A Pilot Randomized Controlled Trial Assessing the Feasibility and Acceptability of Family-Partnered Delirium Prevention, Detection, and Management in Critically Ill Adults: The Activating Family Caregivers in the Identification Prevention and Management of Delirium (ACTIVATE) Study

Abstract

OBJECTIVES (BACKGROUND):

Delirium remains frequently undetected by healthcare providers; partnering with family may be a novel way to identify and manage delirium. This study explores the feasibility of a family-administered intervention for delirium prevention, detection, and management.

DESIGN:

Pilot randomized controlled trial.

SETTING:

Two Canadian ICUs.

SUBJECTS:

Patient-family pairs (dyads) were included. Eligible patients had no primary brain injury, a Richmond Agitation-Sedation Scale score of greater than or equal to –3, and were expected to remain in the ICU for at least 24 hours to complete all study assessments.

INTERVENTIONS:

Dyads were randomly assigned to either standard care (control) or the intervention, which included delirium education and family-administered checklists with prevention/management strategies and a detection tool (“Sour Seven”).

MEASUREMENTS AND MAIN RESULTS:

Outcomes included feasibility indicators, enrollment and completion rates, and psychological outcomes (Generalized Anxiety Disorder-7, Patient Health Questionnaire-9, and the Kessler Psychologic Distress (K-10) scales in the ICU and at 1- and 3-month follow-ups. Between January 2020 and June 2023, during the height of the COVID pandemic, 197 patient–family pairs were approached, with 64 (32%) consenting to participate; participation required both the patient and a family member. Despite recruitment challenges, 75% of families completed in-ICU questionnaires, and 38% completed all follow-ups. The family members in the intervention group demonstrated increased delirium knowledge compared with baseline delirium knowledge, and engagement in prevention strategies, with 8 of 19 (42%) family members identifying delirium using the Sour Seven. At the 3-month follow-up, seven family members showed significant anxiety, and five showed significant depression. Observations of ICU rounds revealed limited delirium discussions.

CONCLUSIONS:

This pilot study demonstrated the feasibility of family-administered delirium care in ICU settings. However, the likely impact of the COVID-19 pandemic cannot be overlooked. The study faced recruitment challenges and demonstrated the difficulties of family involvement in delirium care, particularly during restricted family presence. A full evaluation of effectiveness requires a hypothesis-testing trial with procedural adjustments to streamline data collection and strengthen family-care team partnerships.

TRIAL REGISTRATION:

Clinicaltrials.gov (NCT04099472).

KEY POINTS

Question: Is a family-administered intervention to improve delirium prevention, detection, and management feasible and acceptable in the ICU?

Findings: In this pilot randomized controlled trial across two Canadian ICUs, 197 patient–family pairs were approached, with 64 (32%) consenting. Despite recruitment challenges, 75% completed in-ICU questionnaires, and 38% completed all follow-ups. The intervention group showed increased delirium knowledge and engagement in prevention strategies.

Meaning: Family-administered delirium care in the ICU is feasible, but challenges in recruitment and integration into ICU workflows highlight the need for procedural refinements and structured family-care team partnerships in future trials.

Delirium, an acute disturbance in attention and awareness, characterized by a fluctuating course, attention deficits, and severe disorganization of behavior (1), occurs in up to 70% of patients admitted to an ICU (2–4). Delirium is associated with adverse outcomes, including extended ICU and hospital stays and increased mortality, and is associated with the development of long-term physical dysfunction, cognitive impairment, or mental health disorders collectively recognized as post-intensive care syndrome (PICS) (2, 5–8). In addition, family members of critically ill patients are at risk of developing PICS-family, experiencing psychological or emotional challenges due to their ICU experience (9–11). Delirium can be deeply distressing, both for patients-who often retain vivid memories of their symptoms-and for families who witness it (12, 13).

Delirium has been recognized as a potentially modifiable risk factor for long-term cognitive impairment in patients and psychiatric symptoms in family (14, 15). Efforts to improve early identification and prevention of delirium through non-pharmacological interventions have gained momentum, leading to the implementation of standardized delirium detection programs worldwide (16–18). Despite delirium’s impact on patient and family outcomes and initiatives to increase delirium awareness and prioritization of its detection and prevention, or management, delirium is often missed by healthcare providers (12, 19, 20).

