Comparing rehabilitation outcomes for patients admitted to the intensive care unit with COVID-19 requiring mechanical ventilation during the first two waves of the pandemic: A service evaluation

Abstract

Objectives

To compare rehabilitation outcomes of patients admitted to the intensive care unit with COVID-19 and mechanically ventilated during wave 1 and 2, receiving two different models of physiotherapy delivery.

Methods

Adults admitted to the intensive care unit between October–March 2021 (wave 2) with a confirmed diagnosis of COVID-19 and mechanically ventilated for >24 hours were included. During wave 2, rehabilitation was provided by physiotherapists over five days, with only emergency respiratory physiotherapy delivered at weekends. Rehabilitation status was measured daily using the Manchester Mobility Score to identify time taken to first mobilise and highest level of mobility achieved at ICU discharge. Outcomes were compared to data previously published from the same ICU during ‘wave 1’ (March–April 2020) when a seven-day rehabilitation physiotherapy service was provided.

Results

A total of n = 291 patients were included in analysis; 110 from wave 1, and 181 from wave 2. Patient characteristics and medical management were similar between waves. Mean ± SD time to first mobilise was slower in wave 2 (15 ± 11 days vs 14 ± 7 days), with overall mobility scores lower at both ICU (MMS 5 (Step transferring) vs MMS 4 (standing practice) (4), p < 0.05) and hospital (MMS 7 (Mobile > 30 m MMS) vs MMS 6 (Mobile < 30 m MMS), p < 0.0001) discharge. Significantly more patients in wave 2 required ongoing rehabilitation either at home or as an inpatient compared to wave 1 (81 % vs 49 %, p = 0.003).

Conclusion

The change in physiotherapy staff provision from a seven-day rehabilitation service during wave 1 to a five day rehabilitation service with emergency respiratory physio only at weekends in wave 2 was associated with delayed time to first mobilise, lower levels of mobility at both intensive care unit and hospital discharge and higher requirement for ongoing rehabilitation at the point of hospital discharge.

Implications for clinical practice.

• •The provision of physiotherapy across seven days was associated with shorter times to first mobilise.
• •Earlier mobilisation in wave one was associated with higher mobility levels at both intensive care unit and hospital discharge.
• •Delayed mobilisation on the intensive care unit was associated with an increased requirement for ongoing rehabilitation at the point of hospital discharge.

Introduction

The COVID 19 pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARs-Cov-2) has seen healthcare services experience unprecedented operational challenges (Maves et al., 2019). At the peak of Wave 1 (12th April 2020), 3301 patients occupied mechanically ventilated beds and 21 687 patients occupied hospital beds in the United Kingdom (UK) (https://coronavirus.data.gov.uk/details/healthcare). The initial National Health Service (NHS) response to the pandemic meant all non-essential outpatient services were stood down and operations were cancelled. This strategy allowed re-deployment of outpatient and ward based healthcare professionals to intensive care, supplementing the existing workforce to meet the increasing demands within Intensive Care Units (ICU) (Goh et al., 2020). This redeployment was not without cost. As hospitals emerged from wave 1, cancer waiting times had reached an all-time high and significant backlogs had accumulated for both urgent and non-urgent surgery (England, 2021, Macmillan, 2020).

As the UK moved into the second wave of the COVID-19 pandemic, additional pressures were present in comparison to wave 1. Due to the increased backlog, there was a significant need to maintain surgical pathways and a push to continue outpatient services which had now been established in new novel or remote formats (Zampino et al., 2021, National Institute for Health and Care Excellence, xxxx). Consequently; as NHS Trusts endeavoured to maintain these services, staff that were previously re-deployed during wave 1 were no longer available to support ICU. In addition, the accumulating psychological burden of the pandemic saw high levels of psychological distress amongst frontline healthcare workers, including post-traumatic stress disorder, anxiety and depression (Gemine et al., 2021, Marcomini et al., 2021) with resultant sickness absence from work (Greenberg et al., 2020). This further increased the challenges of meeting the ICU demands during the second wave; which actually exceeded the peak of wave 1; with 4076 patients occupying mechanically ventilated beds on 22nd January 2021 and 39,254 inpatients on 18th January 2021 (https://coronavirus.data.gov.uk/details/healthcare).

