Delivering a psychosocial program for older people living in retirement homes during the Covid-19 pandemic: A process evaluation and recommendations for community interventions.

The Covid-19 pandemic lockdown regulations caused retirement homes to temporarily ban in-person visitation potentially increasing the mental health risks of older people. An opportunity arose for a multistakeholder community collaboration to design a mental health program for older people. To evaluate the process of delivering a 12-week psychosocial program aimed at preventing loneliness, countering boredom, and providing older people in restricted settings with education about Covid-19 during the lockdown, in Durban, South Africa. A qualitative retrospective design was used. Data from two focus groups and six semistructured individual interviews conducted with stakeholders (volunteers, social workers, and residents) postproject were analyzed using reflexive thematic analysis. Stakeholders had varied experiences of the project, in terms of content, processes of engagement, and implementation, resulting in five themes. The study concluded with recommendations. A strong need exists for multistakeholder community collaborations when implementing a program where the context restricts physical access.

INTRODUCTION

Globally, to curtail the spread of the novel coronavirus disease (Covid‐19), countries responded with relevant legislation, guided by the World Health Organization (WHO). South Africa swiftly responded by declaring a State of Disaster on March 15, 2020, only 10 days after the first confirmed case, followed by a nationwide hard lockdown from March 26, 2020 (Abdool Karim, 2020). As the national decision‐makers managing the pandemic response, the South African National Command Council implemented physical distancing (Abdool Karim, 2020). Despite the lockdown's backdrop of beneficence, the pandemic exacerbated psychological distress and symptoms of mental illness (Holmes, 2020; Nguse & Wassenaar, 2021), such as anxiety and depression (Rajkumar, 2020) due to feelings of social disconnection (Santini et al., 2020).

Older people (60+ years) were especially vulnerable to contracting Covid‐19 because of advanced age and medical comorbidities (Zhou et al., 2020). Many older people were confined to their homes or residences with no visitors and limited in‐person interaction (Heid et al., 2021), contributing to loneliness (Tomaz et al., 2021). A further risk factor for loneliness was anticipated among older people living in retirement homes, who generally report higher levels of loneliness (Drageset et al., 2011). As a well‐studied phenomenon, loneliness is documented to increase the mortality and morbidity of older people and contribute to the risk of dementia (Freedman & Nicolle, 2020). In an attempt to mitigate loneliness, numerous interventions have been implemented with varying degrees of efficacy (Jarvis et al., 2020).

Further, following consideration of bridging the digital divide, mental health interventions hold the potential as a preventative measure against loneliness, especially in the absence of face‐to‐face or in‐person contact (Jarvis et al., 2020). Online social connections proved valuable protective measures against loneliness in the pandemic for some population groups (Tomaz et al., 2021). However, during lockdown in low‐ and middle‐income countries (LMICs), online interventions are not easy to implement due to a range of pre‐existing social conditions such as poverty, unemployment, lack of access to the internet, smartphones, and/or computers, including an inconsistent supply of electricity (Nguse & Wassenaar, 2021; Pillay & Barnes, 2020). In South Africa, the pandemic added to the burden of an overstretched public mental health system struggling to attend to historical and contemporary distress, such as trauma and violence (Naidu, 2020).

Against this background, an opportunity arose to help meet some of the mental health challenges by designing a program for older people living in retirement homes in Durban, South Africa. Mindful of the potential negative mental health effects of enforced social isolation on older people living in retirement homes, a collaboration was developed by three organizations to collectively co‐design a program to help alleviate loneliness, boredom, and decrease anxiety among older people living in such settings.

A nonprofit organization dedicated to serving older people in Durban, South Africa (NPO‐1), engaged with a fellow national nonprofit organization committed to mental health support (NPO‐2), for assistance in offering psychosocial support to their residents. The residents were expected to experience bouts of loneliness, anxiety and depression during the initial few weeks of the national lockdown. Given the uniqueness of the situation and the lack of solid evidence to provide clear direction for the delivery of such a program, NPO‐2 further collaborated with two nursing researchers from the University of KwaZulu‐Natal, based on their experience with older populations and loneliness interventions. This resulted in Project C.A.R.E. (Caring for the Aged in Retirement Environments).

The article is not an evaluation of the program per se but a retrospective evaluation of key stakeholders' perceptions of the process of delivering the program. The primary research question for each group of stakeholders (volunteers, social workers, older people) did not focus on the program's effectiveness in reducing boredom, loneliness, or anxiety per se, but instead asked, “What were your experiences of the C.A.R.E. project?”

Donabedian (2002) draws attention to the important contribution of process components on the quality of a project. A process evaluation was considered useful because the rich dynamics of collaborating on community projects often goes unreported in the academic quest to determine the effectiveness of outcomes. While effectiveness is important, the dynamics involved in designing and implementing interventions reveals equally interesting lessons for community psychology (Dadwal et al., 2017). Interpersonal and institutional processes can enable or hinder community collaborations as they intersect with diverse and unanticipated events, changes in plans, or the evolving needs and choices of multiple stakeholders and their varied expectations (Margolis & Runyan, 1998; Marston et al., 2020). In the urgent context of an unprecedented pandemic, and given the numerous obstacles we encountered in rolling out this intervention, a process evaluation was considered beneficial. It was believed a useful way to better understand the ethics and politics of well‐intended rapid interventions, but perhaps not always capable of being collaborative and participatory in the truest sense due to time constraints and a fast‐tracked design and rollout process. We hoped that a process evaluation would allow us to share lessons and recommendations for delivering psychosocial programs in restricted community contexts, such as future lockdowns or pandemics.