Families are often at the bedside and may notice subtle changes in their loved one’s awareness and behavior, aiding in the identification of delirium symptoms (21–24). Furthermore, previous studies highlight the potential of family-administered delirium prevention, detection, and management to enhance early delirium detection and improve patient and family outcomes. For instance, families can be educated about delirium, empowering them with the knowledge needed to actively participate in delirium care (24, 25). Families can contribute to delirium detection by using a family-administered delirium detection questionnaire (21–23, 26). In addition, recent studies demonstrated that family-administered delirium interventions can effectively reduce the prevalence of delirium (12, 27, 28). However, although family involvement in delirium care is promising, its benefits and risks remain uncertain with evidence from studies that are heterogeneous in design and outcomes. Furthermore, there is equipoise regarding the role of families in delirium management—although some evidence suggests potential benefits, the overall impact on patient, family, and provider outcomes remains unclear. This underscores the need for feasibility studies to assess not only the practicality and effectiveness of this approach but also its potential challenges and limitations. This pilot study aimed to assess the feasibility and acceptability of a family-administered delirium prevention, detection, and management intervention.

MATERIALS AND METHODS

Trial Design

This study was pre-registered on ClinicalTrials.gov (NCT04099472) and approved on August 27, 2019, by the University of Calgary Conjoint Health Research Ethics Board (REB19-1000; Study title: Family-administered delirium prevention, detection, and management in the critically ill: a randomized control trial). Procedures were followed in accordance with the ethical standards of the responsible committee on human experimentation (University of Calgary REB) and with the Helsinki Declaration of 1975. Due to COVID-19-related family presence restrictions and research embargoes, protocol modifications were made and reported using the Consolidated Standards of Reporting Trials extension for randomized controlled trials (RCTs) Revised in Extenuating Circumstances (29, 30) (eTable 1, https://links.lww.com/CCX/B528). The full study protocol has been previously published (31).

Participants

We recruited patient and family (defined as family, relatives, or friends) pairs from two general systems ICUs within a publicly funded healthcare system (Alberta, Canada). The ICU Liberation bundle was implemented in all Alberta adult ICUs in September 2016 and included audit and feedback until 2019 (16, 32). Reported compliance with intensive care delirium screening checklist (ICDSC) documentation for eligible ICU days was 83% (32). During the study period, family presence was restricted and evolved over time. Initially, no visitors were permitted except in exceptional circumstances (e.g., end of life). This later shifted to allowing one designated family caregiver per patient, and eventually allowing more family caregivers as public health restrictions eased.

Eligible families were 18+, able to provide informed consent, had the ability to communicate with research staff (fluent in English, no hearing or visual impairment that precludes communication), and were present in the ICU to participate in delirium prevention/management and detection. Eligible patients were aged 18+; eligible for delirium assessment with a Richmond Agitation-Sedation Scale greater than or equal to –3 (i.e., not –4 or –5); no primary brain injury; could provide informed consent (both patient and family member; surrogate consent for patient possible); had the ability to communicate with research staff; and were expected to remain in the ICU for at least 24 hours. Due to COVID-19 research prioritization, patients with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) were excluded from March 15, 2020 to June 7, 2021.

Standard recruitment, which included a patient partner with ICU experience, was not feasible due to restrictions. Instead, research staff—introduced to patients and families by the care team—conducted recruitment. Consent discussions included an infographic (Appendix 1, https://links.lww.com/CCX/B528) co-designed with patient partners. Participants were assigned to the intervention or control group (1:1) using computer-generated randomization (eTable 2, https://links.lww.com/CCX/B528).