We previously published rehabilitation outcomes for patients admitted to our hospital with severe COVID-19 requiring mechanical ventilation during ‘Wave 1’ (McWilliams et al., 2021). The redeployment of staff at this time allowed the physiotherapy service to adopt a 7 day working pattern, facilitating respiratory treatment and rehabilitation across all 7 days. Whilst we presented data to demonstrate the time to commence rehabilitation was delayed due to the severity of illness, this 7 day working model meant rehabilitation remained possible within the ICU, and led to increased levels of mobility at critical care discharge (McWilliams et al., 2021). During ‘Wave 2’, the lack of redeployed staff meant physiotherapy had to revert to its pre-pandemic working model of delivering rehabilitation 5 days a week from Monday to Friday, with only limited access to physiotherapy for emergency respiratory care provided at the weekends. We hypothesise that this lack of access to additional rehabilitation at weekends may have impacted rehabilitation outcomes.

We aimed to evaluate any differences in rehabilitation outcomes for patients admitted during wave 2 to our previously published results.

Objectives

To compare rehabilitation outcomes for patients receiving two different models of physiotherapy delivery (7 days vs 5 + 2 days).

Materials and methods

This was a single centre, prospective, service evaluation comparing outcomes for patients admitted to ICU between March and April 2020 (wave 1) and October 2020 and March 2021 (wave 2) with a confirmed diagnosis of COVID-19. This project constituted an observation of standard care delivery with no randomization and thus met the definition of a service evaluation under the National Health Service Health research authority guidelines (http://www.hra.nhs.uk/researchcommunity/ Before you-apply/determine-whether-your-study-is-research/). As such, ethical approval was not required, and because all outcome measures are collected as part of routine care, the need for consent was waived. The project was registered as an audit on the trusts clinical audit registration and management system (CARMS-17976). Participants were followed up until acute hospital discharge. This study is reported in accordance with the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) guidelines (Vandenbroucke et al., 2007).

Setting

The Queen Elizabeth Hospital Birmingham is a quaternary-level acute care hospital, with one of the largest co-located ICUs in Europe. Before the COVID-19 pandemic, the standard critical care capacity for this unit was 75 beds; however, with surge planning, the overall capacity was increased to>200 for both waves of the pandemic. At the peak of the 2nd COVID-19 wave, the ICU cared for 153 patients (COVID and non-COVID-19 patients) compared to a peak of 164 during wave 1. As per wave 1, capacity was increased through a variety of measures including caring for two patients per bed space and reduced specialist staffing ratios. Clinical pressures and staff shortages related to the COVID-19 pandemic meant it was not feasible to replicate the model of staffing utilised during wave 1. Differences in the models of care are outlined in Table 1 .

Table 1.

Differences in staffing ratios; pre-pandemic, Wave 1 and Wave 2.

Pre-pandemic ICU	Wave 1 (March – April 2020)	Wave 2 (October 2020 – March 2021)
Nursing ratio -1 ICU nurse to 1 level 3 patient	Nursing ratio -1 ICU nurse to 4 bed spaces supported by 3 redeployed non-ICU nurses	Nursing ratio -1 ICU to 3 bed spaces supported by 1 redeployed nurse and 1 redeployed medic
Consultant cover1: 10	Consultant cover -1:35 beds	Consultant cover -1:20
Physiotherapy cover -5 days service -8am – 5pm -Ratio 1:7 -Weekend emergency respiratory ratio 1:10 (no rehabilitation delivered by physiotherapy staff) -Emergency respiratory on call overnight	Physiotherapy cover -7 days service -8am – 8pm -Ratio 1:7 -Emergency respiratory on call overnight	Physiotherapy cover -5 days service -8am – 5pm -Ratio 1:8 -Weekend emergency respiratory ratio 1:10 (no rehabilitation delivered by physiotherapy staff) -Emergency respiratory on call overnight

ICU, Intensive Care Unit.

