2.1 Approach to searching

Searches were preplanned, with a comprehensive search strategy developed to identify all available studies. Therefore, this review was designed to promote inclusivity.

2.2 Electronic search strategy

A scoping search identified that the lack of a unifying term for PID was problematic. With help from a specialist librarian, a comprehensive and sensitive search strategy was developed. A copy of the search strategy for OVID Medline can be found in Table S2. Key words and Medical Subject Headings terms for the three main topic areas (adults living with dementia or cognitive impairment, their carers or health professionals; PIDs; qualitative research or mixed-methods research) were combined, according to the inclusion criteria.

2.3 Inclusion criteria

Inclusion and exclusion criteria were determined a priori. The population of interest included adults living with dementia or cognitive impairment, their carers and HCPs. The term ‘carers’ is used throughout the review to refer to adult informal and unpaid carers, usually family or friends of the PLWD.¹ We included studies conducted in high-income settings (according to the World Bank's income classification) and included community, care home, primary care and hospital.

The intervention of interest was a PID, defined for the purposes of this review as a document assisting essential care through the communication of key information about the PLWD (such as their preferred name), their preferences and routines. Documents with a primary aim of communication of clinical information between HCPs or focussing on end-of-life care or advance directives were excluded. PIDs included within a toolkit of interventions (referred to here as multicomponent studies) were included for completeness. These studies were analysed separately in parallel to single-component intervention studies as the barriers/facilitators may differ due to the presence of additional interventions and the difficulty in extracting data specific to the PID as opposed to other intervention components.

Qualitative, quantitative and mixed-methods studies reporting barriers and facilitators were included. We included published studies and reports, and excluded those that had only been published as conference abstracts or unpublished PhD or Masters theses. Studies had to be published in English (the language spoken by the review team) to ensure that themes were appropriately identified, understood and represented. Date limits were applied from 2010 to the present, as the Royal College of Psychiatrists recommended PID use following their 2010–2011 National Audit of Dementia.⁹ In addition, the Alzheimer's Society developed and launched ‘This is me’, their PID in 2010,¹⁶ and this was endorsed by the Royal College of Nursing in 2010.

2.4 Data sources

The following databases were searched: CINAHL, HMIC, MEDLINE, PsycINFO, Scopus, Web of Science to cover a range of healthcare, social science and nursing journals. Grey literature databases EThOS, OpenGrey and GoogleScholar were also searched as we felt that relevant data might be available in websites and project reports of dementia-related charities. Reference lists of relevant studies were hand searched for further articles or resources that might have been pertinent. In three cases, no full text accompanied relevant abstracts and so authors were contacted. Two responded, but full-text articles did not fulfil the eligibility criteria. Databases were searched between 2 July 2020 and 9 July 2020, and an updated search was conducted between 7 March 2022 and 9 March 2022. OVID was searched in 2022, in place of searching Medline, PsycInfo and HMIC separately.

2.5 Study screening

Records identified from database and website searching were imported into EndNote Online. Records were deduplicated and titles clearly irrelevant to the research question were excluded by the lead author. The lead author reviewed abstracts and full texts against eligibility criteria. The reasons for exclusion were documented. A second reviewer independently reviewed 10% of abstracts and full texts. Disagreements were discussed. Records eligible at abstract review were reviewed at full text using the same process.

2.6 Data extraction

Data from the results sections of primary studies were extracted. We used a data extraction table, which included quality score, research question, study methodology, results (including themes identified) and limitations.¹⁷ Data extraction tables were completed by the lead review author; 10% were independently checked by the second author for consistency.

2.7 Quality appraisal

Studies were quality appraised using the checklist from ‘Methods for the development of NICE public health guidance’¹⁷ to assess their robustness of their conduct. The lead author appraised each paper for quality, and two papers were independently assessed by the second author for consistency. Each paper was then assigned an overall assessment score (++, +, −). There are no validated mechanisms to select studies based on their quality score for qualitative reviews,¹⁵ and so all studies were included irrespective of score. In addition, the priority for the review was generating breadth of opinion including discordant observations.

