How Can Quality of Dementia Care Be Measured? The Development of Clinical Quality Indicators for an Australian Pilot Dementia Registry

Abstract

Background:

A clinical quality registry (CQR) for dementia provides benefits to those living with dementia and their carers by improving the quality and experience of care through benchmarking and monitoring patient outcomes. CQRs use data collected to form clinical quality indicators (CQIs) through which variations in clinical processes and outcomes between different services and jurisdictions can be highlighted.

Objective:

This modified Delphi study aimed to develop CQIs for a pilot Australian CQR for dementia and mild cognitive impairment. These CQIs are based on evidence, patient and caregiver experience, and clinician perspectives across the trajectory of care from diagnosis to end-of-life.

Methods:

An initial list of indicators from existing dementia registries, academic literature, and clinical practice guidelines was synthesized. A working group of clinicians and registry experts further refined these indicators. A panel of experts comprised of a consumer, a carer, clinicians, consumer organization representatives, and academics. The experts participated in three phases of the modified Delphi study: 1) online survey for scoring importance and validity, 2) a one-day face-to-face discussion, and 3) final survey round to assess importance, validity, and feasibility.

Results:

The panel assessed 33 CQIs and confirmed a final set of 18 indicators. The CQIs mapped to the domains of quality of diagnosis, quality of management, access to services and supports, and potentially preventable complications. These CQIs will be tested initially in memory clinics and inform the data collection processes for the Australia Dementia Network Registry (ADNet).

Conclusion:

A dementia CQR is fundamental to ongoing monitoring and development of good quality and consistent care across Australia.

Keywords

Alzheimer’s disease clinical quality registry dementia modified Delphi study

INTRODUCTION

As the global population ages, the number of people living with dementia is projected to increase [1]. Global prevalence estimates predict 48.1 million people will have dementia in 2020, increasing to 90.3 million in 2050 [2]. Since 2013, dementias, including Alzheimer’s disease, have become the second leading cause of death in Australia, after ischemic heart disease [3]. Clinical care for people with dementia consists of early, supported, and accurate diagnosis including identification of reversible or contributory factors; provision of information to both patient and carer; opportunity to consider and initiate treatment with cholinesterase inhibitors for Alzheimer’s disease and Lewy body disease; and strategies to equip people with dementia and their caregivers in planning for the future [4]. A number of lifestyle and psychosocial interventions may also improve the quality of life and clinical course for both patient and carer throughout the entire trajectory of the disease [4].

Gaps in the care of patients living in Australia with dementia have been reported. Delays in confirmed diagnosis, which can occur up to three years past initial symptom recognition [5], may hinder benefits of available interventions and support, increase uncertainty and anxiety related to undiagnosed symptoms, and preclude planning for the future. Further, there is variation in access to dementia services, particularly between metropolitan and regional areas in Australia due to limited health professional availability in rural and regional areas [6].

Clinical quality registries (CQRs) monitor the natural history and care processes of specific diseases or health interventions [7]. Like all clinical registries, they collect a minimum dataset (MDS) of information related to individuals diagnosed with a particular disease, undergoing a particular procedure, or using a medical device [8]. A key aspect of clinical quality registries in particular, is their development of specific measures—clinical quality indicators (CQIs)—that measure and monitor the quality and processes of care [9]. The MDS of a CQR also includes items that allow for differences in patient case mix (such as demographic information and comorbidities), thus enabling risk adjustment of benchmarking and ensuring equitable comparisons across institutions [8, 10].

The Australian Council on Healthcare Standards states that CQIs should “screen, flag or draw attention to a specific clinical issue ... they are designed to indicate potential problems that might need addressing ... they are used to assess, compare and determine the potential to improve care. Indicators are therefore tools to assist in assessing whether or not a standard in patient care is being met” [11]. Essentially, CQIs are developed and used to assess health care processes and outcomes, and to monitor, evaluate, benchmark and improve the quality of patient care and interventions that impact patient outcomes [12].

CQRs for dementia or cognitive impairment exist in a number of countries around the world, most notably in Sweden (SveDem), Denmark and Norway [13 –15]. SveDem has been in operation since 2007 [13]. By 2012 it was estimated to have captured at least 36% of incident dementia cases from both specialist memory services and general practitioners, and by 2018 had registered over half of all prevalent cases of dementia [16].

