This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Research Protocols, is properly cited. The complete bibliographic information, a link to the original publication on http://www.researchprotocols.org, as well as this copyright and license information must be included.
The delivery of optimal care depends on accurate communication between patients and clinicians regarding untoward symptoms. Documentation of patients’ symptoms necessitates reliance on memory, which is often imprecise. We developed an electronic diary (eDiary) for adolescents and young adults (AYAs) with cancer to record symptoms.
The purpose of this paper is to describe the utility of an eDiary designed for AYAs with cancer, including dependability of the mobile application, the reasons for any missing recorded data, patients’ adherence rates to daily symptom queries, and patients’ perceptions of the usefulness and acceptability of symptom data collection via mobile phones.
Our team developed an electronic symptom diary based on interviews conducted with AYAs with cancer and their clinicians. This diary included daily severity ratings of pain, nausea, vomiting, fatigue, and sleep. The occurrence of other selected physical sequelae was assessed daily. Additionally, patients selected descriptors of their mood. A 3-week trial of the eDiary was conducted with 10 AYA cancer patients. Mobile phones with service plans were loaned to patients who were instructed to report their symptoms daily. Patients completed a brief questionnaire and were interviewed to elicit their perceptions of the eDiary and any technical difficulties encountered.
Overall adherence to daily symptom reports exceeded 90%. Young people experienced few technical difficulties and reported benefit from daily symptom reports. Symptom occurrence rates were high and considerable inter- and intra-patient variability was noted in symptom and mood reports.
We demonstrated the utility of an eDiary that may contribute insight into patients’ symptom patterns to promote effective symptom management.
Adolescents and young adults (AYAs) with cancer experience numerous sequelae during treatment [
Current practices to determine symptom occurrence in clinical trials require that patients accurately recall their symptoms during clinic visits over the prior 1 to 6 weeks and report them to their clinicians. Clinicians then record this information in the medical record and researchers abstract the information to a database for analysis. Omissions or errors at any of these steps result in an underreporting or misrepresentation of symptom occurrence. Moreover, recent data suggests that the accuracy of symptom appraisal and recall depends on the use of short recall periods [
Alternatively, symptom diaries in paper format that are completed daily can be used to circumvent the challenges described above. However, patient adherence with these diaries is low, ranging from 11% to 56% [
The collection of AYAs’ symptom data in real-time using electronic diaries (eDiaries) could avoid poorly recalled data. Patients’ and clinicians’ interests in using eDiaries are high [
While initial eDiary research involved the use of personal digital assistants [
AYAs with cancer have unique needs related to frequent and complex exacerbations of their disease- and treatment-related symptoms that could affect their abilities to maintain eDiaries. The lack of an eDiary geared to the symptoms and developmental needs of AYAs with cancer prompted us to develop a mobile application, titled the Mobile Oncology Symptom Tracker (mOST). To do so, we assembled a multidisciplinary team of 2 nurses and 3 software developers from GoMed Solutions, one of whom is a pediatric oncologist. We elicited input from 15 AYA oncology patients from two University-affiliated children’s hospitals in the inpatient or clinic settings. Semi-structured interviews were used to determine their use of technology, and ideas on the types of symptoms to monitor, frequency of reporting, and preferences for icons and graphics. In addition, clinicians and research experts were consulted during the development of the application. The symptoms selected for daily monitoring were those that occur frequently during treatment of cancer [
The purpose of this paper is to describe the utility of mOST tested on a group of 10 oncology patients, ages 13 to 21. The specific aims of this pilot study were to determine the dependability of the mobile application, the reasons for any missing entries and patients’ adherence rates to daily symptom queries for multiple symptom reporting. In addition, patients’ perceptions of the usefulness and acceptability of symptom data collection via mobile phones were evaluated after a 3-week trial assessment period.
In this descriptive, longitudinal study, we evaluated a mobile phone-based electronic symptom diary in a convenience sample of AYAs with cancer. Patients were eligible to participate if they were 13 to 21 years of age, could understand English, gave assent or consent to participate, had not participated in the prior application development interviews, and were receiving chemotherapy (for initial therapy, relapsed, or refractory disease). The study was approved by the Human Subjects Committee at the University of California, San Francisco.
