Abstract
Introduction
Fatigue is a frequent and debilitating symptom that contributes to poor quality of life for people receiving peritoneal dialysis. Ecological momentary assessment using mobile technology (mEMA) is a novel survey technique that can collect symptom data in real-time and has not been trialed in a peritoneal dialysis cohort. The study aimed to explore real-time fluctuations and associations between fatigue, mood, and physical activity using mEMA.
Methods
Adults receiving peritoneal dialysis completed fatigue and mood scales, via a mobile application (app), 5 times daily for 7 days and, concurrently wore an accelerometer. A feasibility questionnaire was completed on the eighth day.
Results
Forty-eight adults completed the study. Within-day fatigue fluctuations were observed with severity lowest during mid-morning to early afternoon and peaking at bedtime. Associations between fatigue and mood were observed with a 1-unit change in mood score conferring a 5.2-unit change in fatigue (P < 0.01). Higher volume of physical activity was associated with lower fatigue and enhanced mood. Overall adherence to the app-based surveys was 73% with most participants reporting mobile phones and the mEMA app being easy to use.
Conclusion
People receiving peritoneal dialysis experience within day and day-to-day fluctuations in fatigue that appear highly variable. Higher fatigue severity was associated with poorer mood and lower physical activity levels with future studies required to explore if physical activity-based interventions could be a potential strategy for the management of these symptoms. Furthermore, mEMA, and mobile phones, were feasible to capture symptom data with potential to be employed in future research or, as part of improved care.
Keywords: ecological momentary assessment, fatigue, feasibility, mood, peritoneal dialysis, physical activity
Graphical abstract
Fatigue is a debilitating symptom for people receiving dialysis, with a reported prevalence of 55% to 89% for this population group.1 Fatigue is defined as a subjective sense of weakness, lack of energy, and tiredness, which can be conceptualized on a continuum of exhaustion and tiredness on one end and energy and vitality at the opposite end.2 The significance of fatigue was highlighted in an international consensus process where people receiving peritoneal dialysis identified fatigue as the highest priority outcome (of 36 identified) that required research attention.3
Fatigue occurs due to a combination of factors for people living with kidney failure, including treatment burden, physical deconditioning, poor mood, and complications of living with kidney failure.4 Both fatigue and poor mood can contribute to physical inactivity, with the contrariwise also possible, suggesting a likely bidirectional relationship.4,5 This potential complex interrelationship remains unqualified in people receiving peritoneal dialysis. Prevalence and between-person relationships of mood, physical activity, and fatigue have been reported for people receiving peritoneal dialysis6; however, little is known about the within-person daily fluctuations and relationship between these key health indicators. If fatigue is to be a target of consumer-driven kidney care, the factors that are amenable to intervention and that impact fatigue, such as physical activity and mood, urgently need investigation.7,8
A limitation of previous research involving people receiving dialysis has been the use of methods that asked people to recall their fatigue (and other symptoms) experience over a preceding time period (i.e., recall instruments), rather than at a specific moment in time. This potentially exacerbates the risk of recall bias and prohibits the ability to perform within-person temporal analysis. This is problematic, because it is likely that people receiving peritoneal dialysis experience within day, and day-to-day, variations in fatigue and other symptom experiences.4
Ecological momentary assessment is a survey method that allows for collection of data in real-life environments (ecological) and in real-time (momentary), typically using mobile phone technology.9 The use of ecological momentary assessment has not been investigated for people receiving peritoneal dialysis. This study aimed to address the knowledge gap by employing ecological momentary assessment via a mobile phone application (terming it mEMA) to collect symptom data, and accelerometry to explore physical activity of people receiving peritoneal dialysis. The overarching research aims were to explore in real-time how fatigue and mood levels fluctuate throughout the day, and to investigate the bidirectional and temporal within-day associations between fatigue, mood, and physical activity levels in people receiving peritoneal dialysis. A secondary aim was to explore the feasibility of using mEMA in this population.
Methods
A detailed description of the study protocol has been previously published.10 This study employed an intensive longitudinal study design, with assessments completed over a 7-day period for each participant.
Protocol Development
Consumer input was a feature of the protocol development involving people receiving peritoneal dialysis, a diverse range of health professionals and experienced renal researchers. Two people currently receiving peritoneal dialysis trialed the final study protocol.
Study Participants
Adults (aged ≥18 years) living with kidney failure and receiving peritoneal dialysis, and incident patients (after a 3-month stabilization period) were invited to participate. Inclusion criteria stipulated as follows: that participants must (i) be able to give written consent, (ii) be independently ambulant (including the use of assistive walking devices [e.g., a cane]), (iii) be able to understand English, (iv) have an active e-mail (or willing to create one), and (v) be competent in the use of mobile phone applications (or willing to learn). Mobile phone ownership was not essential; a loan phone (iPhone 7S, Apple Inc., CA) was provided upon request.
Recruitment Strategy
A detailed description of recruitment strategies has been previously published.11 Peritoneal dialysis physicians and nursing staff provided potential participants with study information / consent form or, a link to a video (https://youtu.be/ppkb9ZxVPNs) explaining the study, either in-person or via postal or email distribution. Upon patient consent, participants were offered either an in-person visit from the lead investigator to set up the mobile application or, were provided written and video tutorial resources to complete setup autonomously. Participants received an $AU100 gift voucher for participation.
