Knowledge

A range of studies explored staff knowledge about factors which can precipitate agitation, confusion or distress during enhanced observation. ^{Reference Schroeder48–Reference Wray and Rajab-Ali56} Staff described understanding that patients could not always regulate their behaviour owing to underlying conditions, such as dementia or traumatic brain injury. ^{Reference Schroeder48,Reference Squire and Hessler49} Several publications also suggested staff insight into patient behaviours indicative of unmet needs. Efforts to identify potential pain, sensory discomfort or hunger were described. ^{Reference Case50,Reference Flaherty and Little52,Reference Vancil, Michael, Wagoner, Farkas, Evans and Pasley55,Reference Wray and Rajab-Ali56} Despite this apparent understanding, potentially avoidable shortfalls in care were reported nonetheless. Problems such as delaying a person’s medications were thought to exacerbate patient agitation. ^{Reference de Jong, Weselman, Kitchen and Hill51}

Skills

Skills underpinning high-quality and safe care during enhanced observation were also discussed. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Case50–Reference Flaherty and Little52,Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57–Reference Handley, Theodosopoulou, Taylor, Hadley, Surr and Goodman66} Studies indicated that enhanced observers showed attunement skills through timely identification of and response to signs of patient deterioration, ^{Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57–Reference Cook, Palesy, Chenoweth and Lapkin59} while also recognising how their own practice could provoke patients. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Cook, Palesy, Chenoweth and Lapkin59,Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60} For example, participants reflected that ineffective communication, or simply the experience of enhanced observation itself, could inadvertently cause patient distress. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Cook, Palesy, Chenoweth and Lapkin59} Examples of poor attunement were also presented. For instance, staff had sometimes carried out invasive procedures on distressed patients without explaining what they were doing, ^{Reference Cook, Palesy, Lapkin and Chenoweth28} suggesting that informed consent was not sought first.

Staff skills in supporting patients to regulate their actions and behaviours represented a further theme. ^{Reference Case50–Reference Flaherty and Little52,Reference Wilkes, Jackson, Mohan and Wallis61–Reference Handley, Theodosopoulou, Taylor, Hadley, Surr and Goodman66} Efforts to enhance quality and safety included a focus on the adequacy of psychotropic medications prescribed to those under enhanced observation. ^{Reference Case50,Reference de Jong, Weselman, Kitchen and Hill51,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63} For example, medication side-effects were considered as contributory factors to falls. ^{Reference de Jong, Weselman, Kitchen and Hill51} Other studies described the use of pro-active medication reviews to appraise how medications contributed to agitation or delirium. ^{Reference Case50,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63}

Further literature reported on enhanced observer skills in de-escalation in response to agitation or distress. ^{Reference Flaherty and Little52,Reference Naughton, Allen, Myers, Wills, Fehin and Jacob62,Reference Flynn, Prescott and Jinks64,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65} De-escalation training and individualised activity planning interventions were evaluated. ^{Reference Naughton, Allen, Myers, Wills, Fehin and Jacob62,Reference Flynn, Prescott and Jinks64} After implementation of such initiatives, higher proportions of surveyed staff stated that they were ‘fairly confident’ or ‘extremely confident’ in using de-escalation techniques (80% v. 55% at baseline), ^{Reference Flynn, Prescott and Jinks64} whereas mean rates of pro re nata antipsychotic prescription showed trends towards reduction (5.5% v. 9% at baseline). Although the absence of statistical testing in these studies precluded definitive assertions about the impact of interventions, further research involving ethnographic observations described positive changes in the wider ward atmosphere following effective use of activity-based de-escalation techniques. ^{Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65}

Attitude

Staff attitudes and beliefs in relation to the provision of high-quality, safe enhanced observation were often referenced. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Schroeder48,Reference Case50,Reference Flaherty and Little52,Reference Jones, Aylward and Jones53,Reference Vancil, Michael, Wagoner, Farkas, Evans and Pasley55,Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57–Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Naughton, Allen, Myers, Wills, Fehin and Jacob62–Reference Waszynski, Milner, Staff and Molony74} Humanity was conveyed when participating staff expressed compassion and understanding of patients’ feelings, ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Schroeder48,Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57–Reference Cook, Palesy, Chenoweth and Lapkin59} and patients responded positively when staff showed awareness of their needs and prior life experiences. ^{Reference Cook, Palesy, Lapkin and Chenoweth28} Efforts to achieve balance between tolerance of potentially risky behaviours by confused patients – including removal of intravenous lines – versus use of restrictive measures were also described. ^{Reference Flaherty and Little52} However, other studies suggested an excessive use of restrictions. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Handley, Bunn and Goodman67} Examples included observations of patients being prohibited from standing up, ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65} without engagement from staff to establish the source of their agitation.

