Acute Respiratory Failure Survivors’ Perspectives on Patient-Centered Communication: A Mixed Methods Study

Abstract

IMPORTANCE:

Many acute respiratory failure (ARF) survivors develop psychiatric sequelae. Inpatient psychologic distress is an important risk factor.

OBJECTIVES:

Evaluate ARF survivors’ perspectives on patient-centered communication and its relationship to the inpatient psychologic experience.

DESIGN, SETTING, AND PARTICIPANTS:

A mixed methods prospective observational study involving one survey and a semi-structured interview administered within 6 weeks of discharge to ARF survivors from two hospitals in the same system (August 2024 to December 2025). ARF was defined as greater than or equal to 24 hours of mechanical ventilation, high-flow nasal cannula, or noninvasive positive pressure ventilation. Fifty-six survivors were enrolled; 20 participated in interviews. Twenty-four survivors (43%) were male, 28 survivors (50%) were identified as Black or African American, 25 survivors (45%) were White, and three survivors (5%) were Hispanic. ARF was most commonly due to respiratory (50%), cardiovascular (23%), or gastrointestinal (7%) disease. Forty-three participants (77%) were mechanically ventilated.

ANALYSIS:

Qualitative interviews were analyzed using grounded theory. Quantitative measures included the “Heard and Understood” (H&U) measure, the Hospital Anxiety and Depression Scale, the Impact of Events-6 Scale, a 0–100 visual analog scale measuring perceived health, and the EuroQol-5 dimension quality of life measure. Associations between H&U measure responses and mental health outcomes were tested using the Wilcoxon rank-sum and chi-square tests.

RESULTS:

Patient-centered communication was characterized by meaningful interpersonal relationships, transparent and responsive communication, effective healthcare dynamics, and timely care. Communication perceived as patient-centered improved psychologic well-being, uncertainty tolerance, and engaged shared decision-making. Twenty-one of 56 (38%) participants felt “completely” H&U. Those who did had lower anxiety (4 [1–6] vs. 7 [5–10]; p < 0.01) and depression (6 [5–10] vs. 10 [7–13]; p = 0.02) scores, greater perceived health (80 [70–90] vs. 60 [43–78]; p < 0.01), and greater health-related quality of life (0.84 [0.62–0.94] vs. 0.55 [0.29–0.72]; p < 0.01).

CONCLUSIONS:

Patient-centered communication may improve psychologic well-being in ARF survivors while admitted and after discharge.

KEY POINTS.

Question: This mixed methods prospective cohort study evaluated the perspectives of acute respiratory failure (ARF) survivors on patient-centered communication and its relationship to psychologic well-being (feeling “heard and understood” [H&U]).

Findings: Interpersonal relationships, communication content and style, and care processes, including healthcare dynamics and timeliness of care, were important elements of patient-centered communication. Survivors who felt H&U reported more positive emotional states, reduced uncertainty, and increased engagement with shared decision-making. Of 56 ARF survivors, only 21 survivors (38%) felt “completely” H&U.

Meaning: Patient-centered communication is associated with improved inpatient well-being and may be a target for efforts to improve post-ARF mental health.

A total of 800,000 patients survive treatment in the ICU for acute respiratory failure (ARF) annually in the United States (1). Alarmingly, one in five survivors will experience new posttraumatic stress symptoms and one in three will experience anxiety or depression symptoms (2–4). These disorders are associated with reduced quality of life, increased risk of self-harm, and increased 5-year mortality (5–8). There is no consensus on the optimal approach to improving post-ARF mental health symptoms (9–13).

In contrast to post-ARF physical and cognitive syndromes, studies seeking to identify modifiable risk factors for post-ARF mental health disorders have found no association with clinical disease severity, respiratory support type, or sedation (14, 15). To date, the only identified risk factor from the inpatient setting is a negative psychologic experience during hospitalization, related to highly stressful ICU events, intense fear, and perceived discomfort (7, 16, 17). How best to mitigate these negative psychologic experiences is unknown, but a small body of literature in other populations suggests that patient-centered communication may reduce illness-related anxiety and stress (18–23).

