Background

This report summarises the findings of an evaluation conducted by Eastern Academic Health Science Network (AHSN) and Health Innovation Manchester (HInM), the AHSN for Greater Manchester, on behalf of the national Innovation Collaborative for digital health.

The collaborative is a learning and support system for health and care professionals delivering technology-enabled innovations to support people at home.

It is commissioned by NHS England and delivered in partnership with the AHSN Network, with HInM as lead AHSN.

Executive Summary

About the report

Purpose 

The report presents findings from an evaluation of a chronic obstructive pulmonary disease (COPD) virtual ward that falls within a virtual hospital managed by South and West Hertfordshire Health and Care Partnership. The COPD virtual ward is one of two virtual wards that fall within the South and West Hertfordshire Health and Care Partnership Virtual Hospital, the second being a heart failure virtual ward. 

It aims to inform the potential wider adoption of the virtual ward model across the UK and understand the model’s potential to support people with other health conditions. It also considers the success of South and West Hertfordshire Health and Care Partnership Virtual Ward objectives to improve patient care, clinical outcomes, healthcare utilisation, and patient and staff satisfaction.  

Overview 

A virtual hospital can be defined as an all-encompassing group of virtual wards that cover a range of conditions, allowing patients to receive the care they need at home safely and conveniently, rather than being in hospital. Supporting patients at the place they call home, this model of care can include using remote monitoring apps, technology platforms and medical devices such as pulse oximeters. Support may also involve face-to-face care from multidisciplinary teams based in the community.  

Patients are admitted to the virtual ward upon discharge from an inpatient bed and monitored continuously using technology, with regular contact from a multidisciplinary team. Patients can also be admitted to the virtual ward from A&E or a GP home visit. The evaluation sought to understand seven areas of interest: 

• the typical characteristics of virtual ward patients 
• virtual ward usage patterns 
• potential of the virtual ward to reduce secondary care length-of-stay 
• potential of the virtual ward to reduce secondary care readmissions 
• virtual ward patient experiences and views 
• staff experiences and views on virtual ward implementation, delivery and impact  
• indicative virtual ward cost-benefit in terms of immediate and long-term impact on healthcare utilisation.   

Method 

The evaluation adopted a mixed method approach involving quantitative, qualitative and economic analysis. It utilised data recorded routinely by South and West Hertfordshire Health and Care Partnership clinical systems and by the continuous monitoring technology, supplemented with information from staff interviews and virtual ward patient surveys. 

Quantitative analysis involved a three-arm cohort comparison, designed to increase the possibility of determining the incremental impact of service features. 

• Historic control: patients admitted to West Hertfordshire Teaching Hospitals NHS Trust pre-pandemic from December 2018 to June 2019 inclusive (n=365). 
• Multidisciplinary team only: patients assessed by the multidisciplinary team during the intervention window but not onboarded to the virtual ward from December 2021 to June 2022 inclusive (n=248).  
• Virtual ward patients: patients assessed by the multidisciplinary team and onboarded to the virtual ward from December 2021 to June 2022 inclusive (n=46). 

The historic control time period was chosen as the base comparator because it is the last full year prior to COVID-19 and the December to June window is determined as directly comparable to the intervention window, accounting for seasonal effects and winter pressures. 

Additionally, the trust’s peak for NHS England performance against waiting time benchmarks for emergency patients peaked in 2019, when a pilot project named Senior Medics Assessment Review and Treatment (SMART) was running. This involved redesigning the patient pathway to ensure that patients had the input of a medical consultant at an earlier point than had previously happened. 

SMART helped to move the trust’s four-hour performance nationally from the lower to the upper quartile and also won the 2020 Healthcare People Management Association Academy Wales Award for Excellence in Organisational Development. This means 2019 data represents the most stringent comparator and a genuine gold standard against which new projects should be measured. 

The Virtual Hospital Programme at West Hertfordshire Teaching Hospitals NHS Trust was initiated as part of a broader piece of transformation work which includes the asthma, bronchiectasis and COPD multidisciplinary team. All airways disease admissions are discussed here and the experience of GPs, specialist nurses (both hospital and community-based), respiratory consultants, physiotherapists, physiologists and palliative care consultants are brought to bear. Virtual hospital decision-making is endorsed through this meeting as the senior decision-making platform. Exploring the multidisciplinary team-only outcomes attempts to isolate the multidisciplinary team-effect from the virtual hospital effect.

Key findings  

Overall, it is evident the virtual ward model represents a safe, effective alternative to inpatient care; improving outcomes, positively impacting patient experience and delivering cost efficiencies compared to traditional care pathways.  

Analysis shows a positive impact on patient care across almost all measures, although some notable limitations should be considered including the size of the intervention cohort and the rapid nature of this evaluation. 