Retrieved research also offered insights into person-centred enhanced observation. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Case50,Reference Jones, Aylward and Jones53,Reference Vancil, Michael, Wagoner, Farkas, Evans and Pasley55,Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Naughton, Allen, Myers, Wills, Fehin and Jacob62–Reference Waszynski, Milner, Staff and Molony74} Some studies reported examples of staff members incorporating patient preferences into care planning, ^{Reference Case50,Reference Jones, Aylward and Jones53,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Handley, Theodosopoulou, Taylor, Hadley, Surr and Goodman66,Reference Broski, Eckels, Blomquist, King and Lineberry68} such as identifying familiar music to play. ^{Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63} Such findings were in sharp contrast with cases in which the enhanced observer role was passive and custodial before implementation of change initiatives, and planning of care according to therapeutic goals was not prioritised. ^{Reference Jones, Aylward and Jones53,Reference Flynn, Prescott and Jinks64} Nevertheless, a range of interventions to support patients to engage in diversional activities were described; these aimed to avert agitation or deterioration ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Naughton, Allen, Myers, Wills, Fehin and Jacob62,Reference Flynn, Prescott and Jinks64,Reference Handley, Bunn and Goodman67,Reference Eeles, Thompson, McCrow and Pandy70–Reference Waszynski, Milner, Staff and Molony74} and the need for further restrictions. ^{Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71} Such practices were not ubiquitous according to the included research; other studies suggested limited pro-active use of de-escalation activities and little emphasis on patient engagement, with observations seemingly conducted in a task-focused manner. ^{Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65}

Confidence

Staff confidence in their ability to support distressed patients in the least restrictive manner, while maintaining self-control, was seldom explored within the included publications. ^{Reference Schroeder48,Reference Flynn, Prescott and Jinks64,Reference Jin, Novik and Saravay75} There were limited descriptions of the ways in which enhanced observers remained calm and avoided provocation in difficult situations. ^{Reference Schroeder48} However, staff confidence in positive risk-taking was addressed. Psychiatrists with consultation-liaison training seemed to be more confident in holding risk and were less likely to request enhanced observation than psychiatrists without such training. ^{Reference Jin, Novik and Saravay75} Other literature reported that staff lacked confidence in using physical restraint, despite receiving training. ^{Reference Flynn, Prescott and Jinks64}

Ethical sensitivity

Staff ethical values, beliefs and practices relating to enhanced observation were also detailed. ^{Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Squire and Hessler49,Reference Jones, Aylward and Jones53,Reference Kerr, Verner and Traynor58,Reference Wilkes, Jackson, Mohan and Wallis61,Reference Flynn, Prescott and Jinks64–Reference Handley, Bunn and Goodman67,Reference Foye, Simpson and Reynolds76,Reference Talley, Davis, Goicoechea, Brown and Barber77} Evidence from ethnographic research suggested that enhanced observation was at times used to safeguard the smooth running of ward routines, by freeing up other members of staff to undertake routine tasks. ^{Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65} A new descriptor was developed to reflect enhanced observers’ recognition and appreciation, or lack thereof, of their responsibility to care for people with a diverse range of needs in the acute general hospital setting. A body of literature suggested that staff disliked being allocated enhanced observation responsibilities, ^{Reference Flynn, Prescott and Jinks64} felt that caring for people with additional needs such as mental health problems was not their responsibility, ^{Reference Wilkes, Jackson, Mohan and Wallis61,Reference Foye, Simpson and Reynolds76} and regarded enhanced observation as an unskilled task akin to ‘babysitting’. ^{Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Handley, Bunn and Goodman67}