Patient-centered communication has been defined as communication emphasizing the patient perspective, acknowledging psychosocial context, and allowing for shared power and responsibility between clinicians and patients (24). However, the specific interactions in a real-world clinical setting that lead ARF survivors to perceive communication as patient-centered are unclear. In this mixed methods study, we evaluated ARF survivor perspectives on what constitutes patient-centered communication and how it relates to psychologic well-being, conceptualized as the extent to which they felt “heard and understood” (H&U) (25). In exploratory quantitative analysis, we also assessed the proportion of ARF survivors reporting patient-centered communication and their characteristics.

MATERIALS AND METHODS

Patient Population

Participants were recruited from an ongoing single-center prospective cohort study enrolling individuals of 18 years or older treated in the ICU for ARF and discharged home (NCT06401239; Appendix 1, https://links.lww.com/CCX/B688). ARF was defined as greater than or equal to 24 hours of mechanical ventilation, noninvasive ventilation, or high-flow nasal cannula. This definition has been used in prior work evaluating the long-term impacts of critical illness (11, 26–28) and was selected because specific etiologies of ARF have not been shown to predict poor mental health outcomes (3, 29, 30).

Outcome Measures

All measures were administered within one month of hospital discharge via an online survey (Research Electronic Data Capture [31]) with the option for in-person or telephonic completion.

The Heard and Understood (H&U) Measure

We measured perceived patient-centered communication using the validated, single-item self-reported H&U measure (25, 32–34). Adapting the original measure to reference the hospital stay, we asked: “During your hospital stay, how much did you feel H&U by the doctors, nurses, and hospital staff?” (five-point scale: “Not at all” to “Completely”). To align with previous work using the H&U measure and similar communication measures in which a systematic upward bias has been noted (32, 35, 36), responses were dichotomized as “completely” vs. less than “completely” H&U.

Mental Health and Quality of Life Measures

Participants completed the Hospital Anxiety and Depression Scale (HADS, license number 2316045), the Impact of Event Score-6 (IES-6), the EuroQol-5 dimensions (EQ-5D, registration number 59569), a health-related quality of life measure, and a 0–100 visual analog scale measuring self-reported perceived health (28, 37–39). The HADS and IES-6 scores were analyzed as continuous outcomes (higher scores representing higher symptom severity). EQ-5D responses were converted to a continuous outcome based on a U.S. valuation scale as previously described (40).

Quantitative Statistical Analysis

A convenience sample of all parent study participants enrolled at the time of the current analysis was analyzed. As this is the first study to use the H&U measure in this population, we performed exploratory descriptive analysis.

Patient demographic and illness characteristics (including age, sex, self-reported race, highest level of formal education, area deprivation index of home zip code, history of anxiety or depression, maximum Sequential Organ Failure Assessment score on ICU admission day, respiratory support, presence of delirium, exposure to sedation, and hospital and ICU lengths of stay) were summarized as median (interquartile range [IQR]) and counts (proportions), as appropriate, for the cohort overall and stratified by H&U measure response. Differences were compared using the Wilcoxon rank-sum test and chi-square test, as appropriate.

For study participants with a single missing response on the IES-6 or HADS, missing values were handled via mean substitution. A missing response on the EQ-5D excluded the participant from the description for this measure.

Qualitative Methods

All survey participants were invited to an interview via telephone call after survey completion, up until completion of the first 15 interviews. Subsequently, survey participants were purposively sampled for invitation to interview based on their response to the H&U measure in order to obtain varied representation of H&U measure responses among interviewees. Interview recruitment continued until thematic saturation was achieved (no new themes in three consecutive interviews), estimated a priori to require 20–25 interviews (41).

A multidisciplinary team developed and iteratively refined an interview guide informed by published literature (24, 32, 33, 42, 43). The interview guide domains were 1) healthcare team communication experiences, 2) the meaning of feeling H&U, and 3) the impact of communication on mental health and recovery (team details and guide in Appendices 1 and 2, https://links.lww.com/CCX/B688).

Audio-recorded interviews occurred via telephone or videoconferencing within 6 weeks of discharge. Interviews were transcribed using a HIPAA-compliant platform and verified by the first author to ensure verbatim transcription. Inductive coding started after three interviews. The first five transcripts were coded independently by two coders, a codebook was developed, and codes were then applied by the first author and independently reviewed by the second author. Disagreements were resolved by consensus. NVivo 15.3.1 (Lumivero, Denver, CO) was used for data management.