Demographic Information
Data on participant age, employment status, height and weight, dialysis history, comorbidities, and biochemistry data (hemoglobin, albumin, urea, and creatinine) were collected from medical and laboratory records.
Study Variables
The data collection schedule is outlined in Figure 1.
Figure 1.
Data collection schedule.
Ecological Momentary Assessment Mobile Software
Ethica Data (Ethica Health Kitchener, Canada) mobile phone software was used to collect fatigue and mood data 5 times each day.
Fatigue Assessment
The Visual Analogue Scale to Evaluate Fatigue Severity (VAS-F)12 scale involves 18 items with 2 built in subscales (fatigue = 13 items; energy = five items). Each item asks respondents to pick a number between 0 and 10 representing how they currently feel, along a scale between 2 anchors (e.g., from “0 = not at all tired” to “10 = extremely tired”). Higher total in each subscale indicates more severe fatigue (0–130) and increased energy (0–50).
Mood Assessment
The Visual Analogue Mood Scale13 asks the participant a single question to “Pick a number from 0 to 10: at the moment I feel.” The participant then picks a number on the screen between two anchors (e.g., from “0 = happy” to “10 = sad”).
Accelerometry
Physical activity levels were measured concurrently with fatigue and mood assessments via a wrist worn GENEActiv (ActivInsights, Cambs, United Kingdom) accelerometer at a measurement frequency of 100 Hz. This was worn on the nondominant wrist.
Feasibility and Acceptability
Participants completed a study-made survey built into the mobile application, regarding feasibility and utility of mEMA and mobile phones. The survey contained 9 questions where participants answered that they “strongly agreed,” “agreed,” “were neutral,” “disagreed,” or “strongly disagreed” with the statement and 5 free-text questions. The survey was adapted from previous studies14,15 with themes including ease of use, convenience, modifications for future trials, and whether participants consider mEMA to have potential to assist in the management of their condition. Feasibility was further assessed by the total number of prompts answered, lag-times (i.e., participants’ response time from mobile application prompt to answering of the prompt) and time taken to complete the survey sessions.
Data Cleaning and Processing
Height and weight obtained from medical records were used to calculate body mass index. Medical history data were condensed to capture multiple conditions under a single bodily structure or system (cardiovascular, pulmonary, neurological, cancer, mental health, muscle/joint/bone, or other conditions) except for hypertension and diabetes. Area-level (residential postcode) socioeconomic status was classified using deciles of The Australian Bureau of Statistics Index of Relative Socio-Economic Disadvantage, with decile 1 representing the most disadvantaged areas and decile 10 representing the least disadvantaged areas.16 The Visual Analogue Scale to Evaluate Fatigue Severity assessment scores were sorted into the 2 subscales (fatigue and energy) for each timepoint. Accelerometer data were downloaded through GENEActiv PC Software Version 3.2 (ActivInsights, Cambs, United Kingdom) and imported into custom software built into MATLAB R2019a (MathWorks, Inc., Natick, MA). Physical activity intensity thresholds established by Esliger et al.17 were then automatically applied to produce minutes of sedentary, low, moderate, and vigorous activity. Total activity (light, moderate, and vigorous summed) and moderate-vigorous physical activity (MVPA) in minutes were the activity variables used in analyses. Activity data were manually trimmed to reflect the amount of activity completed; (i) in the 30-minute window preceding an answered prompt, (ii) in the 30-minute window following an answered prompt and, (iii) between 10 AM and 4 PM as well as 1 PM and 7 PM daily. Feasibility and acceptability survey data were pooled to reflect the percentage of participants for an answer given to quantitative responses and transcribed verbatim for qualitative response. Lag times (minutes) and percentage of prompt answered for each time period of each day data were extracted from the Ethica Data online interface.
Data Analysis
Descriptive statistics were used to describe the cohort demographics. Multilevel linear regression modelling (with participant as a random intercept) was used to explore the following: (i) if fatigue, energy, and mood fluctuated throughout the day; (ii) the associations between fatigue, energy, and mood; (iii) the temporal associations between physical activity completed in the 30-minutes preceding a survey prompt (wakeup scores excluded from this analysis) as well as fatigue, energy, and mood scores at the time of the prompt; (iv) the temporal associations between fatigue, energy, and mood scores and the 30-minutes of activity completed following the survey prompt (bedtime scores excluded from this analysis); and (v) the temporal associations between fatigue, energy, and mood scores at wake up and subsequent activity levels of the same day (1 PM–7 PM); and conversely, the activity levels during the day (10 AM–4 PM) and subsequent fatigue, energy, and mood scores at bedtime. Minutes of MVPA and total physical activity levels were both considered in analyses. Age, dialysis vintage, body mass index, socioeconomic status, biochemistry data, diabetes, cardiovascular, pulmonary, mental health, and muscle/joint/bone conditions were included as confounders. Descriptive statistics were used to report the quantitative data of the participant feasibility and acceptability survey and to describe the adherence (i.e., proportion of prompts answered) and mean response times. Free-text responses underwent qualitative content analysis to categorize the data. This systematic process involved assigning codes, using subjective interpretation, with themes subsequently developed.18 There were ∼106 eligible people within the clinical catchment at the commencement of recruitment with a sample size of 55 participants targeted. This target gave ∼60% power to detect a moderate effect size and ∼90% power to detect a large effect size, based on the number of study days and prompts per day outlined in the protocol above, assuming that, on average, participants would respond to 75% of the prompts; and conservatively estimating that between-subject variability in the momentary experience of fatigue is no higher than 50%. All analyses were conducted using SPSS Statistical Software Version 25.0 (IBM Corp. Armonk, NY).