Eligible patients were identified with assistance from pediatric oncology clinics and in-patient advanced practice nurses. Research staff approached these patients to determine interest in the study. If patients expressed interest, the research staff informed them of the study procedures, risks, and benefits. Patients 18 years of age or older and the parents/guardians of younger patients signed written, informed consent. Younger patients gave written assent. Between March and April, 2011, 11 out of 13 patients approached agreed to participate in the study (response rate of 85%). Patients who refused were too busy or too ill to participate.
Patients completed a demographic form that included their age, race/ethnicity, primary diagnosis, date of diagnosis, number of relapses, prior mobile phone use, and approximate household income. An 8 GB iPhone 3GS, charger, and earphones were loaned to each participant during the study. The monthly service plan included: 450 minutes of peak time voice minutes, and unlimited data, text messaging, and evenings/weekend calling. The mOST application was downloaded on each phone. Patients received an instructions booklet and a tutorial on the application and phone use, expectations of the frequency of symptom reporting, and methods to notify research staff of system malfunctions. They were instructed to report their symptoms at the end of the day. A single daily report was designed to minimize patient burden. Patients could only access the system between 3 pm and midnight, as reports earlier in the day might not accurately reflect all of their experiences for the day. They could program 2 reminder messages to input daily entries with audible alerts, customizable for text choice and time selection. Data were delivered to a secure website, listed by study number only, and could be downloaded to database and statistical packages for analysis. The dates and times of data submission and data upload were coded.
Severity ratings of 5 disease or treatment-related symptoms were assessed daily: pain, nausea, vomiting, fatigue, and sleep quality. A body diagram was used to indicate the location(s) of any pain experienced. A visual analog scale (VAS) was selected to assess pain intensity (ie, no pain or worst pain) based on a consensus statement compiled by pediatric pain researchers [
Validity of the CAS was previously evaluated by eliciting pain ratings from 30 children in an emergency department (ED) who reported pain [
To assess nausea, we included the Pediatric Nausea Assessment Tool (PeNAT), a valid and reliable instrument to assess nausea among pediatric oncology patients [
No single-item validated instruments were available to assess vomiting, fatigue, or sleep among AYA patients. The number of vomiting episodes was assessed with the single forced-choice query of: none, 1 time, or 2 or more times. Fatigue was assessed with 2 queries from the Fatigue Scale-Adolescent [
The occurrence of diarrhea, constipation, fever, numbness/tingling, mouth sores, dizziness, and headaches were assessed daily. These symptoms were included due to their frequent occurrence among pediatric oncology patients and their variability during treatment [
Patients received $1 credit towards a department store gift card for each diary entry. In addition, those who completed ≥ 90% of the assessments in a 21-day period were given a $50 gift card. If patients attempted to report symptoms but experienced a system error, the attempt was also applied towards the gift card credits.
The principal investigator or research assistant conducted 10 to 15 minute interviews with patients at the completion of their evaluation period to obtain their perceptions about using mOST and any technical difficulties they experienced. In addition, they were asked to carry out a “think-aloud” exercise [
Descriptive statistics were used to characterize demographic and clinical characteristics as well as rates of symptom occurrence. The total number of system errors was calculated during the 21-day study period. Data entry adherence rates were determined by calculating the number of symptom reports over the 21 days of the study. One-way analysis of variance was used to evaluate the differences in patients’ weekly adherence rates. The interview transcripts were analyzed by 2 researchers, with verification by 1 of the senior investigators.
The demographic and clinical characteristics of the AYAs who participated are described in
The patients’ baseline technology use is described in
Demographic and clinical characteristics of participants (N=10).
Characteristica,b | n (%) | |
|
||
|
Male | 6 (60%) |
|
Female | 4 (40%) |
|
|
|
|
Hispanic White | 3 (30%) |
|
Non-Hispanic White | 1 (10%) |
|
African American | 1 (10%) |
|
Hispanic-Other or not specified | 5 (50%) |
|
|
|
|
Leukemia/Lymphoma | 6 (60%) |
|
Bone tumor | 3 (30%) |
|
Sarcoma-other | 1 (10%) |
|
|
|
|
None | 5 (50%) |
|
One or more | 5 (50%) |
|
|
|
|
Inpatient | 4 (40%) |
|
Outpatient | 4 (40%) |
|
Both | 2 (20%) |
aAge in years (SD) = 18.2 (2.9)
bTime since initial diagnosis in months (SD)= 12.2 (15.3)
Technology use among participants.