When developing a CQR, it is important to develop quality indicators that reflect the local practices, standards, and interventions to ensure appropriate participation from stakeholders [17]. A number of health professionals from different disciplines are involved in the care of the person with dementia across different care settings (community, acute hospital, residential aged care), hence a consensus process among these experts is needed to define quality of care for a person with dementia [18, 19]. Further, the MDS is largely developed to facilitate measurement against the CQIs; therefore, to minimize the data collection burden of a CQR, the number of CQIs should be kept to a minimum to enhance sustainability [20].

In 2017, Monash University received funding to assess the feasibility of a dementia CQR in Australia, including through the establishment of a pilot registry. This pilot registry is developing and testing the methodology for a national dementia CQR (the Australian Dementia Network (ADNeT) Registry). Data collection for the ADNeT Registry is proposed to be obtained from publicly funded memory clinics and private specialist consulting rooms (baseline data), follow up data from carers and patients, and from data linkage. In this paper, we report on a modified Delphi process used to identify and develop a feasible set of CQIs suitable across the clinical trajectory for patients with dementia and mild cognitive impairment (MCI), considering evidence, patient experience, and clinician perspective for the pilot dementia registry. The Delphi process uses a structured approach to integrating views from a range of experts [21]. Unlike other international dementia registries, MCI will be included in the ADNeT registry and in the pilot registry. The rationale for the inclusion of people with MCI is three fold: firstly, 15% of people with MCI presenting for assessment aged 65 or older will go on to develop dementia within two years [22], secondly dementia medication trials are targeting individuals with early disease and therefore inclusion in the registry will facilitate recruitment of people with MCI into clinical trials and finally, it is important to document the care and follow up pathways—including subsequent development of dementia—for these individuals.

METHODS

A modified Delphi study was conducted. A Delphi study is a systematic and structured process for reaching consensus amongst key stakeholders via a series of ‘rounds’ that use questionnaires [21]. The modified Delphi approach incorporates a face-to-face meeting to facilitate discussion and consensus [23]. This modified Delphi study consisted of six phases, including 1) preliminary indicator development from existing evidence, 2) initial refinement of the indicators, 3) an online survey of dementia experts, 4) a face-to-face meeting of the expert panel, and 5) a second online survey; and 6) further refinement of indicators by the CQI working group.

Phase 1: Preliminary indicator development

Four authors (DA, EP, MG, SR) compiled a set of draft indicators based on existing dementia registries, clinical and best practice guidelines, and the academic literature. These items were categorized as measures of processes or outcomes of care across five domains: pre-diagnosis, diagnosis, management, follow-up, and patient outcomes (complications and disease progression). The draft indicators were presented every two to four weeks to a CQI working group, which comprised a geriatrician, a gero-psychiatrist, registry experts, and a statistician.

Phase 2: Refinement of indicators

The CQI working group reduced the overall number of draft indicators via a process of consolidation (collapsing similar indicators) and by applying the Agency for Healthcare Research and Quality levels of evidence [24] with indicators graded as A (good research evidence) and B (fair research evidence) being prioritized.

Delphi consensus rounds

An expert panel of clinicians, policy makers, researchers, consumers (a caregiver of a person with dementia and a person living with mild cognitive impairment) and two representatives from a national consumer organization were purposively identified through the academic and clinical national networks of the research team. The project manager (KK) emailed individual invitations to participate in the Delphi process.

Phase 3: Online survey 1

The traditional Delphi technique is conducted via self-administered questionnaires over two or more rounds with no face-to-face meeting. This is to prevent the situation where a panel member may dominate the consensus process [21]. To ensure that a panel member did not dominate the consensus process, the first round of the modified Delphi was a self-administered survey. The CQI working group created a questionnaire of the draft indicators using the online survey platform Qualtrics^TM. The project manager sent explanatory documents with definitions, descriptions and evidence for each of the indicators to expert panel members. The expert panel were instructed to rate each indicator on a 9-point Likert scale for importance (1 being not important and 9 being extremely important). Participants were provided the option to select “unable to comment” on any indicator they believed was outside their area of expertise or experience. The survey also included free text entry fields where participants could provide comments or suggest changes to the proposed wording. The last question of the survey asked participants to nominate potential new indicators for the next round of the Delphi.

Median ratings and a disagreement index (DI) were calculated for each indicator based on the round 1 online survey results. The DI was developed by RAND and is a continuous scale measuring variation in expert ratings [25]. A DI of 0 represents complete agreement whereas DI≥1 indicates significant disagreement. The following statistical criteria were used for short-listing of proposed indicators: 1) a median score of at least 7; and 2) no to low disagreement according to the Inter Percentile Range Adjusted for Symmetry (IPRAS).

