Older Individuals and Digital Healthcare Platforms: Usage Motivations and the Impact of Age on Postadoption Usage Patterns

Digital healthcare platforms have enabled patients to receive healthcare in ways that were impossible previously—for example, by providing a “safer” way to meet, as underscored by the Covid-19 pandemic. This article investigates whether older and younger primary care users display behavioral differences on digital healthcare platforms. The article adopts a mixed-method approach in which one-way ANOVA analysis on a sample of 152 000 patient journeys was combined with qualitative interview data. The findings highlight significant differences in usage between elderly and younger patients. The elderly spends more time during use—for example, during anamnesis, onboarding, and in queues. We also outline how the key antecedent factors that are most central to platform usage, such as perceived usefulness, perceived ease of use, digital maturity, and trust, play out in the elderly user context. The study contributes to the nascent literature on digital healthcare platforms and the postadoption usage of information and communication technologies by the elderly. The article also outlines research implications in the area of DHPs and mHealth for elderly users, and it discusses the practical implications for both platform owners and healthcare professionals, where platform design and information management are particularly important for elderly users.

healthcare sector [1], [2] and pave the way for new types of innovative services. Digital healthcare platforms (DHPs) draw on ICTs, allowing patients to interact remotely with healthcare professionals and to receive care in new ways (via computers, tablets, and smartphones) [3]. DHPs are a form of mHealth, defined as "the use of mobile and wireless devices …to improve health outcomes, health care services, and health research" (National Institutes for Health). DHPs offer patients a variety of mHealth services accessible through a single interface, thus enabling interaction with many health professionals. Hence, DHPs comprise both a specific platform technology architecture and a platform-based business model [4].
While the platformization of healthcare is in its infancy, DHPs have recently gained market share, particularly in primary care settings. For instance, in Sweden, DHPs gained momentum in 2015 and have accounted for approximately 10% of all primary care visits before the Covid-19 pandemic [5]. Covid-19 accelerated their usage further [6]. Indeed, leveraging digital health technologies plays a central role in responding to Covid-19 [7], [8]. Further demonstrating their significance, there are over 40 online healthcare service providers in China [9].
Elderly patients, who represent the largest share of healthcare consumption, are potential beneficiaries from DHPs. The world's population of 60-year-olds and older is expected to increase dramatically by 2050 [10], [11]. Population ageing is recognized as a grand challenge, and a pressing social and business issue [12]. Whilst managing this area remains underresearched, the difficulties involved in taking steps to address older individuals have been recognized [13]- [15].
In health care, population ageing places increased pressure on primary and specialist care resources across countries [16]. In principle, DHPs can help mitigate this challenge [9]. Indeed, DHPs have the potential to facilitate independent living [17], which is a key concern in the elderly population [18]. Indeed, in Norway, a key innovation strategy is to facilitate healthy ageing for older individuals by helping them to stay in their homes [19]. Thus, DHPs could relieve pressure on physical primary care services whilst offering easy access to follow-up consultations for patients with chronic diseases [11], [20].
Despite these opportunities, challenges are evident in addressing the needs of elderly individuals who remain significantly under-represented in DHP usage [21]. In Sweden, despite constituting approximately 25% of the total population, only 4% of people over 60 used such services in 2018 [22]. This may be partly explained by prior findings of e-Health studies, which identified barriers confronting the elderly in the adoption process that result from lower levels of education, perceived complexity, and a lack of prior experience and guidance [1], [11], [17], [20], [23]. Prior literature has also highlighted the divide in DHP usage between the generations [24].
However, there is uncertainty, both in theory and practice, as to whether and how DHPs could improve care for the elderly, and to what extent. While it is increasingly acknowledged that "the elderly" is a heterogeneous group who should not be viewed as "laggards" responding negatively to new technology [25], prior research has provided few insights into attitudes to DHPs let alone adoption and postadoption usage patterns. Instead, extant studies have focused either on mHealth adoption antecedents among physicians [9], [26], [27], or on factors influencing the propensity of younger individuals to use DHPs, such as online healthcare communities (which provide users with health information and/or peer-to-peer communication) [28].
Hence, despite the grand visions tied to the potential of e-Health and DHPs [29]- [32] including their potential to address needs among the elderly, we know very little about how elderly individuals use DHPs. Consequently, this reduces our ability to develop realistic expectations on the role of DHPs in the lives of the elderly, and it limits the scope for adapting ongoing developments to their needs. The lack of insight into actual user patterns applies to the literature on platforms in general, where studies typically aim to predict usage by exploring intentions (e.g., [33]- [35]) rather than unpacking the actual usage. These studies also examined younger individuals and failed to address individuals aged 65 and over.
Against this background, we address two gaps in the literature. First, little attention has been paid to the behavior of the elderly after the initial adoption of DHPs and whether there are problems in the process of receiving care that can spark digital disengagement [36]. Second, there is research on the elderly's perception of various ICTs [37]- [39]. However, few studies have paid specific attention to DHPs, which have begun to proliferate over recent years [1], [22], [23]. The purpose of the article is to investigate the behavior and needs of elderly primary care patients in DHPs. We address two research questions: RQ1: What are the differences in usage between elderly and younger primary care patients in digital healthcare platforms?
RQ2: Which factors are the key antecedents that influence usage among the elderly?
Based on an analysis of a comprehensive dataset (152 000 patient journeys) from a DHP, we contribute new insights into the impact of age on actual and postadoption usage patterns. We identify differences between elderly and younger users, as well as within the elderly user population studied (RQ1). Our qualitative study and analysis (RQ2) complement these insights by providing an understanding of the underlying reasons for usage among the elderly identified in RQ1 and by identifying important issues to consider in the future development of DHPs to better address their key concerns. These insights contribute to the broad literature on ICT in healthcare (including mHealth), which lacks empirical and theoretical insights into the postadoption DHP usage patterns and key aspects influencing the intention of the elderly to use DHPs. Our work opens several research avenues in this area.

