Daily users

Daily users - Mobile app score (69 and less; 70 and more)

To compare the symptom severity scores at start and end phase in users with different mobile app scores, the sample was first split into: 1. daily users with the mobile app score of 69 and less and 2. daily users with the mobile app score of 70 and more. The analyses were then performed on each sample separately. Specifically, as the data did not deviate from normality, a series of t-tests was conducted. The only exception were “on average” variables in the subsample of those scoring 69 and less, for which certain deviation was inspected. Thus, only in this case a non-parametric equivalent to t-test was used (i.e. related samples Wilcoxon signed-rank test).

Daily users – score of 69 and less

In users with the mobile app score of 69 and less, the change in symptom severity was not significant (Table 11). Although the average symptom severity was reduced for all three measures, the change did not reach statistical significance.

Table 11

Comparison of the symptom severity in daily users with the score of 69 and less

Symptom severity	M_start	SD_start	M_end	SD_end	t (W)	df	p
right now	5.75	2.59	5.00	2.05	1.924	19	.069
on average	6.40	1.82	5.70	2.15	-1.925	19	.054
at worst	8.45	1.67	7.75	2.10	1.543	19	.139

Daily users – score of 70 and more

In users with the mobile app score of 70 and more, the change in symptom severity were significant, regardless of whether the symptom severity concerned “right now”, “on average” or “at worst” (Table 12).

Table 12

Comparison of the symptom severity in daily users with the score of 70 and more

Symptom severity	M_start	SD_start	M_end	SD_end	t	df	p
right now	6.42	3.09	5.25	2.30	2.880	11	.015
on average	7.17	2.29	5.83	1.75	3.546	11	.005
at worst	8.75	1.76	7.67	2.15	2.600	11	.025

Regular users

When the analyses were conducted on the subsample of regular users (those who calculated their score daily, almost every day, and 3-4 times a week), significant differences were found on all of the eight health concepts (Table 13). Specifically, the health scores were higher in the end phase, suggesting improved health in the end, compared to the start phase.

Table 13

Regular users: Comparison of the eight health concepts between the start and the end phase

Health concept	M_start	SD_start	M_end	SD_end	t	df	p
Physical functioning	67.95	29.40	71.67	28.24	-2.702	128	.008
Role functioning/physical	27.71	38.43	46.90	40.75	-5.761	128	.000
Role functioning/emotional	32.56	39.41	56.33	42.04	-5.615	128	.000
Energy/fatigue	31.09	20.78	44.22	24.52	-8.288	128	.000
Emotional well-being	52.56	18.81	64.12	18.50	-7.445	128	.000
Social functioning	43.60	25.96	57.46	30.28	-7.083	128	.000
Pain	50.29	25.70	64.34	26.93	-7.186	128	.000
General health	41.98	20.21	51.40	20.52	-9.051	128	.000

When we tested the differences concerning symptom severity, significant differences were found, regardless of the measure used (i.e., “right now”, “on average in the past few weeks”, “at worst in the past few weeks”). In other words, the symptom severity was significantly reduced in the end phase, compared to the start phase (Table 14).

Table 14

Regular users: Comparison of the symptom severity between the start and the end phase

Symptom severity	M_start	SD_start	M_end	SD_end	T	df	p
right now	6.15	2.20	5.36	2.18	4.230	104	.000
on average	6.62	1.82	5.67	2.12	5.817	104	.000
at worst	8.42	1.83	7.56	2.04	4.590	104	.000

Symptom severity in regular users – mobile app scores of 70 and below and 71 and above

To compare the symptom severity scores at the start and the end phase in regular users with different mobile app scores, the sample was first split into: 1. regular users with the mobile app score of 70 and less and 2. regular users with the mobile app score of 71 and more. The analyses were then performed on each sample separately.

Results showed that all the differences were statistically significant (Table 15). That is, regardless of whether the users had a score of 70 and below or 71 and above, the symptom severity decreased in the end phase, compared to the start phase.

