Consensus on acute behavioural disturbance in the UK: a multidisciplinary modified Delphi study to determine what it is and how it should be managed

Background Acute behavioural disturbance (ABD) is a term used in law enforcement and healthcare, but there is a lack of clarity regarding its meaning. Common language should be used across staff groups to support the identification, prioritisation and delivery of care to this group of patients. The terminology currently used is inconsistent and confusing. This study aimed to reach a consensus on the criteria for identification and management of ABD, and to agree when other care pathways or guidelines might be more appropriately used. Methods A modified Delphi study with participation from stakeholder organisation representatives was conducted in January–April 2023 online. In round 1, statements were generated by participants in response to broad questions. Participants then rated their level of agreement with statements in subsequent rounds, with statements achieving a consensus removed for inclusion in the final derived consensus statement. Non-consensus statement responses were assessed for stability. Results Of 430 unique statements presented for rating, 266 achieved a consensus among 30 participants representing eight stakeholder organisations. A derived consensus statement was generated from these statements. The median group response to statements which failed to achieve a consensus was reliable (Krippendorff’s alpha=0·67). Conclusions There is a consensus across stakeholder organisations that ABD is not a separate entity to agitation, and guidance should instead be altered to address the full range of presentations of agitation. While the features of concern in this severely agitated group of patients can be described, the advice for recognition may vary depending on staff group. Criteria for recognition are provided and potential new terminology is described.


INTRODUCTION
Acute behavioural disturbance (ABD) is the term used by emergency care providers in the UK to describe 'an altered physiological and psychological state'. 1 There may be unrecognised life-threatening illness leading to the behavioural disturbance, and management is complicated by the potential increased risk of restraint-related death.
4][5][6] This also makes it challenging to hold organisations to account and drive improvement in patient care. 7

WHAT IS ALREADY KNOWN ON THIS TOPIC
⇒ Acute behavioural disturbance (ABD) is a term used in the UK to describe presentations of severe agitation, distress and signs of physiological deterioration of unknown cause.⇒ There is a lack of consensus on ABD recognition criteria in policing and healthcare, whether there is value in the use of ABD terminology and when alternative guidelines may be more appropriate than ABD guidance.⇒ There is a lack of validated UK criteria on which to base management guidelines.This has led to many guidelines using research published on 'excited delirium'-a condition which is not recognised in the UK.

WHAT THIS STUDY ADDS
⇒ This study is the first to bring together stakeholder organisations in the UK to achieve a consensus on the value of ABD terminology, criteria for recognition and response to severely agitated patients at greatest risk, and when the use of current guidelines on ABD is most appropriate.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
⇒ This study provides a clear consensus that ABD is not a separate entity to agitation, but there are criteria which can be used to identify agitated patients at greatest risk of poor outcomes.Specific terminology should be used to identify this group and provide a common language regarding prioritisation and management strategies.Consideration should be given to using new terminology such as 'redflag agitation' to describe the most severely agitated patients at the greatest risk of physical health emergency.

Original research
Use of the term ABD is further complicated by blurred boundaries with excited delirium (ExD), which is a term predominantly used in North America and which (as cogently described by McGuinness and Lipsedge in their 2022 paper) is widely considered problematic, with no proven pathological basis. 8There have been calls to replace the term ABD with a new descriptor due to concerns about conflation with ExD, but until the patient group in question is clearly defined, any new descriptor faces the same risk of conflation due to a lack of evidence on which to base advice regarding the identification, risk stratification and management of patients presenting with severe agitation. 9everal UK organisations have produced guidelines in an attempt to meet their responsibility to provide care for people with behavioural change, but the challenge of identifying people at risk of deleterious outcomes still remains. 1 3 4 6 It is clear that until a consensus is achieved, from which an evidence base can be generated, there will continue to be concerns that consequent guideline application (including use of force and sedation) may be racially biased. 7A lack of a UK-wide consensus definition adversely affects the ability of the services and systems involved to consistently apply guidance, and prevents standardised identification of the cohort of patients on whom UK research should be based, with the aim of improving care.
Establishing objective criteria for recognition of ABD is challenging, as the requirement for patients to be exhibiting an altered physiological state to identify ABD is often at odds with the inability of clinical staff to be able to measure physiological parameters due to the level of agitation.Concerns have been raised that, as a consequence, police and clinicians may be unsure of when to recognise a presentation of ABD or may apply ABD guidance to all agitated patients.Additionally, it is unclear who is responsible for recognising ABD; and whether it is a law enforcement or healthcare term, a presentation or a diagnosis.This has led to guidelines being produced by multiple professional bodies, with different objectives and perspectives.
This study aimed to reach a consensus on the criteria for identification and management of ABD, and to agree when other care pathways or guidelines might be more appropriately used.
In reaching a consensus, it is hoped that there will be benefits for service users, as appropriate care pathways are followed more consistently and with reduced subjectivity; for staff, as clarity around clinical decision-making improves, and for future researchers, as clear ABD criteria are established.

