This descriptive, case-matched, comparative study collaborated with a larger study that investigated cardiac patients’ use of mobile technology and variations among age groups after adjusting for education, employment, and confidence in using the mobile technology. The larger study surveyed 282 English-speaking CR patients on the mobile technology use in nine hospital and community sites across metropolitan and rural New South Wales, Australia [39]. This study enrolled 30 English-speaking patients from the large study to match with a separated Mandarin-speaking group and reported on multilingual groups recruited from this study site that has not been published previously.

The study was conducted in a metropolitan teaching hospital in South Eastern Sydney Local Health District (New South Wales, Australia). The selected health district represented approximately 12% of the New South Wales population; with 37% born in countries outside of Australia and 27% in a non-English-speaking country, the selected health district comprised the most diverse population [40]. Of all, China-born residents constituted the largest proportion of the population from a non-English-speaking background, followed by people born in Greece and Indonesia. Of those born in countries outside of Australia, approximately 10% reported that they either do not speak English well or at all [41].

We recruited a stratified and matched convenience sample in this study. The inclusion criteria were as follows: (1) the presence of a cardiac diagnosis, such as angina, myocardial infarction (MI), ischemic heart disease (IHD), valve surgery, coronary artery bypass graft (CABG), or percutaneous coronary intervention (PCI), the absence of severe comorbidities, and physically suitable to be referred to the exercise-based group CR program; and 2) could speak and understand adequate English or Mandarin for consent and questionnaire processes. Patients with a neurocognitive disorder were excluded from the study. We matched each Mandarin-speaking patient by age (within ±5 years) and gender with a patient from the other two linguistic backgrounds to minimize the demographic variability across groups. Furthermore, the sample size was predetermined to be 30 per group per previous study protocols [42,43].

We used a previously developed checklist to collect sociodemographic and clinical data [44]; the information comprised participants’ age, gender, country of birth, ethnicity, home language, education, marital and employment status, and admission diagnosis.

Then, we developed the survey based on previously validated and used questionnaires where possible; the list of the most common devices (ie, smartphones, computers, and tablets) and mobile features (ie, browsing the internet, text messages, emails, and social media) was based on previously determined parameters [45,46]. The survey comprised 11 questions overall, and the questionnaire was pilot-tested in a small sample (n=15) of cardiac patients similar to the sample. Moreover, the content of the questionnaire was reviewed and amended to improve the ease of use, accuracy, and specificity. Furthermore, the questionnaire was used in a larger study including 282 English-speaking patients [39].

Questions regarding each device were clarified using an illustration (Textbox 1). Most questions (questions 1, 3-9) were in the checklist format, where respondents ticked the technological devices or features that they (1) used, (2) used confidently, (3) would like to learn, and (4) used for health purposes. In addition, we developed a question (question 2) on self-efficacy in using a new computer program based on the speed with which participants could learn a new computer program and used a 4-point scale anchored with responses “very slowly” (1) and “very quickly” (4). Furthermore, we included two open-ended questions (questions 10 and 11) where respondents could provide additional information on the mobile app they used along with details of the health-related app. Notably, respondents who answered the first question with “not using any technology” were not required to complete the remainder of the questionnaire.

Finally, the questionnaire was translated into Mandarin by a certified translator and back-translated for verification. A minor amendment was made to item seven that inquired about the videoconferencing use—WeChat or QQ was surveyed instead of Skype or FaceTime because it was more popular in Mandarin-speaking communities.

