The core goal of this study is to assess the usability of smartwatches among the older adult communities in an LMIC, like Bangladesh. The first step of the study was to identify interest levels and engage with communities of older adult individuals, as detailed in the following sections. Following that, a survey process of 3 cycles was conducted using purposive sampling techniques by recruiting individuals from different socioeconomic and educational backgrounds, lifestyles, and educational backgrounds. This study cycle includes (1) a short-term survey, (2) a long-term survey with 3 data points (DPs), and (3) a focus group interview with 2 DPs. The final step of the study includes survey analysis and finding key insights from all the cycles for designing a specific older adults–age-centric health monitoring smartwatch. Figure 1 presents a flowchart of the methodological framework of the study.

An institutional review board (IRB) was approved by the North South University Scientific Regulatory Committee, and all participants were asked to complete a consent questionnaire. At the start of our study, we surveyed older adult individuals aged 60 years and older from our local community (friends, families, neighbors, and service staff) to understand their interest in using smartwatches for health monitoring. Then we contacted 5 old-age homes in and around Dhaka City and visited them to assess the interest level of participation. Initially, 2 old-age homes and a group of janitors (from residential areas of the authors) agreed and were recruited in the short-term survey study. However, due to internal management issues at one of the old-age homes, the consecutive surveys were conducted with the participants of one old-age home called Probeen Nibash (managed by the Ministry of Social Welfare in Dhaka, Bangladesh). We procured approval from the management of Probeen Nibash to conduct the study at their center and then talked individually to residents to explain the purpose and format of the long-term study. Based on their consent and the inclusion-exclusion criterion, selective participants were recruited from this old-age home. Similarly, several janitors were approached from the authors’ local communities, and those who willingly consented were recruited to continue the long-term study. The participants were between 60 and 103 years old, male and female, and purposefully sampled from diverse backgrounds. The participants were divided into 2 groups based on socioeconomic and educational backgrounds.

Older adults (older than 60 y) who were generally healthy and able to walk >50% time around on their own or with the use of an assistive device (such as a walker) were included. Older adults with restricted mobility (like in a wheelchair >50% time) or conditions that may limit the use of devices, such as blindness and cognitive decline, were excluded.

Before the first survey, a preliminary questionnaire session was conducted with each participant to collect demographic and baseline health-related information, including age, gender, occupation, educational qualification, and medical history of the past 6 months. The objective of this preliminary questionnaire was to evaluate their eligibility (and health condition) and interest in the study. Older adults (older than 60 y) who are generally healthy and able to walk >50% of the time around on their own or with the use of an assistive device (such as a walker) were included. In contrast, older adults with restricted mobility (like in a wheelchair >50% of the time) or conditions that may limit the use of devices, such as blindness and cognitive decline, were excluded. Aging people with severe cognitive dysfunctions were also excluded. The eligibility criteria included a set of subsectioned questionnaires to assess the participant’s health and mental health conditions, technological knowledge, and experience with modern technology to ensure their suitability for our study. From these observations, we found that most participants had no previous experience using technologically advanced devices like any smart wearable, and none had ever used a smartwatch.

Each participant was provided with an Oraimo Watch 4 Plus smartwatch after providing written informed consent to participate in the study. A detailed user guide was also supplied to ensure that participants clearly understood how to operate the device. At the time of the study, the device was commercially available at an approximate cost of US $20.06. The Oraimo Watch 4 Plus was selected due to its affordable price and the availability of standard health-monitoring features commonly found in other commercially available smartwatches.

