Factors Affecting Psychological Distress in Family Caregivers of Critically Ill Patients: A Qualitative Study

Family caregivers—individuals who support the patient—play a central role in critical care survivorship, helping with health management and service use.^1,2  Experiencing a critical care hospitalization can cause psychological distress in family caregivers. This distress, recently termed post–intensive care syndrome–family, has been defined as family member anxiety, depression, posttraumatic stress disorder, and complicated grief following a patient’s stay in the intensive care unit (ICU).³ 

However, family members’ psychological distress, including anxiety and a lack of sleep affecting decision-making, also occurs during a patient’s hos-pitalization.^4–7  These psychological symptoms may affect caregivers’ ability to engage with the clinical team, their long-term psychological health,^8,9  and their ability to help manage a patient’s condition after hospital discharge.³ 

Several studies have identified factors associated with psychological distress of family caregivers during critical care hospitalization. Interventions based on these data, however, have had only a modest effect on the development of psychological distress among family caregivers.^10–15  Many prior studies also provide a limited view of how different factors contribute to caregiver psychological distress because they focus primarily on family uncertainty, informational satisfaction, and decision-making with clinicians about the patient’s medical condition.^16,17  Caregiver-specific factors such as involvement in decision-making¹⁸ ; caregiver burden¹⁶ ; lack of social,¹⁹  informational,^16,20  and religious²⁰  support; and dissatisfaction with staff communication^20,21  and the health care system¹⁸  are all associated with development of anxiety, depression, and acute or posttraumatic stress disorder in caregivers of critically ill patients and have been widely studied in other caregiver populations.^22–24  However, these factors are underexplored and underaddressed for caregivers of critically ill patients. Broadening our conceptualization and deepening our understanding of factors influencing the development of caregiver psychological distress may lead to greater insight for addressing these factors and developing effective interventions.

We conducted a theory-informed qualitative investigation to expand our understanding of psychological distress among family caregivers during a critical care hospitalization. Nursing theories of individual and family management^25,26  focus on the effects of patient and caregiver health status, family factors, resources, environmental factors, and the health care system on how family caregivers manage and respond to health conditions. Drawing on these theories, we sought to determine (1) how different factors contribute to psychological distress among family caregivers of patients in the critical care setting and (2) whether the factors that contribute to psychological distress among family caregivers differ by level of anxiety. Our primary objective was to better understand sources of family caregiver psychological distress during a patient’s critical care hospitalization to facilitate identification of potentially modifiable factors to mitigate its development and impact.

This analysis used an existing data set from a parallel mixed-methods study to examine causes of distress among family members of ICU patients (see Harlan et al²⁷  for more details of the parent study). Our data set was composed of semistructured interview transcripts and survey data. We used the Consolidated Criteria for Reporting Qualitative Research guidelines during manuscript preparation.²⁸  This study was approved by the institutional review board of the University of Michigan Medical School (HUM00118718).

Patients in the ICU who received mechanical ventilation for a minimum of 72 hours were identified through the medical record. A member of the study team purposively recruited in-person adult family caregivers of these patients from 1 adult critical care medicine unit and 1 adult surgical ICU at an academic medical center while the patients were still in the ICU. Following a brief study introduction, family caregivers were invited to participate. If interested, participants completed an informed consent process that addressed the purpose of the study, study processes, and how the resulting data would be used. Once this process was complete, a 1-time semistructured interview and survey were conducted in a private room at the medical center by a critical care physician and a nonclinician sociologist (T.S.V., J.M.). Of the 84 family caregivers approached, 40 participated from February 2017 through January 2018. A sample of 40 participants was considered sufficient to explore in depth the experiences of family caregivers, according to the concept of information power.²⁹  Reasons for declining participation included conflicts with interview timing, circumstantial stress, and no reason given. The mean interview length was 45 minutes (range, 20–89 minutes).