Ethics Approval
Ethics approval was granted from the Central Adelaide Local Health Network (Reference No. 13245) and University of South Australia (Protocol No. 203339) human research ethics committees, and the study was registered on the Australia New Zealand Clinical Trials Registry (Trial ID: ACTRN12620001316998).
Results
Recruitment remained open from November 2020 to October 2022 with 48 participants completing the study. One person was withdrawn due to hospitalization, for reasons unrelated to the study, on day 2 of their 7-day observation period. Their data were not included in any analyses. Demographic data are presented in Table 1. One participant reported currently smoking. All participants were receiving automated peritoneal dialysis with 16 participants having prescribed dialysate in the peritoneal cavity during the day.
Table 1.
Participant demographics
| Variable | Mean (SD) (n = 48) |
|---|---|
| Age (yr) | 61 (13.5) |
| Gender (% male) | 65 |
| Dialysis vintage (mo) | 12.5 (6.3, 21.8)a |
| Previously received hemodialysis (% yes) | 29 |
| Currently employed (% yes) | 33.3 |
| Medical conditions (% yes) | |
| Hypertension | 90 |
| Diabetes | 46 |
| Cardiovascular condition | 42 |
| Muscle, bone, or joint condition | 38 |
| Mental health condition | 22 |
| Pulmonary condition | 19 |
| Neurological condition | 10 |
| Cancer | 8 |
| Other | 25 |
| Cause of chronic kidney disease (% yes) | |
| Diabetes | 25 |
| Polycystic kidney disease | 11 |
| Alport syndrome | 8 |
| Hypertension | 4 |
| Other | 21 |
| Unknown | 31 |
| Weight (kg) | 84.4 (23) |
| Body mass index (kg/m2) | 28.7 (7.3) |
| Hemoglobin (g/l) | 108 (12.2) |
| Albumin (g/l) | 30.6 (4.5) |
| Urea (mmol/l) | 21.7 (5.6) |
| Creatinine (umol/l) | 744 (223) |
Median (25th percentile, 75th percentile).
Within Day and Day-to-Day Fluctuations and Associations Between Fatigue, Energy, and Mood
Relative to mean wake up fatigue levels, fatigue was less severe from mid-morning to early afternoon (10 AM–1 PM: −9%) before increasing later in the afternoon (4 PM–7 PM: +14%) and at bedtime (+27%) (Figure 2). Mean energy scores showed a similar pattern with more energy reported in the mid-morning to early afternoon (10 AM–1 PM: +11%) and less energy reported later in the day (4 PM–7 PM: −11%) and at bedtime (−17%). No significant fluctuations were reported for mood, with average levels ranging between −5% and +2.4% relative to mean wake up score. Relative to the same timepoint the previous day, day-to-day mean score fluctuations ranged from 0.2% to 19.7% for fatigue, 0.4% to 10.3% for energy and, 0% to 28.2% for mood. Mean score of variables across each day and timepoint are available in Supplementary Table S1. Significant associations between fatigue, mood, and energy were observed with a 1-unit change in mood conferring a 5.2-unit change in fatigue (95% confidence interval: 4.55–5.93; P < 0.01) and a 0.65-unit change in energy (95% confidence interval: 0.41–0.89; P < 0.01). A 1-unit change in energy score conferred a 1.5-unit change in fatigue (95% confidence interval: 1.34–1.64; P < 0.01). No confounders were significantly associated with the outcomes in any of the fluctuation or association analyses.
Figure 2.
Within day fluctuations of fatigue (a), energy (b), and mood (c). Data presented as mean (95% CI) ∗P = <0.01 ∗∗P = <0.05 (all fluctuations are relative to wake-up score). Fatigue Score: 0 (No Fatigue)–130 (Severe Fatigue); Energy Score: 0 (No Energy)–50 (High Energy); Mood Score: 0 (Happy)–10 (Sad)
Temporal Associations Between Physical Activity Completed 30-minutes Before a Survey and Subsequent Fatigue, Energy, and Mood
Estimates of the acute relationship of physical activity with fatigue, energy, and mood are reported in Table 2. Total activity and MVPA completed in the 30-minutes prior to a survey prompt were related to subsequent fatigue and energy scores. For example, every 1 minute of MVPA completed prior to a survey was associated with −0.65-unit change in fatigue (i.e., lower fatigue severity). Higher estimates were reported for MVPA compared to total activity, demonstrating that higher intensity activity was associated with larger differences in fatigue and increased energy. Full analysis results, including associations with confounders are available in Supplementary Table S2.
Table 2.
Acute relationships between physical activity and symptoms in the 30-minutes prior to a survey
| Physical activity variable | Symptom | Estimate | 95% CI | P value |
|---|---|---|---|---|
| Moderate-Vigorous | Fatiguea | −0.65 | (−1.11, −0.06) | 0.01 |
| Physical Activity | Energyb | 0.31 | (0.13, 0.49) | <0.01 |
| (per 1-minute increase) | Moodc | −0.01 | (−0.04, 0.03) | 0.63 |
| Total Activity | Fatiguea | −0.53 | (−0.74, −0.32) | <0.01 |
| (per 1-minute increase) | Energyb | 0.19 | (0.12, 0.26) | <0.01 |
| Moodc | −0.01 | (−0.03, −0.00) | 0.04 |
CI, confidence interval.
