Effectiveness of a tablet‐based intervention for people living with dementia in primary care—A cluster randomized controlled trial

Primary care physicians (PCP) play a key role in the care of people living with dementia. However, the implementation and practicability of the German S3 Dementia Guideline in primary care remain unclear. The main objective of the present study was to evaluate an intervention for improving guideline‐based dementia care in primary care.


| INTRODUCTION
5][6] For example, a meta-analysis from Germany found that, although healthcare costs for people living with dementia living at home are lower compared to the costs of people living with dementia who are institutionalized, higher total societal costs can be found among people living with dementia at home due to the higher informal care time. 4To address the dementia challenge, the World Health Organization (WHO) strongly recommends strengthening national dementia healthcare delivery. 7][10] A recent study reported between eight and 15 primary care visits per year among people living with dementia. 11In our study, we found that people living with dementia visited their PCP on average almost three times over a 3month observation period. 12[15] These barriers include lack of knowledge and training in dementia care and lack of time, 16,17 as well as lack of cross-sectorial collaboration. 16,18There is an urgent need to support PCP with the healthcare delivery for people living with dementia and improve primary care for dementia.
Evidence-based guidelines provide recommendations for diagnostics, treatment and care and therefore serve as one promising tool to foster optimal healthcare delivery. 19In Germany, the German S3 Dementia Guideline (GDG) 20 provided such evidence-based recommendations for dementia.1][22] However, the implementation of evidence-based guidelines remains deficient. 23Novel approaches are urgently needed to improve the implementation of guidelines in clinical practice.Technology-based approaches represent a promising strategy to foster health care, 24 especially for dementia. 25,26For example, in Germany, the DelpHi study provided empirical evidence for the efficacy of a computer-based intervention-management system in primary care of dementia (Thyrian et al., 2017).Although health technologies in dementia have received a great deal of attention in the last years, randomized controlled trials on technology-based health care provision for people living with dementia in their home remain rare.

| Aim of the present study
The main aim of the present study was to evaluate the effectiveness of a tablet-based intervention for improving adherence to dementia guideline recommendations in a cluster-randomized controlled trial (cRCT) in primary care.The primary hypothesis was that the tabletbased intervention would lead to an increase in guideline-based adherence after 9 months, compared to the control group (care as usual).Further, effects on secondary outcomes for people living with dementia and their informal caregivers were assessed, including quality of life, neuropsychiatric symptoms, activities of daily living, general health status, depression, and informal caregiver burden.obtained from PCP, people living with dementia (or, if a person living with dementia was not authorized to sign anymore, from the holding power of attorney or another legal guardian), and informal caregivers prior to data collection.Most of the caregivers in the present study were relatives (for more details see Table 1) and are therefore referred to as informal caregivers.The primary focus of the present study was on PCP and their patients with dementia.The intervention, which involved the DemTab application, also offered tools and resources for informal caregivers.However, due to the fact that the majority of our informal caregivers were not patients under the care of the participating PCP in our study, we were unable to directly address the care and treatment they received.The DemTab study was conducted according to the principles of Good Clinical Practice and the Declaration of Helsinki.Ethical approval for the DemTab study was obtained by the ethics committee of the Charité-Universitätsmedizin Berlin (EA1/085/19).The CONSORT guideline for cluster trials to design and report the trial was followed.The study was prospectively registered with the ISRCTN registry (trial registration number: ISRCTN15854413) and a study protocol was published. 27

| Setting and participants
Due to the cluster-randomized design, the study sample was determined in two steps.First, PCP were recruited, who then recruited their people living with dementia.A detailed overview of the recruitment strategies and rates can be obtained elsewhere. 12To sum up, PCP were recruited through three recruitment rounds which included the dissemination of advertisements and calls for participation in primary care related networks, cold calls of randomly drawn PCP from a database of the Statutory Health Insurance Physicians in Berlin, and face-to-face recruitment of PCP in their general practices in Berlin.Inclusion criteria for PCP were as follows: (1) operating as PCP, (2) meeting technical requirements (Internet connection), (3)   willing to participate in a training, and (4) signed cooperation agreement.Overall, n = 629 PCP were contacted for recruitment, of which n = 32 PCP agreed to participate and provided informed consent.This represents an overall recruitment rate of 5%.Successfully recruited PCP recruited potentially eligible people living with dementia and informal caregivers from their practice.For this purpose, information material and leaflets were made available for PCP.PCP were asked to appeal to their patients, provide them with information on the DemTab Study and asked for permission for sharing patients and/or informal caregivers contact details with the research team.Eligible patients were then approached by the research team via phone in order to provide a detailed description of the study for each participant.Inclusion criteria for people living with dementia were as follows: (1) diagnosis of dementia (according to ICD-10, F00-F03, G30, G31.0 and G31.82), (2) home-dwelling, (3)   informal caregiver available, and (4) signed informed consent (by person living with dementia or a person holding power of attorney).
Solely individuals that were already diagnosed with dementia by a healthcare professional prior to the study's beginning were included in the present study.Baseline data has shown that the majority of participants of the DemTab study were diagnosed by a specialist (such as neurologist or psychiatrist, 55.2%), 33.3% of participants were diagnosed by a PCP and 11.5% of participants were diagnosed in a different health facility. 12Exclusion criteria for people living with dementia were (1) other mental and behavioral disorders (according to ICD-10: F10-29, except for F10.1, F17.1 or F17.2, F32.2 and people living with dementia in shared homes, an ambulatory home facility for people living with dementia, caregivers from shared homes were also included in the study.Exclusion criteria included a planned absence longer than 8 weeks during the study period.Additionally, health service research in Germany frequently presents data in the context of quarterly billing.Previous studies have documented the quarterly number of PCP visits by people with dementia. 28Dreier and Hoffmann 29 reported an average of 3.9 PCP visits per quarter for people with dementia.To ensure that participants made multiple PCP visits during the intervention period, we opted for a duration of three quarters, equivalent to 9 months.Further, a recent scoping review aiming at identifying and classifying digital tools used in dementia care found that in 86% of studies, the intervention duration was less than 12 months. 30e DemTab App comprises multiple components and functions.