2.8 Analysis and synthesis

A thematic synthesis was conducted.¹⁵ This method draws on concepts used for thematic analysis in qualitative research in primary studies as a method of identifying and developing themes within the data. This is an accepted method to review qualitative studies as the links between the primary studies and final synthesis are maintained.¹⁸ Analyses were conducted separately for papers describing single- and multicomponent interventions. First, data from the results sections of papers were inductively coded using a line-by-line method according to the meaning and content by the lead author, supported by qualitative data software NVIVO v12. Direct quotes were all coded, as were author statements.¹⁹ Codes were structured (in tree form) or in free form (without hierarchical structure). New codes were created as necessary as we progressed through the papers and similar, or related, codes were grouped. If research was reported in a journal article as well as a research report, both were included. A second reviewer read and coded two of the papers. Although the initial analysis was conducted by the lead author, the analysis was reviewed in discussion with all review authors. The descriptive inductive themes were further developed by grouping codes with similar meanings. As barriers and facilitators often described the same phenomenon from different angles, they were analysed together; however, it was noted whether the quote illustrated a barrier or enabler.

Descriptive themes were then mapped onto the COM-B model for behaviour change¹² to develop an analytical framework for the barriers and facilitators to healthcare passport use, from the perspectives of PLWD, carers and HCPs. The COM-B model is used within the behaviour change wheel, which was designed specifically to support behaviour change interventions.¹²

3.1 Physical capability

Attending training was predominantly discussed by HCPs, with difficulties attending training identified as a barrier,²⁷ and embedding training into existing education seen as a facilitator.²⁷ Completion of training was believed to enable a common understanding that would help HCPs keep the PID near the patient, and not throw the form away²⁷—carers also commented on the need for a shared appreciation of the PID.²⁵

The disease process was a barrier to PLWD using the PID, due to the physical and cognitive manifestations of dementia^{10, 27, 28} or comorbidities. PLWD may forget, or have difficulties writing.^{10, 28} These challenges were greater for PLWD who lived alone or those dependent on carers who were already overwhelmed.¹⁰

3.2 Psychological capability

HCPs identified that their own awareness of the PID is crucial,^{10, 23} for example, paramedics' awareness was seen to be necessary to transfer information between healthcare settings.²⁷ HCPs who were aware would ensure that the PID is ‘kept with the patient’,²⁷ and not ignored or thrown away.²⁷ However, Leavey et al.²⁸ described PLWD/carers with hesitant attitudes towards the PID, resistant to change despite increased awareness. HCPs, carers and PLWD had suggestions for increasing awareness, for example, using local team meetings,²⁷ promotional material in patient-facing areas¹⁰ or TV advertisements.¹⁰

Studies showed that remembering the PID is crucial to its use; for example, HCPs commented that PLWD need to take it with them.¹⁰ Suggestions to help HCPs, carers and PLWD to have the PID available included having it available on a mobile phone¹⁰; reminder stickers or posters in healthcare or care home settings^{10, 27}; and alert systems on healthcare computers,²⁷ and that carers could remind HCPs to use the PID.^{23, 24}

Forgetfulness was a barrier as the PID was sometimes left at home,^{10, 28} not transferred between the care home and hospital²⁷ or, once in the hospital, was forgotten and not used.^{23, 27} There was concern that PLWD would forget the existence of the PID and carers may be too stressed about the admission to remember it.¹⁰

3.3 Physical opportunity

Accessibility of the PID was a key enabler to PID use. The availability of blank forms in the community enabling PLWD and carers or care home staff to complete before hospital admission,^{23, 24} and completing the PID at the PLWD's home^{10, 22} were facilitators. Electronic access to blank PIDs (e.g., on a computer or a mobile app) was seen as helpful by PLWD, carers and HCPs;^{10, 27} however, some preferred paper formats.¹⁰ Some studies scanned completed paper PID into electronic systems, which was found to be helpful.^22-24 Paper PIDs were found filed away in patient notes²²; this was overcome by keeping them in a defined location.^{27, 30, 31}

PLWD,¹⁰ carers^{10, 25, 28} and HCPs^{10, 27, 28, 30} all reported that lack of HCP time was a barrier to PID use. Some PIDs were more onerous to complete than others: shorter documents were generally preferred,²⁷ whereas longer documents were seen as too cumbersome^{10, 28} and people were discouraged by the time required for PID completion.¹⁰ Furthermore, keeping the document up to date was seen as onerous.¹⁰ PLWD,¹⁰ carers^{10, 25} and HCPs^{22, 27} all reported competing priorities as a barrier, although the activities being prioritized varied. Policy changes such as the introduction of the 4-hour rule in the emergency department where patients should be admitted or discharged within that time frame were considered to be barriers.²⁷ PIDs that were easy to use were reported to have increased chance of use than those that were complex.^{10, 22, 25, 27, 29-31}