The interview guide focused on family caregivers’ experiences in the ICU, including perceived sources of distress, and was pilot tested with 9 members of Michigan Medicine’s Patient and Family Centered Care Program and refined.²⁷  Interviews were audio recorded and transcribed.

Data collected through the survey included family members’ demographic information and symptoms of psychological distress. Specifically, participants completed the Hospital Anxiety and Depression Scale (HADS) and the Posttraumatic Stress Disorder Checklist–Civilian Version.³⁰  Patient-specific data, such as age, sex, days in the ICU, and days receiving mechanical ventilation at the time of the interview, were obtained from the electronic health record.

We chose to specifically focus on anxiety because previous studies indicate that anxiety is an early symptom of psychological distress among caregivers in critical care.⁴  We determined caregiver anxiety level by using the HADS administered at the time of the interview, resulting in 3 participant groups as defined by the validated tool: participants with anxiety (HADS scores ≥ 11), participants with borderline anxiety (HADS scores 8–10), and participants with low anxiety (HADS scores ≤ 7).³¹  (Note that in this study we refer to HADS scores ≤ 7 as low anxiety rather than noncases as in the HADS, because a low score does not mean that anxiety is absent.) Additionally, during preliminary analysis, we noted that all participants who had depression or post-traumatic stress disorder also had anxiety or borderline anxiety, further supporting our decision to stratify the sample by level of anxiety as an indicator of level of psychological distress. Patient and caregiver characteristics and caregiver psychological distress outcomes were summarized overall and by caregiver anxiety level.

Interview transcripts were coded and analyzed using a deductive and inductive directed content analysis approach.³²  We constructed an initial code-book using constructs derived from our conceptual framework based on nursing theoretical models for family management of conditions: the Individual and Family Self-Management Theory²⁵  and the Self and Family Management Framework²⁶  (see Figure). Both theories identify factors that serve as facilitators and barriers to how individuals and families manage health conditions. These include patient and caregiver health condition factors, personal or lifestyle factors (eg, knowledge and beliefs), the health care system (eg, navigation, access, and clinician communication), family and environment factors (eg, family functioning and home and work environments), and resources (eg, housing and financial security and social support). Additionally, we looked for new patterns and codes inductively. For this analysis, we focused on caregivers’ perceptions of how different factors influenced their psychological distress. We also distinguished experiences of distress (eg, anxiety and stress) from experiences of less severe emotions (eg, sadness and disappointment) by considering the language and the narrative.

Two of the authors (A.C.B., L.E.W.), began by coding 1 transcript together line by line to establish agreement on code definition and application. Next, the 2 authors coded transcripts individually, compared code application, discussed and resolved disagreements, and finalized the codebook (see Supplemental Table, available online only at ajcconline.org). This process was done with 8 of the remaining 39 transcripts (20% of the total 40 transcripts), after which 1 of the authors (L.E.W.) completed coding the remaining 31 transcripts using the finalized codebook. Dedoose software was used to organize and analyze the data.

Once all 40 transcripts had been coded, 2 of the authors (A.C.B., L.E.W.) reviewed code reports to identify patterns in the data. A matrix was used to examine the data by caregiver anxiety level according to HADS score. Similarities and differences in perceptions across the 3 groups were identified. Rigor was ensured by searching for disconfirming cases and reconciling coding discrepancies. No disconfirming cases were found.

Forty family caregivers related to 39 patients participated in interviews. Family caregivers were most commonly adult children, followed by spouses, parents, siblings, and other family members. By anxiety level, 15 family caregivers (38%) had anxiety, 11 (28%) had borderline anxiety, and 14 (35%) had low anxiety. Patient and family caregiver characteristics can be found in Table 1. No major demographic differences by level of anxiety were noted.

In our qualitative content analysis, we found that family caregivers’ distress language focused on the patient’s health status, the health care system, family functioning, and housing and financial security, reflecting major domains within our conceptual framework (see Figure). We did not find additional themes outside the framework. By comparing family caregiver groups by level of anxiety, we gained deeper insights about these various sources of psychological distress for family caregivers while patients were hospitalized. Moreover, experiences and perceptions appeared to differ within each domain by level of anxiety.