A. Promises and Perils Tied to DHPs
DHPs have received increased attention from scholars in recent years [40]. The rise of ICT and advances in software development has paved the way for DHPs in the healthcare sector. In this study, we refer to DHPs as platforms where patients and care providers can interact and where interactions, transactions, and information exchange between these two groups are facilitated [41]. A DHP typically incorporates a set of mHealth services, which are available to the user through one interface. Value creation and delivery is contingent on a network of participants, and on the interactions and information exchange between them [42], [43]. We focus on DHPs in primary care. They typically offer a range of services primarily targeting patients with milder ailments-such as urinal infection, colds, and eczema-and include real-time video meetings with primary care personnel, real-time chat, and asynchronous messaging through which patients are diagnosed, prescribed with medicine, or transferred to other care units or levels.
As to the platform provider, it is important to distinguish between technology-related roles and business-related roles [4]. In the technology-related role, a provider designs the architecture of its platform by defining components and their interfaces. In the business-related role, a provider determines the means to attract end users and service providers (e.g., patients, health professionals, and third-party app/device providers). In the case of DHPs, these two roles may be fulfilled by one actor or distributed across different actors (technology vendors provide the platform software, whilst a private or public care service provider undertakes the business role).
In principle, there are multiple advantages with such platforms. DHPs can reduce the pressure on physical care and increase the availability of primary care services [44]. This holds the potential to address the problem of increasing strain on physical health facilities and concern over their resilience [45]. From a patient perspective, they can receive care with greater flexibility, mobility, and portability [46], alongside treatment and diagnosis at convenient times [44]. From the perspective of healthcare professionals, DHPs have allowed them to receive, evaluate, and store data for better diagnosis and decision making [47]. Moreover, this indicates that elements of the healthcare sector are being transformed from a one-way communication stream [16] to a more continuous and interactive process for patients [48]. In consequence, platforms offer opportunities to improve sustainability in healthcare service systems [49].
Despite their potential in several countries, DHPs have also received much criticism, especially those operated by private actors. Criticism has included the insouciant prescription of antibiotics, the draining of public healthcare sector resources (human and financial), and the "treatment" of healthy patients [25], [50]. Furthermore, not all diagnoses are fit for treatment through a DHP. From a legal perspective in Sweden, digital providers must comply with the law of accessibility [51]. All individuals, regardless of background, disability, and age, should have the same opportunities to access services. This further promotes the need to understand how various types of patient use and perceive DHP services so that they can address a wider population.
1) Unpacking Platform (Postadoption) Use: While studies of postadoption usage in relation to mHealth and DHPs (and on platforms in general) are scarce, the extant literature provides some useful definitions that we use as points of departure in our exploration of DHP usage among the elderly.
It is necessary to distinguish between adoption and usage. Adoption is the stage where a technology is initially selected for use by an individual or organization [52]. Usage is the stage after adoption where an individual or organization utilizes the technology [53]. This distinction is crucial since adoption does not necessarily lead to usage. Continued use results from the performance that users experience in use relative to their initial expectations [54]. Yet, many studies on platforms claim to study "use" but actually study "intentions to use" or adoption and, thus, treat this as a binary or outcome variable [33], [55].
Unlike concepts of preadoption and adoption, postadoption has been further defined, albeit loosely. For some research, postadoption represents the "continued use" [56], "continuous and repeated usage" [57], and "continued adoption or discontinuance" [58] of an innovation. These studies demonstrate that research on postadoption has largely focused on continued use, with few having examined discontinuance.
Our study draws on Ng [59] who outlines key aspects of postadoption behaviors (in the case of Twitter discontinuance). Drawing on Rogers' innovation adoption process, she suggests that postadoption behavior is not simply a binary distinction between use and nonuse but a range of levels of engagement with and disengagement from an innovation [60]. Hence, it is incorrect to assume that an individual who once used a DHP will always continue to do so, or that individuals who have stopped using the service will never return to it (temporary discontinuance) [53]. Drawing on these insights, we aim to explore postadoption usage in terms of the engagements an elderly user has with a DHP during a consultation (RQ1).
2) Unpacking Antecedents of Adoption and Use of DHPs: While research on the attitude of elderly users to DHPs is scarce or nonexistent, research on the antecedents of the adoption of ICT by individuals in general provides important inroads into our inquiry on how elderly users perceive DHPs (RQ2). At a general level, usefulness (the value users experience by utilizing a technology and how well it meets their desires in relation to existing services [61]) and ease of use (the perceived effort users must mobilize to utilize the service [2], [62]) are emphasized in the vast literature on technology adoption and usage employing the technology acceptance model (TAM) [31], [63]- [66]. The theoretical starting point for our study is the TAM, which has been used extensively to predict usage [36], [67]- [71]. Furthermore, scholars have used TAM to theorize factors that enhance and counteract adoption and usage [31], [63]- [66].
Studies have also identified external factors that patients have considered in relation to ICT in healthcare. Here, satisfaction with the treatment has been shown to influence attitudes [72]. Digital experience in terms of technology anxiety is also a factor that influences adoption and usage behavior concerning digital health-related services [11], [23], [31], [67]. Technology anxiety is a negative emotional response that centers on the fear or discomfort people experience when they use or consider using a technology [23]. Here, research shows that people with abundant digital experience have a more positive attitude to embracing digital healthcare [17], [31].
Several dimensions of trust would appear to be critical. These are trust in the care providers in general [63], [72] and trust in the way in which they handle information (perceived security and confidentiality) [1], [26], [73]. Social influence is another factor that impacts digital care adoption and use, with friends, family, and care providers playing an essential role in influencing patient acceptance of health-related services [36] for both preadoption and future use.
In our study, the influences of ageing are pertinent to DHP use. Ageing is multifaceted in nature, consisting of biological, cognitive, and social aspects [13], [74]. Biological and cognitive ageing in particular hold the potential to influence use. The elderly (defined as 60 plus) is a heterogenous group [31], [75] but, as a whole, this age cohort differs from younger generations in important respects. For example, age is a significant factor affecting digital disengagement and an individual's perception of ICT [37]. The influences of ageing offer several potential explanations for this difference. First, biological changes may influence eyesight, hearing, strength, dexterity, and mobility [76], [77]. For instance, impaired vision and hearing are more common as individuals age [20]. Visual impairment hinders individuals from distinguishing between different colors [38], and impaired muscle function can make it a challenge to manage small buttons [39]. These changes may present significant usage challenges.
Second, ageing influences cognitive abilities, thus reducing individual mental capacity and information processing [13], [78]. Chronological age has been shown to influence attitudes to and use of ICTs in general and in healthcare. Cognitive load is a major concern for the elderly since they are more prone to cognitive problems, disabilities, and chronic diseases compared to younger generations [79]. Furthermore, deteriorating memory and reasoning can negatively impact the ability to learn to handle ICTs [39]. Whilst memory and reasoning problems will not affect all elderly individuals, such problems are more prevalent in this population. This may influence the perceived ease of use of and trust in DHPs, resulting in technology anxiety.
Elderly individuals will have learned about new technologies at a later age compared to younger individuals, making them less accustomed and comfortable with new technologies such as DHPs [80]. Consequently, they may experience reduced confidence in managing the technology [38]. This is also related to demographic factors. Retirement is strongly associated with nonuse or ex-use of technology, where the elderly are six times less likely to be online than employed individuals [81]. Retired people with less income sometimes find it a problem to pay for equipment. In terms of the social aspects of ageing, they may lack the influence of colleagues to which they had formerly been exposed at work [67], [81]. Combined, these factors, related to the technology anxiety and social influence identified in ICT adoption studies, may result in negative attitudes to seeking and receiving digital care [1], [31].
To conclude, our literature review reveals several factors that, alongside the characteristics of an ageing population, may impact usage and, thus, preclude continuous use or trigger the decision to disengage. Yet, there is limited knowledge on how these factors play out in the context of DHP usage.