The study protocol was approved by Northern Sydney Local Health District Human Research Ethics Committee (LNR/15/HAWKE/450). All patients were screened for the eligibility by a CR staff (a clinical nurse specialist) or the bilingual researcher (LZ) upon their admission to a cardiac ward of the hospital or upon referral to the outpatient CR programs at the study site between April and September 2016. Those who fulfilled the inclusion criteria were approached by the CR staff or LZ to participate in the study and provided them with information and time to consider participation. Mandarin-speaking patients were approached by a bilingual Mandarin-speaking CR staff member or LZ. All staff members were trained in using the questionnaire to ensure a standardized approach. Finally, patients who provided written consent were surveyed in this study. The CR staff or LZ collected demographic and clinical data of enrolled patients, and any uncertainty regarding diagnoses was clarified using the medical records. Notably, the questionnaire was self-administered. Of 134 CR patients approached, 10 declined because of the lack of interest, with the final response rate of 92.5%. We surveyed 94 patients from English and “other” language-speaking background for the ongoing matching purpose; of these, 60 participants were matched and enrolled in the final data analysis.

The responses in Mandarin were translated into English by LZ for data entry. Data analyses were performed using IBM SPSS, version 24. In addition, means, SDs, frequencies, and percentage were used to present the demographic and clinical characteristics of the study cohort. Frequencies and percentages were used to describe technology device and feature use, confidence in use, and use for health. Furthermore, categorical variables were reported as a percentage within a language group and tested for differences across language groups using chi-square tests.

We used generalized linear mixed model analysis (GLMM) to ascertain whether the language spoken at home correlated with the mobile device and feature use. In addition, GLMM was used as patients in each language group were selected to be matched for age and gender. Each group of three (one from each language group) was assigned the same group identification (ID; 30 in all) besides a unique individual ID. As outcomes (ie, whether specific types of technology or features were used) were dichotomous, we selected the binary logistic regression function.

Adjusted models comprised age, gender, years of education, marital status, and employment status in the model along with language spoken at home. Furthermore, we explored devices and mobile features that were reported the most prevalent in participants’ report or those that had the highest potential for CR interventions. Then, we assessed correlates of using devices (ie, smartphone, computer, and tablet) and mobile features (ie, the internet, emails, apps, and social media). Finally, the internet use for health was categorized into individual items, including sharing health information, accessing information about general health, medication, and lifestyle. In this study, odds ratios (OR), 95% CI, and P values are reported, and alpha=0.05. P ≤.05 was considered statistically significant (two-tailed).

In this study, the final sample comprised 90 patients (mean age 66.7, SD 13 years; range 46-95 years); of these, 53.3% (48/90) were males, 55.6% (50/90) completed high school, and 63.3% (57/90) were not in the workforce. More than half of the participants were admitted with CHD (52/90, 57.7%), with the leading procedures or diagnoses being PCI, angina, and MI (Table 1). In the “other” language group, the most common languages spoken at home were Greek (7/30, 23.3%), Arabic (6/30, 20%), and the remainder comprised Macedonian, Vietnamese, Hungarian, Italian, Russian, Indonesian, Portuguese, Philippine, Japanese, Samoa, French, Bulgarian, and Czech language. We observed no significant difference in education, marital status, living arrangement, and employment status across the three home language groups.

Most participants (82/90, 91.1%) reported using, at least, one of the following devices: computers (desktops or laptops), tablets, smartphones, text/voice-only phones, and activity trackers (Figure 1). Mobile devices, such as tablets, smartphones, text/voice-only phones, and activity trackers, were used by most participants(n=79, 87.8%), the most common of which were smartphones (57/90, 63.3%), followed by tablets (28/90, 31.1%), and text/voice-only phones (24/90, 26.7%). In addition, 33.3% (8/24) of text/voice-only phone users displayed their interest in learning to use a smartphone in the future. The mean score on how quickly one could learn a new computer program was 2.16 (SD 1.0), with 1 representing “very slowly,” 2 representing “fairly slowly,” 3 representing “fairly quickly,” and 4 representing “very quickly,” suggesting that the participants on average could learn a new computer program but might require time.

In this study, the three language groups were similar in the mobile technology use, except for the smartphone use. The proportion of smartphone users in the Mandarin-speaking group was significantly higher compared with “other” language-speaking group (80.0% vs 53.3%; P=.03). The confidence in the current mobile technology use was similar across groups, except for the confidence in using text/voice-only phones. A larger proportion of participants in the “other” language-speaking group were only confident in using text/voice-only phones compared with the Mandarin-speaking group (51.9% vs 17.2%; P=.006).