• The virtual ward demonstrated favourable healthcare utilisation outcomes compared to the other cohorts, with an observed reduction in both inpatient length-of-stay, arising from increased clinical confidence, and a reduction in the number of repeat readmission events.  
• Virtual ward patients at risk of deterioration are more likely than patients discharged without the virtual ward support to be identified in a timely way and escalated appropriately back into an acute inpatient setting.  
• Patient experience was positively impacted, with the majority reporting feeling well prepared for their transition to the virtual ward and then safely cared for in the community. 
• Staff gave positive feedback and clear recommendations for continued improvement, including developing additional patient information, future development of the technology pathway to support virtual consultations routinely, and enhanced referral routes. 
• The virtual ward demonstrates a positive benefit-cost ratio based on both initial set-up costs and recurring resources required to sustain the model, suggesting a favourable economic case.   

Findings by area  

Typical characteristics of virtual ward patients 

The virtual ward patient cohort tended to be older than the multidisciplinary team-only and historic control cohorts, with more than 80% aged 60 or over (38 of 46).  Dyspnoea, eosinopenia, consolidation, acidaemia, atrial fibrillation (DECAF) scores, which predict acute exacerbation of COPD, ranged between 0 (low risk) and 3 (high risk) for virtual ward patients. 

Those high-risk patients with a DECAF score of 2 or 3 suggest patients of a higher acuity were safely supported for early discharge using the virtual ward model than would otherwise occur without continuous monitoring.   

COPD was the primary diagnosis for 52% of virtual ward admissions. A different respiratory condition was the primary diagnosis for an additional 30% (82% in total), with COPD often featuring as a secondary diagnosis. 

Virtual ward usage patterns 

On average, approximately eight patients (8.3) per month were onboarded to the virtual ward. They were supported by one or more of four main methods of contact: a telephone call to record readings, telephone clinical review, virtual consultation, or face-to-face healthcare professional visit.  

The evaluation was based on 46 individual patients totalling 50 admissions to the virtual ward between December 2021 and June 2022 inclusive. Several patients were admitted to the virtual ward on multiple occasions.   

Virtual ward patients received four reading calls, one clinical review call, fewer than one virtual consultation and fewer than one face-to-face visit per day on average. This is in accordance with the virtual ward standard operating procedures planned call regime. Patients with a DECAF score between 0 and 2 received a similar number of visits and calls. 

The predicted length-of-stay for patients entering the virtual ward was 14 days, however for those admitted to the virtual ward: 

• the average virtual ward length-of-stay was 7 days
• 86% had a length-of-stay of 10 days or fewer (n=43)
• nearly half had a length-of-stay of 6 to 10 days (n=24, 48%) 
• 14% had a length-of-stay of more than 10 days (n=7) to a maximum of 22 days. 

Patients were discharged from the virtual ward to one of four discharge destinations. Of 50 admissions: 

• 12% were discharged with no further West Hertfordshire Teaching Hospitals NHS Trust action (n=6) 
• 66% were discharged for ongoing care by community care, Central London Community Healthcare NHS Trust (n=33) 
• 10% were discharged for re-discussion at the multidisciplinary team meeting (n=5)  
• 12% were re-admitted to a physical bed (n=6). 

The most common discharge destination for patients under each of the four DECAF scores was to ongoing care by Central London Community Healthcare NHS Trust (CLCH). Those with a DECAF score of 2 were least likely to be re-admitted to a physical bed. 

Potential to reduce secondary care length-of-stay  

The average (mean) physical hospital bed length-of-stay in secondary care was lower for virtual ward patients, suggesting the virtual ward model enables earlier supported discharge into the community. It was observed at: 

• 5 days for virtual ward patients 
• 1 days for multidisciplinary team-only patients 
• 2 days for historic control patients 

Potential to reduce secondary care readmissions  

Almost half of virtual ward patients (46%) were not readmitted to hospital as a secondary care emergency for COPD or any other cause within 90 days. Of the remainder: 

• 43% had at least one emergency readmission within 30 days 
• 4% had at least one emergency readmission within 31 to 60 days  
• 7% had at least one emergency readmission within 61 to 90 days 

For COPD causes specifically, the majority of virtual ward patients (70%) were not readmitted to hospital as an emergency. Of the remainder: 

• 20% were readmitted as an emergency within 30 days  
• 4% were readmitted as an emergency within 31 to 60 days 
• 7% were readmitted as an emergency within 61 to 90 days 

Non-emergency readmissions for COPD-specific or any other cause were moderately higher among virtual ward patients compared to the multidisciplinary team-only cohort.  

• 54% of virtual ward patients were readmitted for a COPD-specific cause compared to 48% of multidisciplinary team-only patients. 
• 30% of virtual ward patients were readmitted for another cause compared to 23 per cent of multidisciplinary team-only patients. 

Virtual ward rates are driven by identification of patients at risk of deterioration and appropriate escalation back into an acute setting. This accounts for 15% of all virtual ward patient readmissions occurring within 1 day or less of discharge from the virtual ward.  