Table 15

Comparison of the symptom severity in regular users

Mobile score	Symptom severity	M_start	SD_start	M_end	SD_end	t	df	p
70 and below	right now	6.13	2.22	5.44	2.28	2.916	77	.005
	on average	6.53	1.86	5.69	2.31	4.094	77	.000
	at worst	8.38	1.96	7.68	2.16	2.980	77	.004
71 and above	right now	6.27	2.27	5.05	1.94	4.533	21	.000
	on average	7.05	1.81	5.59	1.53	5.575	21	.000
	at worst	8.50	1.54	7.23	1.80	4.537	21	.000

Symptom severity in regular users – mobile app scores of 69 and below and 70 and above

To compare the symptom severity scores at the start and the end phase in regular users with different mobile app scores, the sample was first split into: 1. regular users with the mobile app score of 69 and less and 2. regular users with the mobile app score of 70 and more. The analyses were then performed on each sample separately.

Results showed that all the differences were statistically significant (Table 16). That is, regardless of whether the users had a score of 69 and below or 70 and above, the symptom severity decreased in the end phase, compared to the start phase.

Table 16

Comparison of the symptom severity in regular users

Mobile score	Symptom severity	M_start	SD_start	M_end	SD_end	t	df	p
69 and below	right now	6.12	2.20	5.44	2.31	2.779	74	.007
	on average	6.53	1.83	5.68	2.31	4.042	74	.000
	at worst	8.39	1.96	7.67	2.16	2.930	74	.005
70 and above	right now	6.28	2.32	5.08	1.87	4.648	24	.000
	on average	6.96	1.95	5.64	1.63	5.431	24	.000
	at worst	8.48	1.61	7.32	1.86	4.529	24	.000

Part 2: Semi-structured group interview

Methodology

To gather further insights into Envol app usage and individual experiences, as well as to validate and contextualize findings from the survey results, a semi-structured group interview was conducted. Enhancing survey findings with qualitative research is particularly useful when the objective is to gain a better understanding of a complex topic such as wellbeing, and to understand health behavior in context (Todres et al., 2009). In particular, semi-structured interviews are suited for gathering qualitative data because, unlike self-administered surveys, they allow for discussion and examination of new topics that develop during the data collection process. Moreover, semi-structured interviews offer a lot of flexibility in terms of how questions are asked and how responses are categorized, as open-ended questions can encourage a variety of responses (Adams, 2010). This form of data collection decreases the risk of bias caused by the researcher's preconceptions and enables the use of elaboration probes to inspire participants to talk freely about a given topic (Patton, 2014). Preparation for the semi-structured group interview consisted of a list of guiding questions on chronic illness experiences and Envol app usage perceptions.

Participants

As part of the aforementioned survey, participants were also asked to record their interest in participating in further discussion via a semi-structured group interview. A total of 3 participants from the survey cohort formed the final interview group. The three participants, all female, came from diverse backgrounds and life circumstances, including recent pregnancy, unemployment, chronic pain, chronic illness, chronic exhaustion, and COVID-19-related social isolation.

Themes identified

Several themes emerged during the semi-structured group interview, including: 1) the need for hope, 2) the app’s visualization feature for facilitating gratitude, 3) the versatility of the app’s meditation feature for wellness goals, and 4) the app’s ease of integration into daily living.

The need for hope

All three participants highlighted the need for hope in their individual wellness journeys. This need for hope anchored their search for different wellness tools for minimizing their day-to-day suffering, including exhaustion, chronic pain, and stress. One participant aptly noted: “When you're in the pain cave, you know, you can't necessarily see your movement forward, you just pray and hope.” Another participant highlighted the need for resilience in facing daily chronic pain and illness: “You just have to embrace the suck, you know, and just go with it, and do the best you can to stay positive.” Having different tools such as Envol was a source of encouragement and optimism for participants to minimize their pain and therefore, manage their symptoms better.