Intervention

The intervention (detailed in the published protocol (31)) included a control arm (standard care with and ICU delirium pamphlet) and an intervention arm [delirium education and family-administered delirium prevention/management (daily checklist with non-pharmacological interventions) and detection (Sour Seven)] for up to five ICU study days. Study measures are outlined in eTable 3 (https://links.lww.com/CCX/B528). The Sour Seven is a seven-item screening tool developed for use by untrained informal caregivers and demonstrated good agreement when compared with a reference standard (21). It includes seven key behavioral and cognitive indicators of delirium, such as altered level of awareness, reduced attentiveness, fluctuation, disordered thinking, disorganized behavior, impaired eating/drinking, and difficulty in mobility (23) (Appendix 2, https://links.lww.com/CCX/B528).

Outcomes

The primary outcomes were the feasibility of enrollment and questionnaire completion, to inform the design of a larger trial. Given the recruitment rates from our prior studies on family-administered delirium detection only (75%) (21) and delirium education intervention only (45%) (25), we anticipated a 40% recruitment rate and in-ICU survey completion of 75% (21, 25, 33), and anticipated a loss of ~33% of participants at 3-months. Research staff collected data using article surveys or direct Research Electronic Data Capture (REDCap; Vanderbilt University, Nashville, TN) while the patient/family was in the ICU. All follow-up data were collected using a link to REDCap surveys. Gift cards were given for each study milestone reached: 1) days 1–5 questionnaires ($10), 2) 1-month questionnaires ($10), and 3) 3-month questionnaires ($10).

Family Member Outcomes

Family member experience: Families’ needs, barriers, and coping strategies were assessed using the following validated questionnaires: Critical Care Family Needs Inventory (CCFNI) (34), Barriers to Care Questionnaire (BCQ) (35, 36), and Coping Strategies Scale (CSS) (37). These surveys were completed once, upon enrollment.

Family ICU Delirium Prevention, Detection, and Management: Families’ delirium knowledge was assessed using a 21-item Caregiver ICU Delirium Knowledge Questionnaire (CIDKQ) (38). The CIDKQ was completed upon enrollment, after delirium education (intervention group only), and at 1- and 3-month post-ICU discharge. Families assessed delirium symptoms in patients using the Sour Seven (23) daily, for a maximum of 5 days (5 total) when convenient for the family member (i.e., any time of day). Research staff interpreted the Sour Seven results (i.e., delirium/no delirium) and shared this with the family and bedside nurse.

Mental health outcomes: Psychological distress was measured using the Generalized Anxiety Disorder-7 (GAD-7) (39), Patient Health Questionnaire-9 (PHQ-9) (40), and the Kessler Psychologic Distress Scale (K-10) (41) at enrollment, daily in the ICU (up to 5 d), and at 1- and 3-months post-discharge. The Delirium Burden Scale (DEL-B) measured the level of distress related to symptom burden, emotional strain, and situational challenges due to delirium (42) and was completed daily in the ICU (maximum 5 d).

Round observations: Sour Seven results were added to the existing registered nurse (RN) and respiratory therapist’s ICU Liberation (Pain, Agitation/Sedation, Delirium, Immobility, Sleep Disruption) (16) rounds sheet to prompt reporting of family-assessed delirium status. Research assistants audited 10% of participants’ bedside rounds to assess compliance (Appendix 3, https://links.lww.com/CCX/B528).

Sample Size

We based the potential sample size on the ability to detect a feasibility problem that might interfere with a subsequent hypothesis-testing trial, such as recruitment rates, intervention delivery, and loss to follow-up. This was informed by enrollment rates in previous studies (33). As such, we aimed to recruit 30 patient and family pairs per arm.

Randomization

Patients were randomized with a 1:1 allocation ratio.

Data Analyses

We reported continuous variables using means (sd) or medians (interquartile ranges [IQRs]), depending on the distribution. As this was a pilot study, it was not powered for hypothesis testing. Feasibility was assessed via recruitment rates, retention rates, and protocol adherence. Secondary analyses examined PHQ-9, GAD-7, and K-10 scores, clinically relevant symptoms of anxiety/depression, and prevalence at ICU discharge, 1-, and 3-months post-ICU. Family delirium knowledge and burden (CIDKQ, DEL-B scores) were compared pre- and post-intervention. Intervention coherence (study days, questionnaire completed) was summarized by the patient and family pair. We analyzed data using R, Version 4.3.0 (R Core Team, 2023, Vienna, Austria) (43). One researcher trained in qualitative methods (K.D.K.) analyzed free text from round observations using qualitative content analysis (44) using NVivo 14 (Lumivero, Denver, CO), aiming to capture a general sense of the topics discussed rather than identifying specific themes or subthemes. A second researcher (K.F.) reviewed the coding to ensure accuracy. Discrepancies were discussed until both researchers agreed the codes reflected the content of the free-text notes.