Lower fatigue score indicates lower fatigue severity.
Higher energy score indicates increased energy.
Lower mood score indicates improved mood.
Temporal Associations Between Fatigue, Energy, and Mood and Subsequent Physical Activity Completed 30-minutes Following a Survey
Fatigue and energy were related to the amount of MVPA and total activity recorded in the 30-minutes following a survey. Higher fatigue severity was associated with reduced amounts of physical activity whereas higher energy was associated with higher amounts (Table 3).
Table 3.
Acute relationship between symptoms and physical activity in the 30-minutes following a survey
| Symptom | Activity variable | Estimate | 95% CI | P value |
|---|---|---|---|---|
| Fatigue | Moderate-Vigorous Physical Activity | −0.01 | (−0.02, 0.00) | 0.08 |
| (per 1-point increase) | Total Activity | −0.04 | (−0.06, −0.02) | <0.01 |
| Energy | Moderate-Vigorous Physical Activity | 0.04 | (0.01, 0.07) | 0.01 |
| (per 1-point increase) | Total Activity | 0.14 | (0.07, 0.21) | <0.01 |
| Mood | Moderate-Vigorous Physical Activity | −0.02 | (−0.15, 0.12) | 0.80 |
| (per 1-point increase) | Total Activity | −0.05 | (−0.36, 0.26) | 0.75 |
CI, confidence interval.
Temporal Associations Between Physical Activity Completed Throughout Day and Fatigue, Energy, and Mood Scores at Wake-Up and Bedtime
Physical activity performed between 10 AM and 4 PM was not related to fatigue, energy, or mood at bedtime. There was a trend (nonsignificant) for higher fatigue and mood score at wake up to be associated with less physical activity participation later in the day (1 PM–7 PM). Full analysis results, including associations with confounders are available in Supplementary Table S3.
Feasibility and Acceptability
The lead author set up the mobile app for 29 participants, and 19 completed set up autonomously via instructions (written and video instruction n = 17; written instruction only n = 2). The loan phone was used by 4 participants. No technical problems from the mobile app occurred. There were 15 incidents of technical error related to the accelerometers resulting in missing data for: 1 day (n = 4), 2 days (n = 6), or 7 days (n = 5). Exit survey results highlighted that most participants found the mobile app easy to use and reported no difficulties using a mobile phone (Table 4). Most participants who enrolled and set up the mobile app autonomously found the written and video instructions easy to follow. Overall adherence across the 7-day period was 73%. Event-based prompt adherence was 77% and semi-random prompt 71% (first prompt answered 57%; second prompt answered 14%). Mean compliance across each day and within day timepoints were similar. The overall mean completion time for a survey was 1 minute 59 seconds. Of note, mean completion time illustrated a downward trend from day 1 (2 minutes 45 seconds) compared to day 7 (1 minute 37 seconds). Mean response time to the semirandom prompts was 6 minutes 1 second. Daily adherence, response and completion time data are available in Supplementary Table S4.
Table 4.
Exit survey quantitative data summary (data presented as % of cohort [n = 48])
| Question | Strongly disagree | Disagree | Neutral | Agree | Strongly agree |
|---|---|---|---|---|---|
| I had difficulty using the mobile phone | 60 | 30 | 10 | - | - |
| I found the mobile app easy to use | 2.5 | - | 2.5 | 40 | 55 |
| The number of mobile app notifications and reminders to complete surveys was annoying | 30 | 37.5 | 32.5 | - | - |
| I keep my mobile phone close to me | - | 7.5 | 22.5 | 47.5 | 22.5 |
| I found the questions difficult to read on the mobile screen | 52.5 | 35 | 10 | 2.5 | - |
| Answering the survey questions (picking a number between 0 and 10) was easy to do | 2.5 | 2.5 | 5 | 45 | 45 |
| I had difficulty remembering to answer the survey questions at wake up | 20 | 35 | 27.5 | 12.5 | 5 |
| The mobile app instruction sheet was easy to followa | - | - | 6 | 47 | 47 |
| The mobile app instruction video was easy to followb | - | - | 11 | 58 | 31 |
| I had difficulty remembering to answer the survey questions at bedtime | 27.5 | 42.5 | 15 | 10 | 5 |
| I was able to hear (or feel) the mobile app notification when it was time to complete a survey | - | 10 | 15 | 57.5 | 17.5 |
| The 7-day duration of this study was too long | 32.5 | 42.5 | 25 | - | - |
| Mobile apps could help me manage my kidney condition | - | 5 | 27.5 | 50 | 17.5 |
Based on 19 participants who were sent the instruction sheet.
Based on 17 participants that were sent the instruction video.
Exit survey qualitative responses (Table 5) highlight that the majority of participants had a positive experience, found the app easy, quick, and simple to use, with few recommendations for future modifications.
Table 5.