Other literature examined whether staff ethical values were aligned with patients’ best interests. ^{Reference Squire and Hessler49,Reference Jones, Aylward and Jones53,Reference Kerr, Verner and Traynor58,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Talley, Davis, Goicoechea, Brown and Barber77} Introduction of an enhanced observation framework for assessment of a person’s care needs was thought to challenge staff to reflect upon their own practice. ^{Reference Jones, Aylward and Jones53} By contrast, further research offered examples of staff members and teams choosing not to respond to patient needs and behaviours. These studies reported on enhanced observers walking away from the person they were assigned to support, ^{Reference Talley, Davis, Goicoechea, Brown and Barber77} or avoiding use of physical restraints despite risks of harm to staff and other patients. ^{Reference Squire and Hessler49} In the latter example, decision-making regarding the use of enhanced observation was influenced by concerns about referrals for care placements being rejected during discharge planning, rather than being solely based on clinical need. ^{Reference Squire and Hessler49} Finally, further research showed active deception of people with dementia during enhanced observation. ^{Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65} For example, nursing staff were observed to lie to a patient about the type of medication being administered (a sedative). ^{Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65}

Teamwork

Themes of teamwork fluency during enhanced observation were central in the identified literature. ^{Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Squire and Hessler49,Reference de Jong, Weselman, Kitchen and Hill51,Reference Jones, Aylward and Jones53–Reference Kerr, Verner and Traynor58,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65–Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71,Reference Voetelink, Trout, Murphy, Rogers, Harris and Ramagnano73,Reference Foye, Simpson and Reynolds76,Reference Concannon, Carr, Doherty, McInerney, Birrane and Kearney78–Reference Grealish, Chaboyer, Darch, Real, Phelan and Soltau83} Staff testimonies indicated gaps in information-sharing about patient needs, including risks of violence. ^{Reference Squire and Hessler49,Reference Kerr, Verner and Traynor58,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71} Other studies suggested that although most surveyed enhanced observers received handovers when commencing their shift, these handovers lacked information needed to respond to crises. ^{Reference Handley, Theodosopoulou, Taylor, Hadley, Surr and Goodman66,Reference Graham, Eaton, Jeffrey, Secher-Jorgensen and Henderson79} Moreover, a wealth of evidence indicated that enhanced observers felt poorly supported by colleagues; for instance, studies described a lack of respect ^{Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71,Reference Ahmed, Cayea, Tackett, Bellantoni, Massare and Ward80} and reassurance, ^{Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57} and non-response to calls for help. ^{Reference de Jong, Weselman, Kitchen and Hill51} Likewise, ethnographic research described how responsibility for decision-making about enhanced observation continuation was often held by individual staff members, who feared personal reprisals should a patient come to harm after the intervention had been discontinued. ^{Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65}

Positive examples of collaborative teamwork during enhanced observation were also identified. Several publications detailed efforts to ensure that enhanced observers felt part of a team. ^{Reference Rachh, Wilkins, Capodilupo, Kilroy, Schnider and Repper-Delisi54,Reference Graham, Eaton, Jeffrey, Secher-Jorgensen and Henderson79} Other settings introduced opportunities for enhanced observers to share their reflections with senior nursing staff, informing decision-making about continuation versus discontinuation of enhanced observation. ^{Reference Jones, Aylward and Jones53,Reference Voetelink, Trout, Murphy, Rogers, Harris and Ramagnano73}

Tensions within clinical teams about the remit of the enhanced observer role were evident. A lack of shared responsibility for use of more person-centred care approaches was thought to shape enhanced observation practice. ^{Reference Handley, Theodosopoulou, Taylor, Hadley, Surr and Goodman66} Studies detailed how enhanced observers were often interrupted by registered nurses and asked to attend to other patients or tasks. ^{Reference Kerr, Verner and Traynor58,Reference Handley, Bunn and Goodman67,Reference Grealish, Chaboyer, Darch, Real, Phelan and Soltau83} Measures to address this challenge included introduction of a leadership role to oversee enhanced observer workflow. ^{Reference Rachh, Wilkins, Capodilupo, Kilroy, Schnider and Repper-Delisi54}

