Emotional Distress, Social Support, and Functional Dependence Predict Readiness for Hospital Discharge in a Prospective Sample of Cognitively Intact Cardiac Arrest Survivors

Abstract

Aim:

To inform screening, referral and treatment initiatives, we tested the hypothesis that emotional distress, social support, functional dependence, and cognitive impairment within 72 hours prior to discharge predict readiness for discharge in awake and alert cardiac arrest (CA) survivors.

Methods:

This was a secondary analysis of prospective single-center cohort of CA survivors enrolled between 4/2021 and 9/2022. We quantified emotional distress using the Posttraumatic Stress Disorder Checklist-5 or PROMIS Emotional Distress – Anxiety and Depression Short Forms 4a; perceived social support using the ENRICHD Social Support Inventory; functional dependence using the modified Rankin Scale; and cognitive impairment using the Telephone Interview for Cognitive Status. Our primary outcome was readiness for discharge, measured using the Readiness for Hospital Discharge Scale. We used multivariable linear regression to test the independent association of each survivorship factor and readiness for discharge.

Results:

We included 110 patients (64% male, 88% white, mean age 59 [standard deviation ± 13.1 years]). Emotional distress, functional dependence, and social support were independently associated with readiness for discharge (adjusted β’s [absolute value]: 0.25–0.30, all p < 0.05).

Conclusions:

Hospital systems should consider implementing routine in-hospital screening for emotional distress, social support, and functional dependence for CA survivors who are awake, alert and approaching hospital discharge, and prioritize brief in hospital treatment or post-discharge referrals.

Introduction

Cardiac arrest (CA) survivorship refers to the long-term health and well-being of survivors. For many survivors, this includes the complex, enduring experience of post-CA emotional, social, functional, and cognitive sequelae.¹

In the United States system of care, patients resuscitated from CA are typically admitted to the intensive care unit (ICU) for management of hypoxic-ischemic brain injury and/or extracerebral organ failures. Survivors of early post-CA critical illness are then transferred out of the ICU setting where additional medical care is offered to address the causes and lingering consequences of CA, secondary prevention, and discharge planning. Because of pressures to shorten length of stay during the acute hospitalization, patients are serially assessed for readiness for discharge and are often transferred to a post-acute setting with significant variability (home, inpatient rehabilitation, skilled nursing facility or long-term acute care facility)^2,3 while functional and cognitive recovery are still ongoing. Throughout the acute hospitalization, there is no standardized process of providing information on CA survivorship, anticipatory guidance, or outpatient specialty referral (e.g., emotional, social, physical, cognitive) for post-CA support.¹ This differs from the support received by other populations that experience a sudden catastrophic event with chronic, multi-domain sequelae (e.g., myocardial infarction, stroke).¹ As result, many CA survivors report feeling unprepared to manage survivorship challenges – particularly due to lack of anticipatory guidance and resources for addressing their post-CA sequelae prior to discharge.^4–9 To address this, both United States-based and international organizations have called for standardization of in-hospital screening, referral to post-acute treatment, and post-discharge surveillance.^1,10–13

Despite efforts to allocate in-hospital resources to improve survivorship, there are a lack of data that describe the impact of specific domains of survivorship (e.g., emotional, social, functional, cognitive) on readiness for hospital discharge. In two prior studies, psychological symptoms, but not functional or cognitive impairment, predicted survivors’ responses to a single-item measure of perceived global recovery both at discharge and 6-month follow-up.^14,15 However, no study has examined the association of each survivorship factor with a robust measure of readiness for discharge.

To address this gap, we tested the hypothesis that emotional distress, social support, functional dependence, and cognitive impairment each predict readiness for hospital discharge in a sample of consecutively enrolled, awake and alert CA survivors approaching hospital discharge within 72 hours. Identifying which factors impact CA survivors’ readiness for discharge can lead to more specific approaches to in-hospital screening, treatment and referral to post-acute treatment.