Patient health status, a major domain in our theoretical framework, is a known contributor to distress for family caregivers. Sudden change, worsening, and lack of change in the patient’s condition were common experiences across all anxiety groups (Table 2). Family caregivers with anxiety (HADS scores ≥ 11) often used terms such as anxious and stressful to describe the experience of waiting for recovery or for answers about the patient’s medical condition. Setbacks or complications during the patient’s care most often caused distress, as did waiting for news about the patient’s health status, perceived lack of pain control, the patient’s appearance while critically ill, and unexpected use of the ventilator. Family caregivers with borderline anxiety (HADS scores of 8–10) similarly used distress language, but with less frequency, to describe the effects of unknown recovery, appearance, and unexpected sedation. Family caregivers with low anxiety (HADS scores ≤ 7) did not express their concerns using distress language but rather used language expressing emotion and disappointment (eg, “mind overloaded,” “frustrated,” “disappointed,” “confused,” “sad,” “angry”). Although family members with low anxiety and those with higher anxiety expressed similar concerns about the patient’s condition, family members with low anxiety tended to describe actions taken to process the concerns (eg, prayed with the patient or called another family member).

Aspects of the health care system, including communication, relationships with staff, and waiting to receive information on the patient’s condition, were a source of distress to family caregivers across all anxiety levels (Table 3). Additionally, anxious family caregivers described a lack of information about health care system processes, including staff roles, staff transitions, and discharge plans, as well as distress related to concerns about patients receiving poor care. Family caregivers in the borderline anxiety group reported feeling that their own wishes regarding care were overlooked and that not understanding the reasons for the medical team’s care decisions contributed to their distress. Family caregivers with low anxiety also had concerns related to communication and care quality, but they mentioned these concerns less often during the interviews.

Within the family and environment domain, family functioning was a specific source of distress for family caregivers across all anxiety levels (Table 4). Although tension and conflict between some family members was present in all groups, many anxious family members also expressed feeling responsible for the well-being of other family members, staying “strong” for them by avoiding displaying emotion, and helping with patients’ dependents. Family members with anxiety and borderline anxiety described themselves as the “go-to” person for information on the patient. Some participants with borderline anxiety reported experiencing pressure from other family members to suppress their own feelings for the sake of others and said that other family responsibilities contributed to their distress. Family caregivers with low anxiety more often described their role as that of mediator when dealing with conflict between family members and said that they served as a communicator to help other family members accurately understand the patient’s condition.

Family members with anxiety described housing and financial issues as sources of distress more often than did those with borderline and low anxiety (Table 5). Several family members with anxiety expressed concern about housing and financial security (eg, dependence on pay as caregivers or dependence on the patient’s ability to work or collect Social Security benefits) and stress due to medical bills and costs related to the patient’s hospital stay (eg, hotel and food). Some participants with borderline anxiety also described future financial fears, such as 1 participant being “scared to death” that the patient would die because the patient was the breadwinner. Family caregivers with low anxiety largely did not use distress language around financial concerns. Participants with low anxiety discussed finances in terms of having already been involved with patients’ finances, knowing patients’ financial wishes, and having already made arrangements accordingly.

The findings from this study provide detailed insights about how family caregivers experience distress related to patient health status, the health care system, family functioning, and housing and financial security. These 4 domains are reflected in our conceptual framework. Personal and lifestyle factors did not surface in our data as independent contributors to distress but were reported along with the other 4 factors (eg, personal knowledge of patient health status or the health care system). Additionally, we found differences in how these 4 domains might contribute to distress among family caregivers by anxiety level. Aspects of these identified domains are potentially modifiable and could be addressed by the care team or by interventions to mitigate the development and impact of caregiver psychological distress.