This survey aimed to determine short-term adherence and compliance to the protocol using the smartwatch device among the participants. Short-term surveys were conducted in 3 places after the participants were distributed the smartwatch for 14 days, with 27 participants from varying socioeconomic and educational backgrounds, as presented in Table 1. Among them, 15 out of 27 (56%) were male participants, and 12 out of 27 (44%) were female, with an average age of 60.55 (SD 7.38) years. In contrast, 13 out of 27 (48%) of participants belonged to a higher socioeconomic background, while 14 out of 27 (52%) belonged to a lower socioeconomic background. This survey involved a comprehensive questionnaire with 19 questions, including 15 multiple-choice and 4 short-answer questions. The following key questions were asked of the participants during the short-term survey:

After completing the short-term survey, the long-term survey had to be proceeded with fewer participants. The participants of one of the old-age homes were excluded from the survey due to internal management issues at the organization, leading us to proceed with just one old-age home’s participants and a group of janitors. Also, a few participants from these 2 groups felt disinterested in continuing after the first DP of the long-term survey, primarily due to health issues and inability to adhere to the study guidelines. The first DP, long-term survey-1, was conducted with 19 participants, while long-term survey-2 and long-term survey-3 were conducted with 18 participants due to the sudden death of one of the participants from long-term survey-1. Each long-term survey was conducted at an interval of 10 days. Therefore, this survey section lasted for 30 days. As described in Table 1, the long-term survey-1 was conducted with 15 out of 19 (79%) male and 4 out of 19 (21%) female participants, 12 out of 19 (63%) belonging to a high socioeconomic background, and 7 out of 19 (37%) from a low socioeconomic background. In contrast, long-term survey-2 and long-term survey-3 were conducted with 16 out of 18 (89%) male participants and 2 out of 18 (11%) female participants. Additionally, a minimum of 8 out of 18 (45%) of the participants belonged to high socioeconomic backgrounds, while the other 10 out of 18 (56%) were from low socioeconomic backgrounds. The survey questions of the long-term survey were correlated to the short-term survey, and some variations in question types were used to improve assessment and analysis. The aim of this survey session was to find insights into usability factors of the smartwatch over a long-term period.

The key questions of the long-term survey included:

Based on the socioeconomic background, 2 focus groups were formed among the participants of the long-term survey. However, some participants did not show interest in participating in these sessions due to various reasons. Group A included participants from higher socioeconomic backgrounds, while group B included participants from lower socioeconomic backgrounds. The first focus group survey was conducted 14 days after completion of the long-term survey (round 3), and the second focus group survey was conducted 14 days after the first focus group survey. These focus groups were repeated in 14-day intervals to investigate if the participants’ responses were consistent. More importantly, the focus groups were stratified across socioeconomic backgrounds to understand if and how different backgrounds might influence their experiences with the survey. The first focus group survey was conducted with 13 participants, while group A comprised 7 out of 13 (54%) participants, and group B comprised 6 out of 13 (46%).

In contrast, during the second focus group survey, there was 1 dropout due to illness, consisting of 12 participants, with 5 out of 12 (42%) representing group A and 7 out of 12 (58%) from group B. Each focus group session was conducted for approximately 1 and a half hours at the primary location of the participants to ensure convenience and comfort. Participants were asked structured questions during the sessions, and their responses were collected individually. A casual conversation was facilitated alongside the structured format to explore their thoughts and expectations regarding wearable technologies. The entire session was audio-recorded with participant consent to capture all responses accurately.

The first short-term survey of the study was conducted to determine compliance levels among the participants from varying socioeconomic and educational backgrounds. The survey was designed with a variety of questions mentioned in the short-term survey section. A significant variation in responses was observed.

Responses to the first question, regarding the duration of smartwatch usage, are portrayed in Table 2. Notably, except for 1 individual, all participants used the smartwatch for at least 3 hours daily, with most participants using it for some time (defined as 3‐4 h), most of the day, and all day, highlighting the interest of the participants in exploring the smartwatch’s features and usability.

The second question examined the health features participants tracked. Most of the participants used or tried using multiple features. As illustrated in Table 3, the most frequently tracked features were heart rate monitoring (“Heart Rate”), step counting, and oxygen saturation level monitoring (SpO₂). Less frequently tracked features included sleep quality, calorie burn, respiratory rate monitoring, and stress level monitoring. Participants indicated that “Heart Rate,” “Step Count,” and “SpO₂” features were easier to access and less complex to use than “Sleep Quality,” “Calorie Burn,” “Respiratory Rate Monitoring,” and “Stress Level Monitoring,” which were described as more challenging to understand and monitor effectively.