The presence of carers was important: barriers to PID use included PLWD who were alone or lacked family,^{22, 27} and facilitators to its use were the presence of nominated individuals to take responsibility for the PID.^{10, 22, 24, 27, 28, 31} HCPs observed that they too were well placed to complete PIDs including in conjunction with carers.^{22, 27}

Organizational-level factors were noted to influence PID use, for example, prioritization of health service response to the Ebola crisis or health service reaccreditation requirements.²⁷ Staff perceptions of small numbers of PLWD being admitted to hospital represented a barrier,²⁹ but proactive identification of PLWD overcame this.²⁷

3.4 Social opportunity

PIDs were seen as able to both cause²⁷ and reduce stigma,¹⁰ and suggestions to reduce stigmatization included informing PLWD that the PID is commonly used.¹⁰ Embedding the PID into organizational norms and incorporating the PID into existing activities helped its use, for example, as patients were readmitted, bringing their existing PID with them.²⁴ HCPs commented on a lack of process as a barrier, overcome by facilitators including keeping the PID with the PLWD throughout their journey,^{24, 25, 27} ensuring that the patient would be discharged with their PID²² and clarifying roles and responsibilities.²²

A culture of collaboration enables PID use. HCPs stated that team working^{27, 29} and using existing communication networks including communication with carers^{26, 27, 29} were beneficial. Good leadership promoted PID use, through clinical champions, modelling desired actions, local implementation teams and senior management support.^{27, 29} Considering PIDs as a process of continuous improvement supported its development through regular review of the initiative,²⁷ sharing of barriers²⁷ and cultivating a supportive environment and culture.^{10, 24, 30}

3.5 Automatic motivation

The desire and willingness of PLWD and carers to use the PID influence the acceptance of PIDs. The PID was seen as a reminder of illness by PLWD¹⁰ and an intrusion into normality by carers.²⁸ For those with mild dementia, the PID was seen as a future necessity once dementia had progressed^{10, 22, 27, 28} and the PLWD/carer had come to terms with the diagnosis.¹⁰ Carer stress^{10, 27, 28} encompassed carer health and well-being and the carer's capacity to cope. The strength of the carer–PLWD bond appears to be important in terms of PID use: where the bond was strong, PID use was better,²⁸ but PIDs were used less where carers were overwhelmed or upset.^{10, 27}

The attitudes of HCPs, carers and PLWD towards the PID were important. Some carers considered the PID a chore.^{10, 28} HCP buy-in was important,^{22, 29} and some HCPs considered the PID as someone else's job.²⁷ Concerns were raised about the PID: carers were worried about data protection,²⁸ and HCPs feared for the legal status of PIDs.²⁸ PLWD and carers were anxious or embarrassed to write in the PID.^{10, 28}

3.6 Reflective motivation

The predominant theme within the synthesis was the value of the PID. PLWD, carers and HCPs had all imagined the positive impact that the PID could have on PLWD,¹⁰ on carers¹⁰ and on HCPs.^{22, 25, 26, 30, 31} For PLWD, the PID would improve patient-centred care,^{10, 22, 23, 25-30} help retain their identity²² could reduce behaviours that challenge^{22, 25-27, 29} and improve their comfort. Carers' stress could be reduced through their knowing that their loved one would receive improved quality of care,²⁷ and that their loved one would be cared for in their absence.^{10, 28} HCPs found the document time-saving,^{22, 26, 30} reduced workload,³¹ helped undertake the job^{10, 22, 24, 27, 31} and improved the quality of care that they were able to give.^{22, 25-27, 29, 30} HCP confidence improved with continued PID use.^{27, 29, 30} Communication between PLWD, carers and HCPs was improved,^{26, 27} for example, where the PLWD was unable to communicate for themselves.^{10, 25} There was an impact on the healthcare organization as well, with reductions in specialist nursing²⁷ and complaints,²⁹ and improved restructuring of the ward.²⁶ Where barriers were identified, it was because the benefits of the PID had not been observed,^{26, 27, 30} scepticism that the PID would be used²⁸ or that the benefit of the PID did not outweigh the burden of using it.¹⁰

Prior experience of similar initiatives influenced receptiveness to PID use, causing PLWD or carers to self-censor,²⁸ engage²⁷ or disengage^{10, 25, 28} with PIDs.