Phase 4: Face-to-face meeting

The expert panel then convened face-to-face and was presented with the results of the ratings of the draft indicators from the first online survey. During the face-to-face meeting, the panel discussed each proposed indicator and refined the wording (where applicable), after which each panel member re-rated the indicators on importance and feasibility using the 9-point Likert scale. Panel members were also provided with the opportunity to propose additional indicators, and to provide written comments in relation to the proposed indicators.

Following the face-to-face meeting, median rating and DI were calculated for each indicator. Specifically, the median rating score among the panel, along with proportion scoring “Not important (median score 1–3), “Uncertain median score 4–6), and “Very important (median score 7–9), disagreement index, and final recommendation (unsure or appropriate) were presented back to the panel via the second online survey.

Phase 5: Online survey 2

A small number of proposed indicators, which had not scored clearly ‘in’ or ‘out’ at the face-to-face meeting, were presented back to the expert panel via a second online survey in Qualtrics^TM. The panel members were asked again to rate the indicators in relation to feasibility and validity. New indicators proposed at the face-to-face meeting also were rated for importance, feasibility and validity during this second online survey. The median rating and disagreement index were calculated for each of the proposed clinical quality indicators.

Also during this second online survey, for some proposed indicators that were considered important but likely to be incompletely collected, the panel was asked to indicate their agreement (either ‘yes’ or ‘no’) as to whether these items were suitable for inclusion in the MDS, rather than as a key clinical indicator. Items proposed for the minimum dataset that scored a rating of over 50% ‘yes’ responses were included.

Phase 6: CQI working group refinement of indicators

The CQI working group refined and consolidated the final set of indicators and presented these to the Delphi Expert Panel for endorsement.

RESULTS

Eighteen experts agreed to participate in the Delphi study. Figure 1 details the characteristics of the panel members and the number of members participating in each round of the study. Sixteen panel members completed survey one, 15 panel members attended the face-to-face meeting with one attending via Zoom web conference technology, and 14 panel members completed the second survey. Table 1 details the median rating and levels of disagreement for each of the indicators across the three rounds of the Delphi. A description of the process can be found in the Supplementary Material.

Fig. 1

Process to develop CQIs for the pilot dementia registry.

Table 1

Delphi Process Results: Indicator (summary) numerators, median rating scores and disagreement results

^*denominator differs; Green, no disagreement in rating; Orange, disagreement in rating.

A summary of discussion via open-ended comments in the questionnaires and the face-to-face meeting provide an insight into the challenges in developing a set of CQIs for dementia and MCI. They are provided below for each of the domains: pre-diagnosis, diagnosis, follow up and management, and outcomes.

Pre-diagnosis

The panel members identified the item relating to assessment within six months of raising concerns as being difficult to measure in practice. Most panel members recognized that the carer would be reporting the duration of symptoms and questioned the validity of this measure as it is highly subjective and there is significant variability in definition of what constitutes “raising concerns”.

“It may be difficult to establish when the concern was first raised with a GP. This is often done over several visits.”

Diagnosis

Panel members recognized the lack of guidelines and uniform standards for cognitive testing, and specific timelines for assessments and specialist consultations.

“Lack of guidelines and uniform standards in terms of testing.”

Some items were considered important but not sufficiently embedded in clinical practice to be appropriate for a CQI (e.g., sub-classification of dementia type, dementia severity); however, by retaining as a data item in the registry and providing regular feedback to clinicians, these items may become routine clinical practice. A number of panel members commented that the diagnosis and assessments of dementia or MCI should be made by a suitable specialist (e.g., geriatrician, neurologist, psychiatrist, gero-psychiatrist) to exclude other causes of symptoms and provide information for future planning and access to support and care services. However, it was acknowledged that not everyone accesses a specialist and receives a formal diagnosis.

“We aren’t capturing the multitudes of people out there that are labelled with a diagnosis of “dementia” without any proper assessment - eg those that attend a hospital or their GP and either mention memory problems or appear confused and have “dementia” added to their diagnosis list without any real investigation to base that diagnosis on.”

Follow up and management

Overall, the panel members believed that the management indicators had a low level of evidence and would require reporting from people with dementia and caregivers, which could compromise validity.

“Memory clinics in Victoria are not funded to provide ongoing review once dementia diagnosis is reached - so would not be able to provide this information.”

