Virtual Post-Intensive-Care Rehabilitation for Survivors of COVID-19: A Service Evaluation

Aim: The objective of this study is to evaluate the safety, utilisation, and effectiveness of a novel, virtual rehabilitation programme for survivors of SARS‑CoV‑2 infection (COVID-19) and intensive care admission. Methods: A service evaluation was performed. Adults admitted to a United Kingdom intensive care unit with COVID-19-induced respiratory failure and surviving hospital discharge were invited to an eight-week rehabilitation programme. The programme consisted of virtually delivered exercise classes and support groups led by critical care physiotherapists and follow-up nurses. Results: Thirty-eight of 76 eligible patients (50%) agreed to participate, of which 28 (74%) completed the rehabilitation programme. On completion of the rehabilitation programme, there were significant improvements in exercise capacity (one-minute sit-to-stand test; 20 stands vs. 25 stands, p < 0.001), perceived breathlessness (Medical Research Council dyspnoea scale; 3 vs. 2 p < 0.001), shoulder disability (Quick Dash; 43 vs. 19 p = 0.001), anxiety (Hospital Anxiety Depression Scale; 4 vs. 3 p = 0.021), depression (Hospital Anxiety Depression Scale; 4 vs. 2.5 p = 0.010), and psychological distress (Intensive Care Psychological Assessment Tool; 3 vs. 2 p = 0.002). No adverse events or injuries were recorded during the programme. Conclusion: It is feasible to recruit and retain survivors of COVID-19-induced respiratory failure for virtual post-intensive-care rehabilitation. It appears that the virtual rehabilitation programme is safe and improves physical and psychological morbidity.


Introduction
Critical illness's physical and psychological sequelae are well documented and collectively known as postintensive care syndrome (PICS) [1]. PICS is considered to affect up to 50% of intensive care unit (ICU) survivors and is recognised as a public health burden due to its lasting disability [1].
Muscle weakness and reduced exercise capacity contribute to delayed physical recovery due to an ongoing cycle of reduced physical activity and the propagation of sarcopenia and frailty [2]. Neuropsychiatric complications such as anxiety, depression, post-traumatic stress disorder, and cognitive impairment are also commonly reported [1,3]. The lasting burden of PICS leads to an impairment in quality of life, exacerbation of long-term conditions, and increased healthcare utilisation for patients and their relatives [1][2][3].
Acute respiratory failure following a SARS-CoV-2 infection (COVID-19) has resulted in a sudden surge in ICU admissions worldwide [4]. The physical and psychocognitive consequences specifically associated with COVID-19 have been recognised, but not yet fully understood in relation to PICS [5][6]. Patients with severe COVID-19 are known to require extended periods of mechanical ventilation, deep sedation, neuromuscular blocking agents, and prolonged intensive care admissions, all of which are common risk factors for the development of PICS [7]. Other risk factors associated with PICS include comorbidities such as diabetes, hypertension, and chronic lung disease, a similar profile to that seen in severe cases of COVID-19 [7].
Rehabilitation has been recognised as an essential component in the recovery from PICS and COVID-19 [8][9][10]. However, many rehabilitation and outpatient services have faced significant disruption during the pandemic [10][11]. Urgent work must ensure appropriate rehabilitation services are delivered, and novel ways of safe working should be established and protected [10][11].
The post-intensive-care follow-up service at our institution formerly consisted of a weekly rehabilitation class and support group led by ICU physiotherapists and ICU follow-up nurses in the hospital's outpatient department. Due to the demands of the pandemic, the follow-up service was suspended to enable the repurposing of staff and resources for frontline emergency services. Although some outpatient services have since reopened, national laws regarding social distancing and infection-control procedures have restricted face-to-face group activities.
To meet the needs of a rapidly increasing post-intensive-care population amidst the unprecedented circumstances of a pandemic, services were required to adapt and work virtually [10][11]. The concept of telerehabilitation has been shown to be possible for COVID-19 patients, but not specifically for COVID-19 survivors of critical care admission [12][13]. In July 2020, the critical care follow-up team re-established and developed an innovative, virtual post-intensive care rehabilitation programme.
This service evaluation aimed to assess the safety, utilisation (recruitment, retention, and adherence), and potential impact on the recovery from physical and neuropsychiatric sequelae of an innovative virtual postintensive-care rehabilitation programme for survivors of COVID-19.
This article was previously presented as a meeting abstract at the 2021 European Society of Intensive Care Medicine (ESICM) Lives Conference on October 4, 2021, and the 2021 Association of Chartered Physiotherapists in Respiratory Care (ACPRC) ACPRC Virtual Conference on April 23, 2021 [14][15].

Design
This was a single-centre service evaluation and has therefore been reported in accordance with Revised Standards for Quality Improvement Reporting Excellence (SQUIRE) guidance [16].