Barriers were identified regarding the purpose,^{23, 25} use^{25, 28} and ownership^{10, 28} of the PID by HCPs, carers and PLWD. There was a belief amongst PLWD and carers that the PIDs were not needed, for example, that HCPs would already be sharing the information,²⁸ or that the carers would be present for any hospital admissions for the PLWD.²⁷ Carers may not have understood the purpose as they reported confusion around the perspective in which the PID should be completed.²⁵ There was uncertainty over who had ownership of the PID¹⁰ and who would have access.²⁸ Carers and HCPs provided suggestions on improvements to the PID to improve its relevance.^{30, 31}

3.7 Multicomponent interventions

Themes from multicomponent analysis corroborated those identified from the single-component studies; for example, in physical capability, the manifestations of dementia were identified as a barrier to PID use by carers³⁷ and HCP training improved PID use.³⁸ Carers described difficulties understanding the language of the PID.³⁷ Clear ownership and responsibility was a facilitator.^{34, 37}

Similar to single-component interventions, accessibility of PID, for example, with electronic adaptation³⁴ or a defined location³⁷ represented enablers. Redundancy of the PID was noted in some contexts,³⁷ with a suggestion to adapt the PID to the local environment. Lack of time was reported as a barrier.^{36, 37} HCPs reported a fear of losing a PID during transfers between wards and suggested inclusion with other documentation and clarity of responsibility.³⁶ PIDs were perceived as normalizing dementia care.³⁷ The documents would also enable carers to feel included in care.³⁷

Carers were concerned about information required for the PID³² and whether the tool would be used.³⁷ HCPs were perceived as disinterested by carers³⁷; yet, HCPs reported feeling satisfied at being able to provide good care through using the PID.³⁴ Concordant with single-component intervention findings, the predominant theme was that of value of the PID with the potential to improve personalized care^{33-35, 37}; support the PLWD's unique identity^{34, 37}; and reduce or prevent responsive behaviours.³⁷

4.1 Relevance to practice

The COM-B model is used for behaviour change intervention development.¹² Some findings provide a direct opportunity to improve the use of PID, for example, ensuring that blank forms are available. Where nonmodifiable barriers have been identified by this review, such as the progressive nature of dementia, understanding interrelationships between wider barriers, facilitators and behaviours can aid intervention development to identify opportunities to mitigate some of the effects of a nonmodifiable barrier, for example, by considering the timing of the PID, supporting and empowering the carer or considering completion of the PID as a collaborative effort that included HCPs.

4.2 Role of the carer

The importance of supporting the carer was identified in this review: ‘the carer is the lynchpin’.¹⁰ Carers are required at key moments in PID use: authorship, ownership, bringing the PID to appointments and reminding HCPs. The PID is unlikely to be used where carers are overwhelmed or disengaged. However, carers have been called the ‘invisible second patients’, subject to significant strain, burnout and depression.⁴⁰ Caring duties may prohibit working, with economic consequences.⁴¹ Thus, any intervention to improve PID use should have both PLWD and their carers at their core. Supporting carers is a key component of the NICE guidelines.²

4.3 Healthcare professionals

HCPs form one perspective in this review; however, different HCP cadres will have different interactions with the PLWD, carer and the PID, and observe different benefits and challenges. Healthcare assistants and nurses provide much of the ‘hands-on’ essential care for inpatients and they should feel empowered to utilize the PID (e.g., through knowledge, access to the PID)⁴²—but they do not feel empowered to bring about change.⁴³ Other cadres of HCP are more influential in establishing organizational culture (e.g., those in positions of leadership and management) to enable and value nursing staff who spend time with PLWD to get to know them,⁴⁴ for example, through PID.