C.A.R.E. project background and content

Project C.A.R.E. occurred within global and national uncertainty about evolving pandemic‐related laws, regulations, and public health directives. This rapidly changing context required design adaptations from the initial conceptualization; some of which was related to the pandemic, such as remaining alert to changing government regulations about indoor gatherings of people, while other adaptations were unrelated to Covid‐19. For example, our original idea was to implement a WhatsApp intervention, but this ran into problems with access to mobile phones and data availability. The intervention itself was not conceptualized as a research study from the outset and was rolled out “on the ground” as a community‐based collaboration; hence the process evaluation only began once the program was complete.

Project C.A.R.E. became a 12‐week psychosocial program for older people (60+ years), living across different retirement homes in Durban, managed by NPO‐1. It began as part of a three‐way collaboration between two NPOs and the university. Project conceptualization and design was carried out primarily by volunteers from NPO‐2, working online and drawing from research evidence on alleviating loneliness. This planning stage took place during a “hard” lockdown; meaning that direct contact with the residents (older people) was not possible and by necessity, mediated by the social workers from NPO‐1, who worked in the retirement homes. Conducting an intervention without direct access to residents and in a context where many residents did not have easy access to mobile phones and data caused collaboration, design, and implementation delays. Volunteers had to find a creative and paper‐based form of communication to support residents.

Consequently, volunteers delivered a “Brown Paper Bag” to residents each week, containing Covid‐19 information sheets and activity packs with Covid‐19 information, cognitive stimulation, emotional inspiration, and psychological support. Examples of content included “Wear a mask and wear it correctly” (Week 1), “Physical distancing—not social distancing” (Week 4), “Healthy eating = healthy immune system” (Week 9), and “Stay hopeful” (Week 12) (Table 1). Activities were varied and randomly assigned to the different weeks (Table 1). The paper‐based program was designed around the choice for it to be done alone (through personal preference or if in quarantine) or in a group with other residents. There is inconsistent evidence for addressing loneliness in older people through groups or individually (Jarvis et al., 2019); however, in the context of the pandemic, synonymous with physical distancing, the group option was seen as able to counter loneliness, while individually the program could intercept the risks of boredom (Brodeur et al., 2020) which is associated with loneliness. The fear of contracting Covid‐19 is strongly correlated with anxiety and depression (Ahorsu et al., 2020), necessitating verified Covid‐19 information, with the additional benefit of it being conversation topics among the residents.

Table 1

Contents of the weekly “Brown Paper Bag” delivered to residents.

Week	Covid‐19 information	Cognitive activity
1	Wearing a mask correctly	Breathing exercises
2	Hand washing	Brain teasers
3	Using sanitizers	Word search
4	Social distancing; safe & effective use of disinfectant	Learning sign language (alphabet)
5	Busting Covid‐19 myths	Learning sign language (alphabet) contd.
6	Signs and symptoms of Covid‐19	Crossword puzzle
7	Keeping emotionally positive during the pandemic	Mindfulness and gratitude
8	Keeping social connections during the pandemic	Sudoku puzzle
9	Building nutritional defenses and importance of healthy eating	Coloring therapy
10	Exercise	Learning sign language (numbers)
11	Importance of good sleep	Simple yoga
12	Staying hopeful	Coloring therapy

All 38 residents who participated in the program were older people (m = 72.5, age‐range: 63–86 years), primarily women, from diverse cultural and ethnic backgrounds, and were living in one of five NPO‐1 retirement homes in Durban (two additional homes opted out of the intervention). A designated social worker was tasked with rolling out the program at each site and was a weekly point of contact for the volunteers from NPO‐2 who telephonically checked in once a week on the delivery of the program. The social workers were the liaison between the volunteers and the end‐users, that is, residents in each home who accepted the social workers' invitation to participate. Participation was voluntary. The social workers determined the date to commence, and this varied widely, with the first group starting in July 2020 and the last group starting in October 2020, using a condensed version. Therefore, the completion dates varied widely, with the last group completing at the end of November 2020.

METHODS

Study design

This was a retrospective, qualitative process evaluation of Project C.A.R.E. and an exploration of key stakeholders' related perceptions of the project. Retrospective process evaluations are used when researchers assess interventions after they have taken place (Patton, 2021). As a community project that did not commence as a research study, the lack of baseline measures necessitated a postproject evaluation. A qualitative approach offered the opportunity to understand the nuances of contextual issues involved in implementing the program as perceived by different stakeholders (Hammarberg et al., 2016). There have been increasing calls in gerontology for more qualitative research in environments of aging (Phoenix, 2018). Frequently, intervention studies measure the outcome in quantitative terms such as proving hypotheses or outcome effect size; however, lesser described are the processes involved to arrive at the outcomes; hence the significance of a process evaluation. Outcomes alone cannot measure a program's quality (Donabedian, 2002). Of course, evaluating processes would unearth data about outcomes, even though this was not the primary reason for conducting the evaluation.

Sampling strategy and sample

Convenience and purposive sampling yielded information‐rich data from three categories of stakeholders (residents, social workers, and volunteers).

To ensure diversity of perspectives from different residences, the social workers (as gatekeepers to the residents), invited everyone to participate. We ensured that the sample contained at least one resident from each residence. Ultimately, our invitation was accepted by a convenience sample of older people living in NPO‐1 and directly exposed to the program (n = 6), including one male resident (aged 73) and five women (aged 63–73).