| Intervention
A detailed description of the DemTab App can be obtained elsewhere. 27For PCP, the main function included a checklist to support PCP in guideline-based care provision of dementia.This checklist is developed as a conversation-guide and is based on main recommendations of the German Dementia Guideline. 20The checklist was were blinded until after the collection of baseline data.However, due to the tablet-based intervention, blinding was no longer possible for follow-up.A detailed description of all variables and measures can be found elsewhere 27

| Primary outcomes
Adherence to the German S3 Dementia Guideline (AGDG) was assessed as self-report from both PCP and informal caregivers, using a 23-item and 19-item checklist, respectively.The checklists were developed based on the GDG 20 and previous empirical research on guideline-based dementia care. 22,33,34 Aherence to the GDG was self-reported by each PCP and each informal F I G U R E 2 Screenshots of the DemTab app.From left to right: home screen for PCP, listing all their respective apps, start screen of the DemTab app, and the home screen for people living with dementia and their caregivers.
LECH ET AL.
caregiver for each participating person living with dementia (at patient level) by assessing adherence to specific recommendations of the guideline ("Do you follow the following recommendation?",yes/ no).A simple English translation of the checklists can be obtained from the Appendix X.The internal consistency of the scales of the present study were Cronbachs's α = 0.848 for the PCP checklist and 0.873 for the informal caregivers checklist.The development of the checklists did not include a separate validation study.However, some measures were undertaken and published to ensure the validation of the assessment. 12For example, originally, the checklists were composed in a dichotomous format (yes/no), but included the category "not applicable".Baseline data analysis of the DemTab Study revealed inconsistencies in the selection of the category "not applicable" resulting in uncertainty how and when this category was chosen by participants.Consequently, the impact of the category "not applicable" on the calculation of the final AGDG score was measured with a set of analyses to compare different scoring methods (for more details see 35 ).Finally, as the different scoring methods did not reveal any significant differences, it was decided to recode "not applicable" into missing data.The final AGDG scores from both PCP and informal caregivers were calculated as the sum over all items.

| Secondary outcomes
A variety of health-related secondary outcomes for people living with dementia were assessed.Quality of Life was assessed using the Quality of Life in Alzheimer's Disease questionnaire (QOL-AD) 36 and additionally by the EuroQuol 5 Dimensions questionnaire (EQ-5D-5 L). 37 For the assessment of depressive symptoms we choose the depression in old age scale (DIA-S), a screening scale for depression comparable to the GDS-15 but -as stated by the authors -more appropriate for the use in geriatric patients. 38uropsychiatric Symptoms were assessed with 12 items by the Neuropsychiatric Inventory (NPI).39 Another proxy scale used was the Barthel Index 40 to evaluate the functional state of the people living with dementia.On informal caregivers' level assessments included depressive symptoms measured by the Geriatric Depression Scale (GDS), 41 health-related quality of life measured with the Short Form Health Survey (SF-36) which consists of a physical (PCS) and a mental component score (MCS) 42 and caregiver burden assessed by the Burden Scale for Family Caregivers (HPS).43 Other variables/Covariates: Sociodemographic characteristics such as age (years), gender (female/male/other), education (years of education) were measured from both, people living with dementia and informal caregivers.Further, people living with dementia level of care was assessed based on the compulsory long-term care insurance in Germany (ranging from 1 = low level of care to 5 = high level of care). 44Cognitive status was assessed with the MMSE, total score ranges from 0 to 30, higher scores indicating higher cognitive status.45

| Sample size calculation
The sample-size calculation for the DemTab study was based on a previous study conducted by Vickrey et al. 22 In a cRCT, the study investigated the effectiveness of a dementia guideline-based disease and a statistical power of 80% was estimated.Accounting for the cluster structure of GPs (ICC = 0.03) and a follow-up drop-out rate of 18% (as found by 22 ), the final sample size for the present study was calculated at n = 102 in each the intervention and the control group.