Physiotherapists within our unit assess all patients within 24 hours of admission; delivering respiratory care often termed “chest physiotherapy” and commencing rehabilitation as indicated. To meet the increasing demand during wave 1 of the pandemic, the physiotherapy service was restructured in March 2020 to ensure physiotherapy was available from 8 a.m. to 8 p.m., seven days per week, maintaining a ratio of one whole time equivalent physiotherapist for every seven patients. This was achieved through the redeployment of respiratory competent physiotherapists working within the Trust but not currently working on ICU.

In comparison, during the second wave due to ongoing service pressures and the lack of available staff to re-deploy, physiotherapy had returned to pre-pandemic working patterns of Monday to Friday only, with a slightly reduced ratio of one physiotherapist to eight patients. The additional physiotherapy staff to accommodate the increase in capacity consisted of non-specialist ICU physiotherapists from within the hospital and physiotherapists re-deployed from the Birmingham Children’s Hospital paediatric ICU. Weekend emergency physiotherapy was delivered by a mix of specialist and non-specialist ICU staff, providing respiratory physiotherapy only at a ratio of one physiotherapist to 10 patients. Consequently, unlike during wave 1, due to these reduced staffing ratio’s only limited initiation or delivery of rehabilitation by the physiotherapy team was possible at weekends. This is standard practice in the United Kingdom for trusts not providing a seven days service.

From a broader workforce perspective, during the second wave despite COVID related staffing absence, better consultant and nursing ratios were observed. Mutual aid from neighbouring trusts who did not have COVID pressures, in addition to not having to release ICU consultants to support with the setup of nightingale wards during the second wave made this feasible. The experience of the wave 1 surge and subsequent rates of ICU admission also meant the surge planning for wave 2 was modelled on a smaller number of ICU beds which reduced the cover requirements for the consultant rota and improved consultant to patient ratios.

Ward staffing

Similar to the change in working model observed on critical care, during wave 1 of the COVID 19 pandemic the ward physiotherapy teams moved to a seven day rehabilitation service. This was supported by redeployed outpatient physiotherapy staff. Prior to this only a five day rehabilitation service was offered, with emergency respiratory work only completed at weekends. COVID 19 patients surviving to ICU discharge were transferred to three, 36 bed designated COVID 19 wards staffed at a ratio of one physiotherapist to 10 beds to continue rehabilitation. Once oxygen requirements resolved, patients were stepped down to an inpatient rehabilitation ward, staffed at 1:7, were rehabilitation and discharge planning continued.

During wave 2 the patient pathway remained the same for patient surviving to ICU discharge. However the ward based physiotherapy service reverted to a five day rehabilitation service with emergency respiratory work only at weekends. As a result the staffing ratios improved Monday to Friday to 1:8 on the medical wards and 1:6 on the rehabilitation ward but no physiotherapy lead rehabilitation occurred at weekends.

Participants

Consecutive participants between October 2020 and March 2021 were included in the analysis if they met the inclusion criteria of being adults (≥18 years of age), having a confirmed diagnosis of COVID-19, and being mechanically ventilated for at least 24 hours and who survived to hospital discharge. Outcomes were compared to data previously published from the same centre for patients admitted to ICU between March and April 2020 with COVID-19 (WAVE 1) (McWilliams et al., 2021).