The third and fourth questions addressed privacy concerns and workplace compatibility. In response to these questions, all participants expressed positive views, indicating comfort in using the smartwatch at their workplaces without privacy concerns.

Questions 5 and 6 explored the features participants liked and disliked. Heart Rate emerged as the most liked feature due to its simplicity and ease of use.

Comfort, addressed in questions 7 and 8, was another key area of investigation. As illustrated in Table 4, most participants reported the smartwatch as “completely comfortable,” while a smaller group consisting of 3 (15%) users found it “mostly comfortable,” and another group of 4 (20%) users felt “somewhat comfortable,” suggesting that participants generally did not encounter significant physiological discomfort while wearing the device.

The ninth question examined whether the smartwatch posed religious, social, or cultural barriers for older adult individuals. As illustrated in Table 5, a majority (20 participants) responded “not at all,” while 2 participants reported “slight” barriers, and 2 remained neutral. Only 1 participant perceived the watch as a large hindrance while performing religious, social, and cultural activities.

Participants were also asked whether they found the smartwatch age-appropriate and if they would recommend it to others in their age group. As illustrated in Table 6, approximately 52% (n=13) of the participants considered the smartwatch suitable for their age and indicated they would use it for a longer duration in the future. In comparison, 8 (32%) participants expressed neutrality. On the contrary, 2 participants found it very age-inappropriate and declined to use it in the future, citing concerns about advanced technology for older people. Furthermore, 1 female participant (>70 y old) from an old-age home commented:

Another question explored whether the smartwatch helped participants achieve health goals. A total of 13 (52%) participants believed it was helpful, while 3 (12%) felt it could not independently establish health goals. Moreover, 9 (36%) participants expressed neutrality. Table 7 presents an analysis of these responses.

The final question focused on whether participants found the smartwatch frustrating to use. As illustrated in Table 8, 15 (60%) participants responded negatively, 4 (16%) reported frustrations, and 6 (24%) remained neutral. Frustration was reported mostly because of the age-inappropriate user interface.

The survey also captured unique insights from participants, particularly those with lower levels of education. For example, a janitor noted:

Similarly, another participant of an old-age home resident noted:

Another female participant (62 y old) from an old-age home reported that she felt nausea, dizziness, and joint pain whenever she used the smartwatch.

The overall responses of this survey phase highlight that both groups of participants were curious about the smartwatch. However, they faced some challenges while using it. According to the responses, Heart Rate, Step Count, and SpO₂ were the easiest features to monitor. Also, the watch was comfortable overall and created no barriers while doing regular activities. However, a few interesting comments were received from participants with lower education and socioeconomic backgrounds. Also, a large portion of the neutrality in the usability of the watch in achieving health goals indicates that a group of participants was confused about the appropriate technical knowledge of using the watch. Many participants shared their complexity in understanding the watch’s functionality during the interview. The research team answered that confusion with proper care and ensured clarification about the issues.

The given survey revealed some interesting trends in smartwatch use and responses from participants across 3 phases of the long-term survey. Unlike the former short-term survey, this one included a series of questions categorized into daily usage, comfort level, feature usage patterns, and expected improvements that participants made.

As Figure 2 illustrates, 31.5% (6/19) of the participants used the smartwatch for 12‐20 hours daily during the first phase of the survey, thus being the most common usage duration; this was followed by 21% (4/19) of participants who used it from 9‐12 hours, and exactly the same percentage for those who said they used the smartwatch for 1‐4 hours daily. Smaller use of the smartwatch (daily periods exceeding 20 h) was reported by 10.5% (2/19), while a single participant reported not using it at all. Usage patterns continued shifting through the second phase, with 31.5% (6/19) of participants using the smartwatch for 5‐8 hours daily and 36.8% (7/19) for 12‐20 hours. Only 15.7% (3/19) of participants used the watch less than an hour daily, indicating that almost all participants engaged with the device to some degree. Finally, in the third phase, 20% (5/20) of respondents resumed using the smartwatch in the 12‐20 hours daily category, and 35% (7/20) used it between 5‐8 hours.