Methods

We conducted a secondary analysis of a prospective study examining emotional distress and CA survivorship outcomes in CA survivors treated at a single, high-volume CA center in Pennsylvania (University of Pittsburgh Medical Center Presbyterian Hospital) and by a multidisciplinary post-arrest specialty consult service (the University of Pittsburgh Post-Cardiac Arrest Service; PCAS) that lasted between April 2021 and September 2022.¹⁶ Details on PCAS and enrollment methods have been previously published.¹⁶ Briefly, the PCAS provides care to CA patients from initial resuscitation, up through discharge and post-acute recovery. Trained PCAS clinicians or staff administered measures as part of a push by the PCAS to assess need for additional resources at the time of discharge. We concurrently planned to analyze the new data being collected.

As CA survivors approached hospital discharge (within 72 hours) a PCAS clinician or trained staff provided discharge information and resources pertinent to CA survivorship. It was during this visit that they also administered study measures.

We included survivors who were: 1) awake and alert, as determined by the medical team; 2) English-speaking adults (18 years or older); and, 3) being discharged from their index CA admission. The parent study was approved by the University of Pittsburgh Human Research Protection Office (Reference #: STUDY19020205) under a waiver of informed consent (minimal risk).

We considered the following demographic and clinical characteristics as potential predictors of readiness for discharge: age, sex assigned at birth, race, ethnicity, household income, education level, CA location, presenting rhythm, Cerebral Performance Category score,¹⁷ and pre-arrest medical comorbidities (summarized using Charlson Comorbidity Index (CCI)).¹⁸ We dichotomized the following variables for our regression analyses: race/ethnicity - non-Hispanic White vs Other, as most of the sample was non-Hispanic White; household income - low (<$50,000) vs medium–high (>$50,000); education - completed college vs never completed college; initial rhythm – shockable vs non-shockable.

Predictors of Interest

We assessed the following domains of survivorship: emotional distress, social support, functional status, and cognitive function. In general, in our parent study we opted to include brief measures with good psychometric properties to simultaneously reduce burden on patients and promote rigor of assessment.

Emotional Distress

We measured emotional distress via the PROMIS Emotional Distress Anxiety and Depression Short Forms 4A^19,20 and the Posttraumatic Stress Checklist-5 (PCL-5).²¹ Both PROMIS measures are 4-item scales (range 4–20); higher scores reflect greater levels of anxiety and depression. The PROMIS short form items are taken from their respective PROMIS items banks that have shown good clinical validity in other samples with chronic medical challenges, such as heart failure, cancer, chronic back pain, and obstructive pulmonary disease.²² Further, the two short forms used here have demonstrated good internal reliability and strong convergent validity in a sample of patients with chronic musculoskeletal pain.²³

The PCL-5 is a 20-item scale (range 0–80); higher scores correspond to greater posttraumatic stress severity. We chose the PCL-5, despite being lengthier than the other measures in our battery, because it captures all four DSM-5 posttraumatic stress symptom dimensions,²⁴ which have previously shown to be differentially important in characterizing posttraumatic stress post-CA.²⁵ Along with its prior use in CA samples,^25–27 the PCL-5 is one of the most widely used measures of posttraumatic stress, and a recent systematic review of its psychometric properties across 56 studies found good internal consistency, construct validity, and criterion validity.²⁸

Social Support

We measured perceived social support using the Enhancing Recovery in Coronary Heart Disease Patients (ENRICHD) Social Support Inventory (ESSI).²⁹ The ESSI is a 7-item scale (range 8–34). Higher scores reflect greater levels of perceived social support. The ESSI was developed for the ENRICHD intervention used items from well-validated scales of social support that individually predict mortality in cardiac patients.³⁰ The ESSI was found to have good reliability and validity with the ENRICHD sample and in other cardiac patients.^30,31

Functional Status

We assessed functional status via the modified Rankin Scale (mRS).³² This ordinal scale ranges from 0 (no symptoms) to 6 (death), although by study design no subject could have an mRS = 6. A score of 3–5 corresponds to significant functional impairment. The mRS is one of the most widely used measures of functional outcome in neurocritical care³³ and CA³⁴ and has shown good reliability that increases when paired with structured interviews.³⁵

