Feasibility of a brief, in-patient coping and sleep intervention to reduce potentially preventable readmission of cardiac patients to hospital

Background Psychological distress is prevalent amongst hospital in-patient and may predispose patients to potentially preventable readmissions after discharge. A particularly vulnerable group are patients with cardiac disorders. This study tested the feasibility of a brief cognitive behavioural therapy consisting of an in-hospital coping session and a post-discharge healthy sleep session. Methods Standardised questionnaire were used to assess sleep, coping/distress and wellbeing at baseline (pre-intervention) and one-month post-discharge (post-intervention). Treatment fidelity and acceptability were assessed at follow-up. Participants included 72 inpatients admitted with a cardiac disorder or reported to have a cardiac problem whilst in hospital from a single Australian public hospital. Results Most (83 %) participants found the intervention helpful/very helpful. At baseline prior to admission, almost half of participants (46 %) reported poor wellbeing, 19 % high levels of distress and poor coping, and 47 % sleeping less than 7 h per night. Following the intervention, 45 % of participants with poor wellbeing at baseline had reliable change in wellbeing at follow-up. Conversely, only 22 % of patients with high levels of coping/distress at baseline demonstrated improved coping/distress at follow-up suggesting smaller gains. On average a large 43 min gain in sleep duration was observed post-treatment in patients with poor sleep at baseline. Fourteen percent of participants were readmitted to hospital within 34-days of discharge. Conclusions The coping and sleep intervention was well received with positive outcomes in patients especially those reporting high levels of distress for sleep and to lesser extent coping and wellbeing. Future studies to assess the efficacy of the brief intervention at reducing hospital readmissions are needed.


Introduction
Hospital readmission rates are an important indicator of care [1,2] and account for a significant proportion of hospital expenditure [3].In Australia, for example, preventable hospital readmissions contribute to 6.6 % of all hospitalisations [4].Readmissions are considered to be clinically related to a prior admission and therefore potentially preventable, " … if there was a reasonable expectation that it could have been prevented by one or more of the following: 1) the provision of quality care in the initial hospitalisation, 2) adequate discharge planning, 3) adequate post-discharge follow-up, or 4) improved coordination between in-patient and outpatient health care teams" [5, p. 76].Preventable readmissions are associated with service delivery factors including incomplete treatment, inadequate treatment of the underlying problem, poor coordination of services at the time of discharge and inadequate access to care [6,7].They are also associated with individual factors and, in particular, psychological distress at discharge [8][9][10].Prominent among those vulnerable to psychological distress are patients with cardiac disorders [11][12][13].Patients with cardiac disorders also have a high rate of readmission varying up to 23 % for percutaneous coronary intervention and up to 57 % for heart failure [14,15].In patients with cardiac disorders, psychological interventions are reported to increase treatment adherence [16][17][18][19] and reduce preventable hospital readmissions [9,20,21].However apart for nurse-led programs providing in-hospital education (e.g., the disease process, early symptom recognition, self-care strategies, medication management, nutrition, and lifestyle modification) [22], trials specifically testing psychological interventions have been limited and where programs have been implemented they have been delivered post-discharge [23][24][25].Early intervention is recommended for most psychology problems [26].This raises the possibility that including an intervention arm while patients are in hospital may allow for further gains, but the feasibility of an in-hospital program remains to be tested.
Psychological distress is caused by problems in any one or more of the biopsychosocial determinants of health and wellbeing: healthy environments, developmental competencies (healthy identity, emotional and behavioural regulation, interpersonal skills, problem-solving skills), sense of belonging, health behaviours (sleep, nutrition, exercise), coping, resilience (perception of innate resilience) and illness (physical and psychological) [27,28].The two proximal domains to high levels of distress-sleep and coping-may be potentially helpful intervention targets to improve mood regulation and consequently improve treatment adherence and health outcomes in patients with cardiac disorders.Patients with cardiac disorders typically report poor sleep [9,[29][30][31][32] and problems with psychological adjustment, i.e. coping [33][34][35].They also respond well to cognitive behavioural programs delivered post-discharge at improving sleep [36,37] and coping [38,39].Validated brief interventions have been developed for sleep hygiene [40,41] and coping [27,42], which could be adapted for in-patient delivery in patients with cardiac disorders.
In summary, the primary aim of this study was to assess the feasibility of a brief two-session (1 x in-hospital and 1 x post-discharge) psychological intervention to improve wellbeing, coping and sleep and, thereby, reduce distress and preventable hospital readmissions.The secondary aim was to examine the impact of the psychological intervention on wellbeing, coping, self-efficacy, self-reported sleep quality/ quality and help seeking intentions.