Exit survey qualitative data summary
| Question 1-What was good about the mobile app? | ||
|---|---|---|
| Respondents | Theme | Exemplar quotes |
| n = 26 | Easy, quick, and simple to use | “Easy and simple to use without any hassle” - Male 60–70 |
| n = 4 | Increased their self-awareness | “I had to give thought to my feelings and physical awareness and answer the questions accordingly” - Female, 50–60 |
| n = 3 | A good tool for self or external monitoring. | “Allowing you to give information that will help others, would like to be able to send this information to the nurses so they can monitor me” -Female 50–60 |
| (No Response: n = 15) | ||
| Question 2 - What was bad about the mobile app? | ||
|---|---|---|
| Respondents | Theme | Exemplar quotes |
| n = 20 | Nothing bad about the App | |
| n = 10 | Survey instrument questions were repetitive / similar | “Same questions asked 10 different ways” -Male 60–70 |
| n = 3 | Limited notifications or time to respond | “Not enough notifications, I missed a few, I don’t always carry my phone at home” - Male 60–70 |
| n = 1 | Mobile interface difficult to read | “Need to have my reading glasses with me all the time” Male 70–80 |
| (No Response: n = 14) | ||
| Question 3 - What do you feel could have improved the mobile app? | ||
|---|---|---|
| Respondents | Theme | Exemplar quotes |
| n = 23 | Unsure or nothing to add | |
| n = 4 | Increased flexibility in the study protocol | “Maybe the reminder time frame could be longer” Female 60–70 |
| n = 4 | Survey instruments to capture other factors | “Maybe ask other questions about anxiety or pain” Female 40–50 |
| (No Response: n = 17) | ||
| Question 4 - Were there any reasons for ignoring the mobile app reminders? | ||
|---|---|---|
| Respondents | Theme | Exemplar quotes |
| n = 13 | Otherwise engaged (employment, appointments) | “I could not always get to the app in the required timeframe when at work” -Female 40–50 |
| n = 12 | No reason | |
| n = 7 | Mobile not close enough to hear, see or respond to | “Did not hear it before the survey disappeared” Male 50–60 |
| n = 1 | Unable to engage at times due to fatigue. | “Sometimes I felt a bit tired” Female 70–80 |
| (No Response: n = 15) | ||
| Question 5 - Any other aspects of this study you would like to comment on? | ||
|---|---|---|
| Respondents | Theme | Exemplar quotes |
| n = 4 | No theme (diverse responses) | “I think this tool has potential for increasing awareness about how people feel over time and teaching me how I can cope with the many challenges that are faced on a daily basis when undertaking PD” -Female 50–60 “Something like this to send my weight, blood pressure and blood glucose levels instantly to nurses would be good” -Female 50–60 “I think it is a well-prepared study, although I do wonder what use the results of the survey will provide to the architect of the exercise” -Male 60–70 |
| (No Response: n = 44) | “A longer study would be of more use. I just got used to trying to catch each questionnaire, and the study was finished” -Male 50–60 | |
Discussion
People receiving peritoneal dialysis experienced within day fatigue and energy fluctuations, with higher fatigue associated with lower energy and poor mood. Acute relationships were found between higher physical activity levels and subsequent fatigue (inverse), energy (positive), and mood (positive); whereas higher fatigue severity and lower energy levels were related to lower amounts of subsequent physical activity. This was the first study to explore the use of mEMA in this cohort, with participants supporting its utility and feasibility to explore real-time associations and fluctuations of symptoms and reporting its potential to explore other research areas or assist in improved care and monitoring.
Within day fluctuations in fatigue were, relative to wake up time, at their lowest during mid-morning or early afternoon, and increased as the day progressed, peaking at bedtime. The within-day pattern of fatigue level rising as the day progresses may reflect a typical diurnal response. However, the fatigue levels reported in the current study were higher at wake up time and bedtime when compared to apparently healthy12 and other clinical populations.19 The mean [95% confidence interval] wake-up score (54.5 [36.3–72.6]) in the current study was ∼72% and ∼75% higher than mean wake-up scores reported for apparently healthy (31.6 [26.3–36.9]) and cancer populations (31 [29.7–32.3]), respectively. The mean bedtime score (71.6 [53.4–89.7]) is ∼29% and ∼35% higher than the mean bedtime scores reported for apparently healthy (55.6 [51.3–59.8]) and cancer populations (53 [51.8–54.2]), respectively.12,19 The higher fatigue severity, relative to other cohorts, highlights why fatigue remains a high research priority in this cohort with investigation into underlying factors that influence it warranted.3 The patterns and high variability of fatigue within and between participants highlight the dynamic and complex nature of fatigue experienced by this cohort and, suggests strategies to mitigate fatigue need to be tailored around the fluctuations and the individual.
At the day level, poorer mood was related to higher fatigue and lower energy; and temporally, higher volumes of physical activity were acutely related to lower fatigue and better mood; and conversely, higher fatigue levels were related to lower subsequent physical activity participation. Mood disorders such as anxiety and depression are highly prevalent in dialysis populations,20 with prolonged poor mood a marker of depression.21 The association between depression and fatigue in dialysis populations has been reported, although not conclusive, but appears to be reciprocal.22 Chronic fatigue and mood disturbances are associated with poorer quality of life and mortality in people receiving peritoneal dialysis; highlighting the need for effective management strategies.5 Physical activity-based interventions have shown potential to reduce fatigue severity and improve mood in hemodialysis and other clinical populations.23, 24, 25 Physical activity-based interventional data for people receiving peritoneal dialysis are limited; however, recently developed guidelines support the role of physical activity to mitigate fatigue and enhance mood in this population.26 Given the demonstrated (in this study) and published relationships, the effectiveness of physical activity-based interventions on overall and acute temporal symptom experiences in this population warrants further investigation.
