Exploring the nexus of Social Media Networks and Instant Messengers in Collaborative Type 2 Diabetes care: A Case Study of Port Harcourt, Nigeria

The widespread use of smartphones in Nigeria has resulted in the adoption of social media (SM) and instant messengers (IM) for communication and various creative purposes. These digital platforms have undeniably transformed communication dynamics. However, their effectiveness in the healthcare sector, particularly in managing chronic conditions, hinges upon a complex interplay of multiple factors. Using Port Harcourt (PH), Nigeria, as our case study, we examine their use for Type 2 diabetes (T2D) care. Employing an approach encompassing qualitative and quantitative data collection, we engaged 110 people with T2D (PWT2D), supplemented by interviews with 51 participants, including PWT2D, caregivers, and community pharmacists. The ensuing abductive thematic analysis generated two overarching themes, shedding light on the prevalent use of SM and IMs in Nigeria’s T2D care landscape. This study furnishes actionable insights into improving SM/IMs for managing T2D and other chronic conditions, while advocating for their prudent use in global southern contexts.


INTRODUCTION
New interventions including digital approaches to chronic disease care are emerging [19,38,44].However, in countries like Nigeria, their slow adoption persists due to several factors such as low digital literacy [5,15,50].Nigeria's Information and Communication Technology (ICT) market has improved, driven by an 82% telecommunication subscription rate and 29% internet usage [42].Yet, compared to global northern nations, challenges remain.Fortunately, smartphone technologies such as SM and IMs, show a possible bright spot for accessibility compared to other technologies [32,41,43].While a significant body of literature concerning the penetration and integration of technology within the Nigerian context exists [10,29,48], along with investigations into the patterns of SM/IM use within the same context [3,11,32,47], it is noteworthy to observe a gap in understanding pertaining to how they impact T2D care.T2D, which is characterized by hyperglycemia and its potential systemic effects [28], tends to exhibit a gradual onset, often diagnosed after significant complications [7].Its global prevalence has witnessed a substantial surge, with estimations projecting an escalation from 360 million cases in 2011 to potentially 552 million by 2030 [1,17].In the African context, over 24 million adults struggle with diabetes [1].Nigeria faces a disconcerting scenario as projections indicate a twofold increase in prevalence by 2045, surging from 3.6% in 2023 to a staggering 7.9%.Notably, Nigeria stands as one of the most impacted nations in sub-Saharan Africa, with a sizable portion of the affected population experiencing a poor quality of life and reduced longevity [9,21,25].
This study attempts to explore T2D against the backdrop of online health-seeking behaviors (HSB), and the assimilation of SM and IMs, shedding light on the current state of collaborative T2D care within the Nigerian context.Using PH as a case study, we adopt a decolonial Afro-centred approach to unravel the coordination of diabetes care within informal and community contexts.Two overarching themes emerge from our analysis, encompassing the role of SM in T2D care and the collaborative use of WhatsApp for T2D care.Reflections upon these findings pave the way for actionable digital recommendations tailored for people with T2D (PWT2D) and other stakeholders engaged in the provision of diabetes care.

Data Collection
We engaged PWT2D, community pharmacists (RX), and caregivers to investigate long-term informal/community care for T2D.They had to meet the following criteria: 1) Adults (18+) able to provide informed consent, 2) RX with over 6 months' experience in PH, and 3) Caregivers currently responsible for PWT2D.We obtained 110 valid questionnaire responses from PWT2D and conducted interviews with PWT2D (P1-P23), RX (RX1-RX11), and caregivers (C1-C17) using both digital and non-digital methods.Ethics approval was granted by the University of Port Harcourt, Nigeria.
For the interviews, participants were recruited through snowball sampling and direct recruitment via The Diabetes Care Network (TDCN) -an expanding community comprising of PWT2D, caregivers, and healthcare professionals.Semi-structured interview questions were validated through a pilot study and expert review.Afterward, 36 synchronous chat-based WhatsApp interviews and 15 in-person interviews were conducted with PWT2D, caregivers, and RXs.Using Kish's formula, which considers factors like Nigeria's T2D prevalence and potential attrition [27], we calculated an optimal questionnaire sample size of at least 59 participants for the study.Thereafter, we automated WhatsApp for questionnaire distribution using Twilio (a cloud communications platform), targeting PWT2D who expressed interest via an advertised Microsoft form.This form screened and eliminated ineligible participants, resulting in 27 valid responses.Thereafter, respondents completed the questionnaires within the WhatsApp environment.Finally, questionnaires were shared physically in PH, targeting only PWT2D and yielding 83 valid responses.