Cognitive Function

We assessed cognitive function via the Telephone Interview for Cognitive Status (TICS).³⁶ The TICS scores range from 0–51; lower scores indicate greater cognitive impairment. A score of ≤30 corresponds to global cognitive impairment. The TICS has been used in neurocritical care³⁷ and CA patients^15,38,39 and has demonstrated good reliability and convergent validity with post-stroke patients,^40,41 and high diagnostic validity for dementia.⁴²

Primary Outcome

We assessed readiness for discharge using the Readiness for Hospital Discharge Scale Short Form (RHDS).⁴³ The RHDS evaluates survivors’ preparedness and confidence in managing their healthcare needs upon discharge using an 8-item scale. Subscale scores are averaged for a mean RHDS score (range 1–10). Higher scores correspond to greater readiness for discharge. The RHDS includes 2 items from each subscale of the RHDS full scale (personal status, knowledge, coping ability, and expected support) that has shown excellent reliability and predictive validity of post-discharge utilization of social support in adult medical and surgical patients.⁴⁴

Analysis

We report brief descriptive statistics previously calculated for the parent study¹⁶ and calculated mean and standard deviation of each key predictor and outcome. We treated emotional distress as a categorical variable (positive screen on PROMIS anxiety (>8), PROMIS depression (>8), or PCL-5 (≥31)), social support as continuous, functional dependence as categorical (Rankin ≥3), and cognitive impairment as categorical (≤30 on TICS). We treated readiness for discharge as a continuous outcome.

Model building

We tested the hypothesis that each survivorship domain (emotional distress, social support, functional dependence, and cognitive impairment) independently predicted readiness for discharge.

First, we calculated unadjusted associations between each candidate covariate (i.e., all those listed above and presented in Table 1) and readiness for discharge, using Pearson correlation for continuous predictors and linear regression for categorical predictors. In our final multivariable model, we included demographic and clinical predictors associated with readiness for discharge at a threshold of p < 0.10, as well as each survivorship factor.

Table 1.

Demographic and cardiac arrest survivor characteristics

Demographics	Survivors (N = 110)	Scale Range
Age, mean ± SDa	59.2 ± 13.1
Sex – Female, % (n)	37.3 (41)
Race, % (n)
-Non-Hispanic White	88.2 (97)
-Black	8.2 (9)
-Asian	1.8 (2)
-Multi-racial	0.9 (1)
-Other	0.9 (1)
Income, % (n)
-Low (<$50,000)	48.2 (53)
-Medium ($50,000 - $99,999)	29.1 (32)
-High (>$99,999)	20.9 (23)
-Prefer not to disclose	1.8 (2)
Education, % (n)
-Did not complete HS	9.1 (10)
-Completed HS or GED	36.4 (40)
-Some college	22.7 (25)
-Completed college	23.6 (26)
-Graduate degree	8.2 (9)
Survivor Characteristics
Pre-arrest Charlson Comorbidity Index, median (IQRb)	2 (1–3)
Out-of-hospital arrest, % (n)	78.2 (86)
Initial rhythm, % (n)
-VT/VF	62.7 (69)
-PEA	21.8 (24)
-Asystole	9.1 (10)
-Unknown	5.5 (6)
Survivorship Factors
PTSD score,c mean ± SD	11.3 ± 9	0–80
PTSD screen,d % (n)	6.4 (7)
Anxiety score,e mean ± SD	6.7 ± 3.4	4–20
Anxiety screen,f % (n)	32.7 (36)
Depression score,g mean ± SD	5.9 ± 2.9	4–20
Depression screen,h % (n)	20.9 (23)
Emotional distress screen,i % (n)	36.4 (40)
Modified Rankin scale, % (n)		0–6
- 0	18.2 (20)
- 1	34.5 (38)
- 2	16.4 (18)
- 3	24.5 (27)
- 4	6.4 (7)
Functional dependence,j % (n)	30.9 (34)
CPC score, % (n)		1–5
- 1	53.6 (59)
- 2	17.3 (19)
- 3	29.1 (32)
Perceived social support,k median (IQR)	33 (30.25–34)	8–34
Cognitive function,l mean ± SD	32.7 ± 5.6	0–51
Cognitive impairment,m % (n)	24.5 (27)
Readiness for discharge,n mean ± SD	7.3 ±1.7	1–10