We found that uncertainty about the patient’s health status or prognosis was an important factor in the development of psychological distress, similar to findings of studies among caregivers of neuroscience critical care patients.^33,34  Caregivers with anxiety described distress related to the unexpected nature of critical illness and care, including setbacks in recovery, complications during recovery, waiting for information or recovery, the patient’s appearance, and unanticipated use of the ventilator. Although these topics may surface in family conference settings, attention to caregiver distress might be inadequate or occur late in the course of hospitalization. Caregivers’ feelings of being overwhelmed and their overall emotional state during the critical care stay need to be addressed actively and in conjunction with other communication support.³⁵  Educational or psychosocial interventions that proactively help family members anticipate these common experiences may prevent or mitigate the development of distress.

The finding that the health care system influences the development of psychological distress among family caregivers is not new. Poor communication and problems with information exchange are well known and provide the basis for multiple communication interventions.^12,14,36  However, these interventions often center on in-person family meetings and decision-making. Our participants, on the other hand, described frustration caused by a lack of regular proactive communication (eg, test results, procedure outcomes, and regular updates on the patient’s condition) and a lack of clinician awareness of prior care, conversations, and caregivers’ wishes. Enhanced capability of the electronic health record to document, synthesize, and facilitate interactions with family members using reports, reminders, and messaging for proactive communication with caregivers may ensure that such communication is integrated into the clinical workflow. Caregivers also described a lack of information about health care team processes and confusion about care decisions made. Attention to early interactions between the health care team and caregivers could avoid misperceptions of team members’ roles and clarify communication processes and caregivers’ roles in decision-making, analogous to a preflight safety briefing.

Factors related to family functioning differed by level of psychological distress, reflecting family differences such as cohesion, flexibility, roles, and communication skills.³⁷  Caregivers in all anxiety groups reported challenges communicating illness information to other family members. However, family members with anxiety more often dealt with hostility from other family members, attempted to protect other family members emotionally by avoiding displays of emotion, and took on additional responsibilities of caring for patients’ dependents. Studies of other caregiver populations, particularly caregivers of patients with mental health disorders, also highlight the importance of interventions to address the influence of family functioning on development of psychological distress.^37,38  Our findings indicate a need for further examination and strategies to address aspects of family functioning and related stressors among family caregivers of critically ill patients.

We found that housing and financial concerns during hospitalization were present for caregivers with anxiety and largely absent among those with low anxiety. Prior research has examined financial concerns, which can be related to work (changes in employment), time (time investment required by caregivers), or out-of-pocket spending (direct spending related to the caregiving role), among informal caregivers in palliative care³⁹  but not among caregivers during critical care hospitalization.⁴⁰  However, our results suggest that financial, practical, and informational support (eg, short-term housing) for caregivers in this context is needed.

The data used for this project were solely from caregivers of patients receiving mechanical ventilation at a single academic medical center. However, the caregivers included were diverse with respect to age, education, and relationship to the patient. Future research could include interviews with family members from multiple organizations, targeting both high- and low-resource settings, for a diversity of experiences related to the health care system to potentially determine additional factors. We also did not collect detailed caregiver demographic information, such as socioeconomic status, distance from the hospital, or preexisting health conditions or anxiety, and had limited information about each patient’s condition. Consequently, further investigation is also needed to better understand how these factors influence psychological distress in family caregivers during the ICU stay.

Our study shows that the factors influencing family caregiver psychological distress during the critical care stay are wide-ranging and vary by level of anxiety, yet they fully reflect 4 of the major domains in our theory-informed conceptual framework. Among caregivers with anxiety during critical care hospitalization, factors of particular salience included early unknowns regarding critical illness, health care team processes, rationale behind care decisions made, financial and housing concerns, frequent family dysfunction, and new responsibilities. In contrast, those with low anxiety generally accepted uncertainties related to the patient’s condition and care, acted as mediators for family functioning and communicating information, and appeared to have fewer concerns about finances. Our findings highlight areas for potential interventions by domain to reduce psychological distress, especially among caregivers with anxiety.