RESULTS

The study was conducted between January 30, 2020, and June 30, 2023, with three recruitment pauses due to research embargoes related to the pandemic (March 15 to September 14, 2020; December 4, 2020 to February 22, 2021; and May 6 to June 7, 2021) and one investigator-initiated pause to reassess and remove questionnaire burden based on participant feedback (March to August 2022). Although we engaged patient and family partners during the study design phase, and they did not express concerns about the questionnaire length at that time, the demands of real-world ICU participation during the pandemic prompted us to revisit this issue. During this recruitment pause, we consulted with patient partners, who are members of the research team, and decided to remove three questionnaires-CCFNI, BCQ, and CSS-to reduce the time required by participants. During recruitment periods, a total of 3311 ICU patients were assessed for eligibility. Of these, 2840 did not meet the inclusion criteria. Common reasons for ineligibility were that the patient was not expected to remain in the ICU for a further 24 h (n = 967), had a primary direct brain injury (n = 887), or was positive for SARS-CoV-2 (n = 300) (Fig. 1). Ultimately, 407 patients (12%) met the eligibility criteria. The most common reasons for not approaching patient-family pairs included the absence of family members (n = 172), unavailability of the bedside RN (n = 75), and conflicting co-enrollment with other studies (n = 27).

Figure 1.

Consolidated Standards of Reporting Trials diagram. RASS = Richmond Agitation-Sedation Scale, RN = registered nurse.

A total of 197 patient-family pairs were approached for study enrollment. The RN, as the first point of contact, approached all pairs, with 22 declining participation at this stage. The remaining 154 were then approached by the research team, with 90 declining. In addition, 21 participants did not consent due to needing more time to consider or being unavailable for follow-up, either from evening visits or not returning for subsequent visits. Consent was obtained for 64 pairs, resulting in a 32% consent rate. Four participants withdrew from the study (three from the intervention arm and two from the control arm) due to feeling overwhelmed.

Overall, 75% (45/60) completed in-ICU questionnaires (intervention = 21; control = 24). Families who did not complete the questionnaires directly after the consent process were left with printed copies or a link to fill out at their convenience. Common reasons for incomplete responses included patient discharge before completion (n = 8), difficulty reconnecting with families who visited only in the evening (n = 8), and COVID-19 precautions limiting family visits (n = 2). Of 45 who completed baseline questionnaires, 67% (n = 30) completed the 1-month questionnaires, and 51% (n = 23) completed the 3-month questionnaires. Reasons for loss to follow-up included patient death or ongoing hospitalization, and family being overwhelmed. With 26 months of recruitment (considering all study pauses), the enrollment rate was 2.5 patient and family pairs per month.

Of the 45 family members who reported their own sex, most were female (n = 27/45, 60%). Family members were most often spouses (19/45, 42%) or children (12/45, 27%). Patient sociodemographic and clinical-related information and family sociodemographic-related information are displayed in Table 1 (detailed characteristics are displayed in eTable 4, https://links.lww.com/CCX/B528).

TABLE 1.