Procedure

Physiotherapy assessment was completed within 24 hours of ICU admission for all patients. Physiotherapists within the United Kingdom are responsible for delivering both respiratory intervention and rehabilitation. Respiratory physiotherapy is an umbrella term used to define a variety of interventions including patient positioning, proning, manual hyperinflation and delivering manual techniques to facilitate secretion clearance, and improve ventilation/perfusion matching and lung compliance. To support the wider workforce pressures across critical care, the physiotherapy team also took on increased responsibility to support management of ventilation during the pandemic, in accordance with lung protective ventilation guidelines (Network et al., 2000). This included calculation of targets for lung protective tidal volumes, which were then displayed in the patients’ bed space. Once clinically appropriate a physiotherapist led and coordinated the commencement and progression of rehabilitation between Monday and Friday. Our critical care multidisciplinary team has extensive experience of delivering early and structured rehabilitation, including established safety criteria to commence mobilization, and a protocol to guide progression (McWilliams et al., 2015).

Outcomes

The primary outcome was the highest level of mobility achieved at the point of ICU discharge, as measured by the Manchester Mobility Score (MMS). The MMS is a simple seven-point mobility scale used and validated for assessing mobility levels within critical care (McWilliams et al., 2015). The MMS ranges from 1 to 7, with 7 indicating the highest level of mobility. It was documented daily by the treating physiotherapist immediately after completion of the physiotherapy treatment session Monday to Friday. Physiotherapists retrospectively reviewed patient’s electronic noting on Mondays to ascertain if patient’s mobility status had changed over the weekend. Secondary outcomes included the number of days taken to first mobilize (defined as an MMS of 2 or higher, i.e., sitting on the edge of the bed or higher) and the location of hospital discharge, which was treated as an ordinal variable with categories of Home (No Rehabilitation), Home (With Rehabilitation), or Inpatient Rehabilitation. As per standard care, discharge destination was decided based on discussions between the ward based multidisciplinary team, considering the individuals ongoing rehabilitation requirements.

Data collection

Data for wave 2 were collected mixed prospectively and retrospectively throughout the evaluation period to mirror data collected and published during the 1st wave of the pandemic, using electronic patient records and electronic databases. This included patient demographics, ventilation days, sedation days, renal replacement therapy using continuous venovenous hemofiltration (CVVH) at any point during ICU admission, tracheostomy insertion, length of stay (LOS) for both ICU and the ward, and mortality. Other factors that delayed mobilization and therefore may have contributed to the development of ICU-acquired weakness were also collected retrospectively from patient noting. Specifically, this included data regarding aspects of critical care management, including the use of neuromuscular blocking agents, proning, and the presence of delirium, defined by a positive result on the Confusion Assessment Method ICU (CAM-ICU) at any point during the ICU stay. The presence of ICU-acquired weakness during awakening was defined as a Medical Research Council sum score of <48 (Hermans and Van den Berghe, 2015). Rehabilitation outcomes were collected immediately after physiotherapy sessions and recorded by the treating physiotherapist using the MMS. This reported either the level achieved during the physiotherapy session, or the level of mobility (if any) completed with nursing staff, whichever was the highest. Frailty scores were collected routinely as part of admission assessment using the Clinical Frailty Score (CFS) (Rockwood et al., 2005).

Statistical methods

Continuous variables were compared between the two Waves. A Mann-Whitney U test was used to analyse non-normally distributed data, and reported with medians and interquartile ranges. Metrical normal data was analysed using t-tests and summarised using means ± standard deviations (SDs).Ordinal variables were also analysed using Mann-Whitney U tests, with Fisher’s exact test used for nominal variables.

All analyses were performed using IBM SPSS 22 (IBM Corp. Armonk, NY), with p < 0.05 deemed to be indicative of statistical significance throughout.

Results

Cohort characteristics

Data were available for a total of 291 patients who survived to ICU discharge, comprising of 110 from Wave 1, and 181 from Wave 2. Patient characteristics were similar in the two Waves, with no significant differences detected in any of the factors considered except lower APACHE II scores on admission for patients during wave 2 (Table 2 ). The approach to treatment was also similar (Table 3 ), with similar use of sedation, proning and neuromuscular blocking agents, although tracheostomies were used significantly less frequently in Wave 2 (66 % vs 77 %, p = 0.048).