The panel members highlighted the challenges related to the indicator on prescription of medication due to issues with contraindications, eligibility for the Pharmaceutical Benefits Scheme (government-subsidized medications) and prescription via general practice for ongoing management, which is beyond the scope of the memory clinic specialist.

“Some people have contra-indications for taking the medication or don’t wish to take it. Data need to be interpreted in the context of knowing the proportion that these apply to.”

Lifestyle and allied health referrals (e.g., physiotherapy, occupational therapy) were suggested by panel members to be encouraged via general practice. However, panel members did query whether this indicator would capture whether there was a need for lifestyle and allied health support and how useful this was in assessing quality of care.

“Indicators for the need for physiotherapy referral are very variable - Is this a measure of access? Needs to reflect access where there has been a need - although that can be very subjective”

“I am not sure that it is an absolute for the occupational therapist to assess activities of daily living and instrumental activities of daily living as well as the living environment. Assessment of the whole situation usually occurs as a team approach, or by the family GP in discussion with the family and the person. The physio, continence nurse specialist, movement disorders specialist are also skilled in assessing aspects of daily living abilities and opportunities.”

A challenge highlighted by panel members related to the use of different terminology between states and territories and settings (hospital versus community) and how this could be captured in the wording of the indicators. Issues of stigma and cultural appropriateness of respite care and support services were also discussed by panel members, recognizing that for some cultures respite care may not be the cultural norm.

Outcomes

Most of the panel members did not perceive the proposed item of admission to hospital from the community and discharge to permanent residential care as an indicator of good quality of care. They posited that the reasons for discharge to permanent residential care is multifaceted, for example, it may be an indication of co-morbidities requiring aged care, or may be due to poor outcomes following hospitalization (e.g., due to delirium, falls, or reduced function). Additionally, it may be appropriate for the person to enter residential aged care and this would be an indication of good care not poor care.

Two indicators—Time from diagnosis of dementia to permanent residential care and Time from diagnosis of dementia to death—were removed initially as indicators of quality of care due to low scores for importance and feasibility as they reflect the complexity of patient factors which may be beyond the control of clinicians and the health system. These indicators were reinstated as longitudinal clinical indicators to indicate disease progression. Table 2 presents the final set of summary indicators categorized under domains aligned with both quality of care and disease progression (quality of diagnosis, quality of management, access to services and supports, potentially preventable complications, and disease progression) and the proposed data sources. Table 3 maps the indicators in the international dementia CQRs against the domains of this pilot dementia registry. The majority of existing registries tend to have indicators focused on the quality of diagnosis with one or two indicators for management. SveDem is the exception to this with indicators for aged care and annual service use and clinical data. The indicators generated from this Delphi study map across diagnosis, management, access to services and supports, potentially preventable complications and disease progression.

Table 2

Proposed Clinical Quality Indicators for the pilot dementia registry

CQI (Summary)		Population	Data Source
Quality of Diagnosis
1	Cognitive assessment by HP within 6 months of raising concerns	All patients with Dx of dementia/MCI	Memory Clinic (MC)/Specialist
2	Basic blood tests completed within 6 months of Dx (FBE, U&E, LFTs, B12, folate, Ca, TSH, fasting glucose)	All patients with Dx of dementia/MCI	MC/Specialist
3	Multiple cognitive domains assessed within 6 months of Dx	All patients with Dx of dementia/MCI	MC/Specialist
4	Neuro-imaging completed within 6 months of Dx (CT or MRI)	All patients with Dx of dementia/MCI	MC/Specialist
5	Person with dementia assessed for function within 6 months of Dx (ADLs, IADLs, PADLs)	All patients with Dx of dementia/MCI	MC/Specialist
Quality of Management
6	Persons with Dx of MCI who have cognition reviewed within 6-18 months	All patients with MCI	MC/Specialist
7	Persons with mild/mod AD taking medications	All patients with mild/mod AD	MC/Specialist &annual carer survey
8	Persons with dementia or MCI who have an advance care plan	All patients with Dx of dementia/MCI	MC/Specialist &annual carer survey
9	Persons with dementia who have annual review of function &cognition by a health professional	All patients with Dx of dementia	annual carer survey
10	Persons with dementia who are assessed for changes in behavior in a 12/12 period	All patients with Dx of dementia	annual carer survey
Access to Services and Supports
11	Time from referral to specialist appointment made is within 90 days	All patients with MCI/dementia with referral data and appointment dates	MC/Specialist
12	Persons with dementia in community waiting for home support services	All patients with Dx of dementia in community	MC/Specialist, data linkage &annual carer survey
13	Persons with dementia in community who had an ACAS/ACAT assessment in last 12/12	All patients with Dx of dementia in community	Data Linkage
Potentially preventable complications
14	Persons with dementia/MCI who are admitted to hospital for a non-elective procedure in last 12/12	All patients with Dx of dementia/MCI	Data linkage –hospital data
15	Persons with dementia who are admitted to hospital with delirium in last 12/12	All patients with Dx of dementia/MCI	Data linkage - hospital data
16	Persons with dementia in residential care who are admitted to hospital following a fall in last 12/12	All patients with Dx of dementia/MCI in residential care	Data linkage –hospital data
17	Time from Dx of dementia to permanent residential care	All patients with Dx of dementia	Data linkage
18	Time from Dx of dementia to death	All patients with Dx of dementia	Data Linkage