Participant Characteristics

Characteristics	Patient			Family
	Total (n = 59)a	Control (n = 31)	Intervention (n = 28)	Total (n = 45)	Control (n = 24)	Intervention (n = 21)
Age, yr, median (IQR)	58 (47–66)	60 (47.5–70.5)	50.5 (46.2–62.2)	48.5 (40.7–61.5)	47 (39.7–55)	56.5 (45.2–65)
Male sex, n (%)	40 (66.7)	26 (83.9)	14 (48.2)	17 (37.8)	9 (37.5)	8 (38.1)
Patient clinical information
Medicalb	41 (69.5)	19 (61.3)	22 (78.6)
Surgical	13 (22.0)	8 (25.8)	5 (17.9)
Trauma	4 (6.8)	3 (9.7)	1 (3.6)
Acute Physiology and Chronic Health Evaluation-II score, median (IQR)	18 (14–21)	17.5 (14–21)	18.5 (14–21.2)
Sequential Organ Failure Assessment score, median (IQR)	6 (5–9)	6.5 (5–9)	6.5 (5–9)
ICU length of stay, median (IQR)	12.9 (6.4–20.8)	12.9 (6.3–15.6)	13.9 (8.5–21.5)
Invasively ventilated, n (%)	47 (79.7)	24 (77.4)	23 (82)
Delirium (ever/never)c	28 (47.5)	16 (51.6)	12 (42.9)
Delirium duration, median (IQR)	1.5 (1–8.2)	1 (1–7.5)	4.5 (1–8.5)
ICU mortality, n (%)	4 (6.8)	3 (9.7)	1 (3.6)
Relationship to patient
Spouse or common-law				19 (42.2)	8 (33.3)	11 (52.4)
Child				12 (26.7)	9 (37.5)	3 (14.2)
Parent				9 (20)	3 (12.5)	6 (28.6)
Sibling				5 (11.1)	4 (16.7)	1 (4.8)

IQR = interquartile range.

^aClinical data for one patient are missing from the intervention arm.

^bMedical admitting diagnoses include patients with cardiovascular conditions, respiratory. conditions, infectious diseases/sepsis, change in level of consciousness, gastrointestinal conditions, and burns.

^cFrom the bedside nurse once per shift, Intensive Care Delirium Screening Checklist.

Family ICU Experience

Based on feedback from participants and patient partners on our research team, we evaluated which questionnaires could be removed from the study to reduce questionnaire burden, while maintaining the focus on primary outcomes. The objective of the future, adequately powered trial will be to determine the effect of family-administered delirium prevention, detection, and management in critically ill patients on family members’ symptoms of depression and anxiety, compared with usual care. As such, retaining the anxiety and depression questionnaires in the pilot was necessary to assess feasibility for the larger trial. These results are now shared in Appendix 4 and eTable 5–7 (https://links.lww.com/CCX/B528).

Family ICU Delirium Prevention, Detection, and Management

Forty-five participants (n = 45/60, 75%) completed the baseline delirium knowledge (CIDKQ) questionnaire (intervention: n = 21; control: n = 24). The mean baseline CIDKQ score for the families randomized to the control arm was 14.2 (sd: 3.6, range: 1–20, possible range: 0–21 wherein a higher score indicates more delirium knowledge), compared with 14.0 (sd: 4.6, range: 1–20) for families randomized to the intervention arm. Of the 21 families from the intervention arm who completed the pre-delirium education CIDKQ, 20 completed the post-delirium education CIDKQ. The mean score of the post-delirium education CIDKQ was 17.3 (sd: 3.0, range: 7–20). The means for the treatment groups and subscales (risk, action, symptom recognition) and time points (in-ICU, 1-mo, 3-mo) are displayed in Table 2.

TABLE 2.

Caregiver ICU Delirium Knowledge Questionnaire

Subscales	Control, Mean (sd, Range)			Mean (sd, Range)
	Baseline (n = 24)	1-mo (n = 9)	3-mo (n = 15)	Pre-intervention (n = 21)	Post-intervention (n = 20)	1-mo (n = 13)	3-mo (n = 12)
Total score, possible range: 0–21	14.2 (3.6,25, 3–20)	15.1 (2.8, 8–19)	14.7 (2.7, 10–20)	14.0 (4.6, 1–20)	17.3 (3.0, 7–20)	17.7 (2.1, 12–20)	16.1 (2.7, 10–19)
Risk score, median (IQR, range), possible range: 0–10	7.4 (2.3, 0–10)	7.5 (1.8, 3–10)	7.7 (1.7, 5–10)	7.3 (2.6, 0–10)	8.6 (1.7, 3–10)	8.9 (0.9, 7–10)	8.5 (1.4, 5–10)
Action score, median (IQR, range), possible range: 0–6	4.2 (1.0, 2–6)	4.5 (1.3, 1–6)	4.4 (1.1, 3–6)	4.3 (1.6, 0–6)	5.4 (0.9, 3–6)	5.4 (0.9, 4–6)	5.3 (0.9, 4–6)
Symptom recognition, median (IQR, range), possible range: 0–5	2.6 (1.2, 0–5)	2.8 (1.0, 1–4)	1.6 (0.9, 2–4)	2.4 (1.0, 0–5)	3.3 (1.0, 1–4)	3.4 (1.0, 1–4)	3.0 (1.1, 1–4)

IQR = interquartile range.

