a. Participants

We identified 30,133 patients from electronic health records who attended a mental health appointment (virtual or in-person) at Kaiser Permanente San Jose Adult Psychiatry within the 12 months preceding the study. Of this group, 650 potentially eligible participants were randomly selected by a random number generator. To be considered eligible, participants needed to be between the ages of 18 to 44 years, an inclusion criterion selected to focus on a generation of patients who were comfortable with social media, which we wished to inquire given its increasing adoption as a modality of social connection. Additional inclusion criteria included participants receiving concurrent care with a KP San Jose mental health clinician at the time of the study, having the ability to read, speak, and write conversational English, being able to receive messages requesting participation from the Kaiser Permanente online patient portal, and being physically present in the Northern California Bay Area during the eligibility screen and the study interview. These criteria were chosen to align with our recruitment and interview process, which began with an invitation letter written in English and sent through the online portal, and ended with a phone interview conducted in English, the primary language of the study team members. Exclusion criteria comprised of employment by Kaiser Permanente or its affiliates, active suicidal or homicidal ideation (but having prior ideation would be acceptable), concurrent treatment in an intensive outpatient program or other higher level of care (but having prior treatment at higher levels of care was acceptable), active conservatorship or guardianship by another individual or state wardship, receiving direct clinical services during study participation from any member of the clinical study team, or by investigator discretion.

The 650 potentially eligible participants were sent a secure message through the online patient portal inviting voluntary participation in the study. Approximately 100 of these potentially eligible participants also received a phone call from a study team member to invite participation. Of the 650, six individuals declined participation and 628 individuals did not respond to either secure message or phone call. Those who declined to participate cited reasons such as lack of time or interest in being involved. Sixteen subjects expressed interest and completed the eligibility phone screening interview. Two of the subjects screened were excluded due to having active suicidal ideation, and two were excluded due to not having a current mental health provider in the department. Twelve remaining eligible patients were verbally consented (as documented and witnessed by the study team; wavier of written consent was submitted to the Institutional Review Board) and completed the study interview (Fig 1). Recruitment occurred from 1/6/2023 to 3/3/2023.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. CONSORT diagram.

Eligibility screening and enrollment.

https://doi.org/10.1371/journal.pone.0316582.g001

b. Procedure

To enroll patients, the study team contacted a random selection of 650 potentially eligible patients by phone or secure message. Secure messages included a succinct description of the study, ways to contact the study team if interested in participating in the study, and the option to opt-out of further contact from the study team. An information sheet (S2 Fig) was attached to the secure messages and provided more details about the study, such as informed consent and privacy. Subjects who expressed interest completed a phone eligibility screen. Subsequently, the eligible subjects were scheduled for a phone appointment to complete the verbal consent process, to ask any further questions or opt-out of the study if they wished, and if enrolled, to participate in completing the study. Interviews were semi-structured (see S3 Fig for interview guide), lasted approximately 45 to 60 minutes, and were conducted by two clinical psychiatrists (TC and AS). Upon the interview’s completion, participants were eligible for compensation of Kaiser Permanente branded gift items worth about $30. Enrollment continued on a rolling basis concurrently with data analysis, until the research team deemed that saturation of themes had been achieved. Twelve participants completed the study. The interviews were recorded on Microsoft Teams software, with all communications and data storage occurring behind the Kaiser Permanente firewall. Research data was stored without any personally identifying information of participants. Interview recordings were transcribed by a study psychiatrist with the aid of Microsoft Teams transcription software and reviewed for quality control purposes by a second study psychiatrist before analysis.

c. Data analysis

Study clinicians used sensitizing concepts to guide examination of transcript data through an open, inductive approach, as described previously [16]. Each participant’s transcript was coded independently by two study psychiatrists (TC and AS), who then met and consolidated codes for each participant by consensus. Any specific social supports, social experiences, and/or social connection modalities mentioned were highlighted as potential codes in initial review of transcripts, as well as any experiences of the COVID-19 pandemic on social experiences or mental health. Ideas for how to “prepare socially” for a future pandemic were also coded. All codes were reviewed to find associations or recurring patterns among subjects’ answers that could serve as independent themes. Themes were identified from the coding library and consolidated by both study psychiatrists in discussion with a third study psychiatrist (HH). Quotes were maintained throughout the entirety of analysis to encapsulate the essence of the topic being discussed, relevant emotions, or experiences at hand. When codes from new interviews began consistently mapping to previously identified themes, saturation was considered achieved and study enrollment concluded. Reflexivity was maintained through reflection meetings between study team members throughout interviewing and analysis steps. Our analysis followed SRQR (Standards for Reporting Qualitative Research) guidelines [17].

