The potential for tech-enabled virtual wards to transform the delivery of healthcare has been spotlighted by recent government funding pledges. This paper considers the topic of patient safety in virtual wards, based on the learning from a number of operational deployments by Doccla for a variety of pathways in acute and community settings. The overall message is that patient safety can be secured through following best practice within a culture of transparency and learning.
“Patient safety is the avoidance of unintended or unexpected harm to people during the provision of health care... Patients should be treated in a safe environment and protected from avoidable harm.” - NHS England1
The advancement of Internet-of-Things technology has afforded healthcare the opportunity to improve practice and change how care is provided. Remote Patient Monitoring (RPM) can replace the need for monitoring in secondary care, allowing patients to discharge early from hospital or avoid admission altogether. The patient recovers at home, with a clinician monitoring remotely, potentially many miles away. The benefits of this are abundant. Patients who succumb to a healthcare acquired infection (HCAI) have increased risk of mortality2, and the costs of managing a patient with a HCAI is three times higher3. Research has shown that a more positive perception of one’s environment improves patients’ perception of quality of care4; an adult patient is more likely to feel comfortable in their home environment, compared to sharing a room with people they don’t know and a curtain for privacy. Bed availability has been a long-standing problem,
particularly in winter. RPM frees capacity by ensuring only patients who
require hospital interventions take up a space. Furthermore, in most instances, RPM offers a higher ratio of patients to nurse compared to acute hospital requirements, therefore contributing financial incentives as well as logistical.
That said, the hospital setting with 24/7 medical supervision offers a higher level of safety for patients. Patients are discharged with RPM to a range of environments. The social setting will differ from patient to patient, as does the ability and education of the patients and those around them. The hospital setting offers control of variables that are for the most part, not possible to control in the community. Given that, the question arises about whether it is possible to ensure patient safety outside of the hospital setting. The answer in our experience is an emphatic yes, subject to two main pre-conditions, namely the implementation of effective formal controls and a culture that encourages transparency.
A well-established model for the management of clinical risk in tech-enabled virtual wards is provided by the DCB 0129 and DCB 0160 standards, which are mandated under Section 250 of the Health and Social Care Act 2012. The two standards are complementary views of the same topic, DCB0129 applying to the supplier of the virtual ward solution, while DCB0160 applies to the user.
Figure 1 shows the outline of the documents needed as part of meeting the standards.
In summary, the standards require that risks to clinical safety are formally identified, assessed and mitigated to an acceptable level before any patients are admitted to the virtual ward, and reviewed thereafter in the light of any actual or ‘near miss’ incidents. The standards defined the processes through which the information about the risks and their associated controls are formally handed over from the supplier to the client who will be responsible for implementing the controls.
Table 2. Definition of clinical safety terms.
The risks are assessed against a standard severity rating. Any clinical risk rated higher than 3 against the standard scale is treated as a showstopper that would prevent the deployment of the virtual ward unless it can be mitigated through appropriate controls to reduce the rating.
Doccla has implemented tech-enabled virtual wards for a variety of pathways in acute and community settings (Table 3). The role filled by Doccla across those deployments has gone beyond that of a tech-provider, typically with Doccla taking care of all the non-clinical aspects of service delivery, including:
Given this, the DCB0129 hazard assessment process has been applied to all these activities, not just to the remote monitoring software (which is provided as a cloud-hosted managed service).
Table 3. Sample of pathways currently provided by Doccla in both community and acute settings.
The DCB standards, require a multi-disciplinary approach. Doccla has established a Clinical Safety Committee (CSC), the membership of which includes clinicians, technical specialists, and service delivery heads, with overall responsibility for patient safety resting with the company’s chief medical officer.
The initial hazard assessment was drafted by a Doccla research nurse with input from individual staff members, the draft then being subject to a workshop with the full committee. The first approved version identified 11 potential hazards from 41 separate causes. The hazard assessment has been extended as new pathways and devices, or new versions of software, have been introduced.
Deciding how to best to manage hazards is informed in part by structured engagement with clients, which falls within the handover process specified in DCB 0160. This will include judgements to be made which have cost implications. For example, an obvious set of potential hazards is associated with the use of a patient’s mobile phone to relay vital signs and other health information. In our application of DCB 0129 we concluded that these risks were unacceptable, and so included the provision of a preconfigured mobile phone as part of the virtual ward equipment, the phones including software that allows them to be managed remotely. This increases the set-up cost for the ward, but significantly reduces the scope for operational risks.
The CSC meets monthly with an agenda that includes a number of standing items:
As of February 2022, there had been a total of 14 recorded ‘incidents’ across all of the virtual wards deployed by Doccla. None of the incidents caused any harm to patients, and some of them were trivial in nature, but in the interests of caution, learning and transparency, all were subjected to a full root cause analysis with corrective action to prevent recurrence.
Almost all the incidents were due to ‘human error’ of one form or another, where someone has failed to properly enter or update an item of information. As far as possible, the scope for human error is ruled out by automation—where that is not a practical option, additional checks can be included in standard operating procedures.
Some of the incidents had a more technical character. For example, the batteries in a device supplied to a patient failed after a short time, easily rectified by including spare batteries with all supplied devices. Another was a technical change on a trust’s network that led to the virtual ward being temporarily blacklisted by the trust’s firewall, again with a reasonably straightforward cure and workaround.
Significantly, none of the incidents was strictly related to the remote nature of the monitoring. All of them have parallels with incidents that could occur on a hospital ward.
Working closely with our clients and on some occasions providing clinical support, we have also learnt some lessons that pertain to maintaining patient safety from the viewpoint of the client/clinician. Despite clinical amenities being fewer in the community, it is important to recognise the resources that are available; such as the Patients themselves. RPM offers an opportunity for them to actively participate in their care and responsibility for their health. It is our experience that patients want to be engaged as they wish to avoid hospital admission. This may not be the case for everyone, but there are a great many who are capable, willing and want to have a deeper understanding of their health. Health education for patients will be individual and down to the clinicians’ assessment of understanding, but in our experience, patients are more than capable of learning the basics of what their different readings mean, and which symptoms to be mindful of. Safety-net advice is vital, as are clear explanations and reassurance to ensure the correct actions are taken if red flag symptoms arise. If the patient has family, this has also proven to be a useful resource and provides an additional layer of reassurance should the patient deteriorate. Finally, depending on the local structure of services, using community teams to undertake tasks that must be face to face, such as chest examinations or blood tests, means that patients can remain under the care of RM but still have access to in person clinical tasks without needing to attend ED.
Remote monitoring of patients offers an abundance of benefits for all stakeholders. It is essential to abide by the DCB guidelines for clinical risk management—with in-depth formal safety meetings before deployment and throughout the life cycle of live services—to maintain safety documents assiduously, and to adopt a preventative approach to the occurrence of hazards to reduce clinical risk. We recommend from our experience that clinicians remain open to the resources that are available in the community that can be utilised to support patient safety. Our experience has shown that it is possible to monitor unwell patients in the community safely; but we emphasise that this depends upon a rigorous formal approach to clinical safety management in a culture of transparency and learning.