A. Research Approach and Case Selection
This study followed a mixed-method research design, meaning that qualitative and quantitative approaches to data collection and analysis were used [82]. The quantitative data and associated analysis were aimed at answering RQ1 while the qualitative analysis was instrumental in answering RQ2. We followed an iterative research process where the principles of abduction were applied, which involved multiple iterations between the data and prior literature to validate our results [83].
The study was conducted with a company developing and supplying a DHP that enables interaction between patients and healthcare professionals. Hence, this firm assumed the role of technology-related platform provider. The platform is currently used by some of the largest primary healthcare providers in Sweden. Patients can access the platform from either a computer, smartphone, or tablet and can sign in using a Digital BankID. Interaction with healthcare professionals occurs through chat and/or video meetings. The DHP had around one million patients each year. Of the nearly one million observations, approximately 19.4% of the total usage was represented by patients aged 60 to 74 and only 3.8% aged 75 plus. The elderly were thus under-represented.

B. Data Collection
Primary and secondary data were collected between January and May of 2021. The primary data consisted of 20 semistructured interviews. Secondary data consisted of 152 000 logged patient journeys inside the DHP-namely, the behavior when seeking primary care consultation from initiating care to case closure. The secondary data thus focused on users of DHP services, and nonusers were not investigated. Data collection was organized in two phases.
Phase 1: We conducted exploratory interviews to gain familiarity with the context, setting, and associated database for secondary data (Table I provides descriptive information on all  interviews).
We also sampled data from the DHP consisting of patient journeys that had been processed during the spring of 2020. We limited the sample to patient-initiated primary care. The initial sampling frame consisted of almost one million anonymized observations, all from 2020, of which 152 000 were randomly selected for this study. The remaining observations were excluded due to data storage capacity constraints. Younger generations were over-represented in DHP usage, so we created a stratified sample with an equal number of observations in each strata (38 000 in each age interval) to facilitate statistical analysis and minimize the possibility of type 1 and type 2 errors [84]. We excluded individuals between the ages of 0 and19, since they often receive help from a guardian, which could render misleading information on usage. The final age intervals consist of four groups, 20-39, 40-59, 60-74, and 75+ [85]. Observations from patients in the 60 and above age range are considered elderly in this study [86]. Moreover, women were over-represented in our sample, which may mean that women either have a greater interest in their health [80], [87], or a greater propensity to use DHPs. However, prior research revealed that gender had no significant impact on DHP usage [10]. Finally, all observations included in the analysis were randomly selected from different months to account for potential digital disturbances in the platform throughout the year. Table II provides an overview of the quantitative sample.
Phase 2: The second data collection phase included 17 semistructured interviews with elderly primary care patients. The selection process followed the nonprobability sampling principle [88]. We sampled respondents aged 60 plus who had previously used a DHP. Specific respondents were recommended through the DHP we studied, via the Swedish National Organization of Pensioners, and by posting on social media platforms (e.g., LinkedIn). The interviews used a semi-structured interview guide (see Appendix A). Users were asked about their experience of using a DHP on their patient journey, why they used it, and what perhaps made them hesitant toward further use. Interviews were conducted by phone or video (due to the Covid-19 pandemic), recorded, and subsequently transcribed.