However, computer use significantly differed across home language groups, with more English-speaking participants using computer compared with Mandarin or “other” language-speaking participants (English: 73.3% vs Mandarin: 36.7%, P=.003; English: 73.3% vs “other” language: 43.3%; P=.02). Furthermore, the proportion of participants confident in using a computer was significantly higher in the English-speaking group compared with the Mandarin-speaking groups (73.1% vs 35%; P=.003).

Mobile features most commonly used among participants using mobile device were voice calls (77/82, 93.9%), text messages (54/82, 65.9%), the internet (39/82, 47.6%), emails (36/82, 43.9%), videoconferencing (Skype or FaceTime [WeChat or QQ]; 35/82, 42.7%; Figure 2). In addition, fewer Mandarin-speaking participants used emails (24.1% vs 65.4%; P=.001) and social media (10.3% vs 42.3%; P=.007) compared with English-speaking participants.

Overall, 44.4% (36/81) of the participants who engaged with technology used the internet for health (Figure 3), used most often for sharing health information (35/81, 42.7%) and accessing information about general health (25/81, 30.5%), medication (20/81, 24.4%), and lifestyle (19/81, 23.2%). We observed no significant difference across groups in using the internet for health, except that a higher percentage of English-speaking participants accessed the web-based medication information than Mandarin-speaking participants (38.4% vs 10.4%; P=.02).

After adjusting for age, gender, years of education, marital status, and employment status, Mandarin-speaking participants exhibited increased odds of using smartphones (OR 7.238, 95% CI 1.262-41.522; P=.03) but decreased odds of using emails (OR 0.089, 95% CI 0.016-0.490; P=.006), and accessing the web-based medication information (OR 0.191, 95% CI 0.037-0.984; P=.05) compared with English-speaking participants (Tables 2 and 3). In addition, other factors associated with mobile devices and features use; for an additional year in age, the odds of using smartphones and emails decreased (OR 0.118, 95% CI 0.809-0.961; P=.005; OR 0.104, 95% CI 0.820-0,978; P=.02). Furthermore, participants who were employed exhibited increased odds of using Apps and social media compared with their nonworking counterparts (OR 6.052, 95% CI 1.256-29.175; P=.03; OR 16.455; P=.01). Male participants exhibited decreased odds of using a tablet compared with females (OR 0.163, 95% CI 0.044-0.600; P=.007).

To the best of our knowledge, this is the first exploratory study on the mobile technology use among CALD patients in the CR setting. Overall, the study suggests that technology might provide an alternative secondary prevention delivery strategy in the future to bridge the gap between growing demands and limited resources, as population aging and CVD prevalence continues to rise. Given the increasing use of technology-based interventions in CVD secondary prevention, this study reveals the unique patterns of use among CALD patients. In addition, the findings indicate that a variation in technology use warrants consideration while developing or delivering technology-based CR to these groups. The study determined that CALD patients are not disadvantaged in using certain types of mobile technology; thus, technology-based interventions could offer a potential solution to overcome their barriers to attending CR, such as communication and transportation difficulties. Meanwhile, the technology use patterns among the study groups revealed that selecting appropriate delivery media is essential for reaching different patients groups to improve the CR uptake.

Although several CALD patients are non-computer users, possibly because they had few opportunities to acquire computer skills during their education and work [47,48], they are not disadvantaged in some mobile technology use, especially not in the smartphone use. Consistently, the smartphone ownership is the highest among CALD groups [49], offering a great promise for implementing smartphone-based interventions in these populations. An important principle for adapting health promotion interventions in CALD populations is to determine and address the barriers to access and participation to decrease disparities [50]. Traditionally, patients from CALD backgrounds have been identified among those who are least likely to attend CR programs [51], especially if they do not speak English, do not drive a car, have lower education or income, or have cultural barriers such as embarrassment of participation [14,21,23,51]. Technology-based CR could potentially address these barriers, as the program can be adapted to different languages and is not constrained by facilities, transportation, and time. Furthermore, it can be used in a patient-preferred environment [51] to improve the patient’s participation, engagement, and overall experience [38].