Envol’s visualization feature for facilitating gratitude

All three participants cited Envol’s utility in helping them visualize hope and in turn, facilitate gratitude for daily living. One participant noted: “It's a very visual app, you know, so it's really helped me [in feeling gratitude]. Similarly, another participant voiced: “I think it was very helpful actually, to start this visualization process.” A key theme was the need to have gratitude for the opportunity of life, even in the face of challenges including chronic illness and pain, as a key factor in maintaining resilience and striving towards improved wellness. Another particularly poignant quote from a participant included: “I have to search for something to be grateful for. [The app reminds me] I'm grateful that I have a body that works, you know, I'm grateful that I'm going through this journey, because I'm going to heal myself.” The discussion therefore identified Envol as a key factor in empowering individuals to, despite their suffering, feel grateful for their bodies and empowered in their wellness journey. The distinctly visual user experience of the app was critical to strengthening these resilient characteristics.

Envol’s meditation feature for different wellness goals

All three participants also cited Envol’s meditation feature as a hallmark of the app. Each participant discussed different ways they used the meditation feature for maintaining wellness in daily living. Some of these were general wellness goals; for example, as noted by one participant: “It helps to have these meditations for taking time for yourself, having your affirmations (...) you know, just incorporating all these things to augment my healing.” Another participant cited using the meditations feature for combatting sleeping issues as a new parent: “I think the meditations have helped me a lot, too, particularly for when [it’s helped me to be] able to rest and recover. Sometimes when I wake up during the night and feed [my newborn], I listen to a short meditation, and I can go to sleep faster.” These comments highlight the versatility of the app’s meditation feature for facilitating different wellness goals, and its flexibility for providing healing support for individual needs.

Integration into daily life

All three participants discussed using Envol regularly and relying on it as a daily tool for monitoring progress. The personalization of the app’s metrics was cited as particularly helpful in this regard. For example, one participant noted: “It's being able to see those metrics and have tools that are designed for you (...) tailored to your particular needs that is key.” Another participant likened Envol to “a tool in my daily life” that helped them keep motivated in their wellness journey, further elaborating: “I would use it quite often to breathe or, you know, if I was going through some particular bad patch, I would go and I would breathe and so I would check in with it often. And always first thing in the morning, which was a great thing.” Similarly, another participant noted: “I charted my metrics each day, and my sleep and my food and all of that.” These comments highlight the holistic, all-encompassing nature of the app for integration into users’ daily routines. Participants voiced genuine enjoyment towards using the app regularly for managing their symptoms and keeping themselves accountable towards their wellness goals.

Discussion

The aim of this research was to investigate whether mHealth opens new avenues for chronic illness management, in order to improve long term treatment impacts and overall health. Specifically, the potential of mHealth application use, with a particular focus on the Envol app, was evaluated over a 1-month pilot study.

Underpinned by the IMPT model, the overarching hypothesis was that, with more frequent mHealth use, there will be a greater improvement in symptom severity and overall health. The current results support this hypothesis, reporting that overall, despite how severe symptoms were, regular mHealth usage was associated with improved mental and physical health outcomes four weeks later. This remained the case, despite how severe symptoms were, whether it was assessing symptoms right now, on average and/or at worst. Moreover, qualitative assessment revealed how Envol’s visualization and meditation features were particularly beneficial to users. Moreover, users appreciated the flexibility of the app for integration into daily wellbeing activities and regarded the app as a key component in maintaining hope and resilience in their individual wellness journeys.

These results suggest that mHealth related apps, such as Envol, may benefit individuals’ health, as well as symptom severity, when used regularly (either daily, almost daily or 3-4 times a week). Among individuals who only used Envol daily, there were significant improvements in all health concepts, except physical functioning and role functioning/physical. However, it is worth noting, that while only six out of eight health concepts improved at the statistical significance levels, all concepts did in fact show improvements. As for symptom severity, daily Envol use improved symptom severity ‘right now’ and ‘on average’, but not for when ‘at worst’. Furthermore, when taking Envol scores into account, those that scored ’70 or more’ showed significant improvements in all three severity ratings; however, those that scored 69 or less only showed significant improvements for symptom severity ‘right now’ and ‘on average’, but not ‘at worst’. This implies that, in order to gain maximum health benefits from Envol, efforts to engage with the application and score highly are important - especially for those experiencing symptoms at their worst.