C. Data Analysis
Phase 1: For the quantitative data used to answer RQ1, we relied largely on a one-way ANOVA analysis followed by the Games-Howell posthoc test for each variable that potentially could reveal differences in the usage of elderly and younger patients. The Games-Howell posthoc test was deployed because our dataset did not fulfil the assumption on the homogeneity of variance [89]. The reason concerns the design of the DHP, where patient journeys unfold somewhat differently depending on the search cause, the level of care, and nonmandatory steps, among others. However, sample size was considered sufficiently large so that the mean value or standard deviation was not affected by smaller inequalities in the sample [90]. All other standard assumptions for an ANOVA analysis were met and validated against prior recommended values for skewness and kurtosis [91]. All tests were executed on a 95% confidence level. Consequently, extreme outliers were excluded after carefully looking at boxplots and histograms [84]. For example, anamnesis duration was excluded if it lasted 50 min or more since it indicated that a patient had either quit or paused the consultation. As a precautionary measure, we performed the Games-Howell posthoc test before and after the outliers were removed, and the results proved stable. Finally, we supplemented the findings from the ANOVA analysis with standard bivariate analysis because some variables only took the values of 0 and 1. Table III describes all tested variables on which the age groups were compared.
Phase 2: Thematic analysis was conducted to examine the qualitative interviews, pertaining to RQ2. We followed the sixstep approach of Braun and Clarke [92]: 1) becoming familiar with the data, which involved interview transcription, reading, and taking notes to document interesting findings; 2) generating initial codes, based on repetition, surprising information, and connections to the literature; 3) analyzing the identified codes in-depth and beginning to conceptualize potential themes. The first draft of themes was conducted with the help of the previous literature. A mind map was drawn to sort the codes systematically and was then translated into formal themes and subthemes.
In the fourth step, all identified themes and codes were reviewed more closely. While some themes were excluded due to a lack of similarity or data, others were renamed to obtain wider scope (i.e., digital experience was changed to digital maturity). In step five, we further analyzed and refined our themes, subthemes, and codes to make sure that we had captured the essence of the data. We thus identified the potential antecedent factors underpinning elderly usage in a DHP context. These themes confirmed those from the previous literature and added new dimensions. In the final step, we ensured that the themes, subthemes, and codes accurately represented the data by revisiting the transcripts and checking for intercoder reliability.

A. Differences in Usage Between Elderly and Younger Primary Care Patients, and Among Elderly Patients
The ANOVA analysis revealed a significant difference in usage between younger and elderly primary care patients on almost all variables (see Table IV). The analysis regarding process time yielded the greatest differences. The time spent in minutes during anamnesis was significant (p < 0.001) among all groups and increased in line with the age intervals. Similarly, the elderly spent more time during onboarding compared to the younger generation (p<0.001 among all groups). However, the analysis regarding case duration showed that the youngest and the oldest groups took the longest time, which was unexpected. That said, those in the group 75 plus are significantly different from the age groups of 40 to 59 and 60 to 74 (p<0.001).
Waiting time increased in line with the age intervals (p<0.001 among all groups), except between the two younger groups where no significant difference was found. This is also true of the health centers opening waiting time, where the two elderly groups spent a longer time queueing. Here, a significant difference could be demonstrated among the youngest age group and the remaining three (p<0.001). Precedingly, a significant difference (p = 0.004) was found between the groups 40 to 59 and 75 plus. In sum, the elderly tend to spend more time during anamnesis, onboarding, and queueing, with those aged 75 plus scoring the highest on all measured variables. We also investigated the number of assigned healthcare professionals that each patient was transferred to after the first case contact (meaning that all patients are not transferred on to a subsequent contact). Our results show that patients aged 75 plus tend to be assigned slightly fewer professionals (p = 0.000 among all groups). Our results also reveal a significant difference on the level of p<0.001 among the age groups 20 to 39, 40 to 59, and 60 to 74.
Finally, we investigated the dropout rate-that is, whether patients decide to drop out during the process and do not complete their case. The elderly were over-represented in dropouts before the first onboarding question, during onboarding, when Patients in the DHP were asked to rate the service and their overall satisfaction. Here, those aged 75 plus tended to rate the service slightly lower compared to the other groups (p<0.001). A significant difference was found between the age groups 40 to 59 and 60 to 74 (p = 0.034). However, the differences are small in this regard. We also performed simple bivariate analysis for the dummy variables. Patients aged 40 to 59 expressed the most agreement that they were treated with compassion by the caregiver (94.2%), whilst those aged 75 plus agreed the least. However, differences were small, and 93.0% still agreed that they were treated with compassion. The two groups of elderly tended not to recommend the DHP to others to the same extent as the younger generations. More than 94% of the two younger groups were happy to recommend it, while 93% of those aged 60 to 74 and 90% of those 75 plus. The willingness to recommend the service to others declines with rising age, but the two groups of the elderly still seemed largely positive.