Data Analysis
For quantitative data, questionnaires were manually entered into Microsoft Excel, cleaned, and analyzed in JASP (version 0.16.4), using both descriptive and inferential statistics.However, qualitative data from WhatsApp interviews yielded different media formats and transcripts which were compressed into a zip folder.Transcripts were then converted to a compatible format and imported into NVivo 11 for annotation and further analysis.Simultaneously, verbatim transcriptions of in-person interviews were uploaded to NVivo.Consequently, two distinct qualitative datasets emerged, each subject to separate thematic analysis using a decolonial abductive approach [45].
We started with an inductive process inspired by Braun and Clarke [14].Subsequently, we conducted five coding iterations to discern trends related to the scope of T2D care, current use of T2D technologies, and potential digital needs for managing the condition.In the fourth coding iteration (which was relevant for this discourse), we identified patterns pertaining to socio-cultural influences on digital health intervention (including SM and IM) use.This heuristic process deepened our comprehension of observed data patterns [36].In this study, a codebook facilitated the primary researcher's code selection process, allowing for deduction, reflection, and relevance assessment.Themes emerged from the interrelation of codes, guided by the researcher's reflexivity [8].These resultant themes were then compared across qualitative datasets, yielding two overarching themes.Subsequently, integration with questionnaire data occurred, following Proudfoot's model [34].A significant balance between inductive and deductive logical frameworks, formed the basis of abduction in this work.

FINDINGS 3.1 Poor collaborative use of WhatsApp for T2D care
Questionnaire data indicated WhatsApp as the most used IM (77.27%), with a focus on T2D care remaining unclear.Respondents described their use of WhatsApp in communicating with pharmacists.Most said that they did not (65%), while those who did preferred texts (53%).Additional insights include: 1) PWT2D were more inclined to visiting community pharmacies or drug stores nearest to them.Thus, their proximity to RX did not make them consider having a virtual relationship.They were more likely to have the phone numbers of any healthcare provider that could offer emergency services: "I don't have his number, the pharmacy is not far from here, I can always go there.. " -P17 2) Most participants did not see the relevance of a triadic digital relationship between PWT2D, their caregivers and RXs.But most RXs appreciated the concept of collaborative care: "It is a good idea [collaborative care].. but it is tricky, using WhatsApp or just doing things online"-RX3 3) Some PWT2D said that pharmacists did contact them sporadically to enquire about their health and adherence to medicines, however, most of these conversations occurred through mobile phone calls: "..there is this pharmacy where I bought drugs one time..[]..They still call me.I never know when they will call.. " -P19 4) Some RXs said that patients abused the relationships that they attempted to build with them.RX7 complained about some patients, claiming that they commented on her status and one of them had sent her inappropriate messages.
3.2 Appropriation of SM for T2D care 3.2.1 Facebook offers peer support through groups and pages: Questionnaire data showed that 75.4% favored Facebook for T2D care.Interviewees acknowledged Facebook's importance, citing group participation and following relevant pages for valuable information.As expressed by P2, using Facebook was fuelled by overwhelming responses from Google search: "I joined diabetes care on Facebook because I could see real activity and posts.Many things on Google are hard and too many to understand...but having people..

