Supportive Care in Radiotherapy Based on a Mobile App: Prospective Multicenter Survey

Patient Characteristics

A total of 975 cancer patients presenting for radiotherapy at the above-mentioned institutions for breast or prostate cancer were screened for participation in this survey. Of them, 200 patients owned a smartphone and consented to participate. All the participants were asked to complete a survey prior to the initiation of radiotherapy (T0) and again upon the completion of the treatment (T1). Patient characteristics are presented in Table 1, and a flowchart of the recruitment workflow is illustrated in Figure 1. This prospective survey was approved by the Heidelberg University Independent Ethics Committee on February 15, 2017 (approval #S-007/2017).

Survey Methods

The survey form consisted of a standardized paper questionnaire containing 27 items (Q1-Q27) about smartphone use. The questionnaire was developed by experienced radiation oncologists with the help of a biomedical informatist and biostatistician and was tested on a small group of patients to allow room for clarifications and corrections before its distribution within this survey. The types of questions included multiple-choice questions, requiring the patients to choose one or several answers out of 2, 4, or 5 provided options. Two were polar questions, requiring the patients to choose either “yes” or “no.” Five questions prompted the patients to additionally fill in optional free text. Items assessed the habits of smartphone usage (Q3, Q7, and Q8), assessed technical knowledge and abilities in smartphone usage (Q4-Q6), assessed readiness to use a smartphone app within the context of cancer and radiotherapy (Q9, Q14, Q20-22, and Q27), suggested features for a potential radiotherapy-related mobile app (Q10-Q13), suggested the timeframe of reachability for smartphone notifications (Q23-Q25), and assessed the general attitude toward the different aspects of radiotherapy (Q15-Q19 and Q26). Additionally, patient- and disease-related information was collected. An English version of the survey questionnaire is provided in Multimedia Appendix 1.

Statistical Analysis

To allow for quantitative comparison, a simple scoring method was devised in which the aforementioned areas of interest (AOI) were considered as the subscales of the questionnaire. Within each subscale, every question was weighted by the number of possible answers, and the points were divided equally between the provided answers. Questions Q5, Q10, and Q13 were taken out of the score because they only provided qualitative information. The sum for every subscale was calculated and used for the comparison. Detailed information about the scoring system is provided in Multimedia Appendix 2.

For descriptive analyses, continuous variables were presented as mean (SD) and median (IQR and min and max) and categorical variables as absolute and relative frequencies. Group comparisons were made according to tumor stage, age (below and above 55 years), Karnofsky performance scale index (KPI), previous courses of radiotherapy, treatment setting, and gender [13]. Wilcoxon rank sum test for ordinal scaled variables and chi-square test for categorical variables were used to evaluate potential differences between patients in the mentioned groups. Group differences were assessed for all subscales and, additionally, for all questions included in the calculation for one of the subscales. To evaluate the differences at time points T0 and T1, Wilcoxon signed-rank test for paired ordinal scaled data and McNemar test for categorical variables were used. All statistical analyses were performed using R software (version 3.4.3, The R Foundation for Statistical Computing, Vienna, Austria).

Descriptive Analysis

Of all patients, 73.2% (146/200) indicated that they were using mobile data on their smartphones, and the common usage included social media apps (45/200, 22.5%), picture taking and Web browsing (86/200, 43.0%), or further apps (48/200, 24.0%); 10.5% (21/200) of the patients used their smartphones for voice calls and instant messaging. Only 24.2% (48/200) of the patients indicated having used their smartphones in a health-related context before; 66.2% (131/200) of the patients stated that they never or rarely required assistance in using their smartphones and 63.9% (124/200) estimated their technical skills in this regard to be solid or advanced.

Patients showed a high overall readiness to use a mobile app in the context of radiotherapy: 73.3% (146/200) of all patients judged using a dedicated mobile app for additional supportive care during their treatment as helpful or very helpful. Mobile apps usage was judged especially helpful in providing support for the occurrences of treatment-associated toxicities (163/200, 81.8%, helpful or very helpful). The favored frequencies at which patients would be willing to answer short app-based queries regarding their well-being or general symptoms were weekly (98/100, 50.8%) and as required (37/200, 19.2%). Thirty-two out of 200 patients (16.6%) wished to do this only at the beginning and end of therapy. Concerns regarding data security were voiced by 12.2% (24/200) of patients. These concerns were somewhat more frequent in patients older than 55 years, although the difference was not significant compared with patients younger than 55 years (13.7% vs 10.0%, P=.16).