All the social workers from NPO‐1 who implemented the intervention and worked in the retirement homes, or wanted to give input on the program, opted to participate (n = 8). These were eight women who were all registered social workers.

All the project developers from NPO‐2 and the university involved in designing, supporting, and overseeing the program (n = 7) opted to participate. These were seven women volunteers who ranged from postgraduate psychology students, an NPO manager, and senior academics in health sciences.

Data collection

In keeping with critical approaches to evaluating community collaborations, the voices of all stakeholders are key to understanding real‐world implementation outside the “ideal” conditions of a pre‐designed research study (Marston et al., 2020). Data were collected from various sources and triangulated to increase the depth of information and gain multiple perspectives (Patton, 2021). Two data collection methods were used: individual interviews and focus groups.

The first data set included semi‐structured, individual interviews with six residents from NPO‐1. The individual interviews with residents were conducted by the volunteers involved in the design, implementation, and evaluation of this project. They had not established a relationship with the residents before data collection, which decreased bias. These interviews were conducted telephonically or through a virtual online platform (Zoom) and ranged from 30 to 60 minutes, at no financial cost to the participant. Interviews were guided by questions about the participants' personal experiences in the program and whether they found it beneficial; and if not, what were the challenges they experienced in its implementation, and how could the project be improved for the future. The interviews were conducted four months after the program was completed. Data saturation for themes was achieved after analyzing the fifth interview and confirmed through the sixth interview (Hennink et al., 2019). The interviewers' qualifications included Bachelor's and Master's degrees in Social Sciences and PhDs in Healthcare Sciences, and they were all employed in education or the healthcare sector.

The second data set included firstly a 90 minute semi‐structured, online focus group conducted by the second author (S. R. P.) one month after project completion, with seven volunteers involved in designing and supporting the project (from NPO‐2). Second, S. R. P. conducted a 60 minute in‐person focus group with eight social workers from NPO‐1 5 months after program completion. This clinical psychologist was not involved in the design or delivery of the project and provided an “outsider” perspective to the evaluation of the delivery process. Data collection in the focus groups continued until no new information emerged and therefore continued in both instances for an extended period (Guest et al., 2017). All data were collected in English, as the universal language in the setting. Data were audio‐recorded and transcribed for analysis. No participants received financial incentives.

Data analysis

We used a reflexive thematic analysis approach (Braun & Clarke, 2021). This was an active and iterative process that required multiple readings of the data to generate codes and themes to understand patterns of shared meanings. As Braun and Clarke (2021, p. 594) note, “Themes are creative and interpretive stories about the data, produced at the intersection of the researcher's theoretical assumptions, their analytic resources and skill, and the data themselves.”

The authors and two student volunteers (n = 7) split into two teams. Each team separately generated an initial set of codes, themes, and points for discussion and then came together in several joint research meetings to compare notes. Joint meetings involved reconciling differences of opinion or perspective; generating appropriate phrasing for codes and themes; and deciding whether data saturation was being reached (Phoenix, 2018). Data saturation occurred after the fifth individual interview, followed by a sixth interview. The data collection and analysis process happened concurrently so that this could be a reflexive and iterative process that enabled the generation of meaningful themes, subthemes, and a list of recommendations for future similar interventions conducted with older people.

Rigor

During semi‐structured interviews, both people are embedded in a constructivist epistemology in which researcher subjectivity is part of the co‐creation of knowledge along with the participants (Braun & Clarke, 2021). Scientific rigor in the analysis stage was ensured by adhering to credibility, dependability, and transferability (Lincoln & Guba, 1985). Credibility was ensured by having experienced researchers and mental health professionals as part of the data collection and data analysis process (Shenton, 2004). Triangulation of the three sources of data increased dependability (Lincoln & Guba, 1985). Quality check reviews among the authors and a detailed description of the study context and data collection process ensured trustworthiness and contextual transferability of data (Malterud, 2001). A draft manuscript was sent back to the participants for member checking to improve validity. This manuscript adhered as far as possible to the Consolidated Criteria for Reporting Qualitative Research (COREQ) (Tong et al., 2007).

Ethics

All the principal investigators (M. A. J., S. R. P., L. M. N., N. H., and O. B.) completed the online Training and Resources in Research Ethics and Eva certificate in research ethics and ensured that the project was carried out with a high degree of ethical reflexivity. The ethical principles proposed by Emanuel et al. (2004) served as a guide for this study. The affiliated university's Humanities and Social Sciences Research Ethics Committee (HSSREC) provided ethical approval (HSSREC/00002310/2021). The CEO and Board of Directors of NPO‐1 gave gatekeeper permission. A memorandum of understanding was signed between the CEO and the researchers regarding the data. Fair selection occurred through invitations to be interviewed of all eligible residents who participated in C.A.R.E. and the project development and implementation teams, ensuring every person had the opportunity to participate in the evaluation. Voluntary participation was offered to all participants who could decline or opt‐out of the study. The use of an information sheet and a written consent form that was verbally discussed with potential participants ensured informed consent. Anonymity involving the use of pseudonyms protected the identities of all participants (residents, social workers, and volunteers). A favorable risk‐benefit ratio was ensured through the opportunity of all participating end‐users to participate in the individual interviews and all project development and implementation teams to join the focus group discussions. Referrals to professional mental health services were available if any adverse psychological events occurred during data collection.