Of the 20 patient and family pairs from the intervention group who completed the pre- and post-delirium education CIDKQ, 19 (95%) finished at least one Sour Seven assessment, with a median completion of four Sour Seven questionnaires (IQR: 2–4, range: 1–5). A total of 8 of 19 (42%) family members identified delirium using the Sour Seven. However, without a complementary CAM-ICU or ICDSC assessment, it is unclear whether these patients had delirium. A total of 19 (n = 19/20; 95%) family members completed the caregiver delirium prevention and management checklist. Families completed these checklists a median of four (IQR: 2.5–4) times. The most common delirium prevention and management strategies families reported using were orienting the patient (date/time, location, current events) (n = 16/19, 84%) and watching or discussing the news with their loved one or talking about a favorite memory (n = 14/19, 74%). The frequency of all delirium prevention/management strategies used by families is displayed in Table 3.

TABLE 3.

Family-Administered Delirium Prevention and Management Strategies (n = 19/29, 66% of Families Enrolled in the Intervention Arm)

Strategy	n (%)
Oriented the patient (date/time, location, current events)	16 (84)
Helped with “normal” sleep pattern: Lights on or blinds up and conversation (DAY)	14 (74)
Watched or discussed the news with the patient or talked about a favorite memory	14 (74)
Brought in photos and/or familiar items from home	11 (58)
Helped with patient hygiene (brush teeth, comb hair, cut nails)	11 (58)
Asked a member of the ICU care team about medication changes (another time)	11 (58)
Helped with “normal” sleep pattern: Lights off and quiet (NIGHT)	10 (53)
Spread visitors throughout the day	8 (42)
Played quiet/calm music	8 (42)
Squishy ball	8 (42)
Brought in hearing aids and/or glasses	7 (37)
Read to the patient	7 (37)
Helped with “normal” sleep pattern: Sleep aids (eye mask, earplugs, blanket from home)	6 (32)
Asked a member of the ICU care team about medication changes (during rounds)	6 (32)
Textured ball	6 (32)
Told a member of the ICU care team about delirium symptoms (during rounds)	5 (26)
Told a member of the ICU care team about delirium symptoms (another time)	5 (26)
Brain games or family-assisted physical activity	5 (26)
Puzzles (sudoku, crosswords, word searches)	4 (21)
Coloring	3 (16)
Played cards	3 (16)
Used SmallTalk Intensive Care app to communicate with the patient	1 (5)

A total of 24 (n = 24/31, 77%) participants from the control group and a total of 21 (n = 21/29, 72%) participants from the intervention group completed the DEL-B a median of 3 (IQR: 2–5) and 5 (IQR: 3–5), respectively, times during their ICU stay. The median DEL-B score for the control group was 12 (IQR: 5–18, range: 0–35), while the median score for the intervention group was 10.5 (IQR: 4–18, range: 0–33), of a possible 40 points. Of those families who detected delirium using the Sour Seven in the intervention group the median score was 16 (IQR: 12.7–22, range: 0–33 wherein higher scores indicating families experiencing more distress related to delirium symptoms) compared to a median of 6.5 (IQR: 2.7–11.2, range: 0–30) in those families who detected no delirium.