Note:

^{a =}Standard deviation

^{b =}Interquartile range

^{c =}Posttraumatic stress measured via Posttraumatic Stress Disorder Checklist-5 (PCL-5)

^{d =}Positive PTSD screen = PCL-5 ≥31

^{e =}Anxiety measured via PROMIS Emotional Distress - Anxiety Short Form 4a

^{f =}Anxiety screen = PROMIS Anxiety score >8

^{g =}Depression measured via PROMIS Emotional Distress - Depression Short Form 4a

^{h =}Depression screen = PROMIS Depression score >8

^{i =}Positive screen on any emotional distress measure

^{j =}Functional dependence defined as modified Rankin score ≥3

^{k =}Perceived social support measured via ENRICHD Social Support Instrument

^{l =}Cognitive function measured via Telephone Interview for Cognitive Status (TICS)

^{m =}Cognitive impairment defined as TICS score ≤30

^{n =}Readiness for discharge measured via Readiness for Hospital Discharge Scale

We conducted sensitivity analyses treating emotional distress and cognitive function scales a continuous predictors. To reduce multicollinearity between emotional distress measures, we conducted three sensitivity models; one for each emotional distress measure (PROMIS anxiety, PROMIS depression, and PCL-5) with all other predictors/covariates.

Results

Cohort Characteristics

We enrolled 110 survivors and summarize cohort characteristics in Table 1, most of which have been published previously.¹⁶ Our cohort was on average middle aged (59.2 years, standard deviation ± 13.1), majority male (62.7%), and predominantly non-Hispanic White (88.2%). Nearly half the sample reported a low (<$50,000) household income (48.2%) and an education status of high school diploma or less (45.5%). In terms of clinical characteristics, survivors generally had low comorbidity burden (CCI 2, interquartile range 1–3), had an out-of-hospital CA (78.2%), and had an initial shockable rhythm (62.7%). Regarding survivorship characteristics, over a third had clinically significant emotional distress (36.4%), a similar proportion were functionally dependent (30.9%), one quarter exhibited cognitive impairment (24.5%), though the median reported social support score was high (33 on scale of 8–34; interquartile range 30.25–34).

Model Building

With regard to our consideration of covariates, only out-of-hospital CA location and CCI were associated with readiness for discharge and included in final our adjusted model. No other demographic or clinical characteristic was associated with readiness for discharge.

Predictors of Readiness for Discharge

Our hypothesis was partially supported, such that in our final multivariable model, clinically significant emotional distress (β = −0.25, p = 0.007), functional dependence (β = −0.30, p = 0.003), and social support (β = 0.26, p = 0.02) were associated with readiness for discharge (Table 2) but not cognitive impairment. Neither out-of-hospital CA or CCI were associated with readiness for discharge.

Table 2.

Associations between cardiac arrest survivorship factors and readiness for discharge.

Variable	β (p-value)
Emotional distress	−0.25 (0.007)
Functional dependence	−0.30 (0.003)
Social support	0.26 (0.02)
Cognitive impairment	0.03 (0.75)
Out-of-hospital CA	0.10 (0.27)
Charlson comorbidity index	0.10 (0.27)
R2	0.33

Bold = significant at p < 0.05

Sensitivity analyses showed that anxiety, social support, and functional dependence were individually associated with readiness for discharge (Supplemental Tables 1–3).

Discussion

Our findings partially support our a priori hypothesis in that emotional distress, social support, and functional dependence – though not cognitive impairment - predicted CA survivors’ readiness for discharge. Our results reinforce prior calls for multimodal in-patient assessment and referral to post-acute treatment for CA survivors,^1,10–13 and provide quantitative support to prior qualitative studies^4–9 by identifying three factors that carry particular importance predicting preparedness for survivorship.