Table 2.

Cohort characteristics.

	Wave 1 March – April 2020	Wave 2 October 2020 – March 2021	p-Value
Age (Years)	53 ± 12	56 ± 11	0.101
Sex (% Male)	83 (75 %)	116 (64 %)	0.051
BMI (kg/m2)			0.149*
<20	0 (0 %)	2 (1 %)
20–24	14 (13 %)	13 (7 %)
25–29	42 (38 %)	58 (32 %)
30–39	39 (35 %)	85 (47 %)
40+	15 (14 %)	23 (13 %)
Ethnicity			0.956
White	53 (48 %)	81 (45 %)
Asian	38 (35 %)	66 (36 %)
Black	8 (7 %)	15 (8 %)
Mixed/Other	11 (10 %)	19 (10 %)
Clinical Frailty Score [N = 290]			0.289*
1	23 (21 %)	35 (19 %)
2	32 (29 %)	71 (39 %)
3	35 (32 %)	50 (28 %)
>3	20 (18 %)	24 (13 %)
Hypertension	50 (45 %)	63 (35 %)	0.083
Diabetes Mellitus	34 (31 %)	56 (31 %)	1.000
COPD	5 (5 %)	2 (1 %)	0.108
Asthma	17 (15 %)	24 (13 %)	0.606
APACHE II	16 (13–25)	13 (10–16)	<0.05
Charlson Comorbidity Index			0.696*
0–1	41 (37 %)	64 (35 %)
2–3	49 (45 %)	89 (49 %)
4–5	17 (15 %)	20 (11 %)
>5	3 (3 %)	8 (4 %)

Continuous variables are reported as mean ± SD, or as median (IQR), with p-values from Mann-Whitney U tests. Categorical variables are reported s N (column %), with p-values from Fisher’s exact tests, unless stated otherwise. Data are based on N = 291 cases, unless stated otherwise. Bold p-values are significant at p < 0.05. *p-Value from Mann-Whitney U test, as the factor is ordinal.

Table 3.

Treatment and patient outcomes.

	Wave 1 March – April 2020	Wave 2 October 2020 – March 2021	p-Value
ICU Treatment/Outcomes
Duration of Ventilation (Days)	19 ± 10	21 ± 16	0.406
Tracheostomy	85 (77 %)	119 (66 %)	0.048
Timing of tracheostomy post intubation (days)	12 (9–14)	12 (8–18)	0.548
Prone Position	74 (67 %)	120 (66 %)	0.287
Renal Failure Requiring CVVH	37 (34 %)	51 (28 %)	0.358
Days on Sedation	13 ± 6	13 ± 10	1.000
Neuromuscular Blockade	99 (90 %)	151 (83 %)	0.741
Duration (Days) [N = 240]	7 (4–11)	6 (3–14)	0.848
ICU Acquired Weakness on Awakening [N = 286]	110 (100 %)	163 (93 %)	0.002
Time to First Mobilise (Days) [N = 290]	14 ± 7	15 =/−11	0.286
ICU LOS (Days)	22 ± 11	24 ± 18	0.508
MMS at ICU Discharge	5 (4–6)	4 (3–5)	<0.05*
1	0 (0 %)	3 (2 %)
2	15 (14 %)	27 (15 %)
3	6 (5 %)	27 (15 %)
4	33 (30 %)	52 (29 %)
5	27 (25 %)	49 (27 %)
6	19 (17 %)	19 (10 %)
7	10 (9 %)	4 (2 %)
Post-ICU Outcomes [N = 276]**
Hospital LOS (Days)	32 (23–45)	36 (18–54)	0.629
MMS at Hospital Discharge	7 (7–7)	6 (6–7)	<0.0001*
3	2 (2 %)	7 (4 %)
4	1 (1 %)	13 (8 %)
5	1 (1 %)	4 (2 %)
6	14 (13 %)	70 (42 %)
7	91 (83 %)	73 (44 %)
Discharge Destination			0.003
Home (No Rehab)	55 (50 %)	56 (34 %)
Home (With Rehab)	46 (42 %)	79 (47 %)
Inpatient Rehab	8 (7 %)	32 (19 %)