Leadership

Leadership practices underpinning enhanced observation quality and safety were discussed in a further body of literature, in which the management of staff competence to undertake this intervention was widely examined. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Schroeder48–Reference de Jong, Weselman, Kitchen and Hill51,Reference Jones, Aylward and Jones53–Reference Cook, Palesy, Chenoweth and Lapkin59,Reference Wilkes, Jackson, Mohan and Wallis61–Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Eeles, Thompson, McCrow and Pandy70–Reference Voetelink, Trout, Murphy, Rogers, Harris and Ramagnano73,Reference Graham, Eaton, Jeffrey, Secher-Jorgensen and Henderson79–Reference Shannon, Cruickshank and Grealish90} Training in de-escalation of aggression, ^{Reference Squire and Hessler49,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71,Reference Ahmed, Cayea, Tackett, Bellantoni, Massare and Ward80} behaviour management ^{Reference de Jong, Weselman, Kitchen and Hill51,Reference Cook, Palesy, Chenoweth and Lapkin59,Reference Wilkes, Jackson, Mohan and Wallis61,Reference Sinvani, Strunk, Patel, Shah, Mulvany and Kozikowski84} and caring for people with additional vulnerabilities such as dementia or cognitive impairment ^{Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Jones, Aylward and Jones53} was consistently described as inadequate. Staff suggested that insufficient training and experience in enhanced observation had contributed to instances of inappropriate behaviour management, ^{Reference Cook, Palesy, Chenoweth and Lapkin59} whereas targeted training and mentoring interventions were thought to promote positive practices. ^{Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57}

Ambiguity in role expectations during enhanced observation was apparent, with some settings lacking definitions of staff responsibilities. ^{Reference Wilkes, Jackson, Mohan and Wallis61,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71} Accordingly, several publications described introduction of structured guidance to clarify the scope of the role. ^{Reference Jones, Aylward and Jones53,Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57,Reference Kerr, Verner and Traynor58} Staff also reported that requirements for documenting observations and safety incidents were unclear. ^{Reference Squire and Hessler49,Reference Kerr, Verner and Traynor58,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71} Efforts to address this problem were centred on streamlining of documentation templates and development of enhanced observer checklists. ^{Reference Case50,Reference Jones, Aylward and Jones53,Reference Vancil, Michael, Wagoner, Farkas, Evans and Pasley55}

Management of safety was a second focal leadership theme. Safety within the acute general hospital physical environment was thought to be undermined when emergency call bells could not be reached easily, ^{Reference Broski, Eckels, Blomquist, King and Lineberry68} or by hazards and clutter which contributed to patient falls. ^{Reference de Jong, Weselman, Kitchen and Hill51} Improvement initiatives concentrated on creation of calming environments with adjusted lighting, neutral colours and noise minimisation to promote healthy sleep cycles, ^{Reference Case50,Reference Vancil, Michael, Wagoner, Farkas, Evans and Pasley55,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} or removal of unnecessary or suboptimal equipment. ^{Reference Eeles, Thompson, McCrow and Pandy70}

Staffing level management was another challenge to quality and safety during enhanced observation. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference de Jong, Weselman, Kitchen and Hill51,Reference Jones, Aylward and Jones53,Reference Cook, Palesy, Chenoweth and Lapkin59,Reference Wilkes, Jackson, Mohan and Wallis61,Reference Naughton, Allen, Myers, Wills, Fehin and Jacob62,Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71,Reference Graham, Eaton, Jeffrey, Secher-Jorgensen and Henderson79} Staff were routinely redeployed from different wards to fulfil enhanced observation needs. ^{Reference Cook, Palesy, Chenoweth and Lapkin59} Other studies described how staff often undertook lengthy enhanced observation shifts without breaks. ^{Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71,Reference Graham, Eaton, Jeffrey, Secher-Jorgensen and Henderson79} This contributed to safety risks such as staff falling asleep, lapses in compassion or patient falls. ^{Reference de Jong, Weselman, Kitchen and Hill51,Reference Jones, Aylward and Jones53,Reference Broski, Eckels, Blomquist, King and Lineberry68}

Given these evidenced inadequacies, a new descriptor was introduced to capture the existence of procedures to govern and safeguard the use of enhanced observation within the acute general hospital setting. Studies discussed deficits in processes for initiation, continuation or discontinuation of enhanced observation, or efforts to strengthen clinical decision-making processes surrounding its deployment. ^{Reference Jones, Aylward and Jones53–Reference Wray and Rajab-Ali56,Reference Kerr, Verner and Traynor58,Reference Cook, Palesy, Chenoweth and Lapkin59,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Voetelink, Trout, Murphy, Rogers, Harris and Ramagnano73,Reference Talley, Davis, Goicoechea, Brown and Barber77,Reference Rausch and Bjorklund81,Reference Torkelson and Dobal82,Reference Sinvani, Strunk, Patel, Shah, Mulvany and Kozikowski84–Reference Takatsuka, Nykamp, Speer, Benko, Bart and St Clair89} Qualitative and observational data revealed poor clarity with respect to which clinical professionals (e.g. psychiatrists, ward doctors, nursing leaders) could approve deployment of enhanced observation, when it was indicated, ^{Reference Kerr, Verner and Traynor58,Reference Goldberg87} and when its continued use should be reviewed, ^{Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65} potentially driving its overuse within acute wards and emergency departments. ^{Reference Voetelink, Trout, Murphy, Rogers, Harris and Ramagnano73,Reference Takatsuka, Nykamp, Speer, Benko, Bart and St Clair89} Several publications described introduction of standardised protocols or mandatory assessments to govern assignment of enhanced observation, leading to reductions in its use without apparent impact on safety outcomes such as rates of physical restraint or falls. ^{Reference Vancil, Michael, Wagoner, Farkas, Evans and Pasley55,Reference Wray and Rajab-Ali56,Reference Rausch and Bjorklund81,Reference Colella, Aroh, Douglas, VanBuitenen, Galesi and Castro86}