The study protocol was guided by the Consolidated Criteria for Reporting Qualitative Health Research (44). The Johns Hopkins University institutional review board approved the survey portion of this study (IRB00429340, “Investigating Dyadic Expectations about ARF Survivorship,” April 12, 2024), and determined the interview portion to be exempt (IRB00425406, “Survivor Input on Factors Affecting Certainty and Expectations,” April 23, 2024). All procedures associated with this research were conducted in accordance with the ethical standards delineated in the Declaration of Helsinki 1975. Interview participants were reimbursed $25.

RESULTS

A total of 223 eligible ARF survivors were identified, of whom 56 were enrolled in the survey portion of this study (Fig. 1). Median age of survivors was 55 years (36–65 yr). Twenty-four (43%) were male, 28 (50%) identified as Black or African American, 25 (45%) as White, and three (5%) as Hispanic (Table 1). The three most common etiologies of ARF were respiratory (50%), cardiovascular (23%), and gastrointestinal (7%) disease. Forty-three (77%) had been mechanically ventilated. Of patients who had been mechanically ventilated, sedation administered included fentanyl in 40 of 43 (93%), propofol in 40 of 43 (93%), dexmedetomidine in 33 of 43 (77%), and lorazepam or midazolam in 27 of 43 (63%). Thirty-seven of 56 (66%) of patients experienced delirium during their ICU stay, with a median duration of delirium of 3 days (IQR, 1–4 d).

Figure 1.

Study cohort selection. Flowchart depicting enrollment of acute respiratory failure survivors from two academic hospitals between August 2024 and December 2025. Numbers indicate patients at each stage. A final cohort of N = 56 patients completed the study survey questionnaire, and a subgroup of N = 20 patients completed a semi-structured interview. ARF = acute respiratory failure.

TABLE 1.

Participant Demographics and Illness Characteristics^a

Variables	Cohort Characteristics (N = 56)
Age, yr	55 (36–65)
Male	24 (43)
Self-reported race
Black or African American	28 (50)
White	25 (45)
Otherb	3 (5)
Self-reported Hispanic ethnicity	3 (5)
Area deprivation index of home postal code percentile	55 (33–73)
History of anxiety or depression	14 (27)
Highest level of school completed
Less than high school	7 (13)
High school graduate, general educational diploma, or equivalent	24 (43)
Some college, Associate, or Bachelor’s degree	19 (37)
Graduate-level degree	5 (9)
Medical comorbiditiesc
Anxiety or depression	21 (38)
Chronic pulmonary disease	18 (32)
Asthma	10 (18)
Chronic obstructive pulmonary disease/emphysema	5 (9)
Congestive heart failure	10 (19)
Coronary artery disease	24 (43)
Hypertension	30 (54)
Admitting diagnosis type
Cardiovascular	13 (23)
CNS	1 (2)
Gastrointestinal	4 (7)
Oncology	2 (4)
Respiratory	31 (55)
Otherd	5 (9)
Respiratory support type
High-flow nasal cannula	8 (14)
Mechanical ventilation	43 (77)
Noninvasive positive pressure ventilation	5 (9)
Highest Sequential Organ Failure Assessment in first 24 hr of ICU stay	7 (4–11)
Presence of deliriume	37 (66)
Exposure to any benzodiazepinesf	34 (61)
Lorazepam/midazolam administration, dg	2 (1–4)
ICU LOS, d	7 (3–13)
Hospital LOS, d	14 (8–32)

LOS = length of stay.

^aContinuous data reported as median (interquartile range) and categorical as n (%).

^bIncludes self-reported race of Asian (including Indian), Native Hawaiian or other Pacific Islander, more than one race, unknown, or other.

^cFive most common categories displayed. Full list of assessed comorbidities available in Appendix 1 (https://links.lww.com/CCX/B688).

^dIncludes nephrology/urology, rheumatology, and uncategorized.

^eAssessed as any Confusion Assessment Method positivity during ICU stay.

^fIncludes administration of any lorazepam, midazolam, diazepam, or alprazolam during ICU stay.

^gMedian duration of lorazepam or midazolam administration among recipients (N = 31).

Of 56 survivors, 21 survivors (38%) felt “completely” H&U, 18 survivors (32%) “quite a bit,” seven survivors (13%) “moderately,” seven survivors (13%) “slightly,” and one survivor (2%) “not at all.” Two survivors (4%) selected “I do not wish to answer” for the H&U measure.