The most requested feature of a mobile app was assistance in appointment making for radiotherapy and consultations (176/200, 88.0%), followed by general or specific questions regarding patient well-being during radiotherapy (130/200, 65%). Additionally, patients requested the option to receive answers to questions and information material about diagnosis and therapy. Of all patients, 83.4% (166/200) welcomed the idea of continually using the app during follow-up to stay in touch with the treating physicians, describing this option as helpful or very helpful. The same was true for being contacted by the treating physician if medical warning signs were detected (182/200, 91.6%, helpful or very helpful). A reminder feature for scheduled follow-up examinations was favored by 81.5% (163/200) of patients.

Of the 200 patients, 21.8% (43/200) indicated their timeframe for reachability via smartphone between 7 am and 11 pm to be at least 12 hours; 40.1% (79/200) of patients indicated it to be between 2 and 12 hours. Regarding smartphone notifications about missed calls, instant messages, or push notifications, 75.4% (147/200) of patients stated that they would review those within a maximum timeframe of 2 hours or shorter; 21.5% (42/200) of patients answered “within 12 hours,” and only 3.1% (6/200) would need “2 days or longer.” The same was true for app-specific notifications regarding radiotherapy: the percentages were 67.8% (132/200) for 2 hours or less, 26.2% (51/200) for 12 hours, and 6.2% (12/200) for 2 days or longer.

In addition to the above-mentioned aspects of smartphone and app usage, the survey enquired about the general and organizational aspects of radiotherapy that could be improved using a smartphone app. Regarding the desired frequency of consultations with a supervising physician during radiotherapy, 34.4% (67/200) of patients favored weekly appointments, followed by 29.2% (57/200) favoring appointments “only as required.” Regarding consultations, a waiting period of up to 30 minutes was considered acceptable by 53.4% (106/200) of patients. However, regarding daily radiotherapy, the acceptable waiting period was shorter: 25.7% (50/200) of patients opted for 15 minutes or less and 6.2% (12/200) for even 10 minutes or less. Detailed information about the answers provided to all survey items is illustrated in Figures 2 and 3 and in Multimedia Appendices 3 and 4.

In Figure 2, Q9 corresponds to “Use of a dedicated smartphone app for support during radiotherapy”; Q11 to “Staying in touch via smartphone app during follow-up”; Q12 to “Being contacted about medical warning signs via a smartphone app”; Q20 to “App collecting relevant medical information prior to consultation”; and Q27 to “App-based supportive care in the context of treatment side effects.”

Areas of Interest

For quantitative evaluation, the items in the questionnaire were grouped according to subject to calculate the scores for AOI, as described above. The scores were transformed to represent a value between 0 and 100, where a higher score represents a higher inclination or acceptance toward the use of a smartphone app [14-16]. The scores for AOI as calculated from the analysis of all returned questionnaires filled out at T0 are indicated in Table 2 and Figure 4. The highest scores were achieved for AOI 3 (“readiness to use a dedicated app within the context of radiotherapy”) and AOI 4 (“suggested features of a mobile app”), achieving median values of 71.4 (Q1-Q3=61.9-76.2) and 75.0 (Q1-Q3=75.0-87.5), respectively. AOI 3 and 4 translate most directly into a high acceptance for the presented app-based model of therapy support. A similar median value of 71.4 (Q1-Q3=42.9-85.7) was calculated for AOI 2 (“technical knowledge and abilities”). AOI 5 (“timeframe of reachability”) and 6 (“general attitude”) provided mostly qualitative information regarding reachability and the setting and frequency of medical consultations during treatment. The median score of 58.3 (Q1-Q3=41.7-66.7) for AOI 5 translates into an average reachability within 2 hours during a daily timeframe of 2 to 12 hours for the majority of patients. The score of 33.3 (Q1-Q3=25.0-50.0) for AOI 6 shows a general acceptance for waiting periods of up to 30 minutes for a medical consultation.

Influence of Radiotherapy

For 140 patients, survey data before (T0) as well as after the completion (T1) of radiotherapy were available and were compared to evaluate whether having undergone radiotherapy influenced the attitude of patients. A small but statistically significant decrease of 4.7 points in the median transformed score for readiness to use an app within the context of radiotherapy (AOI 3) was detected at T1 (mean 68.6 vs 65.5, P<.001, Wilcoxon signed-rank test). The question within AOI 3 leading to this difference addressed the favored frequency of answering symptom-related questions posed by the app (Q14). At T1, patients selected “at therapy start and completion” and “only as required” more frequently instead of “weekly” and “every other day.” The helpfulness of app-based support in the context of toxicity was judged slightly lower at T1 (mean 4.04 before radiotherapy vs 3.88 after radiotherapy, P=.03). Regarding the median score for Q12, addressing the helpfulness of being contacted by a physician in case of medical warning signs, the mean/median score was 4.29/4 at T0 and 4.11/4 at T1 (P=.03). This question translated into a small but significant difference of 4.4 points in the transformed score for AOI 4 (mean 79.4 before radiotherapy vs 75.0 after radiotherapy, P=.04, Wilcoxon signed-rank test). For the other AOIs (AOI 1, 2 and 5, 6), as well as for individual questions, no significant difference was observed between the time points T0 and T1. A comparison of AOI scores between the time points T0 and T1 is illustrated in Figure 5. Detailed information regarding the quantitative comparison between answers provided at T0 and T1 for different survey items is provided in Table 3.