Dx, diagnosis; CQI, clinical quality indicator; MCI, mild cognitive impairment.

Table 3

CQIs from international dementia clinical quality registries mapped to the Australian pilot dementia registry domains

Indicator domain	Pilot Australian Dementia Registry CQIs	Denmark Registry [14]	NorKog –Norway Dementia Registry (email from registry custodian)	SveDem –Sweden Dementia Registry [13]	ReDeGi –Dementia Registry Girona –Spain [29].
Quality of diagnosis	Cognitive assessment by HP within 6 months of raising concerns	Percentage of demented patients amongst number referred	Proportion of patients where information from a proxy (usually family) is collected, and whether this was done as an interview, a telephone call or a questionnaire sent by post.	Proportion of patients diagnosed with dementia during last year	approximate date of the onset of symptoms
	Basic blood tests completed within 6 months of Dx (FBE, U&E, LFTs, B12, folate, Ca, TSH, fasting glucose)	Proportion of patients evaluation within 90 days	Proportion of patients where ADL-function and change in cognition and ADL-function (actually IQ-code) is assessed.	Proportion of patients undergoing basic dementia work-up (patient history, blood tests, simple cognitive testing and CT)	date of the diagnosis
	Multiple cognitive domains assessed within 6 months of Dx	Proportion of demented patients assessed with MMSE	Proportion of patients where neuropsychiatric symptoms are assessed (NPI is most commonly used)		DSM-IV-TR diagnostic criteria
	Neuro-imaging completed within 6 months of Dx (CT or MRI)	Proportion of demented patients assessed with IADL-FAQ scale	Proportion of patients assessed for depression (MADRS or Cornell scale for depression in dementia)		additional diagnostic criteria for the subtype of dementia
	Person with dementia assessed for function within 6 months of Dx (ADLs, IADLs, PADLs)	Proportion of demented patients with structural brain scan (CT or MRI)	Proportion of patients with dementia that receives a specific dementia diagnosis		score and date of administration of the Mini-Mental State Examination (MMSE
	Proportion of demented patients where the etiological diagnosis is determined	Proportion of patients with imaging (MR or CT) as part of the assessment			score and date of administration of the Blessed Dementia Rating Scale (BDRS
					score of the Clinical Dementia Rating (CDR)
					Dementia risk factors: past family history of dementia, present diagnosis of hypertension, diabetes mellitus, dislipidemia, stroke, thyroid disease and a past history of depressive disorder
Quality of management	Persons with Dx of MCI who have cognition reviewed within 6–18 months	Proportion of patients with AD, DLB and PDD treated with anti dementia drug	Proportion of patients in which the ability to drive has been assessed and, if necessary, any action has been taken.	Proportion of Alzheimer patients treated with cholinesterase-inhibitors or memantine
	Persons with mild/mod AD taking medications	Control, follow-up or referral after the initial assessments	Proportion of patients treated with antipsychotics in nursing homes
	Persons with dementia or MCI who have an advance care plan	Proportion of patients followed up at least once a year
	Persons with dementia who have annual review of function &cognition by a health professional
	Persons with dementia who are assessed for changes in behavior in a 12/12 period
Access to services and supports	Time from referral to specialist appointment made is within 90 days			Proportion of patients with day-care at diagnosis
	Persons with dementia in community waiting for home support services			Proportion of patients living in nursing homes
	Persons with dementia in community who had an ACAS/ACAT assessment in last 12/12
Potentially preventable complications	Persons with dementia/MCI who are admitted to hospital for a non-elective procedure in last 12/12
	Persons with dementia who are admitted to hospital with delirium in last 12/12
	Persons with dementia in residential care who are admitted to hospital following a fall in last 12/12
	Time from Dx of dementia to permanent residential care
	Time from Dx of dementia to death

