(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Economic impact of a vision-based patient monitoring system across five NHS mental health trusts [1]

['Ciara Buckley', 'York Health Economics Consortium', 'York', 'United Kingdom', 'Robert Malcolm', 'Jo Hanlon']

Date: 2024-12

A vision-based patient monitoring system (VBPMS), Oxevision, has been introduced in approximately half of National Health Service (NHS) mental health trusts in England. A VBPMS is an assistive tool that supports patient safety by enabling non-contact physiological and physical monitoring. The system aims to help staff deliver safer, higher-quality and more efficient care. This paper summarises the potential health economic impact of using a VBPMS to support clinical practice in two inpatient settings: acute mental health and older adult mental health services. The economic model used a cost calculator approach to evaluate the potential impact of introducing a VBPMS into clinical practice, compared with clinical practice without a VBPMS. The analysis captured the cost differences in night-time observations, one-to-one continuous observations, self-harm incidents, and bedroom falls at night, including those resulting in A&E visits and emergency service callouts. The analysis is based on before and after studies conducted at five mental health NHS trusts, including acute mental health and older adult mental health services. Our findings indicate that the use of a VBPMS results in more efficient night-time observations and reductions in one-to-one observations, self-harm incidents, bedroom falls at night, and A&E visits and emergency service callouts from night-time falls. Substantial staff time in acute mental health and older adult mental health services is spent performing night-time observations, one-to-one observations, and managing incidents. The use of a VBPMS could lead to cost savings and a positive return on investment for NHS mental health trusts. The results do not incorporate all of the potential benefits associated with the use of a VBPMS, such as reductions in medication and length of hospital stay, plus the potential to avoid adverse events which would otherwise have a detrimental impact on a patient’s quality of life.

National Health Service (NHS) inpatient mental health services face various challenges that stretch staffing resources such as frequent patient observations, falls, assaults, self-harm and suicide. Often the required staffing levels cannot be met with nurses and healthcare assistants deployed to specific wards (substantive staff). This means that bank and agency staff are required, which can be costly to individual NHS practices [ 1 ]. A vision-based patient monitoring system (VBPMS) is an assistive tool–used by approximately half of NHS England’s mental health trusts–that enables non-contact monitoring of patients and is intended to support staff to deliver safer, higher-quality and more efficient care. We found that the use of a VBPMS results in more efficient night-time observations and reductions in one-to-one observations, self-harm incidents, bedroom falls at night, and A&E visits and emergency service callouts from night-time falls. Our study highlights the potential benefits of a VBPMS both in terms of resource use and patient safety.

Funding: York Health Economics Consortium (YHEC) were funded by Oxehealth. Oxehealth had no role in the study design of the economic model. Oxehealth were involved in the data collection and data analysis and Oxehealth were jointly involved in the decision to publish and reviewed the final manuscript.

Data Availability: The data in this paper were sourced from studies and service evaluations across five anonymous NHS trusts. Due to confidentiality agreements and the dispersed nature of the data, we are unable to designate a single, third-party entity for data access requests. Interested parties may contact either YHEC's or Oxehealth's ( [email protected] or [email protected] ) data departments to arrange for data access (with appropriate approvals) for anyone concerned. As explained, because there was not one single ethics committee for this collected data, and NHS trusts involved currently wish to remain anonymous, we can facilitate this request with the appropriate trust.

Copyright: © 2024 Buckley et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

This paper summarises the potential health economic impact of using a VBPMS to support clinical practice in two inpatient settings (acute mental health and older adult mental health services) using updated cost and clinical data. We demonstrated that the use of a VBPMS could lead to cost savings and a positive return on investment for NHS mental health trusts.

The aim of this analysis is to adapt a previous health economic model with clinical data collected from additional NHS mental health trusts, and to present the results in an accessible format to inform healthcare decision makers. In order to assist healthcare decision makers, the analysis was conducted from two perspectives: NHS and Personal Social Services (PSS) in England and an NHS mental health trust in England. This is because the impact on each of their budgets will be different, given the mental health trust is a local part of the wider NHS, and so will only cover a specific part of the healthcare service. The cost of external treatment outside of mental health trusts including GP visits or emergency hospital attendance are expected to fall on integrated care systems (ICS) who are responsible for planning and budgeting of health services at regional level. The total cost, incremental cost savings, return on investment, cash- releasing and opportunity cost savings of a VBPMS compared to clinical practice without a VBPMS are identified in older adult services and acute mental health services.