The results of the t test for daily usage, depicted in Table 2, showed a lack of significant variance in usage patterns across all phases; this is evidenced by P values ranging from .78 to ≥.99, which reveals that the overall time spent by the participants on the smartwatch remained relatively stable across the course of the study.

In fact, comfort levels improved a lot in the 3 phases, as portrayed in Figure 3. Initially, 47.3% (9/19) of the participants found the smartwatch comfortable, with only 26.3% (5/19) describing it as very comfortable. A small percentage, 5.3% (1/19), described it as highly uncomfortable since the features were complicated and the users did not have enough knowledge about them. However, in the second phase, no one reported discomfort, while 38.9% (7/18) found the smartwatch very comfortable, which marks the positive shift in user experience. In the third phase, 80% (16/20) rated the smartwatch as very comfortable, while only 1 participant reported it as fairly comfortable, amounting to 5% (1/20).

Furthermore, t test results of comfort ratings are insignificant among the phases, with P values ranging between .78 and .95 (Table 9). However, as revealed by the chi-square test, comfort perception improved dramatically over the survey period (Table 9), with a significant association—namely, P=.03 of phases with comfort ratings.

The usage of the features shows some fixed preferences along with some dynamic behavior, indicated in Figure 4 below. The most used feature in all phases remains heart rate monitoring, which shows an increased usage level from 68.4% (13/19) in phase 1 to 75% (15/20) in phase 3. The participants were primarily using other features such as SpO₂ and Step Count, but with a small fluctuation in them. For example, usage of SpO₂ reduced in phase 2 but increased in phase 3, which shows they were testing different features before making a preference.

The respiratory rate function, which had not been used in phases 1 and 2, gained acceptance among 15% (3/18) of participants in phase 3, which indicated increased interest in advanced functions with time as participants got accustomed to using the device.

The results of the t test indicated a significant change in used features over time, especially when comparing DP2 and DP3 (P<.001) and another significant change when comparing DP1 and DP3 (P=.03), which indicated that people have gotten used to using the smartwatch features.

Besides usage, another area that we have explored is the most liked features in each stage of our survey, which can be seen in Figure 5. Heart Rate remains among the most liked features in all phases. As seen in Table 2, t test analysis for most liked features did not reach significance in the comparisons of DP1 with DP2 with a P value of .62, DP2 with DP3 with a P value of .30, and DP1 and DP3 with a P value of .55 for Heart Rate.

For example, in an attempt to capture features such as SpO₂, the participants recorded a higher preference in the third phase relative to the first phase, as can be noted by the SD recorded in Table 9, which stood at 1.24722 (mean 4.33). Some of these features, such as SpO₂, were thus better appreciated when participants were more conversant with the device. Other functions, such as tracking and respiratory rate measurement, were well-received but demonstrated inconsistency in preference, especially in the early phases.

What is interesting, however, is the fact that respiratory rate tracking was considered hard to use but gained acceptance over time, which is in line with the increased usage of advanced functionality in total. The nonuse of some functions in the first phase, with 31.6% (6/19) of participants not using any functions, highlights the learning curve involved in using smartwatches in general, which had declined substantially in phase 3. The above findings indicate how preferences in features change over time, with Heart Rate being a continually popular choice, given other features, such as SpO₂ and respiratory rate monitoring, becoming more popular with increased familiarity.