B. Antecedents for Usage Among Elderly Users
Our results from the analysis of the interview data reveal that the key antecedent factors that influence usage among existing elderly DHP users is perceived usefulness, perceived ease of use, their digital maturity, and their trust in and attitude to digital healthcare. Below, we unpack the meaning of these factors from the elderly DHP user perspective. Fig. 1 provides an overview of the identified codes, subthemes, and themes. 1) Perceived Usefulness: a) Improved availability compared to offline alternatives: The analysis revealed that one of the key drivers of DHP use for the elderly was the very short waiting times and, thus, the improved availability of care. This contrasted with long waiting times to get an appointment at the ordinary primary care center. "You will have to wait forever if you call the ordinary primary care centre, or you will not get an appointment at all. However, here [with digital care] it takes four minutes and then you are in the process of receiving care. The availability was the main difference." R4 Several respondents stated that the effectiveness of seeking and receiving care digitally was the main benefit in comparison to traditional physical care.
"If I just want an answer to something that worries me, it's a very fantastic thing to be able to come forward and get in touch so quickly." R17 This is in line with the arguments put forward in the TAM framework, where the usefulness experienced depends on the difference in the perceived value of existing services [63]. Our findings confirm that the advantage is time, as emphasized by Fichman [93]. However, none of the respondents mentioned money or expertise as factors underpinning perceived usefulness.
b) Uncertainty of service scope-which issues digital healthcare can treat: Many of the elderly agreed that DHPs could be useful but mostly for milder ailments. More complex symptoms need to be treated physically. This may complement our quantitative results where the elderly tend largely to "solve" their cases in a digital environment compared to the younger generations. According to Martínez-Caro [94], the DHP services must be perceived as more advantageous for the patient to use them. However, our findings reveal that not all services need to be perceived as more advantageous to boost usage.
Many interviewees stated that, if they had multiple illnesses or had more complex diseases, they would find it challenging to seek care digitally. "In my case I have medication for high blood pressure, and I have diabetes, you must see the whole picture, doctors have to do it. So, if I need ointment or something similar then it is probably perfectly okay to take it that way but not all ailments I think." R5 Others who have multiple conditions or more complex medical histories state that using digital healthcare is sometimes problematic because they are required to repeat their medical history every time they seek care for milder ailments. This is related to the automated triaging offered by DHPs, which is a way to improve the matching of patients and professionals. Yet, the sometimes quite long self-rating questionnaires were perceived as very burdensome: Nymberg et al. [20] state that DHPs have the potential to facilitate independent living and the sense of feeling secure, especially for elderly people, by enabling easy access to followup consultations for those suffering from chronic diseases. This contradicts our findings where the elderly with a long medical history appear to be struggling to receive the help they need.
"When your generation also turns 70 or 75 and becomes ill and gets hurt and becomes multi-sick, I think it will be very difficult to have digital healthcare … I think so because you need to feel safe and that there is someone who supports you when you sit there and think …I have an acquaintance and good friend of the family who is multi-sick, and he says that I want the doctor home. I take it as a reference." R6

2) Perceived Ease of Use: a) Perception of the patient journey:
It is evident that several respondents had experienced complexity and obstacles in the process of using DHPs. These ranged from downloading the app, information overload, finding information on their test results and appointment details, understanding when the doctor answered, to pure communication difficulties. This may complement the quantitative results showing that the two groups of elderly took the longest time in most stages of the process. It may also explain why the elderly tend to rate the service slightly lower and do not recommend it to the same degree as younger generations.
"I try to go back in memory, but I think I needed to go through certain steps that I thought were a little bit difficult. I think it was finding the answers, the results came from the blood tests, and it was difficult to find where I could see them. So, I had to contact them and ask where I could find it." R14 However, some respondents suggested that it was an easy process to receive care and that the technical part was no major problem.

"I don't think that the technical part after you have entered is a problem. Where I was, the instructions were incredibly clear." R13
b) Degree of physical impairment: Physical impairments have also been shown to affect the usage of DHPs. Some respondents stated that a small text size challenged use due to reduced sight. Furthermore, the importance of creating an easy-to-understand layout was identified, with different colors to assist those suffering from various impairments. To summarize, physical impairments can create a complex situation when using digital healthcare. Such problems reflect the challenges associated with biological ageing (e.g., [76], [77]) that individuals have in utilizing the platform. These findings support the argument that elderly individuals suffering from impairments and cognitive problems may find it problematic to fully utilize digital care [10], [11]. Chou et al. [38] state that problems can be encountered either while signing in during first use or during engagement, which may explain digital disengagement. Building on their findings, our analysis pinpoints the exact steps in the process at which the elderly tend to drop out, and the underlying reasons for their behavior.
3) Digital Maturity: We can infer two dimensions of digital maturity: overall familiarity with digital technology, which captures the elderly's general experience of technology (from their earlier lives) and technological anxiety, which refers to anxiety specific to DHPs. a) Overall familiarity with digital technology: Many respondents stated that they had gained technology experience from their prior working life, which made them comfortable using digital healthcare.
"No, I did not think it was difficult to seek care digitally. Maybe it has to do with the fact that I am used to working with different applications so to say." R6 Others stated that the Covid-19 pandemic pushed them to use and test new technologies, which they eventually became more comfortable using. This finding aligns somewhat with the prior research of papers [17] and [31] who state that previous digital experience makes it easier to embrace new technology.
"Due to the pandemic, I think many people have become comfortable using computers. They are having video calls with their grandchildren. I believe that's something that will continue from now on." R18 b) Technology anxiety: The absence of technological equipment and skills are among the main factors that contribute to the digital divide between the generations. Many of the respondents agree that not everyone has access to computers or a digital BankID and that this becomes further nuanced with increasing age. However, it should be noted that our respondents were referring in general to other elderly people, whilst they themselves possessed the knowledge required to seek care digitally.
"If you have never chatted before or never had a video meeting, this was before the corona, then I can imagine that there are those who find it difficult [using digital care] … I look at my mother-in-law, she does not have a BankID, and we are terribly grateful for that with all the scams, but it also means that she is limited in the number of services. She presses the wrong icons; we know that because we have to reset her phone almost every time we get there." R15 "But then they must be connected, not everyone is and has that experience. Not everyone is connected and sitting behind a computer screen, that's not the case." R20 Some respondents suggested that training and guidelines are crucial for elderly people to start using digital care and that their lack of knowledge is the underlying reason for them experiencing technological anxiety.
"Maybe you should have a course in this [digital care] for older people and show them how it works, so you somehow reach out to these people. You can provide a short course, so people don't get so scared, I think so." R18 4) Trust: We identified two dimensions of trust that are salient to the elderly's reasoning on DHPs: 1) security, which refers to the perceived treatment of information by DHPs, and clear communication, and 2) confidentiality, which refers to the perceived trustworthiness of the communication.
a) Security and confidentiality: One of the key antecedent factors that influences usage by the elderly is perceived trust, which is mostly concerned with security and confidentiality.
"Some [processes] are very illogical and you have to do things several times, it is illogical how to enter your credit card number and it does not feel safe. You experience friction and then you give up." R3 "I rather prefer that they store my personal information and look at my journals so they can make the right diagnosis. I can read my journal on 1177 and it is great if those who treat me also can read it in order to get a full view." R15 Thus, the payment system must feel secure if it is not to influence usage negatively (c.f. [73]). Our results align with Kim et al. [26] who claim that, if the security risks are experienced as high, this negatively influences adoption and usage. Our analysis also suggests that integrity issues could increase with age.
"I had not thought of that, but the physical is probably to prefer, but it is nothing that I have reflected on or have a concern about [processing of personal information online]." R14 b) Clear communication: Several respondents highlighted the need for physical face-to-face interaction to increase trustworthiness and eliminate misunderstandings during the process of receiving care. This insight aligns with the findings of Chong [63].
"I want to communicate face to face, that's how we old people want to do, I think … I think many people feel secure sitting with their nurse or doctor and talking to him or her and being able to explain their problems." R11