Survivors who felt “completely” H&U were older than those who did not (median [IQR] age, 64 yr [50–68 yr] vs. 46 yr [33–58 yr] ; p < 0.01), but did not otherwise differ in measured demographic or illness characteristics (Table 2). Specifically, there was no statistically significant difference in exposure to sedation of any type, exposure to benzodiazepines, or presence of delirium.

TABLE 2.

Participant Characteristics Stratified by Heard and Understood Measure Response^a

Variables	Completely Heard and Understood?
	No (N = 35)	Yes (N = 21)	p
Age, yr	48 (33–59)	64 (53–68)	<0.01
Male	18 (46)	8 (38)	0.78
Self-reported race			0.24
Black or African American	15 (44)	13 (59)
Otherb	2 (6)	2 (10)
White	18 (51)	6 (29)
Self-reported Hispanic ethnicity	1 (3)	3 (14)	0.30
History of depression or anxiety	10 (29)	6 (29)	1
Area deprivation index percentile	57 (35–69)	57 (33–78)	0.47
Highest level of school completed			0.08
Less than high school	2 (6)	5 (23)
High school graduate, general educational diploma, or equivalent	15 (43)	9 (43)
Some college, Associate, or Bachelor’s degree	15 (45)	4 (19)
Graduate-level degree	2 (6)	3 (14)
Respiratory support			0.26
High-flow nasal cannula	4 (11)	3 (14)
Mechanical ventilation	29 (83)	14 (67)
Noninvasive positive pressure ventilation	2 (6)	4 (19)
Highest sequential organ failure assessment score in first 24 hr of ICU stay	8 (4–11)	5 (4–10)	0.24
Exposure to any sedationc	33 (94)	16 (76)	0.12
Exposure to benzodiazepinesd	23 (66)	11 (52)	0.48
Presence of deliriume	24 (69)	13 (62)	0.82
ICU LOS, d	8 (4–15)	6 (3–11)	0.37
Hospital LOS, d	17 (8–39)	12 (8–23)	0.13

LOS = length of stay.

^aContinuous variables reported as median (interquartile range) and categorical as n (%); p values from Wilcoxon rank-sum test for continuous variables and χ² test for categorical variables.

^bOther includes self-reported race of Asian (Including Indian), Native Hawaiian or other Pacific Islander, more than one race, unknown, or other.

^cAssessed for duration of ICU stay; includes exposure to propofol, fentanyl, hydromorphone, lorazepam, midazolam, dexmedetomidine, ketamine, oral opioids, oral benzodiazepines, and other oral agents (hydroxyzine, quetiapine, clonidine).

^dAssessed for duration of ICU stay.

^eAssessed as any documented instance of positive Confusion Assessment Method during ICU stay.

In unadjusted analysis, compared with survivors who did not feel “completely” H&U, those who felt “completely” H&U had lower median [IQR] HADS-anxiety subscale scores (4 [1–6] vs. 7 [5–10]; p < 0.01) and HADS-depression subscale scores (6 [5–10] vs. 10 [7–13]; p = 0.02). They had higher median [IQR] perceived health (80 [70–90] vs. 60 [43–78]; p < 0.01) and health-related quality of life (0.84 [0.62–0.94] vs. 0.55 [0.29–0.72]; p < 0.01). IES-6 posttraumatic stress scores did not differ significantly (9 [8–13] vs. 14 [9–21]; p = 0.07). For reference, the minimally clinically important difference for the HADS-anxiety and depression subscales is 1.7 points, for the perceived health visual analog scale is seven to 10 points, and for the EQ5D utility index is 0.05–0.08 (28, 45–47).

Qualitative Results

The first 27 survey participants were eligible for interview participation; 10 of whom were unable to be reached or scheduled for interview within 6 weeks of hospitalization, two of whom declined, and 15 of whom consented. Among 21 subsequently eligible survey participants, eight were purposively sampled for invitation, five of whom consented to interview. Thematic saturation was achieved after 20 interviews and no further invitations were extended. Interviews occurred a median of 28 days [22–34 d] after discharge. The demographics of interviewed survivors were similar to the overall cohort (eTable 1, https://links.lww.com/CCX/B688).