Study design
This study used a pre-test post-test design with a 1-month follow-up.

Participants
Participants were 72 inpatients admitted for a cardiac disorder or were reported to have cardiac problems whilst in hospital secondary to other medical issues.Patients were recruited from three wards in an Australian public hospital.Inclusion criteria were 18 years or older, medically stable, and able to provide informed consent and participate.Exclusion criteria were mental status impairment and poor English language skills.

Measures
Demographics.Demographics collected were date of birth, sex, and whom the patient lived with (spouse or partner; family members; friends; nursing home; alone; other).
The primary outcomes are feasibility and acceptability of the intervention, explored through rate of recruitment, intervention implementation, treatment fidelity, readmissions, retention and perceived usefulness of the intervention.Acceptability was assessed with a question about use and how helpful the intervention was, and whether participants would recommend it to other patients.The treatment-use questions were: 1) Did you use your coping plan?And if so, 2) was the plan effective; and 3) Did you edit your plan?; and 4) Did you do anything differently after the sleep intervention?
Treatment fidelity.The project coordinator completed treatment integrity checklists after each treatment component using treatment integrity checklists-10 items for Care Collaborate Connect and eight items for the sleep intervention.Intervention steps were rated as not done (0), attempted, but some elements were missing (1) or delivered competently (2).The total competency scores ranged from 0 to 20 for Care Collaborate Connect and 0 to 16 for the sleep intervention.
Reliability of ratings of the need for additional professional support to cope in Care Collaborate Connect part of the intervention was compared against needs determined from self-reported unhealthy coping subscale of the Coping Index.
The secondary outcomes were wellbeing, coping, coping selfefficacy, self-reported sleep quality and quantity and help-seeking intentions.
Wellbeing.The WHO-5 [43], a 5-item measure of subjective wellbeing, was used as it is not susceptible to the reliability and validity limitations of negative psychological states, such as psychological distress [44,45].Responses are rated on a 6-point Likert-type scale, all of the time (5), most of the time (4), more than half of the time (3), less than half of the time (2), some of the time (1), and at no time (0).The total score ranges from 0 to 25, with scores lower than 13 indicating poor wellbeing.The WHO-5 had excellent internal consistency in this sample, α = 0.90.
Coping/distress.The Coping Index (CI) [46] is a 20-item measure of healthy and unhealthy coping strategies.The measure is based on the Health Theory of Coping [27].Each item is rated on a 4-point Likert-type scale: I don't do this at all (0), I do this occasionally (1), I do this often (2), and I do this most of the time (3).The healthy and unhealthy subscale scores are the average of the ten scale item scores and range from 0 to 3. Unhealthy scores greater than 0.75 are excellent predictors of distress indicative of psychiatric illness and need for additional professional support [47] and this criterion was used to identify 'high' distress participants.
Coping self-efficacy.The Coping Self-Efficacy Index (CCSEI) [48] is a 5-item measure of confidence to use healthy coping strategies when distressed.Each item is rated on a 4-point scale from 0 (not confident at all) to 3 (very confident).The total score is the sum of the items, with higher scores indicating better coping self-efficacy.Total scores range from 0 to 15, with higher scores indicating better wellbeing.The CSEI has adequate internal consistency in this sample, Cronbach α = 0.77.

Help-seeking intentions. The Help-Seeking Intentions Scale (HSIS)
[49] is a 7-item measure of how likely participants would seek support from a range of social and professional supports if they were distressed and had difficulty coping in the future, rated on a scale from 0 (not likely at all) to 6 (very likely).The scale was author constructed and demonstrated excellent internal consistency, Cronbach α = 0.81.

Sleep:
Subjective sleep quantity and quality were assessed through a series of questions asking the participant to reflect on typical sleep at home.Questions were average hours of sleep, sleep quality, daytime sleepiness, sleep onset delay, night awakenings, and the impact of poor sleep on mood, concentration and relationships.