The variation in the mobile technology use among CALD groups warrants elucidation and accommodation when developing or delivering these interventions. In addition, evaluating the usage of mobile technology before developing or delivering to the targeted population is imperative. For instance, no overall significant difference has been reported in the internet use for health between CALD patients and others in this study, implying that internet-based interventions could potentially reach CALD as well as English-speaking patients. Evidence from this and other studies suggests that people from CALD backgrounds tend to access the internet more by smartphones rather than computers or laptops [49]. Thus, internet-based programs should be user-friendly for both computer and smartphone users to encourage participation. Furthermore, smartphone users use mobile features differently. For example, Mandarin-speaking patients tend to use emails less compared with English-speaking patients. Thus, email-based communication in CVD secondary prevention might not be feasible among certain CALD groups [47]. Reportedly, selecting an appropriate delivery method for CR interventions is the key to improving participation among CALD patients [52].

Overall, the ubiquity of mobile technology could potentially enable technology-based interventions in the future to fulfill the increasing demand for CR in CALD patients. Presumably, the population aged >60 years will increase from the current 800 million (representing 11% of the world population) to >2 billion in 2050 (representing 22% of the world population) [53], which would result in a tremendous challenge for health care in dealing with the increasingly prevalent noncommunicable diseases such as CHD. Meanwhile, mobile technology ownership has also witnessed an exponential upsurge [54,55]. Unlike the initial digital divide that placed the computer use and internet access beyond the reach of many older and lower-income individuals, the mobile technology has been extensively adopted across populations [54,56]. A well-established interest in technology enabled CR [57,58], implies that this new form of intervention and delivery might provide an alternative to meet the increasing demands [59]. In addition, some preliminary evidence suggest that technology-based CR has the potential for cost-saving compared with center-based CR [37,38]. However, age-related differences in the mobile feature use suggest that voice call- and text message-based interventions could be superior in reaching current older CR patients [57,58]. Furthermore, apps- and social media-based interventions could have great potential for future CR delivery when young users of today become tomorrow’s CR patient population [58].

This study has several limitations. First, the predetermined sample size is relatively small, and the sample was enrolled from a single hospital in Australia, which might limit the generalizability of the findings to the larger CR population. However, this study does provide crucial insights into the future research. Second, this study primarily aimed to assess the role of Mandarin as a home language in the technology use compared with other language groups. Given that English language users experience different challenges to non-English language users, we recruited two samples to compare with Mandarin-speaking patients. The “other” language-speaking group provides scope for comparison of the effects of a home language other than English that contrast Mandarin speakers. In addition, language spoken at home might be a marker for people’s acculturation level and English language proficiency, which were not assessed. Further studies are required to investigate the subgroups of language and cultures within this diverse group. Furthermore, any differences identified might represent the CALD experience in Australia and, thus, might not be able to be extrapolated to CALD populations in other countries. Third, we did not correct alpha for multiple pairwise comparisons and acknowledge the risk of type I errors because of the exploratory nature of the study and the small sample size. Finally, data collection using self-administered questionnaires is subject to recall and social desirability biases. Hence, further studies should complement self-administered questionnaire with objective measures and in-depth investigation of the role of home language and other correlates of technology use.

This study reveals a high usage of mobile technology among CR patients and provides further insights into differences in technology use across CALD patients in Australia. The findings of this study could be used to guide the design and implementation of the technology-based CR. Furthermore, mobile technology-based CR interventions seem promising to patients from CALD backgrounds, and the identification of the relevant technology use is the key to a successful implementation.