The current study found that mEMA was a feasible and acceptable method to capture symptom data from people receiving peritoneal dialysis, and mobile phones were an acceptable platform to facilitate this. This study was the first to employ mEMA with people receiving peritoneal dialysis and the consistent positive responses that the app was “easy, quick and simple to use” is consistent with other studies exploring the feasibility of mEMA with other clinical populations.27,28 The overall adherence rate (73%) is consistent with other feasibility studies exploring mEMA in clinical populations,29,30 although slightly lower than the reported adherence in a study involving people receiving hemodialysis (81%).31 Nevertheless, there is on-going debate about what constitutes acceptable adherence in mEMA studies.32 mEMA offers convenience and potential to explore other areas of importance to people receiving peritoneal dialysis, such as underlying factors that may influence fatigue. Indeed, participants reported that questions relating to pain, anxiety, and other emotions or mood states could improve the app. In addition, participants highlighted the potential of mEMA as a tool for self and external monitoring as further positive aspects. mEMA has been successfully employed as a symptom-monitoring tool to raise both individual awareness of symptoms as well as permit health professionals to remotely monitor (in real-time) clinical patients, improving care.33,34 Future studies should seek consumer feedback to guide the selection of questions and symptom types, and consider more frequent reminder notifications and longer timeframes (i.e., >30 minutes) to respond to a prompt as strategies to enhance adherence.
The study was strengthened by the ability to conduct research remotely with 19 participants completing the study without any face-to-face contact with investigators, potentially reducing burden and enhancing reach. In addition, the use of mobile technology to facilitate mEMA may have assisted in further reducing participant and researcher burden by removing the need for conventional data collection methods (e.g., pen and paper diaries). People living with kidney failure support the use of mobile apps as a platform for research35 and they may be more inclined to participate in research if activities could be completed at home.36
The study was limited by the inability to determine if the fluctuations or coefficient estimates between fatigue, mood, and physical activity were clinically significant, with no published minimal clinically important differences available for the fatigue and mood instruments. However, comparison with other population groups suggests elevated levels of fatigue. It is possible the single 7-day duration was too short to see the true extent of fatigue and mood experience in this cohort. Future studies could aim to explore multiple timepoints to identify if severity and/or fluctuations change over the course of the condition and treatment trajectory.37 Further to this, the short duration also may have increased the potential for a Hawthorne effect, where participants alter a behavior due to their awareness that they are being observed.38 In addition, the 7-day duration may not accurately capture the variability of physical activity levels that can occur (e.g., more active 1 week compared to another), although, the literature reports that 7-days typically captures the habitual activity patterns of an individual.39 Despite having results that may be relatively generalizable to state and national populations within Australia, the demographic characteristics may not be consistent with other countries. Factors such as age, dialysis vintage, and employment status often vary among countries; therefore, the results may not be fully translatable to all peritoneal dialysis populations. Finally, because no familiarization sessions were completed with the app, future studies could employ this training as a strategy to enhance understanding and adherence. However, the feasibility data suggests that participants were able to quickly develop competency upon study commencement.
Conclusion
People receiving peritoneal dialysis experience daily and within day fluctuations of fatigue, with higher fatigue severity associated with poorer mood and decreased energy. These fatigue fluctuations appear highly variable, suggesting that strategies to mitigate fatigue may need to be adaptable to the variable nature of fatigue. Acute relationships were apparent between physical activity, fatigue severity, and mood levels, with future studies required to explore if physical activity-based interventions could be a potential strategy for the management of these symptoms. This study was the first study to employ mEMA technology in this cohort, concluding that it was a feasible and useful tool to measure these important symptoms in real-time and understand the temporal changes and relationships. This work has provided a foundation for future research with this cohort to explore the following: (i) if physical activity-based interventions influence fatigue and mood experience and (ii) the potential of mEMA to capture other symptom experiences (e.g., dyspnea) and for self and remote symptom monitoring.
Disclosure
All the authors declared no competing interests.
Acknowledgments
The authors would like to thank the participants, nurses, and nephrologists within Central Northern Adelaide Renal Transplantation Service for their support and promotion of the research. Funding was secured for this project from the University of South Australia, The Hospital Research Foundation and SA Health, Allied Health Research Collaboration Funding (October 2019). STROBE statement can be found in Supplementary Table S5.
Footnotes
Table S1. Mean scores of each variable across each timepoint and day.
Table S2. Effect of physical activity completed in the 30-minutes prior to a survey prompt on symptoms (a) and effect of symptoms on physical activity completed in the 30-minutes following a survey prompt (b).
Table S3. Lagged association between physical activity completed during the day and fatigue, mood, and energy scores at wake up and bedtime.
Table S4. Within day and day-to-day adherence rates; daily response times; semi-random prompt response times.
Table S5. STROBE Statement—checklist of items that should be included in reports of observational studies.
Supplementary Material
Table S1. Mean scores of each variable across each timepoint and day.
Table S2. Effect of physical activity completed in the 30-minutes prior to a survey prompt on symptoms (a) and effect of symptoms on physical activity completed in the 30-minutes following a survey prompt (b).
Table S3. Lagged association between physical activity completed during the day and fatigue, mood, and energy scores at wake up and bedtime.
Table S4. Within day and day-to-day adherence rates; daily response times; semi-random prompt response times.