"But otherwise, I think video meetings are good, especially if you have any visible symptoms …" R4
The preferred choice was to meet in person, but video meetings were seen as an acceptable complement. Lack of faceto-face interaction can negatively affect usage by the elderly because they are accustomed to the traditional way of receiving care physically.

5) Social Connectedness:
The elderly's answers reflected a desire not to be "out of date" and to keep up with society, its trends, and new opportunities. This desire exerted an influence on usage. In particular, their 1) motivation to learn something new, and 2) exposure to messages in their environment (social influence) that encourage usage. a) Motivation to learn something new: Interest and curiosity in testing and using technology have been shown to positively influence the motives and attitudes driving the use of digital healthcare.

"I like to learn gradually so I am not completely used to the new technology, but then I think it's fun to sit and learn new things." R10
Testing and using new technology was dependent on the motivation to learn rather than chronological age, which confirms the findings of Lagacé [95]. Yet, it is something that is characteristic of some elderly people and can influence use. Indeed, some respondents maintain that, if they see no possibility of receiving care physically, they will reconsider seeking medical support digitally.
"I will probably be hesitant [using digital healthcare again], but of course if I see no other possibility, I will certainly do it." R8 b) Social influence: Our results demonstrate that social influence has both a positive and a negative impact on attitudes to the use of digital care. The positive influence is based on family and friends. Furthermore, for some respondents, TV advertisements were a decisive factor in promoting usage.
"And then it was a colleague who recommended me [digital healthcare] so I used it …" R16 Authorized licensed use limited to the terms of the applicable license agreement with IEEE. Restrictions apply.

"They did so much advertising on TV about [digital healthcare] so I thought I would see what it was like." R17
On the other hand, acute criticism of a number of digital healthcare providers in the media had changed views of the digital healthcare sector for some interviewees, as well as their acquaintances. Peek et al. [36] state that social influence has proven to be the deciding factor in whether a person adopts and uses digital healthcare or not, a finding that is also reflected in our results.
"Then it's been a lot of writings too …. So, it affects mostly those my age, those my age is affected a lot. I can say that they believe it's hmm, morally wrong … And that [private digital healthcare] don't take the hardest cases. My friends who work in health care think that digital care takes the easy cases, and that's how it is of course. So, it's economical and that they treat what is easy." R13 V. DISCUSSION An aging population combined with higher patient expectations and restrictions on public spending have placed great pressure on healthcare services for the elderly in many countries [96], [97]. A grand hope is therefore tied to the potential of DHPs to address the increasing gap between the supply of elderly care and the demand for it. Yet, much of the platform literature focuses on its internal aspects [98] and empirical insights into the attitude of elderly users to DHPs. Research on how they actually use them is lacking [36], and the need to fill this gap has been accelerated by the Covid-19 pandemic [99]. This gap applies to the literature on platforms in general and beyond healthcare. Studies on "user behaviors" on various kinds of platform have explored intentions only-for instance, purchase intentions on social media platforms [33], "continuous use intentions" on sharing economy platforms [55], antecedents of intentions to use online peer-to-peer financing platforms [34], intentions to switch between mobile payment service platforms [35], and exploration of how platform governance mechanisms influence customer "participation" (treated as an outcome variable) [100]. It is perhaps unsurprising, although somewhat alarming, that elderly users are absent from all these studies. Our work takes a first step, providing a broad overview of antecedents for usage intentions among the elderly who are currently DHP users. We also identify differences in postadoption patterns among the elderly, and between the elderly and younger users. Fig. 2 summarizes the key differences between the use of DHPs among the elderly and younger primary care patients, when the "patient journey" is used as an organizing device to identify differences in usage.
The mixed-method research design was particularly valuable because prior studies have mostly applied qualitative approaches [2], [17]. Our results highlight nuances and the meaning of key antecedent factors that affect usage among the elderly: perceived usefulness and ease of use, digital maturity, and what we refer to as "the need for social belonging." Our findings confirm several aspects identified in the prior literature on technology adoption, whilst extending understanding by unpackaging how they play out in the context of DHP usage for the elderly.