Continuous variables are reported as median (IQR), with p-values from Mann-Whitney U tests. Categorical variables are reported s N (column %), with p-values from Fisher’s exact tests, unless stated otherwise. Data are based on N = 291 cases, unless stated otherwise. Bold p-values are significant at p < 0.05. *p-Value from Mann-Whitney U test, as the factor is ordinal. **Excludes patients that were transferred (N = 9), still in hospital (N = 5), or that died in hospital after ICU discharge (N = 1).

ICU, Intensive Care Unit. LOS, Length of Stay. MMS, Manchester Mobility Score.

ICU outcomes

Patients in wave 2 had significantly lower Manchester mobility scores at the point of ICU discharge in comparison to wave 1 (4 vs 5, p < 0.05), representing an ability to stand but unable to take any steps (See Table 3). This reduced mobility level was demonstrated despite the fact patients in Wave 2 had a significantly lower incidence of ICU acquired weakness on awakening (93 % vs 100 %, p = 0.002). Patients in wave 2 were slower to mobilise, spent longer mechanically ventilated and longer in the ICU although none of these achieved statistical significance.

Post-ICU outcomes

In Wave 2, 9 patients underwent intra-hospital transfer from ICU, and one patient from Wave 1 died in hospital after being discharged from ICU. Therefore, these patients were excluded when considering post-ICU outcomes, leaving 276 data sets for analysis (Table 3). For these, both the post-ICU (p = 0.945) and overall (p = 0.629) hospital stays were similar in the two Waves. However, patients from Wave 2 had a significantly lower MMS at hospital discharge (p < 0.001), with 44 % having a score of seven (ability to walk >30 m unaided), compared to 83 % from Wave 1. As such, 66 % of patients from Wave 2 required further rehabilitation after discharge, compared to 49 % from Wave 1 (p = 0.003).

Discussion

This single centre study compares the demographics, clinical status and rehabilitation outcomes of patients admitted to ICU and mechanically ventilated during wave 1 and 2 with confirmed COVID-19. Patient demographics were largely comparable between waves and patients admitted during wave 2 received comparable medical intervention to those admitted during wave 1. This included prolonged periods of sedation, frequent use of neuromuscular blockade and prone positioning. Despite tracheostomy protocols remaining consistent during waves, significantly less tracheostomies were performed during wave 2. This may be explained by a greater understanding of disease trajectory coupled with greater knowledge and experience in medical management, or the lower APACHE 2 scores on admission. Although statistically different when compared to the first wave, the incidence of ICU-AW remained high amongst both cohorts.

The return to pre pandemic physiotherapy staffing ratio’s, with rehabilitation only routinely delivered Monday to Friday had a significant negative impact on rehabilitation outcomes. Patients in wave 2 demonstrated lower mobility levels at the point of ICU discharge in comparison to wave 1. This was despite slightly better nursing and ICU consultant ratios in wave 2. It was noted that there was a delay in initiation of rehabilitation for patients in wave 2, who took on average two days to first mobilise after the cessation of sedation in comparison to just 1 day for those patients in wave 1 (McWilliams et al., 2021). Whilst this did not reach statistical significance, clinically the time to first mobilise has been proved to be an important metric when evaluating the impact of structured rehabilitation programmes (Pun et al., 2019). Importantly, this delay in mobilisation corresponded with patients in wave 2 requiring longer periods of mechanical ventilation, longer periods in the ICU, and ultimately leaving hospital with lower levels of mobility and higher rehabilitation needs (Fig. 1 ).

Fig. 1.

Impact of time to mobilise on patient outcomes. ICU, Intensive Care Unit. LOS, Length of Stay.