METHODS
To achieve a consensus, a modified Delphi study was undertaken with representatives from key stakeholder organisations.Recruitment took place in November-December 2022, and the study took place over three rounds in January-April 2023.

Patient and public involvement
As presentations of ABD are primarily and formally recorded at inquest, but are not captured in UK clinical coding, it was not possible to identify people with lived experience of the presentation.To ensure that the study outcomes were relevant to patients, we invited comment on the aims and design of the study from the relatives of a patient who died after presenting with ABD, and also sought views from the NHS Race and Health Observatory on the objectives and design of the study.We were asked (as a standard requirement prior to the consultation being agreed) to explicitly state in any publication that participation in consultation does not equate to endorsement of any study or its outcomes.These parties were not involved in the recruitment or conduct of the study.
Study participant organisations will receive a report of the study findings for comment and action within their remit areas.

Participant selection
To minimise selection bias, invitations were issued to participant organisations identified as key stakeholders in a Royal College of Psychiatrists report, of which our study group were not authors. 9ased on our external consultation, we also invited participation from toxicologists and two patient advocacy organisations.The organisations are shown in table 1.
Organisations received up to five phone calls (if a phone number was provided) and five emails in an effort to establish contact and to confirm if they were considering the invitation.Organisations were permitted to delegate participation to professional advisory group members.Eight organisations submitted participants, two declined to participate, two acknowledged contact but did not submit participants without offering a reason and one did not respond to attempts to contact them.
Organisations were invited to submit up to five experts for participation.The Consensus on Acute Behavioural Disturbance in the UK (CABDUK) study group were not involved in the selection of these experts and were explicitly excluded from participation.

Original research
Criteria for expertise were suggested as typically being authorship of a publication related to ABD, UK committee-level work on ABD or expert witness (in UK Court) work relating to ABD.However, the requirement was not made absolute as each organisation was likely to have a different perspective on expertise they wished to see represented.

Study process
To maximise expert engagement from across the UK, the study was conducted electronically using Google Forms.Participant anonymity was maintained for the duration of the study, though participants from within organisations may have been aware of each other's participation.
Each round launched with an email.Participants who had not submitted responses by half-way through the allotted time received up to two phone calls.If this failed to establish contact, a text message and a further email were sent.This process was repeated in the final few days of each study round.No financial incentives or rewards were offered for completion.Nonparticipation in prior rounds did not preclude participation in subsequent rounds.
The study was conducted in three rounds.Responses to eight broad questions offered in round 1 were reviewed for uniqueness and duplicates were removed by study group consensus.In round 2, the list of unique statements was presented to participants for agreement/disagreement using a 7-point Likert scale.Statements were removed if they achieved positive or negative consensus to enter the derived Delphi statement.
In round 3, remaining statements were re-presented to participants for voting with the median group score and the participant's personal prior rating if they had participated in the previous round.Statements achieving a positive or negative consensus then entered the derived Delphi statement.