These conclusions are strengthened when controlling for use of other health related applications. For instance, when assessing Envol usage along with other health related applications, such as mindfulness training, we see significant improvements in all health domains, which was consistent regardless of symptom severity. However, when controlling for users who conjunctively used these other applications, and when only assessing those who used Envol, we still saw substantial and significant improvements in most health-related outcomes. Furthermore, when asked whether participants would recommend Envol to their friends and family, 95.8% responded positively. These findings show promise for the use of Envol specifically, in helping individuals manage and improve their health-related outcomes. The findings also indicate that, in order to benefit the most from Envol, daily usage seems to offer the most potential at improving health related outcomes. Indeed, among regular users, regardless of whether users had a score of below or under 70, symptom severity decreased at statistically significant levels compared to prior to Envol usage. This finding was qualitatively validated in the semi-structured discussion, in which participants expressed getting the most out of the app when using it to regularly track and manage their symptoms and keep themselves focused on their wellness goals on a daily basis.

Future Directions

Given the nature of the study, no control group of users who do not use any mHealth applications, or of users who do not have chronic illness challenges, was included. Future studies should include a control group to strengthen our conclusions that personalized mHealth applications provide health benefits for individuals.

Another limitation is the short longitudinal design of four weeks, which limits our conclusions of whether these beneficial outcomes will persist beyond this short time frame. Going forward, it will be crucial to increase the study duration to increase our knowledge on the long term effects of mHealth application use.

Notwithstanding, our present evaluation provides promising results for mHealth apps, as well as Envol specifically, in strengthening resilience and symptom management for individuals with chronic illness.

Bibliography

Adams, E. (2010). The joys and challenges of semi-structured interviewing. Community Practitioner, 83 (7).

Bland, J. S., Minich, D. M., & Eck, B. M. (2017). A systems medicine approach: Translating emerging science into individualized wellness. Advances in Medicine, 2017.

Brown III, W., Yen, P.-Y., Rojas, M., & Schnall, R. (2013). Assessment of the Health IT Usability Evaluation Model (Health-ITUEM) for evaluating mobile health (mHealth) technology. Journal of Biomedical Informatics, 46 (6), 1080–1087.

Buck, H. G., Shadmi, E., Topaz, M., & Sockolow, P. S. (2021). An integrative review and theoretical examination of chronic illness mHealth studies using the Middle-Range Theory of Self-care of Chronic Illness. Research in Nursing & Health, 44 (1), 47–59.

Cajita, M. I., Gleason, K. T., & Han, H.-R. (2016). A systematic review of mHealth-based heart failure interventions. The Journal of Cardiovascular Nursing, 31 (3), E10.

CDC chronic disease. (n.d.). https://www.cdc.gov/chronicdisease/about/index.htm

Chadi, N., Kaufman, M., Weisbaum, E., Malboeuf-Hurtubise, C., Kohut, S. A., Viner, C., Locke, J., & Vo, D. X. (2017). In-person versus ehealth mindfulness-based intervention for adolescents with chronic illness: Protocol for a randomized controlled trial. JMIR Research Protocols, 6 (11), e7700.

Elbers, S., Pool, J., Wittink, H., Köke, A., Scheffer, E., & Smeets, R. (2020). Mobile health app (AGRIPPA) to prevent relapse after successful interdisciplinary treatment for patients with chronic pain: Protocol for a randomized controlled trial. JMIR Research Protocols, 9 (8), e18632.

Elbers, S., Pool, J., Wittink, H., Köke, A., & Smeets, R. (2020). Exploring the Feasibility of Relapse Prevention Strategies in Interdisciplinary Multimodal Pain Therapy Programs: Qualitative Study. JMIR Human Factors, 7 (4), e21545.

Gallant, M. P. (2003). The influence of social support on chronic illness self-management: A review and directions for research. Health Education & Behavior, 30 (2), 170–195.

Hodgkin, P. (2018, April 10). The Emerging World of Online Health Communities. https://ssir.org/articles/entry/the_emerging_world_of_online_health_communities