A. Theoretical Contributions
To the best of our knowledge, whilst previous research has investigated the acceptance of ICTs by the elderly in the healthcare sector [1], [2], [23], few have paid attention to the stage after initial adoption [31], [36]. Furthermore, no studies have investigated usage by the elderly and the underlying reasons for their behavior in the context of DHPs. In addition, prior studies have often failed to distinguish between the concepts of adoption and usage [23]. Our study responds to these gaps in the literature by attempting to unpack usage patterns in a DHP and the antecedents to using DHPs by the elderly. Our findings both confirm, disconfirm, and add new insights to previous established assumptions about the elderly and ICT use.
First, while there are no empirical studies on the elderly and DHPs, research on the elderly and on ICT in general (including the TAM model) suggests that we could expect chronological age (e.g., [13], [78]) (demographics in TAM) to influence postadoption usage patterns of the DHP. Our study (RQ1) partly confirms these general expectations. There were differences between younger and older users. We also augment this research by empirically illustrating the different stages of the patient journey and how these differences materialize. For instance, as regards dropout, Olphert and Damodaran [11] stated that the elderly and other vulnerable groups tend to drop out to a greater extent than others. Moreover, Chou et al. [38] stress that problems can be encountered when signing in during first use or engagement. Here, we contribute by clarifying where in the process the elderly struggle in their use of DHPs and by offering some indication of the potential underlying reason for their behavior. For instance, the elderly experience challenges related to coloring and letter size.
We identify several aspects where younger and elderly users differ. Our findings add to the literature by illustrating that the elderly form a group with many within group variations. In this respect, we deepen knowledge on whether chronological age is a factor affecting usage-as discussed in Lagacé [95]-by highlighting the differences between patients aged 60 to 74 and 75 plus. This represents a challenge to prior literature [1], [2], [37], which has tended to view the elderly as a homogenous group.
Our findings on the antecedents of DHP usage by existing elderly users (RQ2) give added weight to the literature. Older patients perceived the platform as useful, particularly in regard to the advantages of availability. Our results confirm that not all services must be perceived as inherently more advantageous in order to enhance usage. For example, Martínez-Caro [94] have argued that the digital service must be more advantageous than physical services to enhance usage. Hence, there is no single way to look at perceived usefulness. However, our findings reveal challenges in usage that are evident in those who drop out of the process, with older individuals exhibiting a higher dropout rate across most stages. This increased dropout rate may be linked to the challenges presented by cognitive ageing, where increased task complexity (the steps of the DHP journey) acts as an added hinderance. Second, our results uncover challenges for older individuals in using platforms that are pertinent to and heightened by biological ageing. First, patient problems with coloring and letter size reflect the problems of sensory impairment (e.g., [10], [11]) and inhibit use of the platform to seek care. Second, reported concerns over the scope of DHPs are pertinent to biological ageing, where increasingly complex health problems and differing needs (e.g., [19]) that are more common in older individuals can result in a lack of confidence that their health issues can be successfully addressed. This results in a preference for face-to-face care. Our findings uncover relationships between DHP-use factors and biological and cognitive ageing (e.g., [74], [101]). In doing so, we help to explain why use is lower amongst elderly individuals.
Finally, security and confidentially were, surprisingly, not experienced as a major factor influencing usage. The elderly seemed more prone to receive help than to worry over storage of personal data, compared to younger individuals. The implication for the literature is that security and confidentiality do not appear to have a direct impact on usage in the context of DHPs. Hence, these factors may be specific to the adoption phase rather than the usage phase. This adds nuance to previous research by identifying security and confidentiality as key antecedents of adoption [1], [26], [73].
On a more general level, our findings confirm that the two variables in TAM, perceived usefulness and ease of use, have a direct impact on the motivation of elderly individuals to use DHPs [1]. However, demography, technology anxiety, and security and confidentiality did not appear to affect usage as proposed by the literature [2], [11], [73]. This testifies to the need to reconsider and potentially adjust TAM for future DHP studies. We highlight some additional factors that may impact usage (i.e., digital maturity, trust, and the need for social belonging), and we suggest that they represent a direction that could usefully be explored. Our findings are broadly in line with some previously identified factors of acceptance in telehealth [17], [23] and thus confirm their applicability to DHP usage.