A health economic model was developed to assess the impact of using a VBPMS with clinical practice, compared to clinical practice without a VBPMS. Previous research demonstrated the potential cost savings to the health system, driven by a reduction in night-time observations, one-to-one observations, bedroom self-harm, bedroom falls and assaults from one NHS mental health trust [ 7 – 11 ].

A vision-based patient monitoring system (VBPMS) is an assistive tool that supports patient safety by enabling non-contact physiological and physical monitoring [ 7 ]. Oxevision (Oxehealth Ltd, Oxford UK) is a VBPMS fitted in the patient’s bedroom, that uses an infrared-sensitive camera to detect patient location and motion and measures vital signs (pulse rate and breathing rate). Staff can access this information through handheld devices and a fixed screen in the nurse’s station. The VBPMS also sends alerts to these devices if it detects the patient is in a higher risk location and allows staff to view a 15-second, privacy-controlled video feed when measuring a patient’s vital signs to ensure patient safety.

During 2021/2022, over 3.2 million people were in contact with the National Health Service (NHS) in England for secondary mental health care, learning disability or autism services [ 2 ]. There are around 50 NHS mental health trusts in England which provide a variety of services for people with mental health problems as part of the socialised health service in England [ 3 ]. Patient safety is a priority, with NHS acute mental health units and older adult services facing challenges such as patient falls, assaults, self-harm and suicide [ 4 ]. NHS England guidance aims to support those who require acute inpatient mental health care, including older adults, to receive high quality therapeutic care in the least restrictive setting possible [ 5 ]. Staff are recognised as having a key role in providing safe patient care, by giving medicine, using therapeutic techniques and managing potentially dangerous patient behaviours [ 4 ]. In a survey of an NHS acute mental health inpatient ward, 79% of staff (N = 43) reported that they felt a patient could have an incident in their room without their knowledge [ 6 ]. To support patient safety, patients are risk assessed to determine the frequency at which observations should be conducted: usually hourly, every 15 minutes or one-to-one continuous observation for high-risk patients [ 7 ]. Night-time observations, one-to-one observations and managing incidents which cause harm to patients are labour-intensive and can be especially problematic where there are staff shortages. The employment of bank and agency staff (outside of the core ward staff) is expensive and reducing staff time spent on observations would reduce the need to bring additional staffing onto wards. Furthermore, as well as causing harm to patients and staff, incidents can take up a great deal of staff time. As bank and agency staff are often required to increase the staff capacity to perform one-to-one observations, a reduction in this activity would lead to a cash releasing opportunity for the NHS. Reducing time required for night-time observations, managing incidents and undertaking one-to-one observations can also lead to a reduction in opportunity costs, enabling staff resource to be reallocated to improve patient care and support other therapeutic activities [ 8 ].

An additional key summary outcome reported is return on investment (ROI). This is calculated by dividing the net benefit of the intervention (incremental cost) by the cost of the intervention. This value for ROI is presented as a value, where anything over 1 represents a positive return. Further detail for calculating ROI is contained in S4 Appendix .

The primary outcomes generated from the model were cost per occupied bed day, cost per patient, cost per average ward per year and cost to mental health NHS trusts. The economic model also considers whether costs savings are cash-releasing or an opportunity cost saving. Cash-releasing cost savings relate to whether the saving would produce a monetary return, whereas opportunity cost savings result from resources being released which could be used for other activities. The model considers one-to-one observation hours to be the only event that has a cash-releasing cost saving component, because one-to-one observation hours can require additional resource for a ward beyond planned staffing levels. When this occurs, bank and agency staff are used. Therefore, by reducing one-to-one observation hours, both staff time and bank and agency staff time can be saved.

The costs of implementing a VBPMS include an annual license for use of the system, installation costs, cabling costs and staff training costs. The startup costs of implementing a VBPMS were annuitised over 10 years, the anticipated lifespan of the system’s hardware. In line with the one-year time horizon of the model, only one year of this annutised cost was included in the model.

The economic values for each of the relevant metrics were established using the sources listed in Table 2 . Staff costs collected from PSSRU 2022, and the NHS Cost Collection 2021/22 were used for procedures. Costs associated with bedroom falls at night were sourced from a 2017 NHS Improvement report, which were inflated using the NHS Cost Pay and Prices Inflation Indices [ 15 ]. Further information on how staff hours and procedures were calculated across both populations is provided in S1 Appendix .