NHS England advocates a seven day service provision across all NHS services to ensure equality of care for all (Delivering the Forward View: NHS planning guidance, 2016). Despite this, the majority of physiotherapy services continue to provide only emergency respiratory physiotherapy at weekends on critical care. Evidence has demonstrated that an enhanced physiotherapy provision on critical care across five days (Monday to Friday) to deliver early and structured rehabilitation in line with NICE guidelines CG83, can deliver improved patient and hospital outcomes (Monsees et al., 2022, Zhang et al., 2019). However there is limited literature looking at the impact of an enhanced 7 day rehabilitation services across ITU compared to five day rehabilitation on patient outcomes. Case study reports published by NHS improvement (NHS improvement, 2012) involving UK NHS trusts suggest seven day physiotherapy services may result in more consistent rehabilitation on ICU and improved quality, patient outcomes and reduced ICU and hospital length of stay. However a lack of reported methodology, outcome measures and the wider impact of service re-design on patient outcomes are not explored, particularly as some services were re-structured without additional staffing resource. Published work within sub-acute rehabilitation hospitals have demonstrated a reduction in length of stay with the implementation of a seven day service (DiSotto-Monastero et al., 2012). However the generalisation of these finding to an acute ICU setting is questionable. Further research to assess the impact of a properly funded, gold standard seven day rehabilitation service on patient outcomes is warranted.

The delayed initiation of rehabilitation appeared to have an ongoing impact following step down from ICU to the ward. Lower mobility levels at the point of hospital discharge were seen in patients during wave 2 with an increased need for ongoing community rehabilitation. A recent systematic review identified that delayed mobilisation within the ICU setting is associated with poorer functional capacity, reduced walking capacity and reduced health related quality of life at hospital discharge (Arias-Fernández et al., 2018). In response to the ongoing long-term morbidity experienced by COVID-19 survivors, several regional pathways were commissioned after the first wave to support ongoing rehabilitation and facilitate early discharge from the acute setting. There was also intense pressure within the Trust to create capacity due to the desire to re-establish and maintain pre-pandemic service provision. Ultimately, this may have had an impact on the discharge destination of patients, and their ongoing rehabilitation needs. Consequentially, for the outcomes seen in wave 2, patients’ discharge from hospital could be facilitated sooner than was possible in the first wave. However, despite this, post ICU length of stay was equal between the 2 cohorts.

The unit under evaluation has a well-established multi-disciplinary culture of early mobilisation within the ICU and unit understanding regarding the benefits of early mobilisation (McWilliams et al., 2015). Whilst rehabilitation is often led and coordinated by the physiotherapists, actual rehabilitation delivery involves close multidisciplinary working. This is underpinned by a supportive structure that includes multidisciplinary team rounds, utilising the shared expertise of team members to discuss rehabilitation in the context of medical stability, weaning of sedation and respiratory support, management of delirium and other member tasks which may require completion (Bakhru et al., 2016). The wider changes within the critical care workforce would likely have also had a significant impact on these structures. The provision of nursing at a ratio of one specialist ICU nurse to three patients, with support from a non-ICU nurse and a non-ICU medic, would have significantly reduced the overall expertise available to support complex decision making. Whilst a task-based approach to patient care ensured capacity could be increased, it would be unrealistic to expect a redeployed member of staff to perform at the same level as an experienced critical care nurse, with those redeployed describing the change as moving from ‘expert to novice’ or ‘being thrown in the deep end’ (Tang et al., 2021). This changing skill mix, or the associated high turnover of staff seen during the pandemic would negatively affect the established rehabilitation culture in the unit, potentially reducing the likelihood of rehabilitation plans being followed or progressed over a weekend when physiotherapy support was not available.