DISCUSSION

This study identified 18 clinical quality indicators with which to measure quality of care for a pilot dementia registry in Australia. The identified CQIs capture quality of care and patient outcomes across the trajectory of disease and care for patients with dementia (see Fig. 2). While the majority of the final CQIs are related to the diagnosis and management of dementia, there are also CQIs relating to patient follow-up (CQIs 12-13), and intermediate outcomes that may indicate poor quality care including admission to acute hospitals with potentially preventable complications (CQIs 14–16). Unlike other existing dementia registries internationally [15], the CQIs identified in this project map across the continuum of care and across the trajectory of disease.

A CQR for dementia is challenging as the diagnosis process in Australia can be undertaken in multiple settings by different specialists (psychiatrist, geriatrician, neurologist) or by general practitioners [26 –28]. This study found that the ratings of the CQIs from the expert panel members varied largely due to individual views regarding importance. The inconsistency among investigations and clinical assessments, lack of disease-modifying treatment options and standardized care, and minimal existing data make measurement and monitoring of outcomes of patients with dementia and MCI difficult.

Additionally the evidence base supporting strategies of care for patients with dementia is limited [4]. The Delphi process highlighted that the lack of evidence-based effective treatments for management of dementia led to disagreement among panel members regarding the importance of potential indicators. Even more challenging, however were considerations of feasibility in relation to the collection of required data items for many CQIs.

Further, patients are managed primarily in community settings with variable interaction with health professionals, and where clinical data collection systems are not embedded nor aggregated routinely as it is in the acute setting. The baseline diagnosis indicators were considered items that were either collected as part of routine practice or could reasonably be collected or derived. The remaining CQIs will be collected via data linkage to existing primary data elsewhere (e.g., deaths, hospitalizations) or are to be collected de novo from carers. Identifying a source of data for the indicators was a major factor in the discussions at the face to face meeting; including issues around different data collection systems and processes in different Australian jurisdictions. Thus, following baseline data collection, the Australian pilot dementia registry proposes to collect data as reports from patients and carers and by data linkage to pharmaceutical, hospital and nursing home data. Identified medication and procedure related data require additional individual patient consent, and therefore further investigation as to its feasibility and usefulness for registry purposes.

One of the challenges that the Australian pilot dementia registry will face is the challenge of collecting data for the CQIs in a decentralized system that differs from European countries [13 , 29]. A more feasible option is to provide a standard data collection tool for memory clinics as they exist in many Australian states and undertake services including diagnosis, initial management, and referral. The pilot dementia registry will therefore test the feasibility, recruitment methods and data collection processes for the CQIs identified in this study. Examination of whether the data for the CQIs can be collected via data linkage and annual patient surveys, which include a patient reported outcome measure will be undertaken. These findings will inform the ADNeT Registry, the national dementia registry for Australia, which ultimately aims to enable benchmarking to provide information to governments and jurisdictions to plan services and supports for this extremely significant health system issue.

A strength of this study is the diversity in the expert panel members in relation to representation across Australia and from multidisciplinary clinicians, consumers, policymakers, government representatives, and researchers. This diversity is reflected in the breadth of CQIs identified across the dementia care journey. The inclusion of two consumers (a person with MCI and a caregiver) ensured the consumer voice was captured, and taken into consideration in this process. Each round of the modified Delphi achieved a high response rate.

Several excluded CQIs had high levels of disagreement in the importance and feasibility ratings, which may have reflected the diversity of the backgrounds of the panel members, and a potential limitation of the Delphi process. Additionally, due to the limited evidence base for dementia diagnosis and care, expert opinion is a key influence on the inclusion of an indicator. The CQIs have being presented to key registry stakeholders (clinical sites, clinicians, hospital staff) as part of the implementation processes. The ADNeT Registry Steering Committee (established in early 2019) has recently ratified a set of baseline clinical indicators based on those from this Delphi study. The remaining proposed CQIs will similarly go through further external consultation before being endorsed by the registry steering committee.

While we included two consumers in the expert panel (a person with MCI and a caregiver of a person with dementia), these participants were outnumbered by the clinical, government, and research representatives. To address this, the facilitator of the face-to-face session did ensure that the recommendations and suggestions from the consumers were given weight by the rest of the participants. The expert panel also included representatives from the national consumer organization which also facilitated the inclusion of the consumer perspective. Additionally, the online survey phases of the Delphi enabled the consumers to provide input without the influence of other panel members. We attempted to recruit a person with dementia to the panel however we were unsuccessful as the time commitment and tasks required were considered too complex by the people who had registered interest.