Finally, a further new descriptor concerning the types of staff member assigned enhanced observation responsibilities was inductively introduced. Non-clinical hospital staff fulfilling this role included secretaries, housekeepers, chaplains and security personnel. ^{Reference Kerr, Verner and Traynor58,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71,Reference Worley, Kunkel, Gitlin, Menefee and Conway85,Reference Shannon, Cruickshank and Grealish90} Patients were thought to be frightened and distressed about involvement of security staff. ^{Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Shannon, Cruickshank and Grealish90}

Organisational culture

Organisational culture was addressed as a final dimension of quality and safety in enhanced observation practice. ^{Reference Squire and Hessler49,Reference Case50,Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57,Reference Cook, Palesy, Chenoweth and Lapkin59,Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63–Reference Handley, Bunn and Goodman67,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72,Reference Voetelink, Trout, Murphy, Rogers, Harris and Ramagnano73,Reference Ahmed, Cayea, Tackett, Bellantoni, Massare and Ward80,Reference Grealish, Chaboyer, Darch, Real, Phelan and Soltau83–Reference Worley, Kunkel, Gitlin, Menefee and Conway85,Reference Salamon and Lennon88} Several studies discussed practical recommendations that staff had received on reducing use of enhanced observation, including provision of activity resources, ^{Reference Flynn, Prescott and Jinks64} advice on management of pain or removal of unnecessary medical devices, ^{Reference Salamon and Lennon88} or educational efforts to reduce medication-induced delirium. ^{Reference Worley, Kunkel, Gitlin, Menefee and Conway85} Further studies described the involvement of family carers, or lack thereof, in decision-making about use of enhanced observation. ^{Reference Case50,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Handley, Theodosopoulou, Taylor, Hadley, Surr and Goodman66}

Other research examined links between the culture in which enhanced observation was delivered and patient safety. ^{Reference Case50,Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Handley, Bunn and Goodman67,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} Several interventional approaches sought to reduce situations that compromised safety during enhanced observation. ^{Reference Case50,Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Handley, Bunn and Goodman67,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} For example, individualised diversional activities and engagement were suggested as ideas to avert the need for physical restraint. ^{Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} Indeed, anecdotal reports of restraint reduction were discussed in some studies, ^{Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60} and statistically significant differences between the intervention group relative to matched controls were presented in other research. ^{Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} Further publications suggested that similar initiatives may help to address violence or potentially harmful acts such as removal of intravenous lines. ^{Reference Case50,Reference Handley, Bunn and Goodman67}

Organisational safety culture was a final factor to be inductively introduced into the framework. ^{Reference Squire and Hessler49,Reference Case50,Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57,Reference Cook, Palesy, Chenoweth and Lapkin59} Some studies provided evidence that reporting of safety incidents occurring during enhanced observation was actively encouraged, with survey data suggesting improved feelings of safety among enhanced observers. ^{Reference Case50} Other findings pointed to cultures of underreporting, despite the presence of incident reporting policies. ^{Reference Squire and Hessler49} Moreover, reviewed literature presented mixed findings with respect to whether staff felt able to ask questions or raise concerns, ^{Reference Squire and Hessler49,Reference Beccaria, Hoey, Towler, Rees, Brazier and Gordon57,Reference Cook, Palesy, Chenoweth and Lapkin59} with further evidence suggesting that staff regarded leadership action on safety issues as insufficient. ^{Reference Squire and Hessler49}