The events recorded in each of the NHS trusts were night-time observation hours per patient, number of one-to-one observation hours, number of bedroom self-harm incidents (acute adult population only), number of bedroom falls at night (older adult population only), number of A&E visits resulting from bedroom falls at night (older adult population only), and number of emergency service callouts resulting from bedroom falls at night (older adult population only). Table 1 details the events collected from the clinical studies and the population from which data were collected. Where multiple trusts were used, weighted averages were calculated based on the number of bed days in the study period.

Data used to populate the model were informed by numerous sources, including five clinical before and after studies conducted across five mental health NHS trusts. Cost inputs were extracted where possible from publicly available sources such as Personal Social Services Research Unit (PSSRU) and NHS Cost Collection [ 13 , 14 ]. The outcomes from the clinical studies were adjusted within the model to a standardised metric known as ’per occupied bed days’. This adjustment was made to create more universally applicable results, enabling the estimates to be applied to various populations and settings. The data were scaled to an average ward with 16 beds and 90% occupancy. This assumes that the impact of a VBPMS is scalable to different ward sizes; however, the ward sizes ranged from 16 to 24, so were not substantially different across the NHS trusts involved. Further details on the scalability of the results are presented in S1 Appendix .

Ethical approval was obtained from a research ethics committee (REC) for one of the studies associated with this research. The four other studies were service evaluations exempt from REC approval. The authors of this paper can be contacted for further information.

An economic model was developed previously to assess the potential cost savings from the use of a VBPMS in the NHS in England [ 8 , 9 ]. The economic model was updated with additional cost and clinical data for this analysis. The model was developed in Microsoft Excel (Microsoft Corporation) and was structured as a cost calculator. This structure used data to compare the resource use and costs of specified events with and without a VBPMS, across the two populations, for the specified events in each population. Quality of life was not captured in the model, in line with NICE’s approach to the evaluation of medical devices [ 12 ]. A diagram of the model structure is displayed in Fig 1 . Previous analysis provides further detail and the rationale for the approach to the economic evaluation [ 8 , 9 ]. Future technological innovation is not captured within the model, as it is not possible to predict future innovations in this service area. NICEs approach to evaluating medical devices also does not account for future technological innovations.

The time horizon is one year. Therefore, in line with the NICE methods guide, discounting of future costs was not considered [ 12 ]. The VBPMS is based on an annual subscription, meaning a time horizon of one year was appropriate for evaluating its impact.

The results are presented from two perspectives: NHS and PSS in England and an NHS mental health trust in England. The NHS and PSS perspective includes all costs in the model while the NHS mental health perspective did not include costs where any patient needed external treatment outside of the mental health hospital. The model is designed to align with the standards for economic modelling of medical devices set out by the National Institute for Health and Care Excellence (NICE) [ 12 ]. Further details on the perspectives are presented in S2 Appendix , for wider context and audiences less familiar with the NHS in England.

From an NHS mental health trust perspective there was an estimated incremental benefit of £265,376 from the adoption of a VBPMS. The ROI is 9. Night-time observations and one-to-one observations have the same reduction as those reported from an NHS and PSS perspective. There was a reduction of £17,443 in bedroom falls in an NHS mental health trust. This is significantly lower than for the NHS and PSS perspective, as the costs of a bedroom fall are more likely to fall on integrated care systems (ICS), including aspects such as GP visits, litigation costs, or additional funding to account for longer hospital stays.

Cash-releasing savings from reducing one-to-one observation hours which require bank and agency staff were £107,101 with an additional £140,335 representing opportunity cost savings for substantive staff. The VBPMS was estimated to provide a 68% reduction in A&E visits, which would lead to a £18,308 cost reduction within the NHS. The cost of bedroom falls at night was reduced by £142,948 from an NHS and PSS perspective. Additionally, there was an estimated cost saving of £4,461 in emergency service callouts resulting from bedroom falls at night.

The introduction of a VBPMS in an average sized older mental health ward (16 beds, 90% occupancy) was estimated to reduce net costs by £413,651 for older mental health services over a one year period, from the perspective of NHS and PSS. This gives an ROI of 14.04 including both cash-releasing and opportunity cost savings.

The data relating to the impact of a VBPMS on one-to-one observation hours were collated from seven wards across three trusts. The impact on one-to-one observation hours can have a cash releasing impact on trust budgets. The overall reduction in one-to-one observation hours when using a VBPMS together with clinical practice was 40%.