While the consumer voice is important, the end user of the CQI data will be clinicians, researchers and government and therefore the CQIs did need to have high acceptability amongst these stakeholders. Patient reported outcomes (PROMs) were not identified or included in the final CQIs. A separate project examining the inclusion of PROMs and/or PREMs (patient reported experience outcomes) in the CQR is being undertaken with the initial review of possible PROMs [30].

Conclusion

Dementia represents an increasing burden of disease in Australia. With largely no effective treatment and no cure for Alzheimer’s disease and the majority of dementias, patient management is based on supportive care, symptom management, and access to appropriate services across the spectrum of care settings. This study describes the development of a proposed national set CQIs through a consensus process with clinical, registry and consumer representatives for reporting quality of care for people with dementia and MCI via a national registry. The proposed CQIs from this study are currently in the process of being more broadly reviewed among clinicians, persons with dementia and their carers, and other significant stakeholders, and considered as required for the Australian national dementia registry.

The dementia CQR aims to understand the incidence of dementia in Australia, the natural history of dementia, and current quality of care at key points including at the time of diagnosis, early management, access to support services, potentially preventable complications and end of life outcomes. The dementia CQR aims to provide a broad data-based understanding of dementia currently in Australia, and to additionally follow the cohort of registry participants to determine future potential effective interventions that improve quality of life for patients and carers, and outcomes.

A national registry for people with dementia or MCI in Australia that focuses on quality of care will provide meaningful information to people with dementia, their carers, healthcare workers, and health funders regarding current practices and standards of care, and potential gaps in quality care that exist in the system currently. Together with information collected from people with dementia and carers and through data linkage, the registry has the potential to create a valuable resource in our understanding of disease progression and the factors of potential influence. Opportunities to improve care through specific interventions will be able to be monitored through ongoing registry outputs, thus empowering people with dementia, providers and funders with the information that they need to improve outcomes for people with dementia or MCI.

Footnotes

ACKNOWLEDGMENTS

This study was funded by the National Health and Medical Research Council (APP1140485).

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Livingston

, Sommerlad

, Orgeta

, Costafreda

, Huntley

, Ames

, Ballard

, Banerjee

, Burns

, Cohen-Mansfield

, Cooper

, Fox

, Gitlin

, Howard

, Kales

, Larson

, Ritchie

, Rockwood

, Sampson

, Samus

, Schneider

, Selbaek

, Teri

, Mukadam

(2017) Dementia prevention, intervention, and care, Lancet 390, 2673–2734.

Prince

, Bryce

, Albanese

, Wimo

, Ribeiro

, Ferri

(2013) The global prevalence of dementia: A systematic review and metaanalysis, Alzheimers Dement 9, 63–75 e62.

Australian Bureau of Statistics (2016) Causes of Death, Australia. ABS, Canberra.

Guideline Adaptation Committee (2016) Clinical Practice Guidelines and Principles of Care for People with Dementia. Guideline Adaptation Committee, Sydney. https://www.clinicalguidelines.gov.au/portal/2503/clinical-practice-guidelines-and-principles-care-people-dementia.

Speechly

, Bridges-Webb

, Passmore

(2008) The pathway to dementia diagnosis, Med J Aust 189, 487–489.

Australian Institute of Health and Welfare (2017) Rural & Remote Health. AIHW, Canberra. https://www.aihw.gov.au/reports/rural-health/rural-remote-health/contents/access-to-health-services.

Australian Commission on Safety and Quality in Health Care (2012) Logical Design for Australian Clinical Quality Registries. ACSQHC, Sydney.

Evans

, Scott

, Johnson

, Cameron

, McNeil

(2011) Development of clinical-quality registries in Australia: The way forward, Med J Aust 194, 360–363.

Campbell

, Braspenning

, Hutchinson

, Marshall

(2002) Research methods used in developing and applying quality indicators in primary care, Qual Saf Health Care 11, 358–364.

10.

Wilcox

, McNeil

(2016) Clinical quality registries have the potential to drive improvements in the appropriateness of care, MJA 205, S21–S26.

11.

Travaglia

, Debono

(2009) Centre for Clinical Governance Research, Faculty of Medicine, The University of New South Wales, New South Wales.

12.