Question | Response | n (%) |
Do you have a mobile phone? | Yes | 8 (80%) |
No | 2 (20%) | |
Who pays for the service? | Self | 2 (20%) |
Parents | 6 (60%) | |
Not applicable | 2 (20%) | |
How often do you use your phone for voice calls? | Daily | 4 (40%) |
4-6 days/week | 2 (20%) | |
2-3 days/week | 2 (20%) | |
Not applicable | 2 (20%) | |
How often do you use your phone for text messaging? | Daily | 5 (50%) |
4-6 days/week | 2 (20%) | |
2-3 days/week | 1 (10%) | |
Not applicable | 2 (20%) | |
Do you have a computer at home? | Yes | 10 (100%) |
No | 0 (0%) | |
Do you have Internet access at home? | Yes | 9 (90%) |
No | 1 (10%) | |
Is Internet access WiFi enabled? | Yes | 5 (50%) |
No | 2 (20%) | |
Not applicable | 1 (10%) | |
Not sure | 2 (20%) | |
Do you have a page on Facebook or MySpace? | Yes | 8 (80%) |
No | 2 (20%) |
On day 14 of our data collection period, we noted a system malfunction that occurred if patients skipped a day of diary entries. On the subsequent day, the application did not function. During the “beta-testing” period, testers skipped data entries and did not experience this malfunction. An interim solution was devised in which patients deleted the application and reloaded it from a website. Subsequently, we monitored the database daily. If patients missed an entry, we contacted them and provided instructions on how to reload the application. A total of 3 episodes of system malfunction occurred over the study period. The software malfunction was rectified by the end of the study. An error message appeared at the time of upload for 3 patients, but that the entry was transmitted the next time the application was opened.
One patient missed an eDiary entry when he felt too ill to enter data. He missed the subsequent day due to the system malfunction and a third day when he went out of town and left the phone at home. Another patient missed a data entry on a day he discovered that his leukemia had relapsed. He missed 4 consecutive entries during a period when we attempted to reach him. He then reloaded the application and maintained full adherence until the end of his study period. A third patient missed 2 entries due to forgetfulness. A 4th patient reported full adherence to the data entries, but we noted missing data on 1 day with 2 entries the subsequent day, all uploaded on the third day. This finding likely represented Internet access difficulties, but we coded her data as missing a day to be conservative in adherence estimations.
Adherence can be calculated by assessing different causation end points. When data were counted as missing only for omitted entries due to forgetfulness and illness-related concerns of mild to moderate levels (ie, giving credit for days of system error and time spent in the PICU), the overall adherence rate for daily symptom reports during the 21 day study was 97%. When days of system malfunction were included as missing data, the adherence rate was 95%. When all days of missing data were used in the calculation, including the 8 days when a patient spent was in the PICU, the adherence rate was 91%. Adherence rates did not vary among the 3 weeks (F2,27=1.016,
The data collected were used to delineate the potential utility of the eDiary to clinicians and researchers. Summaries of the symptoms and moods reported by individual patients are listed in
In an exit interview, one patient with peripheral neuropathy noted that the highlighting feature was difficult to master. Others stated that the highlighting feature was simple to use. The body diagram used to describe pain locations was divided into 9 regions–head/neck, hands, arms, chest, abdomen, upper back, lower back/pelvis, legs, and feet. Difficulties with pinpointing the precise location of pain were mentioned by 4 patients (eg, the indicator to depict neck pain showed up on the head).
Patients rated the application as either “easy” (2/9, 22%) or “very easy” (7/9, 78%) to use. Daily entries were generally completed in less than 2 minutes. One patient had mild visual deficits and mentioned that the print was difficult to read at times. However, she was able to read the screens out loud during the “think-aloud” exercise. All other patients reported that the print size posed no difficulties. In the exit survey, patients were asked to select or write in adjectives to describe the application. They could select more than 1 descriptor. All of the selected words had a positive connotation (ie, 67% rewarding, 78% valuable, 78% educational, 67% interesting). No one selected challenging, frustrating, time consuming, worthless, confusing, or difficult to describe mOST. All patients who were questioned (n=7) reported that they would recommend the application to others. Selected quotations regarding why they would recommend the application to others are included in
Patients’ reports of physical symptoms.