Indicators of poor quality of care

Evidenced quality issues centred on established indicators of poor-quality care, ^{Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Haak, Gerstenberger and Reddy69,Reference Eeles, Thompson, McCrow and Pandy70,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72,Reference Talley, Davis, Goicoechea, Brown and Barber77,Reference Weiss, Reddy and Anton91} such as in-hospital mortality. ^{Reference Eeles, Thompson, McCrow and Pandy70,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} Specialist cohorting units focusing on delirium prevention, diversional activities or nursing staff education were used in efforts to address this challenge. ^{Reference Eeles, Thompson, McCrow and Pandy70,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} One example reported a statistically significant reduction in ward-level mortality after unit implementation. ^{Reference Eeles, Thompson, McCrow and Pandy70} A second study yielded similar evidence of reduced in-hospital mortality among older adults with cognitive impairment admitted to the intervention unit, compared with patients matched on clinical and demographic variables. ^{Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72}

A further body of research addressed the problem of extended length of stay in acute hospitals for those receiving enhanced observation. ^{Reference Haak, Gerstenberger and Reddy69,Reference Eeles, Thompson, McCrow and Pandy70,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72,Reference Talley, Davis, Goicoechea, Brown and Barber77} Several publications reported no change in length of stay as a result of different interventional approaches. ^{Reference Haak, Gerstenberger and Reddy69,Reference Eeles, Thompson, McCrow and Pandy70,Reference Talley, Davis, Goicoechea, Brown and Barber77} By contrast, statistically significant reductions in length of stay were achieved following the introduction of a multicomponent intervention delivered by patient engagement specialists. ^{Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72}

Unplanned readmission to hospital was another evidenced quality issue. ^{Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Haak, Gerstenberger and Reddy69,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72,Reference Weiss, Reddy and Anton91} Evaluated interventions failed to affect rates of unplanned readmissions, ^{Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} or the studies did not conduct or report clearly on statistical testing. ^{Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Haak, Gerstenberger and Reddy69} Nevertheless, introduction of a delirium checklist for enhanced observers was associated with statistically significant reductions in 30-day readmission rates. ^{Reference Weiss, Reddy and Anton91}

Interpersonal violence and aggression

Safety risks within acute care wards and emergency departments also included problems arising from violence and aggression. ^{Reference Squire and Hessler49,Reference Case50,Reference Cook, Palesy, Chenoweth and Lapkin59,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71,Reference Ahmed, Cayea, Tackett, Bellantoni, Massare and Ward80,Reference Eastwood and Schechtman92} Violence or aggression was typically perpetrated by patients and directed at staff delivering enhanced observation. Incidents of verbal or physical aggression were documented in routinely collected data, ^{Reference Eastwood and Schechtman92} and assaults involving hitting, kicking or choking were detailed in staff interviews and surveys. ^{Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Schoenfisch, Pompeii, Lipscomb, Smith, Upadhyaya and Dement71}

Harm to self

Harm to self, despite the presence of an enhanced observer, was a further theme in the included literature. ^{Reference Case50,Reference Flaherty and Little52,Reference Wilkes, Jackson, Mohan and Wallis61,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Talley, Davis, Goicoechea, Brown and Barber77,Reference Concannon, Carr, Doherty, McInerney, Birrane and Kearney78,Reference Kroll, Stanghellini, DesRoches, Lydon, Webster and O’Reilly93} Studies reported on self-harm, ^{Reference Wilkes, Jackson, Mohan and Wallis61,Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Kroll, Stanghellini, DesRoches, Lydon, Webster and O’Reilly93} and attempted suicides. ^{Reference Kroll, Stanghellini, DesRoches, Lydon, Webster and O’Reilly93} Suicides in the context of enhanced observation were also described, ^{Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Concannon, Carr, Doherty, McInerney, Birrane and Kearney78} including an example of a person dying by suicide while awaiting transfer to mental health services under mental health legislation. ^{Reference Concannon, Carr, Doherty, McInerney, Birrane and Kearney78} Finally, several studies referred to risks posed to patients with confusion or behavioural needs from interruption to their own treatment, for example, removal of drips, lines and tracheostomies. ^{Reference Case50,Reference Flaherty and Little52,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Broski, Eckels, Blomquist, King and Lineberry68,Reference Talley, Davis, Goicoechea, Brown and Barber77}