Ethical considerations
This single-centre (University Hospitals Birmingham NHS Foundation Trust (UHB)) service evaluation used anonymised linked databases (Clinical Audit and Registration Management System [CARMS] Identification Number: CARMS-17954) with a waiver of ethical approval according to the national guidance [17]. We used routinely collected data, conducted within the clinical audit framework; patients were not contacted outside their routine clinical care.

Setting
A major tertiary hospital in Birmingham, United Kingdom (The Queen Elizabeth Hospital Birmingham (QEHB)), with one of the largest co-located ICUs in Europe. Between March 2020 and April 2020, the ICU at QEHB admitted 177 patients with COVID-19, of whom 110 survived ICU discharge [7].

Sample
Consecutive adult patients (≥18 years of age) who were admitted to the ICU with a confirmed diagnosis of COVID-19 between March 2020 and April 2020 and survived to hospital discharge were identified by a senior ICU physiotherapist. Patients were excluded if they had unstable medical conditions contraindicating their safety to exercise, a severe neurological injury impeding their ability to cooperate with rehabilitation, or an inability to provide informed consent for the programme as required by our hospital.
After hospital discharge, patients were invited to participate via telephone by a senior ICU physiotherapist. If patients declined to participate, the primary reason was recorded. Baseline data, including demographics, ventilation days, and length of stay, were obtained from patient notes and electronic health databases.

Intervention
Patients were invited to participate in an eight-week post-intensive-care rehabilitation programme. The programme consisted of a once-weekly exercise class followed by a support group, completed virtually via video call using Microsoft Teams. Patients were able to complete the programme in their own homes using either a smartphone, tablet device, laptop, or computer, with a group of up to six patients on each call. Verbal and written patient information was provided, including the risks and benefits [18].

Exercise Component
A structured, interval exercise approach was adopted, consisting of a warm-up, 20 minutes of circuit-based exercises, and a cool-down, delivered by the class physiotherapist [19]. The programme was designed so patients could complete the exercises from a chair and therefore remain visible on the video screen. Ten exercises were completed in each circuit, with one minute for each exercise completed twice. A combination of upper-and lower-limb exercises aiming to improve strength and cardiovascular fitness were included [19]. Patients were advised of four different intensity levels for each exercise (Appendix 1, Table 3) [18]. Patients were advised to titrate exercise intensity based on their perceived breathlessness, aiming for a score of 3-4 on the modified Borg breathlessness scale (Appendix 2, Table 4) [18][19][20]. An online patient information booklet was provided to each participant, including written instructions and photographs of each exercise, in addition to the Borg breathlessness scale [18].

Support Group Component
Following the exercise component, patients were invited to participate in a support group. The structure of these sessions is based on recommendations from the patient support charity 'ICU Steps' [21]. In line with these recommendations, the groups were supported by a senior nurse with experience in ICU follow-up and a senior critical care physiotherapist. The combination of a nurse and physiotherapist with experience in managing patients during and following critical illness ensures appropriate expertise in both physical and neuropsychiatric recovery. The role of the healthcare professionals was to support new participants and facilitate discussion. Patients were given the opportunity to share their experiences with their ICU recovery. Educational topics such as delirium, sleep, pacing advice, and expectations of recovery were discussed.

Primary outcome measures
The primary outcomes assessed the utilisation of the virtual rehabilitation programme; including recruitment, retention, and adherence rates. Recruitment was defined as the percentage of eligible patients who attended. Retention was defined as the number of patients who completed rehabilitation. Adherence was defined as completing a minimum of 75% of the available virtual rehabilitation sessions. All adverse events were recorded.

Secondary outcome measures
Secondary outcomes assessed the impact of the virtual rehabilitation programme on patient clinical outcomes (Appendix 3, Table 5). Outcomes were assessed by the class physiotherapist during an individual video-call consultation before commencing the programme and within one week of its completion. Physical outcomes included the one-minute sit-to-stand test (1MSTS), the Medical Research Council breathlessness scale (MRC), and the Quick Dash Upper Limb questionnaire (QD). Non-physical outcomes included the Hospital Anxiety and Depression Score (HADS), the Intensive Care Psychological Assessment Tool (IPAT), and the perceived health rating of the Euro-Qol 5 dimension quality of life questionnaire (EQ5D).

Statistical analysis
Data for all outcome measures were assessed using IBM SPSS 26 software (IBM Corp., Armonk, NY). The characteristics of the patient cohorts who attended and did not attend were summarised, with continuous variables reported as means ± standard deviations (SDs) where normally distributed and medians and interquartile ranges (IQRs) reported otherwise. Comparisons between patients who attended and did not attend were then performed using Mann-Whitney U tests for ordinal or continuous variables and chisquared tests for nominal variables. Data for all outcomes post-intervention were assessed for normality using the Sharipo-Wilk test, and non-parametric tests were used as appropriate. Continuous variables were reported as medians and IQRs, and the Wilcoxon matched-pairs signed rank test was used to assess for statistical significance, with p<0.05 deemed to be indicative of statistical significance throughout.