Definition of consensus
There is no established definition for a Delphi consensus. 10We defined a consensus as 75% agreement or disagreement, as by convention many Delphi studies appear to use either 70% or 80% as consensus values.This value was declared in advance in the participant information.
Due to the study including participants from both healthcare and law enforcement backgrounds, and the potential for questions requiring specific clinical or policing expertise, participants were allowed to decline to respond to a question if they felt they had insufficient expertise in that area.

Generation of broad statements
Eight broad questions were designed by the authors to generate participant statements covering the major issues identified with current application of ABD terminology: ► A lack of clear criteria to enable police officers to recognise ABD.► A lack of clear criteria to enable healthcare staff to recognise ABD.► A lack of clear understanding of when alternative guidelines might be more appropriately used than ABD guidelines.► A lack of a clear understanding whether ABD terminology or guidance is applied across a spectrum of severity.The eight questions are shown in table 2. Timing thresholds for the spectrum of severity questions were based on UK ambulance service response targets.

Analysis
Statistics support was provided by the University of Edinburgh Epidemiology and Statistics Core.
For statements which did not reach a consensus in rounds 2 and 3, reliability of median group responses was assessed using Krippendorff 's alpha. 11This was done using the kripp.alpha function in RStudio V.2023.03.0.386 with 95% CIs obtained by bootstrapping.

Response to broad statements
Round 1 (broad statements) generated 430 unique statements for voting once duplicates were removed.Care was taken to leave apparent duplicate statements when subtle differences in expression might have affected statement interpretation.

Consensus statements
Round 2 saw 150 statements achieve a positive consensus, 3 statements achieve a negative consensus and 277 statements fail to achieve a consensus.
Of the 277 statements entering round 3, 101 achieved a positive consensus, 12 achieved a negative consensus and 164 statements fail to achieve a consensus.

Original research
Assessment of the reliability of median group responses for statements which did not reach a consensus in round 2 or round 3 gave a Krippendorff 's alpha of 0.67 (95% CI 0.58 to 0.76), which by convention allows a tentative conclusion of rating reliability to be drawn. 12he list of statements which achieved/did not achieve a consensus is provided as online supplemental table 1.

Derived Delphi consensus statement
The views of the CABDUK Delphi participant group are given in box 1.

DISCUSSION Applicability
This study is the first in the UK to bring together stakeholder organisations across the entire patient journey to clearly establish what ABD is understood to mean in a UK setting, and expectations around recognition and care.The use of formal research methodology serves to increase the credibility of the conclusions.The study authors had no role in selecting each organisation's representatives, and the heterogeneity of the participant organisations should provide assurance that a range of views have been represented.Some of the statements closely mirror current guidance from different UK groups.
A consensus has been achieved across professional groups in what has historically been a contentious area.However, it is important to interpret the consensus statements in the UK's social context and avoid simplistic narratives or the application of these statements without thought to potential harms.

Key findings
From the study consensus, it is clear that ABD is not, in the UK, considered a diagnosis or syndrome, and refers to a presentation of an individual in a state of severe agitation, with numerous

Original research
potential causes.There was support for integrating current guidance for agitation and ABD to provide guidelines which are applicable to the entire spectrum of presentations of agitation.It was felt that the features of concern in ABD can be described, and that (while not applicable to all agitated people) ABD guidance addressed a clear need which was not met by other agitation guidelines.Importantly, there are clear differences in expectations between first responder and healthcare provider use of ABD terminology.First responder recognition is focused on the identification of people for whom physical health management should be prioritised, while the subsequent healthcare focus is informed by a recognition of ABD as a very severe presentation of agitation, with multiple potential causes, some of which may be life-threatening.Recognition of this level of agitation by healthcare providers allows concurrent use of ABD guidance and any other relevant guidelines.
There is a clear desire to limit the use of some features historically associated with ExD (eg, 8.a-glass attraction) but a clear consensus that in spite of stereotyping concerns, some features are important in identifying patients presenting with agitation who are at greatest risk (eg, 21. b. viii-increased pain tolerance).In places, this has led to potentially contentious statements, such as recommending the avoidance of phrases such as 'superhuman strength' (identified as a criterion disproportionately applied to people from black ethnic groups) while simultaneously identifying 'exceptional strength' as a feature which may be identified in presentations of ABD. 13 This study is also the first to identify criteria for return to police custody following a presentation of ABD, opening the door to a more collaborative approach between emergency care providers and custody suites.