Procedure
Ethics was approved by the Central Adelaide Local Health Network Human Research Ethics Committee (number 595).In accordance with the tenets of the Declaration of Helsinki, all participants provided written informed consent for study participation.No stipend was provided to participants.Participants were recruited between February to October 2020.Eligible patients were identified by medical or nursing staff on each ward.Nursing staff obtained permission from eligible patients for the project coordinator to meet with them to discuss the study.The study was described to patients as a study about patient wellbeing and coping.Participants had the option to complete baseline assessment on paper or online.After the baseline assessment, the project coordinator delivered the Care Collaborate Connect intervention (coping intervention) to the patient at their bedside.The sleep intervention was delivered via telephone one-week post-discharge.A mood and coping review were done before the sleep intervention.Follow-up assessments were completed 1-month post-discharge online with the URL emailed to participants (n = 54), on a paper questionnaire mailed to a participant (n = 4), or over the telephone (n = 12); two withdrew before follow-up.
Participants were emailed or mailed up to three times to complete a follow-up assessment.The project coordinator completed Care Collaborate Connect: Suicide Prevention training [42] and was supervised a Clinical Psychologist.The project coordinator completed the self-report treatment integrity checklists after each intervention.

Intervention
Coping.Care Collaborate Connect is a low-intensity psychological intervention that normalises distress and supports healthy coping [27,42].The intervention involves providing psychoeducation about coping, reviewing patients' current healthy and unhealthy coping strategies, and supporting patients to strengthen their coping plan with additional professional supports if needed.Participants are encouraged to record their healthy coping strategies on a My Coping Plan worksheet or the My Coping Plan app [50].Participants and the project coordinator collaboratively decide if the participant has high needs for additional professional support (immediate support), moderate needs (may need ad hoc professional support), or low needs (has few or no unhealthy coping strategies or adequate social and/or professional support).
A mood and coping review were conducted at subsequent sessions.If participants experienced any unpleasant emotions since the coping intervention, they are asked how they coped.A review of the effectiveness of their coping plan follows, and participants are encouraged to strengthen their plan by adding additional supports if it was not effective.Self-directed Coping Planning has been shown to significantly reduce short-term psychological distress and improve wellbeing [45].
Sleep.Sleep hygiene is an effective strategy to reduce daytime sleepiness and improve sleep quality [51,52].The sleep session involved a review of sleep hygiene behaviours based on previous studies that have used cognitive behaviour for insomnia in patients with cardiac disorders [36,37].Participants who had sleep problems were encouraged to identify and set goals for improving sleep hygiene behaviours.As hospital environments disrupt patient sleep [53], the sleep intervention was conducted one-week post-discharge when patients returned to normal routines.

Statistical analysis
Descriptive statistics were used to describe the sample and recruitment, retention, and data completion rates.Completer and noncompleters were compared on outcomes variables using Pearson Chisquare tests for categorical variables or independent samples t-tests for continuous variables.Multiple imputation was used on missing data for outcome variables.The Reliable Change Index [54] was used to assess reliable change on measures from baseline to one-month follow-up.Statistical tests were two-sided with a p < .05significance level.

Recruitment
Recruitment was conducted on weekdays for six months.One hundred and seventy-eight patients were approached by nursing staff, of which 145 (81 %) consented to learn more about the study, and 72 (40 %) agreed to participate.Reasons for non-participation included: not thinking the study was relevant to them, being too tired or unwell, being unable to commit to follow-up, or being too worried about their health to concentrate.Almost all participants (97 %) completed the baseline questionnaire on paper rather than online.Twenty-two per cent of participants did not have an email address for communication, so mail was used for post-discharge appointment setting and assessments.
At baseline, 68 % of the sample had a least one problem associated with mood dysregulation (poor wellbeing, high distress indicated by unhealthy coping or sleep problems).This includes 46 % with poor wellbeing, 19 % with high distress, and 15 % with poor wellbeing and high distress.Two participants had poor wellbeing, high distress, and sleep problems.At 1-week post-discharge, 39 % reported sleep problems.Of those with sleep problems, 5 % had onset insomnia problems, 12 % had maintenance insomnia problems, 3 % had both onset and maintenance insomnia problems, 12 % woke to feel unrefreshed, and 8 % had insomnia and woke to feel unrefreshed.

Intervention implementation
The average duration to deliver the coping intervention was 37.5 min (SD = 10.5;Range = 20-71).Most participants (76 %) had low needs for additional professional support to cope, 24 % had moderate needs for additional professional support, and none had high needs for additional professional support.All participants with moderate needs identified suitable professional supports and included them in their coping plan.
Almost all (92 %) participants attended the sleep session one week following discharge.Only one participant declined the intervention, three were unable to be contacted, one was readmitted to another hospital and was lost to the study, and one withdrew from the study.Of those who reported sleep problems (39 % of participants), most (62 %) developed a plan to improve their sleep hygiene.The average duration of the sleep intervention was 15.5 min (SD = 5.3 min, range = 7-27 min).