Mainz

(2003) Defining and classifying clinical indicators for quality improvement, Int J Qual Health Care 15, 523–530.

13.

Religa

, Fereshtehnejad

, Cermakova

, Edlund

, Garcia-Ptacek

, Granqvist

, Hallback

, Kawe

, Farahmand

, Kilander

, Mattsson

, Nagga

, Nordstrom

, Wijk

, Wimo

, Winblad

, Eriksdotter

(2015) SveDem, the Swedish Dementia Registry - a tool for improving the quality of diagnostics, treatment and care of dementia patients in clinical practice, PLoS One 10, e0116538.

14.

Johannsen

, Jorgensen

, Korner

, Elmo

, Lauesen

, Utzon

(2011) Development of a dementia assessment quality database, Aging Ment Health 15, 40–46.

15.

Krysinska

, Sachdem

, Breitner

, Kivipelto

, Kukull

, Brodaty

(2017) Dementia registries around the globe and their applications: A systematic review, Alzheimers Dement 13, 1414–1421.

16.

SveDem (2019) SveDem. https://www.ucr.uu.se/svedem/in-english,

17.

Eldh

, Fredriksson

, Halford

, Wallin

, Dahlstrom

, Vengberg

, Winblad

(2014) Facilitators and barriers to applying a national quality registry for quality improvement in stroke care, BMC Health Serv Res 14, 354.

18.

Gnanamanickam

(2018) Direct health and residential care costs of people living with dementia in Australian residential aged care, Int J Geriatr Psychiatry 33, 859–866.

19.

Access Economics (2003) The dementia epidemic: Economic impact and positive solutions for Australia. Prepared for Alzheimer’s Australia, Canberra. https://www.dementia.org.au/sites/default/files/20030300_Nat_AE_FullDemEpEconImpPosSolutions.pdf.

20.

Zecchin

, Candelaria

, Ferry

, Ladak

, McIvor

, Wilcox

, Bennett

, Bowen

, Carr

, Randall

, Gallagher

(2019) Development of quality indicators for cardiac rehabilitation in Australia: A modified Delphi method and pilot test, Heart Lung Circ 28, 1622–1630.

21.

Boulkedid

, Abdoul

, Loustau

, Sibony

, Alberti

(2011) Using and reporting the Delphi method for selecting healthcare quality indicators: A systematic review, PLoS One 6, e20476.

22.

Petersen

, Lopez

, Armstrong

, Getchius

TSD

, Ganguli

, Gloss

, Gronseth

, Marson

, Pringsheim

, Day

, Sager

, Stevens

, Rae-Grant

(2018) Practice guideline update summary: Mild cognitive impairment: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology, Neurology 90, 126–135.

23.

Eubank

, Mohtadi

, Lafave

, Wiley

, Bois

, Boorman

, Sheps

(2016) Using the modified Delphi method to establish clinical consensus for the diagnosis and treatment of patients with rotator cuff pathology, BMC Med Res Methodol 16, 56.

24.

Agency for Healthcare Research and Quality, Clinical Guidelines and Recommendations, AHRQ, https://www.ahrq.gov/professionals/clinicians-providers/guidelines-recommendations/index.html.

25.

Fitch

, Bernsteing

, Aguilar

, Burnand

, LaCalle

(2001) The RAND/UCL appropriateness method user’s manual.

26.

Woodward

, Woodward

(2009) A national survey of memory clinics in Australia, Int Psychogeriatr 21, 696–702.

27.

Hansen

, Hughes

, Routley

, Robinson

(2008) General practitioners’ experiences and understandings of diagnosing dementia: Factors impacting on early diagnosis, Soc Sci Med 67, 1776–1783.

28.

Robinson

, Emden

, Elder

, Lea

, Vickers

, Turner

(2008) Multiple views reveal the complexity of dementia diagnosis, Australas J Ageing 27, 183–188.

29.

Garre-Olmo

, Flaque

, Gich

, Pulido

, Turbau

, Vallmajo

, Vinas

, Lopez-Pousa

, Registry of Dementia of Girona Study Group (2009) A clinical registry of dementia based on the principle of epidemiological surveillance, BMC Neurol 9, 5.

30.

Ayton

, Gardam

, Pritchard

, Ruseckaite

, Ryan

, Robinson

, Brodaty

, Ward

, Ahern

(2020) Patient-reported outcome measures to inform care of people with dementia - a systematic scoping review. The Gerontologist, gnz179, 10.1093/geront/gnz179.