As perceived by ARF survivors, the experience of patient-centered communication, or “feeling H&U,” could be described by three major themes: interpersonal relationships (respect and empathy), communication content and style (transparency and responsiveness), and care processes influencing communication (healthcare team dynamics and timeliness). The effects of feeling H&U included increased engagement in shared decision-making, trust in the medical system, emotional well-being, and tolerance of uncertainty.

Interpersonal Relationships

Meaningful interpersonal relationships were a core feature of feeling H&U. Specifically, survivors commonly reflected on the development of relationships in which they felt safe, respected, and seen as a person. Clinicians in these relationships were described as kind, empathetic, and attentive; they shared personal stories and spent time at the bedside beyond what was required clinically.

“Always, for each [clinician], the last question was, ‘Is there anything else I can do for you while I’m here?’” (#15)

“Share who you are with the people in ICU. Find out who they are, where they come from, and why are they in the hospital on a human basis. It’s not your... white gown that is important. It’s who’s inside of it.” (#7)

Although verbal communication was important, nonverbal communication was also foundational to the development of meaningful interpersonal relationships. Survivors favorably recalled clinicians crying with them or holding their hand. In addition, survivors noted that inclusion of their family was essential to their experience of feeling H&U.

“[My wife and mother] stood... singing Bohemian Rhapsody to me... And my mother told me there were times where the doctor came to give updates, and he waited outside the door for them to finish... stuff like that you don’t see on television.” (#11)

Communication Style and Content

Survivors valued transparent, honest, and nonjudgmental communication, in which all parties could be open about their perspectives. Specifically, patients’ ability to share fears and admit vulnerability increased the perception of communication as patient-centered. In addition, conversational responsiveness characterized by appropriate follow-up questions and demonstrated interest in the patient perspective were important.

“[Participant:] I’m tall, like 6’ 2”. And sometimes, I think, people look at me like this big, strong man... like I ain’t supposed to never have nothing wrong with me... [Interviewer:] Even the doctors looked at you that way? [Participant:] Yeah, sometimes I felt that way.” (#17)

When receiving medical information, survivors preferred direct communication about possible outcomes and about the presence of uncertainty. The use of medical jargon, frequently changing care plans, and decisions made without survivor involvement could leave survivors not feeling H&U.

“It was helpful to know that they were trying to keep me in the loop of everything... It was very comforting to know that they were trying to come to some conclusion of what was wrong with me” (#10)

Importantly, patient-centered communication was defined not only by the communication instance itself but by its outcomes. Survivors described seeking behavioral responsiveness demonstrated by observable changes and reported that such responsiveness was frequently absent.

“The only way I can respond to [if I felt heard and understood] is, were changes made after that discussion? And the answer is, no. Nothing substantive... changed as a result of my discussions.” (#7)

Care Processes

Survivors were highly observant of the ways in which care processes and the larger healthcare system context influenced communication. They demonstrated keen awareness of healthcare team dynamics, noting positive experiences related to collaboration and effective information sharing, and negative experiences related to frequently changing teams, conflicting information, and lapses in information transfer.

“Because my care normally involves multiple teams, there’s oftentimes a miscommunication between [teams]... I think coordination is hard for them.” (#4)

“Some doctors had answers, some didn’t. Some had to talk to somebody else, and at times it was like, “Are y’all even talking to each other?” (#5)

The timeliness of care and communication was central to many survivors’ experience of feeling H&U. Specifically, some patients noted that they were less receptive to communication early in the arc of their recovery when they were experiencing lingering cognitive effects of critical illness and, in some cases, sedation. In addition, communication occurring at odd hours of the day, such as very early in the morning, was poorly received. Distinct from verbal communication, instances in which physical needs related to sleep or hygiene were not addressed in a timely fashion frequently arose as examples of not feeling H&U.

“The doctors come in real early in the morning. 2 or 3 of them at a time... and they say, “How are you doing? How are you feeling? Goodbye! See ya! Boom!” (#7)

“I was dry heaving over and over and over, and I’m hitting the button, calling for a nurse... Well, they’re out there... They hear me throwing up, but they’re just sitting in a chair doing nothing.” (#2)

Effects of Feeling Heard and Understood

In describing the experience of feeling H&U, survivors often reflected on the effects of patient-centered communication. First, they noted that feeling H&U was associated with positive emotional states, describing sensations of comfort, happiness, and peace. Such sensations impacted their mindset about recovery, provided hope and encouragement when things were difficult, and increased their trust in the medical community.