Results
Eighty-seven patients out of 110 patients (79%) who were admitted to the ICU at QEHB with COVID-19 between March and April 2020 and survived hospital discharge were screened for their eligibility for virtual post-intensive-care rehabilitation during a three-month period ( Figure 1). Seventy-six of those screened (87%) were eligible for rehabilitation, with 38 patients consenting to the programme (50%). The reasons for declining rehabilitation (n=38) were independent self-management (47.4%), lack of technology access (the patient did not have access to a smartphone, tablet device, laptop, or computer) (28.9%), language barrier (18.4%), and other rehabilitation programmes in place (5.3%). Of the 38 who agreed to participate, 28 completed the programme (74%) (Figure 1). Of those that completed, 82% adhered (n = 23); quantified as attending 75% of planned rehabilitation sessions. Patients attended rehabilitation for a mean of 79 days following discharge from the hospital. No serious adverse events or injuries were recorded during the programme. The baseline characteristics of patients invited to the post-intensive-care rehabilitation programme can be found in Table 1. Patients who attended virtual rehabilitation were predominantly male (79%) with a mean age of 54 years (standard deviation [SD] 11). There were no statistical differences in demographics between those who completed the virtual rehabilitation programme and those who declined ( Table 1).  Following completion of the virtual rehabilitation programme, there was a statistically significant improvement in all physical outcomes (   There was also a statistically significant improvement in all non-physical outcomes ( , and psychological distress (IPAT) were suggestive of no clinical psychological symptoms prior to commencing rehabilitation. Further analysis of the non-physical data showed nine of the 28 (32%) participants had clinical symptoms of anxiety (HADS anxiety score ≥ 8) prior to rehabilitation, reducing to six participants after rehabilitation. Six (21%) participants had clinical symptoms of depression prior to rehabilitation (HADS depression score ≥ 8), reducing to three post-rehabilitation. Six (21%) participants had clinical symptoms of psychological distress prior to rehabilitation (IPAT score > 7), reducing to one post-rehabilitation. Perceived health ratings also showed significant improvement following rehabilitation (EQ5D 60 vs. 76, p=0.001).