HCPs wanted to attend PID training and stated that it would improve their confidence, corroborating research findings that HCPs are not adequately prepared to care for unwell PLWD,⁴³ and the NICE guidelines advising specialized training on ‘understanding the person as an individual’.² Specialized training in dementia care for HCPs is advised,^{2, 45} improves HCP attitudes⁴³ and increases PID use.³⁸ HCPs' attitudes were identified as a barrier to the use of patient-held records in other studies—where the HCP did not understand the purpose of the document, or did not empower the patient as intended.⁴⁶

4.4 Healthcare services

Hospitals are task orientated rather than person-centred.⁴⁷ PLWD, carers and HCPs all identified that PID can promote person-centred care. Therefore, the organizational culture must support and value person-centred care to promote quality of care.⁴⁷ For example, studies have shown that getting to know PLWD on hospital wards was not prioritized by managers or ward staff, in part due to workload constraints.⁴⁴ High HCP workload is a barrier to the use of other patient-held records also.⁴⁶ Yet, although time and workload constraints were identified by this review as barriers, others stated that the PID could be time saving. The healthcare service must provide adequate resources—including staff time and valuing the staff that provide it.^{42, 45} Reprioritization at an organizational level could facilitate better PID use.

Other research has advocated for senior support for personalized care.^{39, 45} Patient safety is inherent to any healthcare organization,⁴⁸ and so this focus could be harnessed by the power of the PID to improve patient-centred care.⁴⁵ For example, Luxford et al. noted a reduction in chemical restraint since the introduction of their PID,²⁹ an important outcome as medications are associated with increased morbidity and mortality.⁴⁹

4.5 Overarching policy

A system-wide approach is needed to embed the PID fully—PLWD receive care from a number of organizations, and PLWD may be transferred between settings. Expectations and understanding of the PID must align for all stakeholders for successful PID use. In accordance with our findings, reviews of other PIDs identified the importance of clarifying in advance the responsibility for the PID, its purpose and process, including embedding the PID within current work practices.^{39, 46}

4.6 Review limitations

Difficulties were encountered during the literature search due to the lack of a unifying term for PID. There may be publication bias, for example, with internal service evaluations and audits of PID not having been published. Thus, some pertinent barriers and facilitators may not have been identified. However, a range of studies, heterogeneous in design, population and PID used, were identified. This review utilized an inclusive approach to garner the breadth of barriers and facilitators.

Four of the 19 studies were of low quality (scoring ‘−’ on quality appraisal), with the remaining 15 scoring ‘+’ and none scoring ‘++’.

There is no standard PID in use, and so the PID used differed between studies. Barriers and facilitators may depend on the PID used (e.g., some study participants complained that the PID was too long²⁸; others stated that the PID was short and easy to use.²⁹ This will limit comparability between studies—but also validates some findings as they are reproducible in different settings.

There is limited external validity of findings; they may need adaptation for consideration to context. For example, cultural differences between healthcare organizations or PLWD demographic groups may influence engagement with PID. Parke et al. had separate focus groups for English- and French-speaking participants in Canada and noted that cultural differences mediated the appreciation of hospital-readiness tools, including PID. Changes to the language used were advised by participants.³⁷

The voice of PLWD remains quiet, especially in studies of PID within toolkits. Data were lacking on the influence of the stage of dementia on the barriers and facilitators to PID use. This may reflect difficulties in recruiting PLWD into research studies.

Finally, while mapping themes to the COM-B model has benefits in terms of identifying future interventions to support PID use, some themes did not fit neatly within any one subcomponent of the model. In some ways, this reflects the nature of the COM-B as the components influence each other (capability and opportunity can influence motivation). In our mapping, we have, for example, mapped remembering to use the PID to the psychological capability subcomponent, but it can also have resonance with physical opportunity if there are resources in the environment that might aid memory. The intersections across subcomponents reflect the limitations of mapping to a model in which components interact.

4.7 Implications for future research

Future research should consider the impact of the SARS-CoV-2 pandemic on PID use in dementia care, as visiting restrictions to homes and hospitals have changed, and so the relative value of the PID may have changed. The impact of PLWD and carer stress, HCP workload and organizational priorities may also have changed the use of PID in practice.

The impact of PID use on actual person-centred care experienced by PLWD should be studied. Increased completion of the PID or increased availability of PID in hospital settings may not translate to actions undertaken by HCP (the ‘know-do gap’).⁵⁰