The future of ABD terminology
Historically in the UK, the use of ABD as a term has been promoted by learning from coroners' inquests held after deaths, in which it has been identified that improved recognition of the risks to health has the potential to prevent future deaths and provide a common language between emergency services. 2This study consensus has recognised that ABD is considered a state of severe agitation, but with features which suggest significant risk to physical health.However, the clear consensus that it is not helpful to separate guidance on ABD and agitation raises the question of the value of using ABD terminology in policing and healthcare as opposed to a new term to identify patients presenting with agitation who are at greatest risk of experiencing a time-critical or physical health emergency.This is particularly important given concerns from advocacy organisations regarding how terminology which emerged from historical use of ExD may be applied. 14o build on the consensus in this study, to foster trust with service users, to clearly establish that ABD is of limited utility as a descriptive term and to move towards agitation guidelines which address the entire spectrum of agitation, the consensus features suggested in this study for the recognition of ABD could instead be described as 'red-flag agitation' (similar to the highest-risk patients with other conditions such as 'red-flag sepsis'), identifying patients at highest risk of poor outcomes and most likely to require the most urgent specialised management strategies.A suggested spectrum of agitation is given in figure 2. This spectrum would provide scope for different suggested care strategies

Box 1 Continued
ii.An abnormality identified in primary survey.iii.Abnormal vital signs, blood glucose or ECG.iv.Confusion

Original research
across a range of presentations; allow dynamic reassessment and minimise anchoring bias in decision-making; clearly establish that ABD is not a diagnosis; and provide common terminology across prehospital and ED settings regarding prioritisation of care, which would have its foundation in the features identified in this study rather than in historical literature.Additionally, this approach was suggested as a solution by the Royal College of Psychiatrists in their 2022 report on ABD.

The management of undifferentiated agitation
When people present with severe agitation, healthcare staff may be unable to safely approach the patient and undertake a standard physical health assessment.Clinical investigations and clinical procedures are frequently not possible without sedation.Conversely, sedation is normally delivered in a highly monitored and controlled environment, with prior patient assessment and clinical optimisation.This is the paradox which ABD guidelines have historically tried to address, providing strategies (such as potent intramuscular sedation prior to the application of clinical monitoring) which differ from standard care.Agitation guidelines will always require adaptation to specific care settings, considering differing expertise, resources, and the level of clinical risk related to unidentified or unmanaged conditions.

Limitations
There are limitations to the Delphi technique which should be noted.A Delphi study is an iterative process which requires a significant time commitment from participants.While there is a desire to continue repeating rounds until a perfect consensus is achieved, this needs to be balanced against the burden on participants, dropout rates with each round, and the desire from participants and their parent organisations for results within clear timescales.Our analysis of non-consensus statements at the end of round 3 allows a tentative conclusion that stability has been reached, but the value of Krippendorff 's alpha was at the lower limit of the conventional range.
Statements were generated by participants from a broad range of professional backgrounds, with differing perspectives on the aspects of ABD they wished to see represented in the Delphi process.Statements ranged from broad perspectives on the validity of ABD through to specific clinical features denoting high clinical risk.
The scope of the Delphi was deliberately broad and was designed to provide the foundation on which future patient care, guidelines and research could be based, rather than providing immediate resolution for every contentious aspect.Individual organisations who participated may feel that the process has failed to appreciate every nuance of their role in the care of people presenting with ABD.Additionally, it should be noted that it was only possible to vote on statements generated in round 1.This precluded exploration of the implications of statements, reasons for non-consensus and the introduction of additional statements to expand the scope of consensus.One consequence of this is that a small number of contradictory statements emerged from questions 5, 6 and 7.For example, the statement 'The person has a raised/elevated HR' was approved for both questions 5 and 6, meaning that this would be included in criteria for an emergency healthcare response (input within 20 min) or an urgent healthcare response (input within 2 hours).In the derived consensus statement, this contradiction was overcome by keeping the 20-minute time frame, which emerged as the only acceptable response to ABD presentations.It should be noted that people with priority symptoms or signs (suggesting a periarrest or cardiac arrest state) would still be prioritised by ambulance services to receive an immediate response.
There were 164 statements which did not achieve consensus, and these largely addressed specific features regarding the identification of ABD.It is possible that a traditional Delphi model including face-to-face discussion would have helped to achieve further consensus.The full list of consensus and non-consensus statements is provided as online supplemental material.