Thematic category A: Factors that influenced the social experience of shelter-in-place

Thematic category B: Types of social support

As participants cited examples of social connections that were helpful to their mental health, three themes of social support were identified:

Baseline socialization patterns

Given the limitations of social activities during the pandemic, self-described introverts felt comfortable and enjoyed being able to stay at home in isolation, while some self-described extroverts reported feeling a loss of social connection that worsened their mental health. Individuals with more introverted tendencies might be more inclined to limit social activities and to avoid crowded gatherings. In addition, introverted individuals might be better equipped to cope with decreased social interaction while maintaining healthy behaviors and mindsets, without resorting to risky behaviors like substance use [18]. On the other hand, extroverts tend to have higher levels of social supports [19] and perceive their social relationships more positively [20], which could also help reduce the extent of loneliness during quarantine. Therefore, our findings suggest that clinicians consider personalizing social recommendations to individual preferences. For example, for an introvert, it can be helpful to emphasize time alone to recharge daily and to include self-care practices for mental wellbeing. For an extrovert, prioritizing interactions with positive individuals that foster healthy activities can be helpful advice for improving mental health.

Health concerns about COVID-19 spread

Individuals with pre-existing health conditions or living with those with pre-existing conditions faced heightened risks during the pandemic, including increased susceptibility to severe illness from COVID-19. As a result, individuals living with someone with chronic disease seem to have adhered more strictly to shelter-in-place measures, even if at the detriment to their own social health. From a prescribing perspective, an informed discussion of health risks and benefits of social distancing should be considered in future pandemic scenarios, including reviewing risks to social wellbeing and how best to overcome these clinical gaps within appropriate bounds of public health guidelines.

Use of electronic and social media communication

Social isolation during the pandemic led some participants to increase reliance on social media platforms to communicate with others. Our study showed a wide spectrum of social media use and its impact on mental health. Some participants reduced their use because it worsened their anxiety, and others found it helpful to continue connecting with others, find a community, and develop their hobbies. A national online survey with 3474 participants found that a decrease in mental health was associated with using social media for purposes to decrease loneliness and for entertainment motives, whereas mental health improved when social media was used for personal contact and maintaining relationships [21]. This suggests that the benefits of using social media may be partially influenced by the purposes of use and can be considered for patients to connect with their friends and family or to engage in their preferred hobbies and activities.

Types of social supports for mental health

The All of Us study conducted by the National Institute of Health found that during the COVID-19 pandemic, several types of social supports were reported that closely mirrored the types identified by our study participants [6]. In particular, tangible support (similar to our instrumental support examples), emotional support, and positive social interactions (similar to our community connection theme) were identified as reducing the odds of depression symptoms in a dose-response manner. Indeed, we found that participants reported significant benefits to their wellbeing whenever they received assistance with tasks (like grocery shopping, medication delivery, or childcare,), or felt there was someone (a friend of family member) that they could speak to anytime, or felt connected to a wider community of people. Much like how clinicians perform reviews of symptoms or inventories of strengths, we propose that clinicians can catalog or checklist these domains of social support with patients to foster mental well-being and resiliency.

Our participants universally noted the importance of having a sense of community and a feeling of closeness with others as an important part of their wellness and resilience. Individuals can be encouraged to foster social connections, supportive relationships, and community engagement to mitigate the negative impact of social isolation on their mental health. We note that many participants found community organizations and structured identity group spaces important to their well-being during the pandemic. Church groups, recovery groups, and communities that form amongst colleagues are just a few examples of social spaces that helped people feel accepted and connected to others. Clinicians can consider these aspects of individuals’ identities and life experiences to encourage building relationships in ways that facilitate access to these social spaces. This concept is already being used in the realm of substance use recovery with the concept of “recovery capital,” or the internal and external sources that are necessary to achieve and sustain recovery [22].

Limitations

This is the first qualitative study focused on patient perspectives of their social experience and the social supports that played a role in their mental health recovery during the early period of the COVID-19 pandemic. The framing of our study with a focus on social support may have attracted participants who felt positively toward their social life, and may not have captured perspectives from individuals who do not have as much social support. The majority of our participants were female, which may reflect the reality that women tend to have greater social connections than men [23], and thus may have been more likely to agree to participate in our study. Interviews were also done during 2023, a few years after the early pandemic era, and thus participants’ reflections may have been different had interviews taken place during the height of shelter-in-place. Nonetheless, we believe the timing of this study and the opportunity for reflection can also provide rich information regarding participants’ views of their social experiences. Lastly, we know that individuals of historically oppressed identities such as people of racial and ethnic minorities, sexual and gender-minority individuals, those with disabilities, rural residents, and immigrants and refugees, are at heightened risk for mental health challenges and loneliness [24]. It would be interesting for further studies to look more closely at the experiences of these populations in order to improve the quality of social prescribing and capture individuals’ specific needs.