Table S5. STROBE Statement—checklist of items that should be included in reports of observational studies.
References
- 1.Ju A., Strippoli G.F.M., Craig J.C., Tong A., Saglimbene V.M., Unruh M.L. Interventions for fatigue in people with chronic kidney disease requiring dialysis. Cochrane Database Syst Rev. 2018;8:CD013074. doi: 10.1002/14651858.CD013074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Stone P., Richards M., Hardy J. Fatigue in patients with cancer. Eur J Cancer. 1998;34:1670–1676. doi: 10.1016/s0959-8049(98)00167-1. [DOI] [PubMed] [Google Scholar]
- 3.Manera K.E., Tong A., Craig J.C., et al. An international Delphi survey helped develop consensus-based core outcome domains for trials in peritoneal dialysis. Kidney Int. 2019;96:699–710. doi: 10.1016/j.kint.2019.03.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jhamb M., Weisbord S.D., Steel J.L., Unruh M. Fatigue in patients receiving maintenance dialysis: a review of definitions, measures, and contributing factors. Am J Kidney Dis. 2008;52:353–365. doi: 10.1053/j.ajkd.2008.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bossola M., Vulpio C., Tazza L. Fatigue in chronic dialysis patients. Semin Dial: Wiley Online Library. 2011;24:550–555. doi: 10.1111/j.1525-139X.2011.00956.x. [DOI] [PubMed] [Google Scholar]
- 6.Turkmen K., Yazici R., Solak Y., et al. Health-related qualıty of lıfe, sleep qualıty, and depressıon in peritoneal dialysis and hemodıalysıs patıents. Hemodial Int. 2012;16:198–206. doi: 10.1111/j.1542-4758.2011.00648.x. [DOI] [PubMed] [Google Scholar]
- 7.Murtagh F.E., Addington-Hall J., Higginson I.J. The prevalence of symptoms in end-stage renal disease: a systematic review. Adv Chronic Kidney Dis. 2007;14:82–99. doi: 10.1053/j.ackd.2006.10.001. [DOI] [PubMed] [Google Scholar]
- 8.Kosmadakis G.C., Bevington A., Smith A., et al. Physical exercise in patients with severe kidney disease. Nephron Clin Pract. 2010;115:c7–c16. doi: 10.1159/000286344. [DOI] [PubMed] [Google Scholar]
- 9.Burke L.E., Shiffman S., Music E., et al. Ecological momentary assessment in behavioral research: addressing technological and human participant challenges. J Med Internet Res. 2017;19:e77. doi: 10.2196/jmir.7138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tarca B., Jesudason S., Le Leu R., et al. Ecological momentary assessment to explore fatigue, mood and physical activity levels in people receiving peritoneal dialysis: a study protocol. Perit Dial Int. 2021;41:502–508. doi: 10.1177/0896860821992243. [DOI] [PubMed] [Google Scholar]
- 11.Tarca B., Jesudason S., Bennett P.N., et al. Research involving people receiving peritoneal dialysis: lessons learned. Other journal article. Ren Soc Australas J. 2022;18:51–56. doi: 10.33235/rsaj.18.2.51-56. [DOI] [Google Scholar]
- 12.Lee K.A., Hicks G., Nino-Murcia G. Validity and reliability of a scale to assess fatigue. Psychiatry Res. 1991;36:291–298. doi: 10.1016/0165-1781(91)90027-m. [DOI] [PubMed] [Google Scholar]
- 13.van Rijsbergen G.D., Bockting C.L.H., Berking M., Koeter M.W.J., Schene A.H. Can a one-item mood scale do the trick? Predicting relapse over 5.5-years in recurrent depression. PLoS One. 2012;7 doi: 10.1371/journal.pone.0046796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lee J.-E., Song S., Ahn J.S., Kim Y., Lee J.E. Use of a mobile application for self-monitoring dietary intake: feasibility test and an intervention study. Nutrients. 2017;9:748. doi: 10.3390/nu9070748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Zang J., Song J., Wang Z., et al. Acceptability and feasibility of smartphone-assisted 24 h recalls in the Chinese population. Public Health Nutr. 2015;18:3272–3277. doi: 10.1017/S1368980015000907. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Australian Bureau of Statistics. Census of Population and Housing: Socio-economic Indexes for Areas (SEIFA) 2016. https://www.abs.gov.au/ausstats/abs@.nsf/mf/2033.0.55.001
- 17.Esliger D.W., Rowlands A.V., Hurst T.L., Catt M., Murray P., Eston R.G. Validation of the GENEA Accelerometer. Med Sci Sports Exerc. 2011;43:1085–1093. doi: 10.1249/MSS.0b013e31820513be. [DOI] [PubMed] [Google Scholar]
- 18.Hsieh H.-F., Shannon S.E. Three approaches to qualitative content analysis. Qual Health Res. 2005;15:1277–1288. doi: 10.1177/1049732305276687. [DOI] [PubMed] [Google Scholar]
- 19.Kober K.M., Roy R., Dhruva A., et al. Prediction of evening fatigue severity in outpatients receiving chemotherapy: less may be more. Fatigue Biomed Health Behav. 2021;9:14–32. doi: 10.1080/21641846.2021.1885119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chilcot J., Wellsted D., Da Silva-Gane M., Farrington K. Depression on dialysis. Nephron Clin Pract. 2008;108:c256–c264. doi: 10.1159/000124749. [DOI] [PubMed] [Google Scholar]
- 21.American Psychiatric Association A. Diagnostic and Statistical Manual of Mental Disorders, vol 3: American psychiatric association; 1980.