Family Psychological Outcomes

Forty-five participants (45/60, 75%) completed the baseline (day 1, pre-intervention) GAD-7, PHQ-9, and K-10 questionnaires. While in the ICU (days 2–5), 40 of 45 participants (88.9%) completed at least one assessment, with a median of 3 completions (IQR: 0–4, range: 1–4). The distribution of GAD-7, PHQ-9, and K-10 scores during the ICU admission and at 1- and 3-months post-ICU by treatment group is shown in Figure 2. Overall, seven family members experienced clinically significant symptoms of anxiety, and five reported clinically significant symptoms of depression during their 3-month follow-up. Clinically significant symptoms are displayed in eTable 8 (https://links.lww.com/CCX/B528), and are categorized by treatment groups, timepoints, and patients with ever/never delirium in Figure 3.

Figure 2.

Boxplots showing median, 25th, and 75th percentiles of the Generalized Anxiety Disorder-7 (GAD-7), Patient Health Questionnaire-9 (PHQ-9), and Kessler Psychologic Distress Scale (K-10) score in families at study enrollment (baseline), during the study (daily for up to 5 study days), and 1- and 3-months post-ICU discharge. Whiskers show the minimum and maximum reported scores. GAD-7 measures generalized anxiety severity (range: 0–21); scores of greater than or equal to 10 indicate clinically significant anxiety, with cutoffs of 5, 10, and 15 representing mild, moderate, and severe anxiety, respectively. PHQ-9 measures depression severity (range: 0–27); scores of greater than or equal to 10 indicate clinically significant depression, with cut-offs of 5, 10, 15, and 20 representing mild, moderate, moderately severe, and severe depression, respectively. K-10 measures general psychological distress (range: 10–50); higher scores indicate greater distress. These data illustrate the psychological burden experienced by families during and after ICU hospitalization.

Figure 3.

The families reporting clinically significant symptoms on the Patient Health Questionnaire-9 (PHQ-9) (depression) and Generalized Anxiety Disorder-7 (GAD-7) (anxiety) across different timepoints (baseline, ICU, 1-mo, and 3-mo) in both the intervention and control groups. Results are further stratified by whether the patient experienced delirium (orange) or no delirium (blue).

Rounds Observations

A total of 19 rounds were observed for six study participants. The mean (sd) length of rounds was 15:00 minutes (sd: 5:59). All rounds were led by an attending physician, with six different attendings participating, representing six different service weeks. Family members were present for 10 (n = 10/19, 53%) of the rounds. Delirium was discussed during 11 (n = 11/19, 58%) of the rounds, with the bedside RN or attending physician initiating conversation on possible signs of delirium. During one round, a family member acknowledged that the patient appeared confused. In two rounds, family members described how factors such as the ICU environment (limited light due to window orientation), medications, and procedures (e.g., continuous renal replacement therapy) affected their loved one’s cognition and mood. Delirium prevention strategies were mostly initiated by bedside RNs and physicians (resident, attending) during four rounds (n = 4/19, 21%). In one instance, a family discussed their loved one’s medication being changed to an antipsychotic. The Sour Seven was never discussed during rounds.

DISCUSSION

This multicenter, pilot randomized controlled trial comparing family-partnered delirium prevention, detection, and management to usual care informs the feasibility of a larger randomized controlled trial. We enrolled 2.5 patient-family pairs per month (45). This recruitment rate was lower than in our pre-pandemic studies involving the same ICU settings and population (21, 25), which may be attributable to pandemic-related barriers such as restricted family presence and increased strain on families and healthcare systems. Despite challenges such as loss to follow-up, we achieved an in-ICU study completion rate of 75%. However, only 19 of 29 (66%) of families randomized to the intervention arm engaged with daily delirium prevention/management and detection activities.

Consistent with previous studies, family members of ICU patients experienced high levels of anxiety and depression during the ICU stay (15, 46, 47). Symptoms persisted at 1- and 3-month post-ICU, exceeding general Canadian population rates (48), underscoring the need for preventive and follow-up interventions for families in addition to patients. Median GAD-7 and PHQ-9 scores were lower in the intervention group than in the control group, while the duration of delirium was longer in the intervention group. However, our study was designed to evaluate the feasibility and acceptability of the intervention and was not powered to test for differences in clinical outcomes. A larger, hypothesis-driven trial is needed to assess clinical impacts.