Accidents and injuries

Accidents and injuries represented a further safety issue, with falls having occurred despite enhanced observation. ^{Reference Cook, Palesy, Lapkin and Chenoweth28,Reference de Jong, Weselman, Kitchen and Hill51,Reference Flaherty and Little52,Reference Rachh, Wilkins, Capodilupo, Kilroy, Schnider and Repper-Delisi54–Reference Wray and Rajab-Ali56,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Haak, Gerstenberger and Reddy69,Reference Eeles, Thompson, McCrow and Pandy70,Reference Voetelink, Trout, Murphy, Rogers, Harris and Ramagnano73,Reference Grealish, Chaboyer, Darch, Real, Phelan and Soltau83,Reference Colella, Aroh, Douglas, VanBuitenen, Galesi and Castro86,Reference Salamon and Lennon88,Reference Eastwood and Schechtman92,Reference Purvis, Kaun, McKenna, Weber Viste and Fedorov94,Reference Skowronsky, Bena and Albert95} Interventional research and quality improvement initiatives typically yielded no changes in fall rates following adaptations in the ward environment, staff training, diversional activities and delirium checklists. ^{Reference Rachh, Wilkins, Capodilupo, Kilroy, Schnider and Repper-Delisi54,Reference Haak, Gerstenberger and Reddy69,Reference Salamon and Lennon88,Reference Skowronsky, Bena and Albert95} By contrast, other studies focusing on prevention and management of delirium reported trends towards reduction in post-implementation fall rates, although either statistical testing was not reported, or statistically significant differences were not observed in these studies. ^{Reference Wray and Rajab-Ali56,Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Weiss, Reddy and Anton91}

Restraint

Physical restraint was recognised in several texts as a risk to patient safety. ^{Reference Flaherty and Little52,Reference Wray and Rajab-Ali56,Reference Waszynski, Veronneau, Therrien, Brousseau, Massa and Levick60,Reference Flynn, Prescott and Jinks64,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72,Reference Colella, Aroh, Douglas, VanBuitenen, Galesi and Castro86,Reference Salamon and Lennon88} Studies described interventions including the introduction of restraint-free cohorting rooms for people with dementia, ^{Reference Flaherty and Little52} or protocols and justification forms to guide decision-making about initiation of observations. ^{Reference Salamon and Lennon88} The authors reported limited use of physical restraints after implementation of these strategies, but the absence of baseline data or a comparator group prohibited comparisons with usual care. Further studies detailed a range of strategies including delirium prevention efforts, improved patient engagement, and addressing pain and discomfort. ^{Reference Wray and Rajab-Ali56,Reference Sinvani, Warner-Cohen, Strunk, Halbert, Harisingani and Mulvany72} These indicated promising trends towards reduction in physical restraints, yet results of tests of statistical significance were seldom presented. ^{Reference Wray and Rajab-Ali56}

Neglect of caring responsibilities

Unsafe staff behaviours and practices, in which staff neglected to attend to patient needs, were the subject of further studies. ^{Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29,Reference Broski, Eckels, Blomquist, King and Lineberry68} Examples included staff being observed to fall asleep during enhanced observation shifts. ^{Reference Broski, Eckels, Blomquist, King and Lineberry68} Another example involved a patient’s constipation having remained undetected over a prolonged period, while staff were seen reading magazines instead of paying attention to this individual’s needs. ^{Reference Moyle, Borbasi, Wallis, Olorenshaw and Gracia29}

Missing patients

Patients going missing, having absconded from hospital during enhanced observation, represented a further safety concern. ^{Reference Pinkhasov, Singh, Chavali, Legrand and Calixte63,Reference Northcott, Mwale, Wyatt, Series, Harrison Dening and Featherstone65,Reference Concannon, Carr, Doherty, McInerney, Birrane and Kearney78,Reference Kroll, Stanghellini, DesRoches, Lydon, Webster and O’Reilly93} In some studies, patients returned without incident. ^{Reference Kroll, Stanghellini, DesRoches, Lydon, Webster and O’Reilly93} However, another study reported that the patient involved unfortunately died by suicide while missing from the ward. ^{Reference Concannon, Carr, Doherty, McInerney, Birrane and Kearney78}

Medication safety

Finally, medication safety problems were also identified, although the data were limited in depth and volume. ^{Reference Talley, Davis, Goicoechea, Brown and Barber77} The literature included a report of a medication error involving a patient assigned an observer. ^{Reference Talley, Davis, Goicoechea, Brown and Barber77} However, details of the nature and circumstances of the error were lacking.