Determinants of mobile technology use and smartphone application interest in cancer patients

2.1 Survey design and patients

We conducted a cross-sectional survey study at the Abramson Cancer Center at the University of Pennsylvania in Philadelphia, PA, and Penn Cancer Network community hospitals in suburban and rural areas of Pennsylvania, New Jersey, and Delaware (Cape May Regional Hospital (Cape May, NJ); Chester Hospital (West Chester, PA); Community Medical Center (Toms River, NJ); Doylestown Hospital (Doylestown, PA); Kennedy Hospital, Kennedy Health Center (Cherry Hill, NJ); Kent General Hospital (Dover, DE); Lancaster General Health (Lancaster, PA); Milford Memorial Hospital (Milford, DE); Monmouth Medical Center (Longbranch, NJ); Pennsylvania Hospital (Philadelphia, PA); and Phoenixville Hospital (Phoenixville, PA) between December 2014 and September 2015. Research staff evaluated the eligibility criteria, approached patients during regular clinical visits, performed informed consent process and conducted the survey study. Patients were required to be 18 years of age or older, have a primary diagnosis of cancer, have a Karnofsky functional score of 60 or greater (ie ambulatory), understand written English, verbally indicate to the research staff that they felt physically well enough to complete a survey at the time of approach, and report experiencing nonzero pain (on a scale of 0-10) in the last seven days. The Institutional Review Board of the University of Pennsylvania and the Scientific Review and Monitoring Committee of the Abramson Cancer Center approved the study protocol and surveys.

2.2 Outcomes

Mobile technologies were defined as smartphones, tablets, or text messaging. Patients reported frequency of use on a 5-point Likert scale—never, less than once/month, less than once/week, at least once/week and daily. We defined regular use as patient reported utilization at least once per week or daily.

Interest in smartphone applications for information delivery format was measured on a 4-point Likert scale, from very unimportant to very important; we defined a given communication method as “of interest” if patients rated it as important or very important.

Patients self-reported date of cancer diagnosis and demographic factors including age, sex, race, education, and marital status. We dichotomized education to high school or less and college or above. We determined cancer type and stage from chart abstractions and dichotomized stage to metastatic and non-metastatic. Patients reported if they had received surgery, chemotherapy, and/or radiation.

2.3 Statistical analysis

We performed statistical analysis using STATA software (Windows version 12.0, StatCorpLP, College Station, TX). We used univariate Chi² testing to identify factors associated with mobile technology use and preference. We then conducted multivariate logistic regression analyses to identify factors associated with use and preference for mobile technology. We incorporated variables that were significant at P = 0.10 in the univariate Chi² analysis. All analyses were two-sided with p less than 0.05 indicating statistical significance in the multivariate model.

3.1 Demographics and clinical characteristics of study participants

Among the 631 participants, mean age was 60.3 years (range 23.1-90.4), 415 (65.8%) were female, 521 (82.6%) white, 312 (53.8%) had non-metastatic cancer, and approximately half (51.8%) were seen in community hospitals (Table 1). Most (n = 427, 68.2%) had completed at least some time in college, and 65.8% (n = 415) reported they were married or currently living with a partner. The most common cancer types were breast (n = 202, 32%), followed by thoracic, hematologic, and gastrointestinal. Nearly half (n = 303, 49.6%) of patients had been diagnosed in the 12 months prior to completing the survey, with 20.6% (n = 126) diagnosed within 12 to 36 months; 182 (29.8%) had been diagnosed more than 36 months before taking the survey. Most (n = 556, 88.1%) were treated with chemotherapy and about half had received surgery (n = 336, 53.3%) and radiation (n = 335, 53.1%).

3.2 Use and determinants of use of mobile technologies

Among 631 respondents, 466 (73.9%) regularly used mobile technologies including smartphones (n = 356, 57%), tablets (n = 240, 38%), and text messaging (n = 418, 66%). Younger patients were more likely to report regular use of mobile technologies (91.4% for age ≤45 years, 89.0% for age 46-55, 78.5% for age 56-65 and 55.1% for age > 65) (Table 2). In addition, patients with at least some college education (79.6% vs 60.8% with high school education or less, P < 0.001), women (76.4% vs 69.0% of men, P = 0.045), and patients seen at an academic hospital (78.9% vs 69.1% of those treated at community hospitals, P = 0.005) were more likely to report regular use of mobile technologies. In multivariate analysis, patients under 45 years old were substantially more likely to use mobile technologies ([AOR] 6.8, 2.8-16.9 95% CI, P < 0.001) than those aged older than 65 years (Table 3). Those with college or above education were also more likely to use mobile technologies than those with high school education or less (AOR 2.3, 1.5-3.5 95% CI, P < 0.001). There was no difference in mobile technology use across races.

3.3 Determinants of interest in smartphone applications

Overall, 242 (39%) patients expressed interest in smartphone applications to learn about supportive care services, with more younger patients reporting this interest (60.0% for age ≤45 vs 28.0% for age >65, P < 0.001). There was no significant difference in interest between genders, with 37.9% of women and 41.0% of men expressing interest in information delivery via smartphone application (P = 0.45). Interest varied by race with 34.0% of white patients and 63.2% of non-white patients expressing interest in information delivery of supportive care services by smartphone application (P < 0.001). Of those with a high school education or less, 32.8% expressed an interest in smartphone application-delivered information compared with while 41.6% of those with college or above (P = 0.036; Table 2).

In multivariate logistic analysis, younger age (AOR 3.2 for age ≤45 compared to >65, 1.7-5.9 95% CI, P < 0.001), non-white race (AOR 3.4, 2.1-5.5 95% CI, P < 0.001), and male gender (AOR 1.6, 1.1-2.4 95% CI, P = 0.016) were associated with an interest in receiving supportive care information through smartphone applications. Interest in information via smartphone app was similar across education groups (Table 3).