Treatment fidelity
The average treatment fidelity score was very high for both the coping intervention (M = 18.7,SD = 1.9) and the sleep interventions (M = 15.2,SD = 1.1).Of the 14 participants who self-reported unhealthy coping strategies indicative of high distress at baseline, 71 % (n = 10) were identified as needing additional professional support during the intervention.Of the participants who self-reported high distress but were assessed as having low needs for additional professional support to cope during the intervention, one participant already had regular counselling.Three additional participants reported high distress in the self-report questionnaires, but this was not identified in the intervention.

Retention
Completion of the two intervention components was very high (92 %), however, there was a 26 % attrition rate by one-month follow-up assessment.Of those who were lost to follow-up, 68 % had poor wellbeing and 26 % had high distress at baseline.Two participants withdrew from the study, and 23.6 % were unable to be contacted at follow-up.Compared with non-completers, those who completed the follow-up assessment had a shorter in-patient stays (t (69) = 3.08, p = .003),better wellbeing (t (70) = 2.40, p = .02)and coping self-efficacy at baseline (t (70) = 2.12, p = .04).There was no significant difference between the groups on unhealthy (t (70) = 0.05, p = .96)or healthy coping (t (70) = 0.40, p = .69).

Perceived usefulness of the intervention
Of those who experienced unpleasant emotions during the 1-week following discharge, 64 % (n = 7) of participants reported using their Coping Plan.Almost half (43 %) found their plan effective (n = 3) and 57 % edited the plan to make it more effective (n = 4).There was no significant effect of baseline wellbeing (χ 2 (1, N = 11) = 0.51, p = .48),or high distress (χ 2 (1, N = 11) = 1.64, p = .20),on the proportion of participants who used their coping plan within one-week.
At one-month follow-up, 14 % of all participants reported the sleep intervention was helpful, and 71 % found it somewhat helpful; 15 % did not find it helpful at all.Of those with sleep problems at baseline, 37 % of participants changed their behaviour post-intervention.There was no significant difference between participants with and without sleep problems on the perception of the usefulness of the sleep intervention (χ 2 (2, N = 51) = 2.25, p = .33)or behaviour change (χ 2 (1, N = 50) = 0.69, p = .41).Most participants (92 %) would recommend the coping and sleep intervention for other patients.

Intervention outcomes
Table 1 shows the means and standard deviations at baseline and 1month follow-up for participants who had poor wellbeing and high distress at baseline.Overall, 45 % of participants with poor wellbeing and 22.2 % of participants with high distress had reliable change in wellbeing at follow-up.By contrast, only a few of the participants with poor wellbeing demonstrated improvement in unhealthy coping strategies (10 %).However, one-fifth of those with unhealthy coping strategies showed reliable change at follow-up.This suggests that high distress patients were reasonably responsive to intervention.Fifteen per cent of participants with poor wellbeing showed improvements in coping selfefficacy compared with none of those with high distress.Finally, none of the participants with either poor wellbeing or high distress showed reliable improvements in help-seeking intentions.For the poor wellbeing group, mean HSI scores at baseline were 17.75 (SD = 5.94) and at one-month follow up were 18.95 (SD = 8.93).For the high distress group, mean HSI scores at baseline were 19.88 (SD = 6.41) and at onemonth follow up were 20.22 (SD = 4.38).
The average sleep duration for the sample was just below the lower recommended limit at baseline (M = 6.82 h, SD = 1.84) with 47 % of participants reporting fewer than 7 h of sleep a night.By follow-up, the average sleep duration was 7.00 h (SD = 1.68).For participants who had fewer than 7 h of sleep at baseline, 8.3 % had positive, reliable improvements at follow-up with their mean sleep duration increasing from 5.27 h (SD = 1.14) to 5.99 h (1.31).