RESULTS

Five core themes were conceptualized from the qualitative data. Each theme relates to interconnected aspects of the design, implementation, and/or evaluation of the delivery of the C.A.R.E. project and speaks to the varied experiences of the different stakeholders involved. These are adapting to shifting contexts; varying methods of program delivery, moving from reluctance to acceptance; benefiting from active participation; and making false assumptions resulting in uneven power dynamics.

Adapting to shifting contexts

C.A.R.E. was designed and rolled out in a continuously shifting context of the Covid‐19 pandemic in South Africa, creating circumstances beyond anybody's control, including uncertainty and confusion about rapidly changing state laws, institutional rules and regulations, and public health policies.

We all had that Covid information overload […] which is maybe a characteristic of a pandemic that it was a new thing which was confusing (Volunteer 5)

The initial design of an mHealth program was influenced by various shifting and unpredictable contexts, compounded by a lack of funding to fully support the rollout of the project, which was designed in a context of crisis and urgency by and for two nonprofit organizations:

And it seemed to be very impossible to try and quickly get funding to try and give people [cellphone] data despite thinking of ideas… And then we almost had to kind of give up on that […] we ended up having then done quite a big turnaround on the design because a lot of people didn't even have the right cellphones or data (Volunteer 7)

The project development team had no direct physical access to the end‐users (residents) because of the retirement homes' safety regulations based on the Covid‐19 regulations. Consequently, they worked through gatekeepers (social workers and management at the residences), but this triggered an increased sense of frustration and lack of control over quick decision‐making:

I think things could have worked a lot faster, but we just didn't have control over that because we were waiting as we were liaising with so many different people, and we needed so many things that were on hold because we were just waiting on other people (Volunteer 1)

As a result of these delays, not altogether uncommon in multistakeholder community partnerships, the rollout of the program was delayed, despite the urgency with which it was initially conceptualized. This resulted in a disconnect between the content in the C.A.R.E. information packs and the evolving needs of the residents.

Ironically, when we made the packs, it was intended for the lockdown period. And because of Covid, we had included Covid information. But in the feedback from my social worker, she said the residents actually hated receiving the Covid information (Volunteer 3)

Three specific buildings could not start the groups at all. The concept was given to people, the invitation was written, the posters were put in […] people refused to come in […] It was about two or three months since we had given them the information and then the formalised program came in. That's why many of the buildings didn't run with the group (Social Worker 5)

It was ok, but I found it a bit boring because these are the things that we already know. I mean, it is all over. Everybody should know it by now (Resident 5)

Conducting a program during a pandemic had unintended consequences for everyone involved. It was “a strange time” and “a tough time,” as one volunteer pointed out because stakeholders had to juggle multiple competing demands, both professional and personal:

I think it wasn't just a strange time for the people at [NPO‐1] and the social workers, I think it was a tough time for all of us as well. It was a strange kind of thing.… because besides dealing with the Covid issues, you have a project running as well, and how do you incorporate that into your daily schedules and tasks? (Volunteer 2)

A defining emotion of these strange times was fear and anxiety, which impacted the residents' ability to engage in the intervention, as they had to adapt to the changed context, which involved dealing with the fear and psychological impact of contracting Covid‐19, and the grief of losing loved ones to the virus.

To come and sit in groups during Covid time, maybe it was their fear. They couldn't ex‐pose themselves; rather take a pack and read in their flat (Social Worker 4)

They were not interested in coming. It happened that one of the residents got Covid, and she was in isolation. So I had four people, one was in isolation, and I was left with three, and they decided that they don't want to meet anymore and I must give them the packs individually in their own homes (Social Worker 8)

Fear of somebody going to the hospital, there was a lot of added stress, and a lot of older people were stressed because a lot of people passed on, and they could not attend the funeral, even family members (Social Worker 7)

The unpredictability of the pandemic and related regulations laid the foundation and provided a lens through which to view the other themes, because they tend to center on the idea of adapting to shifting contexts in different ways.

Varying methods of program delivery

As a result of the shifting contexts, none of the social workers could deliver the project as originally designed. During both focus groups, it became evident that the rollout had to be tweaked and creatively adapted for shifting contexts, depending on the needs of the residents or the unexpected circumstances arising from Covid‐19. “Covid drama” caused disruptions in the smooth running of the groups.

I ran quite a few groups in the area that I work in […] Unfortunately, there was Covid drama, and there was a situation where we thought one of our residents had Covid. Everything was shut down and fell out of place, and there was a Christmas break. So now it's April, and we need to reformulate groups. That's something I have ongoing, so part of it is positive, and part of it is negative. It just depends on what's in place because some of it worked well, and some things we tweaked (Social Worker 7)

The social workers provided feedback that some residents resisted the formalities of joining the groups by either declining to sign consent forms or simply requesting to do the activity packs individually at their leisure:

My residents were interested in participating in the group, but some didn't want to sign when I gave them the consent form to sign, and a few residents signed the form. When we were meeting, more residents were interested in joining. They didn't want to sit with us, but they wanted to get the packs every week when we are meeting. When I didn't give it to them, it would be a problem. That is how I experienced it (Social Worker 1)

Residents indicated different experiences with the various activities and information in the C.A.R.E. pack. One social worker indicated that the residents with whom she was working enjoyed the activities meant for mental stimulation but found the Covid‐19 related information too “basic”:

I felt like that because my people were not interested at all, and I ended up not having a group. I had to do it individually here. I would visit the person and give them the pack of that week. They came only twice, but I had to carry it over 12 weeks […] They were not interested because the information was basic, and it was something they already knew. The only thing that they liked was the activities. They enjoyed doing the activities individually, not in groups (Social Worker 8)

One reason for the variance in delivery was that some social workers already covered familiar aspects of Covid‐19 safety precautions earlier in the year, so parts of the novel activity packs were found to be more beneficial.