Symptom | Percentage of days each patient reported symptoms | Mean of symptom occurrence rates | % of patients reporting symptoms | ||||||||||
Patient | |||||||||||||
|
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|
|
|
|
Inpatient | Both in and outpatient | Outpatient | ||||||||||
Fatigue-Physicala | 100 | 100 | 100 | 86 | 43 | 21 | 90 | 62 | 39 | 5 | 64 | 100 | |
Fatigue-Mentala | 95 | 92 | 88 | 81 | 48 | 16 | 90 | 52 | 44 | 5 | 62 | 100 | |
Nauseab | 100 | 100 | 65 | 24 | 95 | 47 | 75 | 0 | 29 | 38 | 57 | 90 | |
Headache | 95 | 54 | 29 | 81 | 29 | 0 | 5 | 52 | 22 | 76 | 44 | 90 | |
Painc | 81 | 69 | 65 | 43 | 38 | 0 | 10 | 0 | 24 | 33 | 36 | 80 | |
Dizzy | 52 | 46 | 24 | 24 | 29 | 0 | 30 | 5 | 17 | 33 | 26 | 90 | |
Numb | 62 | 8 | 100 | 5 | 0 | 100 | 0 | 0 | 0 | 5 | 10 | 60 | |
Constipation | 29 | 0 | 65 | 0 | 10 | 0 | 55 | 0 | 11 | 0 | 10 | 50 | |
Diarrhea | 29 | 92 | 6 | 5 | 5 | 0 | 0 | 0 | 0 | 0 | 14 | 50 | |
Mouth sores | 0 | 39 | 0 | 5 | 0 | 0 | 5 | 19 | 39 | 0 | 11 | 50 | |
Fever | 0 | 31 | 6 | 10 | 0 | 0 | 0 | 0 | 0 | 5 | 5 | 40 |
aOccurrence rate based on percentage of days that VAS score >30 on 0-100 scale
bOccurrence rate based on percentage of days that patients selected face 2, 3, or 4 on the Pediatric Nausea Assessment Tool (PeNAT)
cOccurrence rate based on percentage of days that patients selected face 2, 3, 4, 5, or 6 on the Faces Pain Scale-Revised (FPS-R)
Patients’ selection of mood descriptors.
Mood descriptor | Percentage of days each patient selected mood descriptors | Mean of mood occurrence rates | % of patients reporting mood | |||||||||
Patient | ||||||||||||
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |||
Inpatient | Both in and outpatient | Outpatient | ||||||||||
|
||||||||||||
Hopeful | 38 | 0 | 65 | 86 | 33 | 26 | 5 | 43 | 22 | 91 | 41 | 100 |
Happy | 67 | 15 | 82 | 0 | 62 | 26 | 5 | 52 | 33 | 24 | 37 | 90 |
Confident | 52 | 15 | 12 | 86 | 5 | 37 | 0 | 10 | 17 | 86 | 31 | 90 |
|
||||||||||||
Frustrated | 62 | 23 | 53 | 10 | 43 | 16 | 10 | 0 | 33 | 86 | 34 | 90 |
Irritable | 0 | 46 | 12 | 52 | 38 | 16 | 15 | 5 | 56 | 71 | 31 | 90 |
Worried | 0 | 46 | 59 | 10 | 19 | 11 | 40 | 10 | 0 | 81 | 28 | 80 |
Anxious | 38 | 8 | 65 | 0 | 19 | 5 | 10 | 10 | 11 | 95 | 26 | 90 |
Sad | 0 | 46 | 47 | 57 | 10 | 11 | 0 | 5 | 6 | 33 | 21 | 80 |
Angry | 19 | 0 | 24 | 5 | 14 | 5 | 0 | 0 | 11 | 48 | 13 | 80 |
Scared | 0 | 31 | 24 | 0 | 5 | 0 | 0 | 5 | 0 | 24 | 9 | 50 |
Lonely | 0 | 0 | 18 | 33 | 0 | 0 | 0 | 0 | 0 | 0 | 5 | 20 |
Participants’ reasons for recommending the use of the mOST application to others.