Several trends were observed based on the feedback obtained from the participants with regard to the clarity of instructions, as presented in Table 10. At the beginning, 78.9% (15/19) found it simple and clear; hence, the reason for the easy adoption. This trend flipped as users continued using the application to explore features that became more complicated. In the second and third phases, the participants began to veer toward the neutral side concerning clarity. This is depicted by the high SD obtained in “Yes” (mean 12, SD 2.44) and “Neutral” (mean 4.33, SD 2.86), indicating that respondents do change with increased usage upon exploring more features. These results stress the need for increasing the clarity of instructions as users proceed to more complex features.

In Table 9, the chi-square test indicates a significant association (P=.04) between survey phase and participants’ opinion on instruction clarity, which signals the dynamic user perception as feature complexity increases.

Figure 6 illustrates that the most difficult feature, especially in the third DP, is stress measurement. This would indicate a possible design or functionality issue with this feature. The same trend occurred for respiratory rate monitoring, which was considered a difficult feature through an increasing number of participants during the course of the experiment. Finally, on the opposite side of the spectrum, Heart Rate was the most favored feature and had no reports of difficulty throughout all points.

Furthermore, t test results indicate significant changes in the perceptions of difficulty regarding difficult to use features, where P<.001 for DP2 versus DP3 and P=.05 for DP1 versus DP3 support that design improvements are required in some features, such as stress measurement and respiratory rate.

After long-term use of the smartwatch, adaptability and comfort were observed among the participants, as illustrated in Figure 7. Alongside, a trend was also observed in the expected improvements from the smartwatch (Figure 7). The data show that the most commonly suggested improvements were blood pressure and diabetes monitoring capabilities from the smartwatch, indicating that older people are mostly interested in critical health monitoring features. Interestingly, some participants initially had no comments on improvements but later suggested new features, though this shift was minimal. The overall low SD across all expected improvements confirms that user needs remained stable, reinforcing the necessity of designing a smartwatch specifically for older adults’ health monitoring.

The focus group interviews constituted the final phase of the study and were designed to provide qualitative insights that contextualize and extend the survey findings. Moreover, 2 rounds of focus group interviews were conducted using identical structured interview guides to enable both cross-group and longitudinal comparisons. Participants were stratified by socioeconomic status and educational background to explore how these factors influenced perceptions, adoption, and experiences of smartwatch-based health monitoring.

Group A comprised older adults residing in an assisted living facility, representing a higher socioeconomic and educational profile; all participants had completed at least a high school education, with the majority holding undergraduate or graduate degrees. Group B consisted of individuals from lower socioeconomic backgrounds with less formal education (below high school level), who were employed in custodial and maintenance roles. This stratification enabled a comparative examination of digital health adoption across differing social and educational contexts.

Qualitative responses were analyzed thematically, with dominant patterns reported alongside illustrative quotations. Descriptive response counts are included to enhance transparency, consistent with Journal of Medical Internet Research qualitative reporting guidelines.

The focus group interviews took place in separate locations and under the same setting to gather similar perspectives. A total of 13 participants participated in the study, including 9 males and 4 females. Group A included 7 participants, of whom 4 were female and 3 were male. Group B consisted of 6 male participants. Both focus group interviews followed the same structured questionnaire to examine potential differences in responses and variability across the groups.

Participants’ first impressions of smartwatches revealed clear group-level differences. Group A participants predominantly expressed positive attitudes, whereas group B participants more frequently articulated curiosity rather than immediate endorsement. A small subset of participants (2 from group A and 1 from group B) expressed skepticism or negative perceptions.

Qualitative analysis identified three emergent subthemes:

For instance, concerns about language and accessibility were raised by a group B participant (“I cannot read English texts well”), while doubts regarding clinical validity were voiced primarily by group A participants (“Technology cannot replace doctors”). Notably, these reservations were more pronounced among participants from higher socioeconomic backgrounds, despite their higher educational exposure to technology—implying deeper analysis and more cynicism in technology trust among users with higher education.