B. Practical Contributions
DHPs have the potential to revolutionize the healthcare sector by increasing accessibility, flexibility, and efficiency. Our practical contributions are pertinent to platform owners (in both technology-related and business-related roles) and healthcare professionals, providing direction on how they could facilitate higher quality usage among the elderly.
Overall, our results point to the following conclusion: the elderly take a longer time with onboarding and anamnesis (RQ1). Yet, they are still motivated to use DHPs (RQ2) due to their perceived superior access (versus physical primary care centers) and growing comfort with technologies over recent years. Their experience is that it is fun to learn something new, and they have tended to respond positively to recommendations for DHP usage from their social contacts. Yet, several downsides are associated with usage, including the perceived use-related difficulties tied to interface design issues (information overload, size, colors), and perceived trust issues. DHPs are seen primarily as a solution for milder ailments generally and are not considered suitable for addressing complex healthcare needs. Hence, DHPs are associated with a tradeoff between availability and their perceived reduced ability to address more challenging and multifaceted health problems. Based on these insights, we recommend the following: Platform owners must integrate adjustable text size into their systems and include a color scale that distinguishes between different sections and buttons. This would facilitate platform usage for those with impaired functions. Our findings reveal the elderly are 1) spending more time in the process when seeking care and 2) have a higher dropout rate in most stages. Therefore, platform owners could arrange an elderly test group and observe their usage at various stages to gain a detailed understanding of their struggles, or even engage in value co-creation [102]. Short videos clearly explaining the different steps throughout the patient journey and how the process works would be beneficial and could lower anxieties [103]. In consequence, the digital anxiety of elderly people can be reduced and the digital divide between the generations minimized.
For healthcare professionals who interact with patients through the platform, it is important to remember that not everyone possesses the same digital maturity. Our recommendation is to make sure that patients are fully informed about how the process will unfold and are made aware where they can source both general information and specific details about their appointments. For digital healthcare providers that also run physical care centers, opportunities exist to explain how the platform works when the patient is at the physical location, given the positive effect of social influence. This is a new phenomenon that is clearly not well understood. To this juncture, DHPs have generally marketed their services as providing access benefits and support for milder ailments-a message that seems to have been accepted and found credible by elderly users. To expand beyond this niche, DHPs need to invest in establishing trust in their capacity to deliver more complex services, communicate how they will deliver these, and what demands will be placed on the user. Collaborating with social actors in the elderly's everyday surroundings could provide a viable path forward. For example, retirement organizations could play a constructive role, involving elderly users in service development and ongoing refinement. While the elderly are not the typical consumers involved in user collaborations, this approach is possible provided that the collaborative process is adjusted to their specific prerequisites [25].
Our study has taken initial steps toward understanding what motivates care consumers to use DHPs. Elderly users use DHPs primarily for reasons of accessibility. Maintaining this advantage in relation to other competing care services is important if existing users are to be retained. Indeed, the promise of DHPs in general is their ability to connect many suppliers with many consumers remotely, thus increasing access and availability. Our findings assist with efforts to improve DHP usage and ease of use. Yet, their ultimate role may remain limited while perceptions persist that the service is essentially of limited scope and incapable of addressing more complex problems.
To increase loyalty and gain market share among consumers with more complex care needs, additional advantages in other quality dimensions will be required, such as more specialist competence, better pre-execution and postexecution services, and continuity. Future research could employ more fine-grained service evaluation models to explore what other advantages platforms might offer to users, including those with more complex needs.
Finally, despite the potential advantages of DHPs, they also present increased cyber risks to healthcare systems, which organizations are poorly equipped to address [104]. These risks apply to, and thus require action from, health care professionals, platform providers, and patients, given the ethical concerns they present and their impact on future policy [105]. For example, cyber attacks are likely to elevate trust and security concerns in DHP use. As DHPs continue to spread, further research is required to address these aspects, particularly from a user and platform provider perspective. On a broader level, there may also be additional risks that place constraints on the intention to use DHP services [106]. This is a significant area that warrants both scholarly and practical attention to provide a basis for devising policy options. Research on platforms suggests that, in general, they typically begin small and then move on to adjacent niches. This is often difficult in healthcare, not least because of the trust issue and the existing regulatory and reimbursement systems that envelop DHPs in their efforts to offer viable public services.

C. Future Research and Limitations
Our study paves the way for future research on the elderly and DHPs. To expand our insights, the elderly could be clustered into other strata for analytical purposes. Indeed, our findings point to a range of more relevant and insightful strata than simply "the elderly," who are currently defined only in chronological terms. Qualitative research is required on the perception of elderly subgroups-for example, with different digital maturity, primary trust bases (perception of the trustworthiness of online/offline care/combinations), other healthcare needs, and different living conditions, education, and income levels. Critically, studies should explore subgroups based on biological, cognitive, and social ageing. For example, whilst the elderly respondents in our study discussed their own use of the platform, future studies should examine both nonusers and users who are reliant on help from others when using a DHP (including retirement home helpers and younger relatives). With respect to social influences, despite their potentially positive effect, elderly individuals are often more isolated. Research is required to understand this influence and how it may be harnessed.
Third, this study is limited to the Swedish primary healthcare context, which is characterized by relatively digitized primary care in terms of the diffusion of electronic medical records, and a high level of Internet and mobile phone use in the population. Sweden also has a history of public care provision. Whilst the number of private actors entering the sector has grown, there is an ongoing debate about this privatization. This most likely has contributed to the politicization of DHPs and, in turn, may have affected the results. Overall, there is clearly great intercountry heterogeneity when it comes to the geography, funding, operations, and organization of health care, and institutional arrangements (laws, rules, norms) [107]. As DHPs spread internationally, we call for further research to investigate differences between nations and to identify other factors that influence usage.

VI. CONCLUSION
This article provides new insights into DHP usage amongst elderly patients. First, we enhance understanding by unpacking the differences in usage between elderly and younger primary care patients. Second, we reveal the key antecedent factors that influence usage, coupled with their meaning as perceived by the elderly. The analysis reveals significant differences between the elderly and younger patients along the different stages of the DHP patient journey. These insights are particularly important in the light of an aging population internationally and the productivity crisis in many national health care systems, where an increased use of DHPs can be part of the solution. The relevance of our study is also underscored by the Covid-19 pandemic. Clearly, DHPs are helpful in circumstances where social distancing is mandatory, being able to usefully cater to the medical needs of special risk groups. Yet, as we show, several areas of improvement need to be further explored and continuously attended to if DHPs are to realize their potential in the elderly care setting.
Did you complete the entire care process, or did you choose to finish before your case was completed? Do you have the knowledge required to use digital care, for example care apps?
r Do you understand icons, where to press, or similar?
Did you think that the symbols were the right size and that it was easy to fill in information? Do you have any ideas or suggestions for improvement on how the platform could be designed to better meet your needs? That is, do you see any potential for improvement in digital healthcare platforms? Is it likely that you will seek care digitally again?