Predictive Factors

To determine possible factors that influenced patients’ attitude toward the usage of a mobile app, we tested several patient characteristics for their impact on survey results. The tested factors were age (<55 years vs ≥55 years), tumor stage (early [defined as tumor stage ≤T2, N0, M0/X] vs advanced), gender, previous radiotherapies, treatment setting (curative vs palliative), and initial KPI (<80% vs ≥80%). Age appeared to have the most sizable impact on the answers provided in the questionnaire, leading to significant differences in usage habits and technical skills as well as reachability. In all cases, younger patients were found to be more inclined toward and versed in more intensive smartphone use (AOI 2, P<.001). Interestingly, the favored frequency of seeing a physician during therapy was higher in patients younger than 55 years (P=.009). An overview of the survey items significantly impacted by the analyzed patient characteristics is displayed in Table 4.

Interpretation of Survey Results

We conducted a prospective and systematic survey regarding the habits and skills in smartphone usage, readiness to use a supportive mobile app during and after radiotherapy, and opinions on suggested functionality for such an app among cancer patients undergoing radiotherapy. Our results showed high overall acceptance levels for the usage of mobile technology in the context of radiotherapy, and the proposed functionality and features were considered as helpful by a large majority of patients.

Moreover, our survey showed that the ability to use a dedicated mobile app for additional supportive care (eg, mobile data on smartphone, reachability, and technical skills) on the patient side were present in almost three quarters of the survey population, and the obtained numbers were consistent across the different subareas of this survey. These results suggest a promising potential for an mHealth approach in the field of radiotherapy, although they also outline the need for careful patient selection to avoid the unnecessary burdening of subgroups that are either not willing or not capable and, thus, will likely not benefit. As the ownership of a smartphone and informed consent were prerequisites to participating in the survey, 975 patients were screened, and a total of 200 patients participated. The possible selection bias introduced by this approach has to be considered when generalizing the survey results to all patients undergoing radiotherapy. On the other hand, the results can be valuable in identifying and describing patient subgroups that are most likely to benefit from additional mobile app–based support.

The quantitative analysis of the survey data allowed us to identify age as the most influential factor for patient willingness with younger patients being considerably more inclined toward the use of mobile technology in the context of radiotherapy. Oncological diagnoses are typically associated with certain age groups (eg, prostate cancer, lung cancer, and different subgroups of head and neck cancer). Thus, a degree of patient selection based on age and age-associated diagnosis seams feasible when offering mHealth-based support. On the other hand, prospective clinical evidence exists, showing that patients with unfavorable characteristics regarding age and diagnosis, such as lung cancer patients, can also benefit. This especially holds true if the design of the electronic health (eHealth) or mHealth app integrates the patient’s next of kin to assist in the usage [17,18].

Several studies have shown that patient compliance plays a special role in the successful implementation of mHealth initiatives [19]. Clinical experience in oncology shows patient compliance to be worse in subgroups with lifestyle-related risk factors (eg, heavy smoking and alcohol abuse), potentially making such patients less eligible for mHealth-based support [20,21]. However, evidence exists that offering mHealth support in addition to close clinical surveillance may improve patient compliance, particularly in the abovementioned subgroups, by providing regular prompts and reminders and facilitating adherence to prescribed exercises or supportive regimens [22-24].

The perceived needs of cancer patients for supportive measures are manifold; they may strongly vary depending on culture, diagnosis, prognosis, and associated symptoms and may, hence, influence the acceptance of such measures [25-27]. Our survey showed that, disregarding few minor points, the acceptance of the proposed mobile app approach was high among patients, irrespective of tumor stage, treatment setting, potential previous radiotherapies, and initial clinical performance. Regarding toxicity-related surveillance, acceptance was significantly higher in patients with a reduced performance status. Nevertheless, as poor-performing patients require a different form of intensified personal care, the use of a mobile app alone may show limitations, particularly in the palliative setting [28].