There was a lot of repetition in what we had done initially out of the crisis because of our gut and our training. The activities were something that a lot of social workers enjoyed. [We] just extracted the activities and proceeded with them. We had already learned how to make masks, self‐isolation, and all the precautions by then (Social Worker 5)

Volunteers observed similar experiences of the variance in the program's delivery.

The one [residence] started late, so it was more of a fast‐track project for them, but it worked really well for them (Volunteer 2)

Yeah, [the social worker] lost a couple of people along the way because [residents] said that they preferred the packs to just be delivered to their rooms… [so] she continued to give it to them every week because that was what they requested but she would've liked for them to join the group (Volunteer 4)

These varying delivery methods were not unexpected, and fidelity to the original program was not vital to a project of this nature. Some aspects were seen as valuable, and others were disposable or flexible. Given the competing demands and shifting contexts of implementation due to Covid‐19 and delays in getting this project off the ground, such variance was anticipated, and the social workers and residents showed agency in tweaking the rollout and content to suit their needs.

Moving from reluctance to acceptance

Some stakeholders showed initial responses of reluctance at being part of C.A.R.E. However, these feelings changed, and they began to identify aspects of the project that contributed to their mental well‐being. In this individual interview with one of the residents, she expressed how her feelings of social connection to the group members helped her overcome her initial trepidations:

Interviewer: […] what was your overall experience of the project?

Resident 1: In the beginning, I didn't really want it because I thought, well how can these people know what I'm feeling. Then eventually it grew on me, I started realizing differently.

Interviewer: Okay. How did you realize differently?

Resident 1: Because they all had something in common with me.

Interviewer: You mean all the group members?

Resident 1: I thought I was the only one who was feeling that way. You know I had fear, and then I realised that everyone had fear and it's not only me. And to overcome that was a big hur‐dle hey (Resident 1)

Another resident's reluctance was due to not wanting “self‐pity” and wanting to maintain her sense of independence. She eventually accepted that receiving inspirational messages was “a fantastic thing” that did not undermine her autonomy.

Resident 3: You know I'm a very strange person. I don't like self‐pity, so I'm not for these things. When I see something happening, and I know something is happening and it's dangerous, I get up and do what makes me healthy so that I can carry on. I mean, I have been through strokes, and I'm still walking, and it wasn't an easy job.

Interviewer: So, can you tell us how reading those inspirational messages made you feel?

Resident 3: It was great because it shows you that somebody is passing on information to you to make you notice. I think it's a fantastic thing (Resident 3)

The volunteers also experienced some reluctance, having to make regular time commitments to meet online every week for the project's duration to assess progress. However, the volunteers eventually derived pleasure from regularly meeting together for a common purpose:

[It] has been a long few months of quite trying times, and I've really appreciated how people have come together, nearly every single Monday, until now as well, too, and I think it's been quite a major thing since it's been on Zoom. And it just shows that it is possible to create something on Zoom […] I'm sure we all felt like there was a time to give up but we still kept going (Volunteer 7)

And I was extremely grateful for being a part of the C.A.R.E. project because it not only helped with my own anxiety of feeling unproductive and the whole uncertain future of when I was going back to work, or when I was going back to university, or what I was even going to be doing in the next week. I knew that I'd always have the C.A.R.E. project on a Monday. So I think for me, that was the highlight of my week (Volunteer 3)

…even though we took forever to get it off the ground, I sort of felt I was making some con‐tribution to the people that were vulnerable because of the pandemic (Volunteer 5)

The shifting contexts and variations in the delivery of the project seemed to create feelings of reluctance to participate or continue with C.A.R.E. However, as the extracts demonstrate, feelings of uncertainty, resistance, or indecisiveness made way for a more positive experience for the residents and volunteers.

Benefiting from active participation

Despite initial reluctance and varied implementation methods, most stakeholders expressed a range of positive benefits from participating in C.A.R.E, including reductions in loneliness due to increased social connections, less anxiety, and fear related to Covid‐19, improved mental stimulation, and inspiration, and better self‐efficacy.

Through discussion with the social workers, the volunteers found that residents benefitted from the program in several ways, especially in terms of using the C.A.R.E. pack as a reason to socialize and connect emotionally.

It provided a space for people to speak where they didn't feel judged. They had this in common, and it was something that they could relate to in terms of addressing their concerns and anxieties So they could speak freely. And it provided a sense of connectedness with the residents as some people never spoke to each other as it was such a large number of residents. But in the group, they now got to speak to people who they had never interacted with before (Volunteer 4)

In some cases, the C.A.R.E. groups were also a springboard for personal growth and confidence among residents and a foundation for starting more sustainable and ongoing groups.