When these occurrences are removed:  

• 22% of virtual ward patients were readmitted for a COPD-specific cause compared to 23% of multidisciplinary team-only patients 
• 46% of virtual ward patients were readmitted for any cause compared to 48% of multidisciplinary team-only patients 

The average number of readmissions per patient was higher in the multidisciplinary team-only compared to the virtual ward cohort, suggesting multidisciplinary team-only patients are more likely to have multiple readmission events (0.87 vs 0.78 average number of readmissions <90 per patient between the multidisciplinary team-only and virtual ward cohorts). 

Virtual ward patient experience and views 

All patients mentioned that being able to be in their own home while being monitored and being able to contact staff was what they liked most about the virtual ward.  

Some commented that this allowed them to feel safe, to feel that people were caring, and that being more relaxed was beneficial to their comfort and breathing issues. 

Generally, patients reported they felt adequately prepared and confident about using the equipment fully or to some extent. However, some reported challenges, which are grouped as those relating to using the equipment and aspects of communication. 

All respondents reported they were satisfied with the way concerns were answered, that they were given adequate information about risks, that the equipment was explained to them, and that they had been involved in decisions about their treatment.  

Respondents suggested that various staff roles had answered their questions, including the monitoring team, specialist nurse, on-call nurse, and respiratory consultant.  

Some provided suggestions for improvements, including: 

• extending out-of-hours advice line 
• closer liaison and more communication with various departments including with GP and social care or a social worker 
• providing a printed leaflet to give to patients “explaining what a virtual ward is and how it operates so they don’t feel they are just being discharged and forgotten” 
• screen the patients more. 

Staff experience and views  

Staff felt there was organisational readiness for the launch and implementation of the COPD virtual ward, highlighting staff and teamwork as a key enabler and critical to success. 

The referral process was described as being very quick, with some issues around making sure patient information like the discharge summary was available to the community team before the home visit on day 1 in the first month of implementation. This was acknowledged as being mainly resolved. 

Some staff mentioned some initial technical issues in the first few weeks that were then resolved. Concerns were expressed over potential problems when the level of care is removed on discharge from the virtual ward and whether the current model encourages self-monitoring and self-care in the most appropriate way for all patients. 

Overall, staff felt the right patients were being admitted to the virtual ward and the criteria, consent process and safety nets put in place for readmittance to the hospital meant the processes were working appropriately.  

Some staff commented that in the first few weeks there were some challenges relating to the appropriateness of patients being admitted to the COPD virtual ward. 

Staff received positive feedback from patients, who felt secure and liked the level of care provided by the virtual ward. Feedback from family and carers to staff was positive, including that the virtual ward enabled them to feel involved and had given them a better understanding of the condition and care.    

Indicative cost-benefit  

Overall, the virtual ward represents a cost-beneficial model for caring for patients with acute COPD exacerbations, based on the impact on reducing length of stay and the number of repeat readmissions within 90 days of discharge.  

The virtual ward demonstrated a positive benefit-cost ratio of £1.12:1 under the lower, more conservative scenario, or £1.45:1 in the upper bound scenario, when considering all costs incurred over the evaluation period. 

The net-present value, the difference between costs and benefits realised, was £6,370 over the evaluation period, or £24,400 per annum with optimism bias correction (OBC) included. 

Cost-benefit analysis outputs per lower and upper estimates, all costs included for duration of the evaluation period: 

Extrapolated over a 12-month period and excluding any initial, one-off expenditure, this ratio increases to £1.25:1 under the lower scenario and to £1.62:1 in the upper scenario with a net present value of £28,930 per annum. The per patient benefit per virtual ward episode is £263 to £656 respectively for each scenario.  

Cost-benefit analysis outputs per lower and upper estimates, recurring costs only, per annum:

Recommendations 

South and West Hertfordshire Health and Care Partnership should continue delivering the virtual ward and explore further expansion of the model into other care pathways outside of COPD and heart failure, considering our suggestions for further future improvements and learnings from initial implementation. 

There are several suggestions for any future analysis that builds on this evaluation. 

• Expanding the size of the intervention cohort to enable exploration of the degree of statistical significance of any observed effects. 
• Increasing completion rates of the patient survey to ensure full representation of the diversity of patient experience in a way which is recognised as reflective of the virtual ward cohort. 
• Involving data from other healthcare services supporting patients on the virtual ward, including through more extensive engagement with primary care. 
• Exploring wider impacts on patient quality of life, including through a cost-utility analysis leveraging recognised health economic techniques and through engagement with patients’ families and carers. 
• Reviewing the number of calls made to patients on the virtual ward and co-design remote interventions with patient groups to reduce staff resources and the cost attached.

Full Report

The full rapid evaluation report can be accessed on the FutureNHS platform, via the Innovation Collaborative for digital health workspace.

To access the report, and other shared learning resources related to digital home care and remote monitoring, you will need to be signed up to the FutureNHS platform and be a member of the Innovation Collaborative for digital health workspace. You will then be able to access the rapid evaluation report.

Publishing reference: PRN00214