Our partial observations revealed that families were present for just over half of the bedside rounds, reflecting a culture of non-inclusion by families either by the managing team or by the families themselves. Although some families reported delirium observations during rounds, overall engagement was limited. This highlights the need to foster an inclusive culture during rounds, aligned with the Society for Critical Care Medicine’s guidelines on family-centered care (49). Although evidence supporting these recommendations is “very low” and is based on small RCTs and observational studies, heterogenous in design and outcomes, promoting family input (e.g., inviting observations about cognition) could improve engagement and outcomes. The feasibility of studies incorporating family engagement, particularly in delirium detection and management, will highly depend on the degree of family involvement and institutional practices. By fostering a collaborative approach, families may feel more like active partners in delirium detection, prevention, and management—an area identified as a priority in our prior research (50).

Recruitment and retention were lower than anticipated, likely due to several factors. The COVID-19 pandemic and restricted family presence affected participation, with a “one designated visitor” policy in the pilot ICUs during the second, third, and fourth pandemic waves in Alberta. Even after policies became more lenient, many eligible patients were missed due to continued family presence restrictions. Another possible factor influencing recruitment and retention is the high-stress environment of the ICU, which may have reduced family members’ willingness to take on additional responsibilities, even those perceived as beneficial (51, 52). The complexity of the intervention, requiring family members to adopt new roles and responsibilities, may have discouraged some participants—a challenge noted in prior family engagement studies, where not all families desired direct involvement in care (53, 54). Importantly, as shown in Table 3, families varied in their use of delirium prevention interventions, underscoring the need for flexible, tailored engagement options to accommodate diverse patient preferences and capabilities.

To reduce participant burden, we removed several baseline questionnaires (CCFNI, BCQ, and CSS), and will exclude the DEL-B in the future. Families will only complete demographic, CIDKQ, GAD-7, and PHQ-9 questionnaires, which should improve completion. Despite prior findings that educational materials enhance delirium knowledge, Table 2 shows little difference in delirium knowledge between groups, suggesting our education intervention may need revision. Improvements could include a teach-back method or an infographic summarizing key delirium information. To improve follow-up, we will implement evidence-informed strategies (55) like text message survey links and study newsletters, as most participants wished to remain informed about study outcomes. We will also shift to a single 3-month follow-up, consistent with other ICU family studies (56–58). Ultimately, improving engagement will require a broader culture shift beyond study logistics. Families must see their input as valued and integrated as a routine and essential part of delirium care (50).

Strengths and Limitations

A key strength of this study is the patient partner involvement from study inception to analysis, interpretation, and publication. This ensured that patients and families who were approached to participate found the study materials and recruitment processes acceptable (i.e., not overwhelming and understandable). The standard recruitment strategy we employ on our team involving a patient partner was not feasible during these circumstances, which may have further affected enrollment and engagement rates (21, 33). Limitations of this study include the use of self-report measures, which are not diagnostic tools that can diagnose anxiety or depression, and challenges with recruitment, as it may have impacted the generalizability of the findings due to potential selection bias or limited participant enrollment. Families who are more engaged, motivated, or have greater resources may have been more likely to participate, which could affect the representativeness of the sample and limit the broader applicability of the results.

CONCLUSIONS

This pilot study demonstrated the feasibility of family-administered delirium care in ICU settings. However, the impact of the COVID-19 pandemic cannot be overlooked. Strict visitation restrictions and the heightened stress of the ICU environment created significant barriers to recruitment and retention, limiting family involvement in ways that would not be expected under normal circumstances. Despite these challenges, families who were able to participate engaged actively with the intervention, reinforcing its feasibility for a larger trial. A full evaluation of effectiveness requires a larger trial with procedural adjustments to streamline data collection and strengthen family-care team partnerships.

ACKNOWLEDGMENTS

We thank the patients and families who participated in this pilot study. This study was supported by a Canadian Institutes of Health Research (CIHR) Transitions in Care team grant and CIHR Strategy for Patient-Oriented Research grant awarded to Dr. Fiest. We also acknowledge Kara Plotnikoff for important contributions as the project manager at the outset of the study.

Supplementary Material

cc9-7-e1287-s001.pdf ^{(1.7MB, pdf)}