. []" -P2
Two participants (a caregiver and a PWT2D) noted that their involvement in Facebook groups had led them to discover online courses they otherwise would not have found: "..so one of these groups, someone posted a link, it was a short free course offered by a UK institution..[]" -C4 Most participants did not disclose whether they took online courses, however, they mostly used Facebook the same way.They received information from their existing networks, targeted ads and Facebook groups and/or communities: "I joined eh the diabetes care people on Facebook and I am on two other groups..You know I am a student....so i even get info from my department[].. " -P8 3.2.2Pharmacists' Passive Facebook Engagement: Most RXs said that they used platforms such as Facebook, Instagram, Twitter, and YouTube.With regards to providing diabetes care through these platforms, a significant percentage of them were mostly observers : "I have two jobs so If I am on Facebook it is to just see what other people are doing.Once in a while I used to post about health things, not even just diabetes... " -RX2 Most RXs claimed to passively participate in Facebook groups like TDCN, occasionally engaging with likes and shares.RX4 believed that she made positive impact with her infrequent shares, while RX9 refrained from doing so due to concerns about misinformation: "There are some popular pages on Facebook.[]...So sometimes I respond to whatever they say, but I do not bother much.I just like to share... Facebook is dead" -RX9 When asked about commenting, RX6 admitted that she only commented when outrageous claims were made or there were lingering unanswered questions.A few of these pharmacists implied that they did not see their engagement on Facebook as impactful because they were not "influencers".They preferred to share personal photographs and achievements online and relegate pharmaceutical care to work hours.

3.2.3
Online Health influencing by some Community Pharmacists: RXs claimed that whenever they decided to actively engage in any online activity, they were intentional about what they shared and/or posted and will usually either tag their sources or clarify that they were specialists in the field.However, only RX13 and RX4 described themselves as "influencers".RX13 claimed that she made a Facebook post every weekend and that her consistency over the past few months facilitated her employment in a better establishment.This sentiment was shared with RX4 who attempted to make a post most days of the week: "I took it upon myself to always post.Not just on Facebook..[]the only thing I have not done is TikTok" -RX4 3.2.4Lax caregiver behavior on social media use for T2D care: Caregivers on the other hand had a different approach to SM.Most caregivers admitted to being active on most SM, however, they rarely thought about using these platforms to assist in the management of any disease condition: "I am on Instagram, Facebook..

.[]I am on many things [laughs] but I am on it for myself []..I do not go online because of my mother's condition, if I see anything related to it ofcourse I will read and share" -C11
A few of them disclosed that they shared diabetes related information on Facebook, some of which were devastating stories or posts that triggered fear.This was not done with the intention of spreading fear of falsehood, but the stories were extremely captivating that they thought it fit for someone else to see it: "Let me not lie [laughs] the last post I shared was very sad.I was scared o...[]...I think I learned one or two from it, but eh I wanted others to see it [laughs]" -C6 Some caregivers claimed that they preferred YouTube as a source of information because they could watch a variety of videos, some of which were easy to replicate.However, they only resorted to using YouTube when under pressure or due to sporadic curiousity: "..only when they ask me to do something or I don't have time to read plenty talk.. " -C4 3.2.5 Learning on Facebook: Participants acknowledged staying on Facebook because it served as an extensive contact list and a source of diverse information beyond healthcare.Caregivers reported encountering informative posts about T2D, often through group memberships or their feeds.Furthermore, PWT2D actively sought information due to their condition.Insights obtained from interviewing PWT2D include: (1) They rarely thought of checking SM for information regarding their disease: "My dear it never crosses my mind...[]." -P2 (2) When it came to seeking information regarding their condition or associated complications, SM ranked poorly compared to calling a physician, family nurse or caregiver, or visiting a nearby community pharmacy.It was implied to be because of the impracticability of having prompt answers to follow up questions, and comments from fellow users could not be trusted: "If I have a problem..[] I don't know why I should think of eh Facebook.. " -P19