Our findings contribute to literature emphasizing need for psychosocial support after CA.^{1–15,25–27,45–50} At least one third of CA survivors experience emotional distress,¹ which is associated with worse clinical and patient-centered outcomes.^{25,26,45–50} As recommended by several professional organizations, hospital systems should consider adopting routine emotional distress screening, surveillance, and treatment referral to prevent these poor outcomes downstream.^1,10–13 Further, by developing community and social care systems and programs in post-CA care, some of the burden can be lifted from hospitals, such as early psychosocial interventions that can teach survivors and members of their support systems how to navigate anticipated post-CA challenges, community support groups, and peer lead support groups. Finally, hospitals and community organizations can collaborate to refer new survivors to the various support organizations already in existence, including those online (e.g., Sudden Cardiac Arrest UK) and any local groups.

Unsurprisingly, functional dependence also predicted readiness for discharge. Indeed, professional organizations have called for multimodal rehabilitation assessment and treatment in-hospital, including cardiac, physical, occupational, and speech rehabilitation.^1,10–13 Additionally, hospital systems should provide appropriate referrals and recovery expectations for ongoing functional challenges before discharge.^1,4 They should also include family members and/or other support people in treatment planning to identify areas they can provide instrumental support to the survivor.^1,4

Our hypothesis that cognitive impairment would predict readiness for discharge was not supported. While surprising, this is consistent with two prior studies that did not find an association between cognitive impairment and CA survivors’ perceived recovery at discharge or at 6-month follow-up.^14,15 There may be several reasons for these findings, for example, survivors may lack insight into their cognitive changes and their impacts, as has been proposed before.^15,51 As such, to gain a more comprehensive understanding of a survivor’s readiness for discharge and adapting to survivorship, future studies can consider examining survivors’ insight, the impact of cognitive impairments on functioning once survivors have returned home and are no longer in the protected hospital environment (and cognitive load is greater), caregivers’ perspectives of survivors’ functioning, and caregivers’ own readiness for caregiving. Another consideration may be that the survivors the cognitive impairments experienced in this particular cohort may not have been severe, and thus may have had a lower impact on our outcome. That said, it may still be important to work with families and provide them with resources prior to discharge (e.g. verbal and written expectations for cognitive changes, instructions for seeking care should cognitive changes manifest, referral etc.) to safeguard against any negative surprises that survivors and families may encounter when they return home.⁴ It is clear from qualitative studies that cognitive changes are a stressor for survivors and family members,^4,5,7 and as such, it will be important to consider comprehensive screening methods (again, including insight and perspective of caregivers) and timing of screening going forward.

Limitations

We reported many of our parent study’s limitations previously.¹⁶ While our sample was regionally representative it lacked racial and ethnic diversity, which limits generalizability. Further, our sample was cared for by a unique post-cardiac arrest service that provided information, resources, and anticipatory guidance throughout their inpatient stay, which is not the norm, and which may have led survivors to feel more prepared than the “typical CA survivor.” Conversely, our study took place entirely during the COVID-19 pandemic, which impacted patients’ experiences compared to what may have been typical before the pandemic. Taken together, both limitations underscore the need to replicate our findings with different populations and temporal contexts. That said, an important strength of our study was our ability to feasibly collect our measures despite the ongoing pandemic. Finally, even after considering clinical characteristics, social determinants of health, and four of the most often reported survivorship challenges,^1,4–9 our main analysis explained a third of the variance in readiness for discharge, suggesting a need to identify other factors associated with our outcome. Potential variables to consider for future inquiries may include quality of patient-provider interactions, presence of and quality of relationship with a caregiver, and insight into cognitive changes as discussed above.

Conclusion

Emotional distress, social support, and functional dependence predicted readiness for hospital discharge in awake and alert CA survivors. Hospitals should consider providing routine assessment and referral/treatment for psychosocial and functional problems prior to discharge and ongoing surveillance post-discharge to offset poor clinical and patient-centered outcomes downstream.

Highlights.

• Emotional, social, and functional status predicted readiness for hospital discharge.

• Demographic and clinical variables did not predict readiness for discharge.

• Hospitals should prioritize screening and treatment of psychosocial factors.