There were a lot of voiceless people who came together, and they developed a sense of comradery amongst each other, and it increased their confidence, and they became voiced people (Volunteer 5)

It also brought about a sense of understanding amongst themselves, and based on the C.A.R.E group, they have actually decided to create a spin‐off group, which will continue even after C.A.R.E. has finished (Volunteer 3)

My residents are still continuing, even though we not talking about Covid anymore, but if you have something stressing you, you can talk about it (Social Worker 2)

In some instances, this helped deepen the relationship between the social workers and their residents:

My social worker said that she hadn't connected with the residents like how she is con‐necting with them now …. she said that because she had some guidance with regards to the C.A.R.E. package and because we were always in contact, she felt like she connected with every‐one on their own levels and formed a special kind of bond (Volunteer 1)

The residents themselves corroborated these positive benefits in their individual interviews:

You know it really made me feel good knowing that other people had the same problem as me and it wasn't just me alone and we could speak openly about it (Resident 1)

I actually looked forward to the meeting and that we could get together and chat because otherwise, you go to your room, isolate and sit there by yourself (Resident 2)

It made me feel special that I'm sitting here alone and I'm talking to somebody who knows about something that is so dangerous like this illness and taking her time to explain the whole thing to me. It makes you feel special (Resident 3)

Resident 5: I liked that I could do it in my room at my own leisure and stuff, you know.

Interviewer: Any particular activity that really was the best for you?

Resident 5: I actually preferred the word search. It kept my mind occupied (Resident 5)

The volunteers also expressed a renewed sense of purpose, productivity, and connection at a time when they were also isolated due to lockdown regulations, demonstrating unexpected benefits for them, even though they were not the primary target of the program:

Also, it was good to connect with new people that I had never met before, especially on a platform like Zoom. Usually, you would meet people face‐to‐face and it was good to do it this way—very different—and we have formed friendships now that we wouldn't have had before (Volunteer 1)

I've always wanted to be of some service and be a help to some people. So, it was fulfilling, and I'd like to continue with future programmes and see how that goes. I've learnt quite a bit as well in terms of processes, protocols, volunteer work—what can be done, what we should do, and what we shouldn't do (Volunteer 2)

Most of the expressed benefits involved improved social connections to other people and a realization that one is not entirely alone and that other people may be able to relate or empathize with the experience of loneliness triggered by the pandemic.

Making false assumptions resulting in uneven power dynamics

Despite the above‐mentioned positive benefits, some stakeholders stressed that along the way, several false assumptions hindered the project's success. For example, despite initial requests by the social workers for assistance in designing a mental health promotion project for their residents, they felt that their expertise was ignored and that project developers (volunteers) made the incorrect assumption that the social workers did not have the requisite skills to put together a similar program.

…the assumption that when people plan stuff, they don't consider that there is expertise, a body of professionals in the organization that are more than capable of doing those things […] We could have done exactly what was given to us on our own. I just wondered why we needed somebody from outside to come in and structure this for us when we could have done it. We were busy with a crisis, so we didn't think and we appreciate it for that (Social Worker 5)

Additionally, a pre‐packaged program was not welcomed by everyone, and some residents resisted participating because they felt excluded from the design process:

[My residents] are the kind of people that you sit with and plan together and come up with topics, and they say “yes, I think this one can work for us,” then they will be willing to participate. If you come with something preplanned, they don't participate (Social Worker 4)

Even the foundational assumption was called in question, that is, the residents were going to experience heightened levels of loneliness due to a lack of visitors during the Covid‐19 lockdown period. Some stakeholders expressed that it was necessary to avoid assumptions about the vulnerability of older people as inherently lonely, lacking Covid‐19 knowledge, or anxious because this ignores their resilience.

the information packs assumed that older people are vulnerable, and they are not savvy with technology […] So, the assumption that older people were vulnerable, lonely and that they needed us to pamper them for me came out […] So maybe sometimes the assumption should not be that older people are victims. They are and should be seen as people who can teach us a thing or two (Social Worker 5)

Everybody was anxious, which is kind of normal but when the residents hear anxiety and depression, they don't want to be part of the group (Social Worker 7)

Interviewer: […] Being alone can sometimes be very daunting. Some people find it frightening, and some people find it sad.

Resident 5: Some people can find it very depressing, but not for me because I've had plenty of work that kept me occupied.

Interviewer: Okay, so being alone during Covid was not depressing for you?

Resident 5: Not at all. (Resident 5)

Residents felt further restricted in their functioning by having to implement a project whose information on Covid‐19 was considered elementary (and, due to delays, had become common knowledge to them).

My residents felt like they were bombarded with Covid information. So, when I did the pack, they asked me, “Do you think that I'm in grade one that you have to repeat this over and over?” So those were the things that I was getting. My residents are very vocal, and they are better than me with a phone. So, they felt like I was making them feel little because I was telling them the same thing, like how do you wash your hands and how to wear a mask appropriately. So, they listened because obviously, I was doing the group for them, but I don't think they took in much (Social Worker 6)

In their urgency to implement the project, some in the project development team were presumptive about the program implementers (social workers) delaying their access to the homes. The program developers presumed that if the social workers had not delayed their access, the project would have had a better outcome.

…. sometimes when I did talk to my social worker, I got the feeling that this was something that was sort of pushed onto them, and they had to make it work. And some of us actually did struggle to get a hold of social workers to the point where it didn't kick off into certain houses (Volunteer 4)

These false assumptions about the needs of the residents, although rooted in a review of the literature and discussions with their social workers as gatekeepers and liaisons “on the ground,” were sometimes misguided and triggered some feelings of resistance to the program. This, in turn, led to tensions among stakeholders, and these uneven power dynamics had to be mitigated as part of multistakeholder collaboration to ensure that all parties felt heard and respected. Tensions were mainly alleviated through direct contact between volunteers and social workers working together one‐on‐one and forming a direct relationship on a personal level. Post‐evaluation processes, including the focus group with social workers and interviews with end‐users, also ensured that all stakeholders had an opportunity to voice their concerns and challenges, which helped to both expose and ease power tensions.