DISCUSSION
This study describes individuals' overarching attitudes toward SM and IM use for T2D care [18,30].Notably, it uncovers the malleability of these attitudes contingent upon stakeholders' needs and age.E.g., most PWT2D fell within 45 and 55 years, belonging to Generation X.Generation Xers, as research indicates, tend to employ SM for educational purposes [16,20,31].Contrastingly, Baby Boomers (born 1946-1964) exhibited greater resistance to leveraging SM for any purpose, particularly in healthcare [4,23,26].We observed that this population preferred in-person consultation with physicians, Google search, and were particularly worried about potential "indoctrination".This was linked to COVID-19-related misinformation, substantiated by recent literature [13,35,46].Conversely, most RXs endorsed digital interventions for T2D care and held favorable views on their potential in collaborative care.They also highlighted the value of SM and IM in disseminating information, particularly in remote areas, as described by Benetoli et.al [12] and other Nigerian academics [6,39,40].Additionally, most caregivers (mostly cis-gendered women [2,33,49]), used WhatsApp for parental communication, but less with RX.Although WhatsApp eased communication barriers, collaborative care was not a priority, echoing Hermansyah et al.'s findings [24], where similar issues hindered caregiver-pharmacist communication.

Design implications for SM/IMs in HCI
Findings from our study provide several design implications that could be relevant in existing SMs/IMs or upcoming platforms targeted at managing chronic conditions in similar global southern contexts.
(1) Optimizing search engines: Many participants found Google searches challenging, overwhelming, and confusing.Consequently, some combined Google with SM for better results.
We recommend prioritizing culturally sensitive search results, diversifying medical information websites for accessibility, integrating virtual health assistants conversant with contextual socio-cultural factors, and developing contextsensitive platforms seamlessly integrated with SM. (2) Strengthen digital health policies: Inculcation of optimal online HSBs and digital collaborative care can be achieved by enforced digital health policies.Within existing ethical constructs, these policies should encourage the use of SMs/IMs and other digital platforms by healthcare providers.People with chronic conditions will imbibe similar habits if they are taught about endless possibilities of SMs/IMs/other digital heath platforms by pharmacists and other healthcare providers that they are in contact with and possibly trust.(3) Optimizing WhatsApp for patient care: As dynamic as WhatsApp is, there are concerns regarding security and trust, specifically within the auspices of healthcare [22,37].Amid evolving digital health policies, Meta can develop a health specific version of WhatsApp.This version could prioritize patient authentication with unique identifiers and integrate seamlessly with other medical platforms.In an optimal SM/IM environment, we advocate for healthcare practitioners to be able to establish work hours and accessibility parameters.This could be facilitated by the addition of message rejection or redirection to individuals who disregard boundaries, instead of delivering offline messages whenever the recipient [healthcare practitioner] is online.This feature not only informs healthcare providers of such infractions but also shields them from a plethora of unread messages, concurrently endowing them with the opportunity to prioritise their conversations.

CONCLUSION AND FUTURE WORK
In this study, we identified factors influencing the adoption of SM and IM for T2D care.Facebook emerged as the preferred SM for healthcare, with WhatsApp as the dominant IM.Factors like age, income, and family values significantly shaped Facebook and What-sApp use.PWT2D aged 45+ preferred physical contact, while RXs supported leveraging SM and IM for pharmaceutical care.Most PWT2D favored traditional healthcare over virtual collaboration on Facebook or WhatsApp.Caregivers and PWT2D relied on physicians for T2D care and preferred platforms offering physicianrelated services.While refraining from asserting overarching conclusions, we advocate for the development of adaptable, healthfriendly versions of the aforementioned platforms, accompanied by contextually attuned medical information frameworks.These design implications could be relevant in similar global southern contexts.Moreover, we propose research on alternate SM platforms like TikTok etc. to identify prospects for enhanced chronic disease care.