Participant | Quotation |
18 year old, Hispanic female | “…it's a good idea to keep like, to keep a list of how you're feeling" |
20 year old Hispanic male | “…it kind of like channels whatever you have wrong with you like out…” |
21 year old Hispanic male | “… because it really helps to see how you've been doing too actually, not just like, you know, go on day by day. But it does help you see and reflect on how you're doing. And if you did something the other day that helped, will help you make you feel better and so on. “ |
19 year old Hispanic female | “…because "It’s a good app and it just makes you -- I know for me it makes me feel better because I can keep track of my symptoms, kind of like a self-comforting type of thing because you know what you’re going through and you’re not scared. And eventually if this does go to where the doctors can see it, it’d be even better because you know that your physician is seeing it and the communication’s there.” |
Mental fatigue and nausea ratings for participant 5. Mental fatigue rating in response to the statement “My mind has felt worn out” on a 0-100 VAS, where higher values indicate more severe fatigue. Nausea rating is response to nausea query on a 0-100 VAS, where higher values indicate more severe nausea. Dashed line represents mean values over the 21-day study period.
This study is the first to demonstrate that the use of an eDiary on a mobile phone platform is a feasible method to collect daily symptom data from AYAs with cancer. The adherence rates in this study (>91%) greatly exceed those associated with pen and paper diaries [
Not all diseases or treatment-related symptoms experienced by cancer patients warrant daily monitoring. However, many symptoms are labile, particularly during chemotherapy [
Researchers can evaluate within-subject changes using the statistical process of multilevel modeling [
Daily reports may not be sufficiently frequent among a group of critically ill patients. Broderick and colleagues [
EMA trials which involve frequent assessments throughout the day have been conducted among AYA patients with chronic pain [
The use of mobile phones to collect symptom characteristics offers many advantages over pen-and-paper diaries for AYAs with cancer. AYAs are extremely comfortable with technology and may become more fully engaged in their care when they are empowered to track symptoms in a manner that resonates with them. AYA may prefer eDiaries to report sensitive information. In one study, adolescents reported more sexual behavior using a computer assessment compared with paper forms [
Currently our team is further evaluating the psychometrics of the mOST application. In this study, patients complete the eDiary daily over a 3-week course. On day 8 of the study they complete a series of symptom questionnaires and a quality of life measure (ie, the Pediatric Quality of Life Inventory) using a 1-week recall. Their daily eDiary responses during the first week will be compared to their recall of events reported on the symptom measures. Once the validity and reliability of the instrument are established, we plan to incorporate it into mHealth symptom intervention studies among AYA patients with cancer.
Although the use of mobile phones in research is considered an expensive endeavor, older models of phones are often available free of charge or for nominal fees. Mobile phone use among young people has been prevalent for many years and the age for acquiring smartphones is rapidly decreasing [
The limitations of this study need to be acknowledged. These limitations include the use of a convenience sample of only 10 patients and a relatively short trial period of 21 days. The patients received a relatively high incentive of up to $71 for their participation in the study, an acceptable practice in the United States considering the patients’ 3-week commitment to the study, but not standard in all countries [
mOST screen shots.
adolescent and young adult
color analog scale
emergency department
electronic diary
ecological momentary assessment
end-of-day
faces pain scale-revised
mobile oncology symptom tracker
pediatric nausea assessment tool
visual analog scale
We would like to thank the patients and their families who helped assisted in this project and the clinic and inpatient staff who supported us in many ways.
This work was supported by a grant from the National Institute of Nursing Research (1P30NR011934-01, PI: Miaskowski) and an American Cancer Society Institutional Research Grant Program. Dr. Baggott is funded by an American Cancer Society Mentored Research Scholar Award (MRSG 12-01-PCSM). Dr. Miaskowski is funded by an American Cancer Society as a Clinical Research Professor.
Mr. Richard Kletter and Dr. Paul Zeltzer developed the software for GoMed Solutions, and thus declare a potential conflict of interest.