Discussion
We assessed the safety, utilisation (recruitment, retention, and adherence), and impact on the recovery from physical and neuropsychiatric sequelae with an innovative virtual post-intensive-care rehabilitation programme in survivors of COVID-19. To the best of our knowledge, this is the first known report of innovative, virtually delivered post-intensive-care follow-up specifically for survivors of severe COVID-19 requiring ICU admission.
This service evaluation has demonstrated that it is feasible to recruit survivors of an ICU admission with severe COVID-19 to a virtual, post-intensive-care rehabilitation programme. During the three-month recruitment period, 79% of patients were approached and offered rehabilitation. Patients were approached an average of 79 days post-hospital discharge, in accordance with national guidelines, which advise followup at two to three months [8]. Twenty-three patients were not offered rehabilitation due to staffing constraints. Patients found the experience and understanding of ICU clinicians valuable; however, it is recognised that the service requires protected resources and funding to meet the demands of the COVID-19 pandemic. A previous survey identified that intensive-care follow-up services are notoriously underfunded and only provided by 27% of UK intensive care units, with less than 10% offering follow-up in the form of structured exercise [22]. Further work is required to ensure follow-up services are equitable and accessible to all ICU survivors.
Of those approached during the period of the service evaluation, 50% participated in the virtual postintensive-care rehabilitation programme. The primary reason for declining was that rehabilitation was not required due to independent self-management. This may be reflective of a working-age cohort (mean age 54 years) and pre-admission Clinical Frailty Score (median 2 = well); however, in our study, the demographics of those who attended and those who declined showed no significant differences. The physical and nonphysical outcomes of those who declined rehabilitation were not investigated in this study, yet previous literature suggests that PICS affects 50% of critical care survivors [1,8]. Predictors of long-term morbidity in survivors of COVID-19 are currently unknown.
Of those who attended, we found 73.6% retention (defined as completion of the programme) and 82% adherence (defined as attending a minimum of 75% of rehabilitation sessions) to an eight-week programme; demonstrating patient acceptability of the programme. The virtual nature of the programme has enabled accessibility to rehabilitation and the ability to overcome the challenges of social restrictions and infection control procedures during the COVID-19 pandemic. However, it is important that rehabilitation services do not discriminate against those who do not have access to technology. Almost one-third of patients declined participation due to a lack of technology access, and 18.2% due to language barriers. Further work is required to ensure rehabilitation and post-intensive care follow-up are accessible to all patients, expanding to all socioeconomic and ethnic regions.
The programme was found to be safe, with no adverse or safety events recorded. The participants who completed the rehabilitation programme demonstrated improvements in all physical and non-physical outcomes. Despite being delivered virtually, with patients self-directing their exercise intensity based on perceived breathlessness, the rehabilitation programme remained effective. This supports existing work that virtual rehabilitation is feasible and effective, yet this is the first known report of virtual rehabilitation specifically for survivors of COVID-19 following critical care admission [12][13][23][24].
The rehabilitation programme aimed to improve physical fitness through an interval training approach. Baseline data identified poor levels of exercise capacity prior to rehabilitation, with the mean 1MSTS score for the group equivalent to the 2.5th percentile for healthy 65-to 69-year-olds, despite a mean age of 54 years [25]. Following completion of the programme, patients demonstrated a significant improvement in exercise capacity (1 MSTS 20 vs. 25, p<0.001). However, it is important to note that although the 1MSTS results improved, the scores remained below the 2.5th percentile for the 50-54 age groups [25]. The 1MSTS is a pragmatic measure of exercise capacity during a virtual assessment [26]. In chronic obstructive pulmonary disease cohorts, the 1MSTS has been found to be a reliable and valid measure of exercise capacity, comparable with the six-minute walk test; a measure that has been utilised in previous ICU research [27].
In addition to physical fitness, the programme also aimed to improve upper-limb strength. Shoulder impairment and its associated functional disability are common in survivors of critical illness [28]. Pronounced upper limb deficits have been found in COVID-19 survivors due to injuries associated with prone positioning [29]. Our cohort demonstrated severe upper limb dysfunction at baseline. On completion of the programme, patients showed a significant reduction in shoulder dysfunction (QD 43.18 vs. 19.31, p=0.001).
In contrast to previous ICU literature, overall, our cohort demonstrated normal scores for anxiety, depression, and psychological distress [1,3]. Despite this, psychological symptoms continued to show significant improvement following the completion of the programme. Psychological morbidity has been known to influence adherence to medical treatment in other populations; therefore, results may only be reflective of those who participated [30]. At the point of contact, many patients reported gratitude for their survival and returned home to their relatives after a period of separation during their hospital admission. This may explain the low levels of psychological symptoms in our cohort. A longitudinal, qualitative exploration of the psychological recovery of COVID-19 survivors should be considered, exploring the unique experiences of COVID-19 patient groups and potential long-term impacts.

Limitations
There are several limitations to this service evaluation. First, this is a non-randomised trial that does not have a control group for comparison. We recognise that this reduces the validity of the study findings as the results may be subject to temporal changes, acknowledging that activity levels between hospital discharge and follow-up were not monitored and that the advice given to patients by ward physiotherapists was not standardised. The results are also open to bias as they only reflect those who consented to participate and are therefore representative of a cohort that is motivated and engaged with rehabilitation. As this service evaluation did not collate qualitative data, the results do not account for the experiences, needs, or preferences of patients. Participant numbers were small, and the service evaluation was performed at a single centre; therefore, results may not be representative of the wider population.

Conclusions
The COVID-19 pandemic has placed overwhelming demand on healthcare services and has subsequently led to the repurposing of resources and staff to bolster frontline emergency departments. As we face the secondary impacts of the pandemic, the optimisation of recovery as a therapeutic objective should be prioritised over patient survival alone. For many survivors of critical illness, their discharge from the ICU is the start of an uncertain journey, facing numerous physical and non-physical limitations. It is essential that critical care rehabilitation and follow-up services are protected in order to optimise patient recovery and quality of life after critical illness.
This service evaluation demonstrates that with protection and prioritisation of rehabilitation, postintensive-care follow-up services can be adapted using innovative and virtual methods. This is the first known report of virtual post-intensive-care rehabilitation for survivors of COVID-19, providing accessible critical care follow-up despite the restrictions faced by healthcare services during the pandemic. This service evaluation has demonstrated that it is feasible to recruit and retain intensive care survivors for virtual rehabilitation and that the programme is safe and effective. For those who attended the virtual postintensive-care rehabilitation programme, there was a significant improvement in all physical and nonphysical outcomes. Further research, including an appropriately powered multi-centre randomised clinical trial with longitudinal follow-up, is required to further validate these findings.

Modified Borg scale
The scale below helps you record how breathless you feel. It is important that you are aware of how to record your breathlessness as this will help to show how you are progressing. It is normal to feel more breathless when exercising, reaching a level of 3 or 4 on the scale below.   [20].