DISCUSSION

Donabedian (2002) draws attention to the important contribution of the process component to the quality of a project. Too frequently, the success of an intervention study is reported through the outcome component in either significant or insignificant statistical results. Examining the C.A.R.E. project's implementation highlighted the importance of the process. This was understood through the five themes conceptualized from the stakeholder interviews discussed below.

Breitenstein et al. (2010) caution against “adapting to shifting contexts.” However, the nature of the pandemic resulted in ongoing international and national adaptations to curb its spread (Abdool Karim, 2020), thereby continuously shifting the context of our program. Shifts were unpredictable and legally enforced, hindering the “ideal” of certainty and smoothness. Our heightened awareness of the impact of the pandemic on older peoples' mental health increased our sense of urgency, but we experienced a lack of control over delays.

“Delivery as intended” describes implementation fidelity, but Breitenstein et al. (2010) draw attention to the problems of diminished fidelity in revised designs compared to initial blueprints. Lockdown restrictions necessitated us implementing a project without former access to the end‐users; a significant challenge to collaborative mental health promotion practices and counter to the call for community‐partnered interventions (Jané‐Llopis & Barry, 2005). Shifting contexts also hindered rapid fundraising, adding to poor fidelity. Bimha (2019) highlights how limited funding is problematic to the process component of project implementation. However, in the context of a pandemic, unprecedented times meant abandoning notions of strict adherence to pre‐packaged interventions, given the inherently unpredictable nature of the rapidly changing context. “Fidelity” then became a concept fraught with tensions.

Further to both the rollout delays and the difficulty linking with the project implementers, the restrictions in accessing the retirement homes prevented direct oversight of the implementation process. The project designers needed to trust the process and let go of “their creation.” The result was selective program modifications by the program implementers and variances across the implementation sites. However, despite debate supporting the need for strong adherence to implementation blueprints (Breitenstein et al., 2010), counter arguments question these ideals and argue for flexibility.

In a crisis, the affected people may not wait for, or be dependent on a project to empower them to meet their information and health needs (Jacobson et al., 2020). However, catalyzing social connectedness might be more challenging, in particular with older people with fewer contact opportunities (Kasar & Karaman, 2021). In this setting, many residents rejected the delayed Covid‐19 knowledge component of Project C.A.R.E. but verbalized satisfaction with the project's social connectedness and personal growth. The project implementers echoed the beneficial effect of the program on social connectivity, and for some, it resulted in sustainable support groups. Numerous studies have highlighted the value of meeting together for a common purpose, to alleviate loneliness (Fakoya et al., 2020; Hwang et al., 2020; Tomaz et al., 2021). In alignment with the sense of solidarity, the project implementers acknowledged the residents' voice in the program's requirements and made adjustments, compensating for their exclusion in the design. This also meant, by necessity, taking into account participants' individual needs, health status, fears, and anxieties about Covid‐19. The need for project modifciations highlight the importance of timeous, united, collaborative striving by all parties to reach a common goal; when steeped in uncertainty and powerlessness due to Covid‐19, offering choices held relevance (Cattan & Ingold, 2003). Of course, these choices would have been better located in the program development stage (Cattan & Ingold, 2003).

The three groups of participants showed varying degrees of reluctance toward the project, but eventual feelings of acceptance. The pandemics' restrictive context required a rapid project rollout, resulting in the implementation of a largely top–down, fast‐tracked program with decisions made by the program designers as they encountered each new stumbling block with little room for buy‐in from and checking‐in with inaccessible participants. Although directive leadership was not used per se, it is suggested to be effective in a crisis, but not without its challenges (Luksyte et al., 2021). Since time was of the essence, this interfered with the process and possibly explains the low uptake from residents, also influencing stakeholders' perceptions of each other. “Good intentions” were met with resistance and tension between stakeholders. Ideally, co‐production should be synonymous with developing a community project, but this was not the case, resulting in false assumptions and uneven power dynamics. The program designers based the motivation for the project on evidence‐based literature (Jarvis et al., 2019) and studies conducted in the geographical context describing the mitigation of loneliness (Jarvis et al., 2019) but falsely assumed an unconditional acceptance of the project by older people.

Finally, in community projects, researcher positionality requires consideration to mitigate and be reflexive about the influence of one's world views on a project (Holmes, 2020). Adopting a critical social theory lens would intercept power dynamics and permit common recognition of work demands, skills, wisdom, and expertise as stakeholders come together in a common arena for co‐production (Marston et al., 2020). It is of interest and linked to researcher positionality, that altruism did not underpin all of the project designers' motivation for delivering the program, but on reflection reciprocity was important, and they benefited from active participation. Tierney and Mahtani (2020) describe the “importance of mattering” as a motivator for volunteering during the Covid‐19 pandemic. Volunteering has protective value to combat social isolation (Tierney & Mahtani, 2020).

While the focus of evaluations is almost always on the intended beneficiaries, there is less written about the benefits to project implementers. Upon reflection, this project seemed to offer the project designers a sense of purpose, productivity, and connectivity, which might have emerged and strengthened in the enduring presence of the pandemic.