“It made me happy that... I’m being heard, that these doctors are trying to understand me and trying to make me feel comfortable” (#16)

“My dad always said, “Hospitals are shit. You don’t want to go to hospital.”... So every time I got sick, I just dealt with it... [Now], if anything occurred again, I would have a hundred percent faith in the doctors, trust in them.” (#2)

Survivors also noted that shared decision-making was augmented in the presence of patient-centered communication, as a result of their feeling informed and that personal preferences had been considered. Feeling H&U specifically helped normalize and alleviate uncertainty.

“Regardless of what the emotional impact was, they armed me with enough information where I could make a… sound decision.” (#11)

Additional representative quotations in Table 3.

TABLE 3.

Additional Representative Quotations of Major Themes^a

Major Theme	Feeling Heard and Understood	Not Feeling Heard and Understood
Interpersonal relationships (respect and empathy)	“[Feeling heard and understood] makes me feel like I’m a person and not just a statistic… or ‘patient number 374’... Their care is very genuine.” (#15)	“Doctors feel like you may or may not understand… I work really hard to know what’s going on with my health... So, for me, it’s a little frustrating.” (#4)
	“It was just incredible to see the relationship that [my husband] was able to build with the nurses and the doctors... I think that that really attests to how well they took care of him.” (#9)	“She talked in a nasty tone of voice and ridiculed me.” (#15)
Communication style and content (transparency and responsiveness)	“I’m all open with the doctors, because if they don’t know, then how are they gonna help me, or understand what I feel?” (#20)	“There were just so many things that they were throwing at me, and then none of them were true... that just had me overwhelmed because I didn’t know what were they gonna do…” (#10)
	“[This doctor] would leave you and go out and fix the issue. And if... when he stopped in to see you again that issue’s still not fixed, he was on it again.” (#19)	“They were surprised that I wasn’t sleeping in the bed. And they really didn’t respond... ‘Why? What can we do to help?’ They weren’t as responsive as they could have been.” (#7)
Care processes (healthcare team dynamics, timeliness of care)	“She was giving us information on what she assumed would happen, and then... would follow up with someone to make sure... It’s nice that everybody has information, or at least someone to go back to, to get accurate information” (#9)	“I had one young lady, while I was asleep, try to come in and draw blood... And I woke up startled, because someone’s trying to prick me with a needle at 2 in the morning while I’m asleep!” (#11)
	“They were always conferring to each other, ‘What was this? What can we do with the medication?’ They bounce ideas off each other, and experiences. I thought that was a very good thing.” (#12)	“There were so many doctors... you’re telling me one thing and then another doctor was telling me something else...” (#8)

^aQuotations categorized as positive and negative examples of themes related to the experience of feeling heard and understood based on interview context. This categorization is distinct from responses to the quantitative “H&U” measure.

DISCUSSION

This study aimed to understand ARF survivors’ perspectives on patient-centered communication and its relationship with inpatient psychologic well-being, a risk factor for post-ARF mental health disorders. We found that communication that was perceived as patient-centered was rooted in the development of respectful and empathetic interpersonal relationships, was characterized by transparency and responsiveness, and required awareness of multidisciplinary healthcare team dynamics. ARF survivors connected patient-centered communication with increased positive emotions, reduced uncertainty, and empowered shared decision-making.

In exploratory quantitative work, we found that only 40% of ARF survivors felt “completely” H&U. In unadjusted analysis, feeling “completely” H&U was associated with lower postsdischarge anxiety and depression scores, and higher perceived health and quality-of-life scores.

This study complements existing literature in several ways. It has recently been recognized that the inpatient psychologic experience may contribute to long-term mental health symptoms after ARF (15). Our findings suggest that patient-centered communication may be an important mediator of this relationship by modifying patients’ emotional state, levels of uncertainty, and feelings of inclusion in decision-making. The finding that these characteristics are modifiable by inpatient clinician communication highlights a potentially impactful opportunity for intervention in the context of increasingly complex medical care and decreasing healthcare trust (48–50).