Next steps
There are key next steps which should be undertaken as a consequence of this study.While they are potentially immediately actionable, this will require funding, formal quality improvement methodology, further research and cross-systems leadership to improve quality of care.Educational materials and clinical guidelines require adaptation to reflect this consensus opinion of UK stakeholder organisations, and remove advice previously given which lacks a meaningful evidence base.
Clinical guidelines should address the full range of presentations of agitation and respective management strategies.Use of the term ABD separately to agitation is not felt to be helpful, but national healthcare and emergency services providers need to agree with any language or terminology changes for first responders, to manage patient safety concerns with regard to consistent language use.
The Delphi criteria will require prospective validation at all points in the service user journey to identify which features perform well in identifying people at risk, identify any other features which have not been identified by this Delphi process and ensure that the established criteria are not subject to bias in their application.The Delphi has provided a consensus against which appropriate or inappropriate management may be measured.

CONCLUSION
It is key that ABD should be understood to be a presentation, not a diagnosis.While there are a number of features which may be identified in the highest-risk presentations of agitation, the triad of being hot to touch (tactile hyperthermia), exhibiting constant or near-constant activity, and extreme agitation or aggression should be the focus of recognition.
Specific terminology should be used by first responders to recognise agitated people who require a focus on management of their physical health, and consideration should be given to using new terminology, such as 'red-flag agitation' rather than 'acute behavioural disturbance' to identify this group of agitated patients at highest risk of poor outcomes, who may require specialised management strategies.
Twitter Christopher Humphries @cp_humphries Acknowledgements The authors wish to thank the participating organisations and their Delphi participants for offering their time and effort.The following participants gave permission to be identified as contributors: AACE-Mark Faulkner; FICM-Monika Beatty, Richard Bourne, Dale Gardiner; FFLM-Alex Gorton, Jason Payne-James, Keith Rix, Stacey Shelley, Margaret Stark; NPCC-Michael Brown, Ben Rowe; NPIS-James Coulson, Paul Dargan, Aravindan Veiraiah, Stephen Wiltshire, David Wood; RCEM-Meng Aw-Yong, Phil Cowburn, Charles Deakin; RCPsych-Rajesh Mohan, Aileen O'Brien, Annabel Price, Nicola Taylor; UKAFN-Thomas Bird, Matthew Peel.Additionally, they would like to thank the family of Douglas Oak and the NHS Race and Health Observatory for their thoughtful comments on the design and aims of the study.
Contributors Substantial contributions to the conception or design of the work (CH, AK, AS, AW, JS); or the acquisition (CH, AK, AS), analysis (CH, AK, AS, AW, JS) or interpretation of data (CH, AK, AS, AW, JS) for the work; and drafting the work

Original research
(CH, AK, AS) or reviewing it critically for important intellectual content (CH, AK, AS, AW, JS); and final approval of the version to be published (CH, AK, AS, AW, JS); and agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved (CH, AK, AS, AW, JS).CH is the guarantor of the work.Competing interests CH-access to University of Edinburgh Read & Publish agreement, reimbursed for travel expenses to speak at conferences for Royal College of Emergency Medicine (RCEM), unpaid member of RCEM Toxicology advisory group, lead author of RCEM guidelines on ABD, cannabinoid hyperemesis syndrome and nitrous oxide toxicity.AW-acted as expert witness in one coroner's inquest related to ABD in 2021, chair of UK EMS Clinical Practice Guidelines Committee, lead of committee for UK ABD guidelines for paramedic practice, chair of RCEM ABD guidelines committee, reimbursed for travel expenses to speak at 2023 RCEM conference.AK, AS and JS-none.