Twenty recommendations for delivering psychosocial programs in restricted community contexts

The Covid‐19 lockdown limited our access to the intended beneficiaries of our program and framed an unusually complex context. Not all community collaborations are wholly successful, and although several positive benefits resulted from this program, there were also some shortfalls. Health programs achieve success when “based on a clear understanding of targeted health behaviors, and the environmental context in which they occur” (U.S. Dept. of Health and Human Services, National Institute of Health, National Cancer Institute, 2005, p. 4). Our context was novel and rapidly changing. As program developers, we often reacted to obstacles based on instinct and experience. The themes that we conceptualized from the data were used as reflection points for us to evaluate the design, process, and implementation of the program. The program took place during a challenging and uncertain context, so it was necessary to reflect on what could have been done better, with hindsight, using feedback from all stakeholders. A reflexive thematic analysis and evaluation process ensured that the research team took a critical approach to the program, highlighting both strengths and weaknesses in design and delivery. Weaknesses were viewed as lessons to be learnt. These became apparent as the themes generated were summarized by theme and used to develop a set of recommendations to aid others in conducting a psychosocial program during the restrictions of a pandemic like Covid‐19 or as a checklist of warning signs to look out for to assist in future programs in similar contexts (Table 2).

Table 2

Twenty recommendations for delivering psychosocial programs in restricted community contexts, such as Covid‐19 lockdown.

Theme	List of recommendations
A. Adapting to shifting contexts	1. Guard against allowing urgency to undermine planning–conducting a thorough needs analysis saves time.  2. No matter how difficult, find a way to connect to potential beneficiaries from the design stage.  3. Plan for uncertainty, delays, and changing information that is bound to happen with a novel event.  4. Carefully consider the feasibility of the project and its impact on design, especially concerning funding needs and human resources, which may be impacted during the crisis.  5.  Ensure that guidelines for open communication channels are built into the project design between all parties.
B. Varying methods of program delivery	6. Plan for flexibility in project implementation that accounts for differences of opinion among developers and participants.  7. Plan for adaptation during the disruption caused by crises and personal challenges, such as illness, death, and increasing stress levels.  8.  Leave enough room for agency in implementing the project “on the ground.”
C. Moving from reluctance to acceptance	9. Work with resistance is a natural part of intervention processes and needs to be reflexively managed. 10. Build connectedness and consistency among team members—structure can help create a sense of control. 11. Recognize the shortfalls of working online and without in‐person contacts, such as online fatigue. 12. Build on emerging benefits and go where the need is.
D. Benefiting from active participation	13. Reinforce good teamwork—including commitment facilitated by a sense of purpose, productivity, meaningful learning, and a sense of connectivity. 14. Promote social connection among stakeholders, especially during periods of isolation. 15. Consider and plan for the program's sustainability even in a short‐term intervention.
E. Making false assumptions resulting in uneven power dynamics	16. Guard against making false assumptions—even in times of urgency, rely on existing evidence, and adequately assess the needs of all stakeholders. 17. Take into account differences within the so‐called “vulnerable populations” and provide wide‐ranging responses from all stakeholders. 18. Ensure the collaboration of all parties involved in the intervention, from the design stage, and ensure that all are working together toward clear and common objectives 19. Ensure ongoing monitoring and conflict resolution—do regular check‐ins with gatekeepers, especially as the environment changes. 20. Consider the actual and psychological drain of working and conducting a program during a pandemic.

Limitations

First, the use of volunteers involved in designing the program to interview the residents may have had both advantages and disadvantages. Optimistically, this may have enabled them a space to fully express how they felt about the program, but it may have had social desirability effects that limited the reporting of adverse outcomes.

Second, this evaluation would have been enriched with the triangulation of other forms of data that ensured anonymity. Third, it may be considered that important nonverbal cues such as facial expressions limited the gathering of data, but it may also have facilitated the residents feeling more at ease to express themselves freely through the increased sense of anonymity.

Future research directions

The “Brown Paper Bag” concept is recommended to be modified through collaborative partnerships, and the study repeated in retirement homes as a pre‐ and post‐design study. More research on retirement homes in LMICs is needed to explore how residents and staff coped during the Covid‐19 pandemic and what strategies were used to prevent loneliness and improve mental health and well‐being. We hope our recommendations are tested in other contexts requiring a mental health promotion intervention involving people whose movements may be restricted through crises, disasters, or war.

CONCLUSION

This study was a retrospective, qualitative evaluation of the processes involved in delivering a 12‐week psychosocial program for older people living in retirement homes in Durban, South Africa, during the Covid‐19 pandemic in 2020. Twenty‐one stakeholders were interviewed in either focus group discussions or individual interviews, and five dominant themes were conceptualized from the feedback data. The lessons learned from these results were further consolidated into 20 recommendations for delivering psychosocial programs during restrictive contexts, such as a pandemic.

Positive mental health, inclusive of social connectedness, is both a personal and a community resource that facilitates healthy approaches and perspectives to living (Jané‐Llopis & Barry, 2005). However, actioning and facilitating positive mental health as a counter to loneliness in older people during a pandemic lockdown requires stakeholders to focus not only on the outcome component but also the process of delivering a community‐based program. The process offers valuable information, and project leaders are accountable to researchers, intervention scientists, community stakeholders, and policy developers to report the findings of the process. The developed recommendations highlight the need for repeated reflections‐on‐actions by all stakeholders as a program evolves. In the implementation process, collaborative partnerships assist in navigating the challenges imposed by a restrictive context while contributing to favorable mental health outcomes. Improving the process of conducting such a program in this context would improve further care of older people living in retirement homes.

CONFLICTS OF INTEREST

S. R. P. and L. M. N. are affiliated to NPO‐2 and M. A. J. became affiliated to NPO‐1 post‐completion of the project and after data collection. The remaining authors declare no conflict of interest.

PEER REVIEW

The peer review history for this article is available at https://publons.com/publon/10.1002/jcop.22876.