The limitations of a five day rehabilitation service meant emergency respiratory physiotherapy and only limited rehabilitation was provided at the weekend during wave 2, by a team consisting of redeployed respiratory competent physiotherapists working within the Trust but not currently working on ICU. Consequently, reduced physiotherapy provision, coupled with an increased proportion of non-specialist staff adversely affected consistency of rehabilitation planning and delivery. In real terms, this likely meant patients appropriate for rehabilitation between Friday afternoon and Monday morning may have only been reviewed from a respiratory perspective. The staffing model adopted during wave 1 allowed maintenance of a more highly skilled critical care workforce across seven days which provided consistency and familiarity with the COVID-19 patient cohort. This importance of this skill mix should not be underestimated, with consistent staffing and cohesive team working associated with improvements in ICU mortality and patient outcomes (Wheelan et al., 2003).

Limitations

Our study has a number of limitations. As a single centre observational study this may not be reflective of other ICU COVID-19 populations. Data was collected mixed prospectively and retrospectively for wave 2 however, it was directly compared to historical data from wave 1. Although populations were statistically similar, aspects of ICU care and management evolved between waves, which could have impacted patient outcomes e.g. pharmacological strategies. In addition, with the emergence of new strains e.g. delta COVID-19 variant (B.1.617.2) and the Kent variant in 2021, what impact this had on patients’ outcomes is beyond the scope of this study. Finally, due to service pressures we were unable to record information relating to the consistency and frequency of rehabilitation delivered during either wave which potentially could have an impact on functional outcomes.

An important consideration when evaluating these findings is the changing knowledge and understanding of COVID-19 over the course of the pandemic. Whilst intensive care treatments such as the use of invasive ventilation and sedation remained broadly similar, our study does not consider the specific impact of other novel therapies which may have evolved into use for the second wave in comparison to the first. The most notable of these is likely to be an increased use of dexamethasone, which was shown to significantly reduce mortality in those patients requiring mechanical ventilation or supplementary oxygen (Recovery Collaborative group, 2021). It could be hypothesised that the increased use of dexamethasone was associated survival of a proportion of patients who may not have survived during the first wave. Whilst survival increased, this may have had a greater impact on associated morbidity and could explain to some extent the worse physical outcomes for those patients evaluated during wave 2. In addition, the use of corticosteroids has been significantly associated with an increased incidence of ICU acquired weakness (Yang et al., 2018). Whilst actual rates of weakness observed were actually lower in wave 2, we did not capture severity of ICU acquired weakness which may have impacted in rehabilitation progress and recovery trajectory.

Conclusion

Patients admitted with SARS Cov-2 during the second wave were comparable to patients admitted during wave 1 in both demographics and their medical management whilst on critical care. The change in physiotherapy staff provision from a seven day rehabilitation service during wave 1 to a five day rehabilitation service with emergency respiratory physiotherapy at a weekend in wave 2 resulted in patients taking longer to commence mobilisation and more dependant at ICU and hospital discharge. This study highlights the benefits of having a seven day physiotherapy service to deliver consistent rehabilitation and should be considered by commissioner’s when reviewing physiotherapy workforce structure. This may also have the added benefit of reducing demand on community and follow up services.. It is worth noting that analysis of the predominant COVID-19 strain between the 2 waves was beyond the scope of this study and therefore any differences in disease progression and pathology are unaccounted for.

Funding

No additional funding was received to complete this project.

Authors contributions

JW helped design the study, conduct the study, analyze and interpret the data, and draft and critically revise the manuscript. AH helped design, interpret the data, draft and critically revise the manuscript. KB helped interpret the data, draft and critically revise the manuscript. CS helped interpret the data, and draft and critically revise the manuscript. DM helped design the study, conduct the study, analyze and interpret the data, draft and critically revise the manuscript. All authors read and approved the final version of the manuscript.

Ethics statement

This project constituted an observation of standard care delivery with no randomization and thus met the definition of a service evaluation under the National Health Service Health research authority guidelines. As such, ethical approval was not required, and because all outcome measures are collected as part of routine care, the need for consent was waived.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.