- 22.Artom M., Moss-Morris R., Caskey F., Chilcot J. Fatigue in advanced kidney disease. Kidney Int. 2014;86:497–505. doi: 10.1038/ki.2014.86. [DOI] [PubMed] [Google Scholar]
- 23.Stroud N.M., Minahan C.L. The impact of regular physical activity on fatigue, depression and quality of life in persons with multiple sclerosis. Health Qual Life Outcomes. 2009;7:68. doi: 10.1186/1477-7525-7-68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Oberoi S., Robinson P.D., Cataudella D., et al. Physical activity reduces fatigue in patients with cancer and hematopoietic stem cell transplant recipients: a systematic review and meta-analysis of randomized trials. Crit Rev Oncol Hematol. 2018;122:52–59. doi: 10.1016/j.critrevonc.2017.12.011. [DOI] [PubMed] [Google Scholar]
- 25.Motedayen Z., Nehrir B., Tayebi A., Ebadi A., Einollahi B. The effect of the physical and mental exercises during hemodialysis on fatigue: a controlled clinical trial. Nephrourol Mon. 2014;6 doi: 10.5812/numonthly.14686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bennett P.N., Bohm C., Harasemiw O., et al. Physical activity and exercise in peritoneal dialysis: International Society for Peritoneal Dialysis and the Global Renal Exercise Network practice recommendations. Perit Dial Int. 2022;42:8–24. doi: 10.1177/08968608211055290. [DOI] [PubMed] [Google Scholar]
- 27.Ramsey A.T., Wetherell J.L., Depp C., Dixon D., Lenze E. Feasibility and acceptability of smartphone assessment in older adults with cognitive and emotional difficulties. J Technol Hum Serv. 2016;34:209–223. doi: 10.1080/15228835.2016.1170649. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Moitra E., Gaudiano B.A., Davis C.H., Ben-Zeev D. Feasibility and acceptability of post-hospitalization ecological momentary assessment in patients with psychotic-spectrum disorders. Compr Psychiatry. 2017;74:204–213. doi: 10.1016/j.comppsych.2017.01.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Forster S.D., Gauggel S., Petershofer A., Völzke V., Mainz V. Ecological momentary assessment in patients with an acquired brain injury: a pilot study on compliance and fluctuations. Front Neurol. 2020;11:115. doi: 10.3389/fneur.2020.00115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Juengst S.B., Graham K.M., Pulantara I.W., et al. Pilot feasibility of an mhealth system for conducting ecological momentary assessment of mood-related symptoms following traumatic brain injury. Brain Inj. 2015;29:1351–1361. doi: 10.3109/02699052.2015.1045031. [DOI] [PubMed] [Google Scholar]
- 31.Abdel-Kader K., Jhamb M., Mandich L.A., et al. Ecological momentary assessment of fatigue, sleepiness, and exhaustion in ESKD. BMC Nephrol. 2014;15:1–8. doi: 10.1186/1471-2369-15-29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Williams M.T., Lewthwaite H., Fraysse F., Gajewska A., Ignatavicius J., Ferrar K. Compliance with mobile ecological momentary assessment of self-reported health-related behaviors and psychological constructs in adults: systematic review and meta-analysis. J Med Internet Res. 2021;23 doi: 10.2196/17023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Fuerst M.L. Digital symptom monitoring improves cancer patients’ care. Oncology Times. https://journals.lww.com/oncology-times/fulltext/2022/01200/digital_symptom_monitoring_improves_cancer.5.aspx
- 34.Beames J.R., Kikas K., Werner-Seidler A. Prevention and early intervention of depression in young people: an integrated narrative review of affective awareness and Ecological Momentary Assessment. BMC Psychol. 2021;9:113. doi: 10.1186/s40359-021-00614-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ju A., Scholes-Robertson N., Johnson D.W., et al. Patient-led identification and prioritization of exercise interventions for fatigue on dialysis: a workshop report. Clin Kidney J. 2021;14:831–839. doi: 10.1093/ckj/sfz200. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Natale P., Gutman T., Howell M., et al. Recruitment and retention in clinical trials in chronic kidney disease: report from national workshops with patients, caregivers and health professionals. Nephrol Dial Transplant. 2020;35:755–764. doi: 10.1093/ndt/gfaa044. [DOI] [PubMed] [Google Scholar]
- 37.Veldhuijzen van Zanten J., Douglas M.R., Ntoumanis N. Fatigue and fluctuations in physical and psychological wellbeing in people with multiple sclerosis: a longitudinal study. Mult Scler Relat Disord. 2021;47 doi: 10.1016/j.msard.2020.102602. [DOI] [PubMed] [Google Scholar]
- 38.McCambridge J., Witton J., Elbourne D.R. Systematic review of the Hawthorne effect: new concepts are needed to study research participation effects. J Clin Epidemiol. 2014;67:267–277. doi: 10.1016/j.jclinepi.2013.08.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Matthews C.E., Ainsworth B.E., Thompson R.W., Bassett D.R., Jr. Sources of variance in daily physical activity levels as measured by an accelerometer. Med Sci Sports Exerc. 2002;34:1376–1381. doi: 10.1097/00005768-200208000-00021. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.