Readmissions
A readmission was defined as an admission happening more than 24 h after the index admission.The 24 h criteria was chosen to limit ambiguous cases and improve data clarity (e.g., a readmission within 24 h might be considered a continuation of the initial hospital stay rather than a distinct and separate event), to remain consistent with other studies that have used the same criteria and the impracticality of providing the intervention to same day patients.Notably and notwithstanding the 24 h criteria, the earliest readmission in those participants not lost to follow-up was 14 days post-discharge.Fourteen percent (10/ 72) of participants were readmitted to hospital approximately within one-month of discharge (14-34 days).Apart for one patient who was readmitted to another hospital and follow-up details are incomplete, the remaining readmissions were for cardiac-related problems.Of the remaining readmissions, 66 % percent (6/9) were previously admitted for coronary heart disease, 22 % (2/9) for heart arrhythmia and 22 % (2/ 9) for other cardiac-related problems.All readmissions were unplanned.There was no significant difference between those who were or were not readmitted to hospital on age (t (51) = 1.14, p = .26),sex (χ 2 (df 1, N = 53) = 0.35, p = .55),in-patient days (t (51) = 0.39, p = .70),whether they were living alone or with someone else (χ 2 (df 1, N = 53) = 0.01, p = .97),use of coping plan (χ 2 (df 1, N = 52) = 1.60, p = .45),baseline wellbeing (t (51) = 0.24, p = .81)or baseline high distress (t (51) = 1.08, p = .29).

Discussion
This study tested the feasibility of a brief sleep and coping intervention that included an in-hospital components to reduce potentially preventable hospital readmissions secondary to psychological distress.Despite expectations, 14 % of participants were readmitted to hospital within approximately one month of discharge-a rate consistent with the current 30-day post-discharge rate of readmissions in South Australian public hospitals (21 %) [55] and the Australian readmissions for cardiac patients (6.3-27 %) [2].Moreover, readmission rates were not predicted by either demographics or outcomes measures.However, the intervention received high acceptance and in the patients with high levels of distress, promising, but modest outcome gains.
Of the participants recruited, fewer than half of participants had poor wellbeing, and only one fifth had unhealthy coping scores indicative of high distress.These rates are similar to previous reports of psychological distress in in-patient populations [20,56], unhealthy coping indicative of mental illness [57] and psychiatric illness in the general population [58].Nevertheless, and regardless of severity, almost all participants found the brief intervention helpful or very helpful.In addition, the retention rate was high, and most participants reported that they would recommend the intervention to others.Finally, consistent with evidence that additional assessment and intervention is of benefit, in the subset of cardiac patients with high distress the brief intervention produced reliable changes at follow-up [28,47].
The delivery of an in-hospital intervention is not without challenges.The learnings from the present study include the importance of scheduling and staff engagement.Fewer than half of the patients approached consented to participate because of short hospital stays, crowded treatment regimens and other timing issues.Where possible prescreening of mood and coping may assist [59].Recruitment was dependent on staff availability and promotion.In future, greater gains may be achieved by adapting the screening process to fit within the standard organisational procedures rather than as an additional task.

Table 1
Baseline and 1-month follow-up mean, standard deviations, and proportion of sample with reliable change on outcomes variables for the participants with poor wellbeing and high distress at baseline.Note: CSEI = Coping Self Efficacy Index, RCI = Reliable Change Index, % RCI = proportion of participants that showed significant improvements.And WHO-5 = World Health Organisation-5.
For example and consistent with recent Australian governmental productivity and mental health review recommendations [60,61].The latter could be achieved by including a routine mood and coping review whilst patients are in-hospital to identify those at risk.Finally, gains were not observed in health seeking intentions for psychological problems.Mental health literacy programs are reported to alter help-seeking attitudes but there is less evidence that they modify help-seeking behaviour [62].This study has several limitations.First, the limited uptake of the intervention by patients and loss of patients at follow-up.Those who did compete the program tended to have shorter in-patient stays and better wellbeing and coping self-efficacy at baseline.This may have influenced the sample representativeness.Second, the limited sample size prevented sub-analysis and it is possible that the efficacy of in-hospital psychological interventions may differ according to cardiac disorder.Third, the intervention was not compared to a concurrently recruited control condition or matched controls, thus limiting conclusions regarding the effectiveness of the psychological intervention.Building on the findings from the present feasibility study, these limitations can be addressed in future research.

Conclusions
This study examined the feasibility of a brief coping and sleep intervention for inpatients with a history of cardiac disorders.The intervention had a high degree of acceptability and demonstrated gains in patients with poor wellbeing and/or high levels of psychological distress especially for sleep and to a lesser extent coping and wellbeing.Despite expectations, hospital readmissions were not reduced.However, it is acknowledged that efficacy studies will be needed to evaluate more fully the Care Collaborate Connect and sleep intervention model.The present findings point to the utility of modifying clinical practice in hospitals to address the biopsychosocial needs of vulnerable patients.