ARF survivors identified multiple factors relevant to patient-centered communication outside the classic definition described by Epstein et al (51). Rather than focusing on patient-clinician interactions, ARF survivors emphasized a global experience of interaction with the healthcare system. Regardless of how well individual clinicians delivered information, negative sleep or hygiene experiences, competing obligations limiting healthcare team responsiveness, and poor inter-clinician communication hindered survivors from feeling H&U. This mirrors previous work linking environmental and organizational factors to the inpatient experience of dignity (52) and suggests the need for recognition of the complex contributors to successful inpatient communication.

Beyond expanding our understanding of what constitutes patient-centered communication, our findings demonstrate that survivors retrospectively assess the patient-centeredness of communication holistically across a hospitalization. They commented on how communication was implemented across the many teams they interacted with from the time of admission through to discharge, rather than focusing on the days spent in the ICU. This is important because few published interventions target inpatient psychologic well-being in critically ill patients; only two address the transition between the ICU and other inpatient spaces (53–57).

Regarding the relationship between perceived inpatient communication quality and psychologic well-being after hospitalization, our exploratory quantitative analysis suggests an association between feeling H&U and improved mental health at discharge, present across multiple measures. This finding aligns with work demonstrating that empathetic communication may reduce anxiety and stress in the clinical setting, both of which represent precursors to posthospitalization mental health symptoms (18–23). In the last decade, increasing attention has been turned to examining the efficacy of curricula and tools geared at improving the quality of inpatient communication (58–61), though the optimal content and timing of such interventions require further study (62).

Three novel areas for investigation and intervention are suggested by the current study. First, while empathetic verbal communication was important, survivors reported poor behavioral responsiveness and follow-through on conversations, suggesting a need to understand clinician-level barriers to aligning talk with action. Second, we found that as survivors transitioned across multiple hospital units and provider teams, they often witnessed poor team communication. Future research on ICU team dynamics, a burgeoning area (63), could examine the intersection between medical team dynamics and patient-team relationships. Finally, inadequate attention to basic hygiene and physical needs undermined patients from feeling H&U. Improved staffing models enabling timely care may improve patient well-being.

Limitations

This is the first study linking ARF survivor perspectives on patient-centered communication with a validated quantitative metric, the H&U measure (64–66). Although an effective starting point, we found that interviewed survivors often shared both positive and negative communication experiences regardless of measure response (eTable 2, https://links.lww.com/CCX/B688). This measure has not been specifically validated in the ARF population, and our experience suggests some considerations. First, given the complexity of the inpatient ARF experience, communication quality varies over time, and repeated measurement may be beneficial. Second, communication experiences were anchored in expectations; evaluating previous healthcare experiences may add nuance when assessing patients’ perceptions of “optimal” communication. Third, survivors often raised concerns related to physical experiences and hospital logistics, not verbal interactions alone; pairing this measure with domain-specific follow-up questions may be beneficial.

Our study has additional limitations. One is the high rate of non-enrollment within the required 30-day post-discharge window (168/223 eligible individuals, 75%). Nonenrollees may have been healthier, having returned to work or normal routines, or, conversely, more burdened by illness. Negative experiences during hospitalization could have also contributed to disengagement. It is difficult to hypothesize about the magnitude and direction of any resulting selection bias. Ultimately, our small sample size limits characterization of associations between the H&U measure and mental health and quality of life outcomes adjusted for relevant confounders. Also, our study was cross-sectional, and prospective work is needed to understand the directionality of the relationship between perceived inpatient communication quality and postdischarge outcomes.

Furthermore, our qualitative work occurred on average one month after hospital discharge and may be subject to recall bias. Our interviews included two patients whose maximum respiratory support modality was high-flow nasal cannula; exploring differences in the perspectives of patients treated with mechanical ventilation vs. noninvasive respiratory support was outside of the scope of this study.

CONCLUSIONS

Patient-centered communication may improve psychologic well-being in ARF survivors while admitted and after discharge. Interpersonal relationships, communication content and style, and care processes are all important contributors to the perception of communication as patient-centered by ARF survivors. An additional prospective, longitudinal study is needed to confirm these findings and better understand the relationship between inpatient communication and post-discharge mental health.

ACKNOWLEDGMENTS

We acknowledge the work of the Investigating Dyadic Expectations about Acute Respiratory Failure Survivorship (IDEAS) study team: Omar Valentin, Rochelle Hall, Deanna Rahman, Milagros Sambula, Victoria Grimaldi, and Ruth Alozie.

Supplementary Material

cc9-8-e1492-s001.pdf ^{(70.9KB, pdf)}