Night-time observation data were collected from one NHS mental health trust in the older adult mental health service. The proportion of patients requiring night-time observations was 90%, with approximately half requiring observations every 15 minutes, and the other half requiring observations every hour. There was a 50% reduction in the time taken to complete night-time observation rounds with a VBPMS. This is assumed to lead to opportunity cost savings, as night-time observations are generally led by substantive ward staff.

The clinical events captured in the model and the reduction of their incidence after the introduction of a VBPMS in older adult mental health services are detailed in Table 5 and have been scaled to per occupied bed day.

From an NHS mental health trust perspective there was an estimated incremental benefit of £89,305 from the implementation of a VBPMS. The ROI from the total benefits value is 3.03. Night-time observations and one-to-one observations have the same reduction as those reported from an NHS and PSS perspective. The cash-releasing savings from the reduction in one-to-one observation hours would be more than the cost of implementing a VBPMS, resulting in a cash-releasing saving per ward. There was an £8,449 reduction in self-harm incidents in the NHS mental health trust. The lower savings reported from the NHS mental health trust perspective is driven by A&E visits associated with self-harm. While A&E visits associated with self-harm drive wider savings in the NHS, these benefits do not represent a saving to the ward.

The cash-releasing savings from reducing one-to-one observation hours requiring bank or agency staff were estimated at £75,895 with £27,038 opportunity cost savings for substantive staff. The cash-releasing savings from the reduction in one-to-one observation hours would be more than the cost of implementing a VBPMS, resulting in a cash-releasing saving per ward. There was a £12,578 reduction in the cost of self-harm incidents with the NHS.

The introduction of a VBPMS in an average sized acute mental health ward (16 beds, 90% occupancy) was estimated to reduce net costs by a total of £93,433 over a one year period, from the perspective of the NHS and PSS. This gives an ROI of 3.17. This indicates that for every pound invested using VBPMS, the NHS will save £3.17, including cash-releasing and opportunity cost savings.

The cost impact of adopting a VBPMS from the NHS and PSS, and NHS mental health trust perspectives is detailed in Table 4 . The total cost of the Oxehealth Service is £29,457 per ward per year, which includes an annual license fee, installation, and cabling costs.

The reduction in self-harm incidents with a VBPMS was calculated as a relative reduction, by comparing the change between the wards with a VBPMS and the control wards. The number of self-harm incidents reduced by 44% when a VBPMS was introduced, compared with clinical practice without VBPMS.

The data relating to the impact of one-to-one observation hours was collated from eight wards across four trusts for data on cash releasing staff activities, and two wards across one trust for substantive staff activities. The impact on one-to-one observation hours can have a cash releasing impact on trust budgets. The overall estimated reduction in one-to-one observation hours when using a VBPMS alongside clinical practice was 24.4% (a weighted average of 26.2% and 20.4%—see Table 3 ). This is split by cash-releasing savings and opportunity cost savings ( Table 3 ).

For the acute mental health services population, night-time observation data were collected from one mental health NHS trust. The proportion of patients requiring night-time observations was 90%, of which approximately 20% required observations every 15 minutes, and 80% required observations every hour [ 16 , 17 ]. There was a 45% reduction in the time taken to complete night-time observations. Night-time observations are generally performed by substantive ward staff (allocated as core staff to the ward). Any savings generated by reducing night-time observations are likely to be opportunity cost saving, i.e., releasing substantive staff time for other activities.

Results from the clinical studies to evaluate the impact of a VBPMS in acute mental health services are detailed in Table 3 . The number of events for clinical practice without a VBPMS have been scaled to per occupied bed day.

Discussion

The implementation of a VBPMS in clinical practice, compared with clinical practice without a VBPMS, is estimated to be cost saving when considering acute mental health services and older adult mental health services. The results indicate that using a VBPMS could lead to an incremental benefit per ward of £93,433 in acute services and £413,651 in older adult services from an NHS and PSS perspective. Additionally, a VBPMS was estimated to be cost saving within a mental health NHS trust, with an overall incremental benefit per ward of £89,305 and £294,848 in acute and older adult mental health services, respectively.