When asked, “Did you find the smartwatch easy to use?” Most participants rated usability as “easy” or “average,” with only 1 participant reporting significant difficulty, as represented in Figure 8. Qualitative responses suggest that initial apprehension transitioned into gradual acclimatization, indicating a learning effect rather than persistent usability barriers. Participants emphasized novelty, visual familiarity with traditional watches, and perceived multifunctionality as positive features.

In light of this question, a further question was asked, “What did you love about it?” In response to this question, some participants answered,

A participant from group B answered,

Therefore, this shows excitement in some older people about using new and “smart” technology, irrespective of their social belonging or understanding of the technology.

While overall comfort ratings were favorable, as illustrated in Figure 9, discomfort related to weight and strap design emerged as a recurring concern, particularly among group A participants. Participants expressed strong preferences for traditional materials (eg, leather or steel straps), suggesting that social acceptability and aesthetic continuity play a role in long-term adoption.

The question, “Would you like your data to be continuously and automatically shared with family members and/or designated caregivers?” received mostly positive feedback. As illustrated in Figure 10, all participants in group A responded positively. In contrast, 4 out of 16 participants from group B preferred that data sharing be enabled only during emergencies, while 2 participants expressed full support for continuous sharing. This divergence highlights differing expectations regarding privacy and data oversight between the 2 groups, as the less privileged (and less educated) group identified interpretability concerns and seemed more aware of not burdening their family with their health emergencies.

In response to the question, “How accurate did the health metrics seem to you?” participants generally perceived health metrics as reliable. Limited engagement with data—not distrust—explained neutral responses, indicating that attention and interpretation, rather than sensor credibility, influenced perceived accuracy.

The second focus group interview involved 12 participants (10 males and 2 females), with 5 participants in group A and 7 in group B. This interview was conducted 14 days after the first, enabling assessment of temporal stability and experiential adaptation.

Initial impressions in interview 2 closely mirrored those from interview 1, demonstrating thematic consistency over time. However, qualitative depth increased, with participants expressing more experience-based reflections rather than speculative judgments. Curiosity remained dominant in group B, while group A responses polarized slightly between positive endorsement and skepticism.

Compared with interview 1, negative perceptions were less frequent and more contextualized, suggesting that prolonged exposure mitigated abstract concerns while amplifying practical considerations.

Usability ratings improved overall in interview 2, particularly within group B, as depicted in Figure 11. The reduction in “average” or negative usability responses indicates learning consolidation rather than mere tolerance. Difficulty reported by 1 older participant (88 y) underscores the influence of advanced age on interaction with wearable technologies, highlighting the need for age-adaptive interfaces.

Figure 12 illustrates that comfort ratings in interview 2 remained stable relative to interview 1, reinforcing the robustness of this finding. Importantly, fewer participants raised new complaints, suggesting that comfort-related perceptions plateaued early and did not deteriorate with extended wear.

A clearer divergence in data sharing preferences emerged in interview 2, as illustrated in Figure 13. All group A participants supported continuous data sharing, whereas the majority of group B participants favored emergency-only sharing. Compared with interview 1, these preferences were articulated with greater certainty, indicating preference crystallization over time rather than ambivalence.

Confidence in metric accuracy increased slightly in interview 2, particularly among group B participants. Overall, the responses were mostly neutral and a bit vague at this phase, again reflecting lack of attention rather than distrust, reinforcing findings from interview 1.

Across both focus group interviews, dominant themes, including usability adaptation, comfort acceptability, data-sharing conditionality, and perceived metric reliability, remained consistent. The repetition of themes across independent samples and time points supports thematic saturation, enhancing analytical power despite modest sample sizes.

Although formal statistical power calculations are not applicable to qualitative focus group data, analytical power was strengthened through:

The observed pattern, where participants from lower socioeconomic backgrounds adopted the smartwatch more readily and expressed fewer aesthetic or conceptual reservations, was robust across both interviews. This repeated contrast increases confidence that the finding reflects a genuine behavioral and perceptual trend rather than sampling variability.