It can be argued that by limiting the present survey to include only prostate and breast cancer patients, a selection bias in favor of patients with favorable prognosis and good clinical performance is introduced, and the generalizability of the results for other cancer patients undergoing radiotherapy could be limited. This potential limitation was accepted to achieve a more homogeneous dataset and ensure timely and systematic survey completion and analysis. However, it should be noted that we included 49.0% of patients with advanced tumor stages, 17.0% of patients with a metastatic disease, and 28.0% of patients who had undergone previous radiotherapy. These patient subgroups feature a different clinical profile characterized by unfavorable prognosis, usually a rapid decline in clinical performance, and a high need for supportive care [29]. The described clinical profile is shared by a majority of patients undergoing radiotherapy for different diagnoses, and the survey results regarding the acceptance of mobile app–-based support did not differ significantly in this subgroup [30-32]. Furthermore, it is of interest that irrespective of age, no further diagnosis-specific influencing factors regarding mobile app acceptance were identified for either breast or prostate cancer patients. This finding, in turn, supports the approach of cautiously extrapolating our results to patients with differing diagnoses.

The helpfulness of app-based supportive care in the context of monitoring potential radiotherapy-induced side effects was judged minimally higher by patients before the beginning of therapy, as was the favored frequency of interaction with the app. These results highlight the importance of providing sufficient information and support before and during the early stages of radiotherapy to address potential fears and worries. Such fears and worries and, thus, the need for additional support are less likely to be observed when the therapy is successfully completed, and the results of our survey accurately mirror this constellation.

Review of the Literature

The management of cancer patients is an area of special interest for the development of new eHealth and mHealth initiatives because they show promising potential, particularly in the context of supportive care and follow-up [7].

Only one other survey focusing on the acceptance of a dedicated mobile app among cancer patients has been published [33]. Overall results showed convincing similarity in both surveys. However, good or very good technical skills and the willingness to send data to the treating clinic via an app were slightly less frequent in the previous survey than in our survey (54.1% vs 63.9% and 48.5% vs 71.1%, respectively). Moreover, younger patients were generally more willing to use mobile technology in a disease-related context. In comparison, our survey described a more homogeneous and precisely selected cohort, focusing exclusively on patients receiving radiotherapy. Consequently, the specific requirements and circumstances related to this course of treatment are more comprehensively examined.

Ruland and colleagues have reported results similar to those of our survey, testing the usefulness and acceptance of a Web-based, multicomponent eHealth app with special focus on patient self-management among breast and prostate cancer patients [34]. The tested app did not focus exclusively on radiotherapy, and it provided supportive care in a more general manner, featuring message boards and general information sections, among other functionalities. Active usage among the regarded cohort was 64%, which is close to the results of our survey, and in this context, age and diagnosis were reported to significantly impact patient usage [34].

The value of an eHealth or mHealth resource is particularly high when tailored to fit the needs of the target patient cohort because patients are more likely to appreciate the immediately relevant and personalized support [34]. Depending on the diagnosis and therapy regimen, the wide heterogeneity among cancer patients and the very specific needs of different subgroups make them a challenging collective to address as a whole. Consequently, existing mHealth projects have typically been addressing either a specific question or a specific subgroup of cancer patients.

A recent randomized controlled trial by Denis and colleagues has shown a significant median overall survival benefit of 5 months for lung cancer patients who were systematically telemonitored using a mobile app based on patient-reported data and using the dynamics of patients’ clinical symptoms for risk stratification and individualized follow-up [18,35]. Similar approaches have successfully been used in the management of toxicities related to head and neck cancer treatment: computerized screening could facilitate the identification of treatment-related toxicities, and telepractice apps or videoconferencing could assist in the delivery of intensive home-based dysphagia therapy [24,36]. Regarding general supportive care, considerable advances have been made in the development of mHealth interventions to address the common issue of fatigue among cancer survivors. A recent meta-analysis identified 9 completed eHealth studies that revealed a significant beneficial effect of eHealth interventions on fatigued patients with improvements in health-related quality of life and depression [37].

Patients undergoing radiotherapy represent a distinct subgroup of cancer patients with special requirements in terms of supportive care. The nature of radiotherapy results in a specific set of therapy-related side effects and medical issues that require surveillance and potential support. Of the symptoms, the most common are dermatitis, nausea, fatigue, and localized toxicities within the respective treatment region [12]. Furthermore, depending on the diagnosis, patients undergoing radiotherapy vary in terms of age and characteristic profiles regarding individual risk constellations and comorbidities. Based on the data reported here, the usefulness and clinical implementation of a mobile app will be evaluated in a prospective trial (OPTIMISE-1; ClinicalTrials.gov identifier #NCT03168048) for which the dedicated mobile app has been designed according to the requirements of the patients who were assessed [38].

Conclusion

To the best of our knowledge, this study is the first to prospectively evaluate and demonstrate a high acceptance of and distinct patient requirements for the use of a supportive mobile app in a large homogeneous cohort of cancer patients undergoing radiotherapy. The reported patient acceptance will serve as a basis for future clinical trials that prospectively investigate the benefits of mobile app–based treatment support in routine clinical settings. The introduction of mobile apps into the clinical routine should be regarded as an opportunity to improve and intensify supportive treatment for cancer patients.