If a VBPMS was implemented across all of NHS England for acute mental health services, with approximately 18,400 beds for mental healthcare available (of which 65% are acute services), this could lead to approximately £69 million in cost savings, assuming 90% average occupancy [17]. Additionally, around half of these cost savings would be cash-releasing to the healthcare system. If a VBPMS was implemented across all of NHS England for older adult mental health services, with around 18% of beds for mental healthcare assigned to older adults, this could lead to cost savings of approximately £89 million [17]. Furthermore, around one-third of these cost savings would be cash-releasing to the healthcare system.

The largest driver of cost savings is the reduction in one-to-one observation hours in both populations. A proportion of one-to-one observations is estimated to be cash releasing for the NHS, as bank and agency staff are often required to carry these out. It is estimated that a VBPMS could reduce the need for bank and agency staff related to one-to-one observations by 26.2% and 40.4% in acute and older adult mental health services, respectively, leading to net cash-releasing savings of £68,854 and £141,452. The reduction in night-time observations across an average ward per year is expected to free up resources which can be reinvested into patient care and engagement, in both acute and older adult mental health services. The total cost saving per year to the NHS in acute mental health services is £7,380 and £29,472 in older adult mental health services. The reason savings were higher in the older adult services is driven by the increased requirement for more regular observations in this population, which were more likely to occur every 15 minutes, rather than every hour. Furthermore, on average, night-time observations took longer in the older adult services, meaning any reduction in night-time observations with a VBPMS has a greater impact.

Self-harm incidents are a significant challenge within acute mental health services. The number of self-harm incidents reduced by 44% with the adoption of a VBPMS. This has a substantial impact on patient safety. Night-time bedroom falls also pose a threat to patient safety for those in older adult mental health services. Falls can often result in A&E visits and require emergency service callouts. The use of VBPMS, compared with clinical practice without a VBPMS, resulted in a reduction of night-time bedroom falls, A&E visits, and use of emergency service callouts.

The economic model intended to capture the potential health economic impact of using a VBPMS to support care in two inpatient settings: acute mental health and older adult mental health services. However, there are likely to be additional impacts which are not captured within the analysis. For example, qualitative evidence suggests that there are benefits to both staff and patient experience through use of the system [6]. In relation to patient safety, the reduction in incidents, such as self-harm, is likely to make patients and their families feel safer. Patients may also benefit from improved sleep, due to staff being able to carry out night-time observations without entering patients’ rooms.

Alongside the cost effectiveness, the implementation of a VBPMS requires patient and staff acceptance. Previous research reported that hospital staff had embraced the VBPMS analysed in this study, as they felt patient care had improved with introduction of the system [11]. For example, 91% of staff in an acute mental health hospital reported that a VBPMS allowed them to make better care and clinical decisions [6]. Staff in an acute inpatient mental health unit reported when surveyed that a VBPMS improved patients’ sleep at night, while also improving staff experience. They also reported a reduction in verbal and physical aggression towards staff [18]. In a qualitative study conducted at a high secure forensic hospital many patients preferred non-invasive monitoring to physical checks as physical checks were deemed annoying and disruptive to their sleep [19]. Other research suggests patients feel reassured of their safety and that their wellbeing improves from using the VBPMS [6,7].

In the older adult population, broken bones and fractures from falls could have longer term costs such as rehabilitation, which are captured in the model. Evidence for reduced length of inpatient stay has yet to be captured; however, a reduction in incidents could potentially lead to a shorter length of stay. Serious incidents such as night-time bedroom falls, and self-harm incidents often require a lengthy internal review and may lead to expensive legal costs. The model captures the cost of litigation; however, it does not capture the costs associated with an internal review. Nevertheless, this is a rare occurrence, with only around 0.2% of falls estimated to involve litigation [15,20].

Similar studies across literature focusing on patient monitoring technologies within the healthcare system report similar findings [11,21–24]. For example, one study reported an estimated annual cost saving of £1.3 million per hospital with the introduction of continuous and intermittent patient monitoring, compared to intermittent monitoring in patients admitted to surgical wards [22]. Similarly, a technology which provided continuous in-hospital patient monitoring was found to improve patient safety, with the fall rate decreasing from 1.8 to 0.6 per 10,000 and reported being more cost effective than traditional methods of patient monitoring, such as individual nurse visits to the patient’s bedside [24]. Although these technologies all differ in their targeted population and use case, they tend to suggest technology assisted monitoring can result in cost-savings for healthcare systems.

[END]
---
[1] Url: https://journals.plos.org/digitalhealth/article?id=10.1371/journal.pdig.0000559

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.