Funding
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research.Refer to the Methods section for further details.
Patient consent for publication Not required.

Figure 1
Figure 1 describes the overall process of the Delphi.

Figure 1
Figure 1 Study flow diagram.AACE, Association of Ambulance Chief Executives; FFLM, Faculty of Forensic and Legal Medicine; FICM, Faculty of Intensive Care Medicine; NPCC, National Police Chiefs' Council; NPIS, National Poisons Information Service; RCEM, Royal College of Emergency Medicine; RCPsych, Royal College of Psychiatrists; UKAFN, UK Association of Forensic Nurses and Paramedics.

Figure 2 A
Figure2A suggested agitation spectrum applicable across the service user journey.ABD, acute behavioural disturbance.
No funding was received for this study, except for open access charges which were funded through the University of Edinburgh/BMJ Publish & Read agreement for UKRI funded researchers.The corresponding author is employed by grant funds from the Medical Research Council [MR/T044802/1]'; and The Centre for Precision Cell Therapy for the Liver which was funded by the Chief Scientist Office of the Scottish Government Health Directorates (PMAS/21/07).For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising from this submission.

Table 1
Organisations invited to participate in the Delphi study and outcomes of those invitations

Table 2
Broad questions presented to participants in round 1 to support the generation of statements for voting 6For patients who are breathing and conscious: What features following a presentation of ABD require urgent healthcare provider input?(within 2 hours) 7For patients who are breathing and conscious: What features following a presentation of ABD permit non-urgent healthcare provider input?(within 4 hours) 8For patients who are breathing and conscious: What features following a presentation of ABD suggest that observation in police custody would be safe?(care provided solely by custody healthcare staff)

Original research Box 1 The derived Consensus on Acute Behavioural Disturbance in the UK Delphi consensus statement Regarding recognition of acute behavioural disturbance (ABD) by first responders
Individual features are insufficient for recognition of ABD.The person should be exhibiting more than one consensus feature.However, the presence of all consensus features is not required.6. Features which may be present in presentations of ABD and identified by first responders include: a.The person is hot to touch.b.The person is sweating profusely.c.The person is exhibiting constant (or near-constant) physical activity.d.The person is unable to sit or stand still.e.The person is exhibiting extreme agitation.f.The person is exhibiting extreme aggression.g.The person's presentation of aggression or hostility may appear atypical.h.The person appears disoriented (to place/person/time).i.The person is exhibiting bizarre behaviour.j.The person does not respond to other people present.k.The person is exhibiting exceptional strength.l.The person is constantly resisting restraint, or is restrained for 15min without resolution or de-escalation.m.The person does not become calmer with verbal deescalation or restraint.n.The person has a raised heart rate.o.The person's behaviour is not explained by other factors.7. Phrases such as 'superhuman strength' should not be used.8.The following features should not be used to identify ABD: a. Attraction to mirrors or glass.Live feed from a police officer's body-worn video to specifically trained individuals in force control rooms may be helpful.13.All police custody staff should be trained to recognise a patient who is at high risk of deteriorating early.In ContinuedBox 1 Continued 12.

settings with healthcare provision (custody suites, ambulance services or EDs) 14
. Clinical staff have additional skills, training and potential opportunities to undertake patient assessment which may lead to a better understanding of the cause of the presentation of ABD, or negate the need to use the term.15.Healthcare professionals have a responsibility to differentiate the cause of the patient's presentation.16.Staff should: a. Have plans regarding the management of patients presenting with ABD until emergency care has arrived, including de-escalation.b.Have access to readily available emergency equipment in environments which may be required to care for people presenting with ABD.c. Be aware that restraint may worsen a patient's condition.d.Understand that the combination of physical health deterioration and behavioural disorganisation is likely to need emergency medical care.17.Observable features leading to recognition of a presentation of ABD should have more detail for healthcare staff.18.The triad of being hot to touch (tactile hyperthermia), Continued