| Randomization
The cluster-randomization of PCP was conducted as a covariateconstrained randomization with the package cvcrand (RefNum) used with the statistics software R (RefNum).For the clusterrandomization, an adjustment for the following variables was applied: urban versus rural location of practice, total number of patients and number/percentage of dementia patients per PCP, age and sex of the PCP.Adjustment for covariates was conducted in a reproducible way with a defined seed, with balance metric l2 and a weighting procedure for the covariates (urban vs. rural location of the practice being the main covariate).

| Baseline characteristics
The majority (61.0%) of participating PCP were female and on average

| Secondary outcomes
For secondary outcomes at the level of people living with dementia and informal caregivers, no significant differences between the two groups were found at follow up when adjusted for the respective baseline values (t0) and in addition for people living with dementia for the MMSE at baseline, and for informal care givers for age at baseline (Table 3).

| POST HOC ANALYSIS/SENSITIVITY ANALYSES
To address two major limitations of the present study, sensitivity sample size of N = 204 remarkably, it was possible that an effect was not detected due to the small sample.As a second sensitivity analysis, a bootstrap approach was applied with the intention to simulate the present effect in the initially intended, for the expected treatment effect calculated necessary sample size of n = 102 for both control and intervention; bootstrap models were run with 500 repetitions for each model and results were pooled.Again, no significant differences between the control and the intervention group could be found for all primary and all secondary outcomes both on the people living with dementia's and the informal caregiver's level, neither for the intention to treat nor the per protocol analysis.

| DISCUSSION
The aim of the present study was to investigate the effect of a tabletbased intervention on adherence to evidence-based dementia guidelines in primary care.We hypothesized that a multi-component technology-based intervention would increase adherence to the guideline in the intervention group compared to a control group practicing standard care.Overall, reported AGDG was high in both groups, especially when reported by PCP themselves.However, we did not find a positive effect of the intervention on the primary outcome AGDG in the intervention group.Regarding the secondary outcomes for people living with dementia (depression, quality of life, ADL, neuropsychiatric symptoms) and secondary outcomes for informal caregivers (depression and caregiver burden) we also found no differences between both groups.
LECH ET AL.
-7 of 14 Clinical evidence-based guidelines are the corner stone of evidence-based health care and can bridge the gap between scientific evidence and clinical practice. 46,47Adherence to trustworthy and rigorously developed clinical guidelines is therefore viewed as one important indicator of quality of care. 48Although developed to aid and inform clinical decision-making, adherence to clinical practice guidelines remains a challenge for many clinicians and effective interventions to increase guideline adherence are lacking. 49A

Cochrane review of tailored interventions for improving clinical
practice including guideline adherence found only few studies with mixed results and small to moderate effect sizes. 50A scoping literature review by Fischer et al. 23  improve guideline adherence in primary dementia care by providing PCP with a comprehensive overview of the current GDG recommendations and by supporting their treatment process through interactive reminders and prompts.However, our results do not support this notion.Disregarding a possible ceiling effect (high selfreported AGDG in both groups at baseline), this may in part be explained by the type of intervention and by problems related to the implementation.
In contrast to our findings, Vickrey et al. 22 found that adherence to a comprehensive set of dementia guidelines almost doubled in an intervention group who collaborated with a trained dementia care manager using an Internet-based care management software system for care planning and coordination.Furthermore, Vickrey et al. 22 also reported a reduction in decline of health-related quality of life in the intervention group compared to the control group receiving standard care.While our study mainly focused on supporting PCP in their treatment process and clinical decision-making, the key intervention component in the study by Vickrey et al. 22 was the care manager who conducted a structured home assessment and thoroughly discussed care needs and priorities with the informal caregivers.Our results suggest that an intervention focused primarily on PCP without including additional staff may not be sufficient to improve PCP-based dementia care.However, in our study we chose a pragmatic approach with an intervention that could be implemented within the existing conditions of real-world outpatient dementia care in Germany.This means that no additional resources other than the tablet application were needed to implement the study.Future studies aiming at improving dementia care in the primary care should also include intervention that aim at improving educational resources and training for PCP.This approach would be in line with previous research, showing that PCP themselves have numerous times acknowledged the urgent need for education on diagnosis of dementia and dementia care. 13,14Recent research and initiatives have shifted their focus toward addressing and narrowing this Informal caregivers (i.e., unpaid family members, friends or neighbors) play an essential role in dementia care, as they often organize and plan most, if not all activities of people living with dementia, including medical appointments and examinations, self-care and daily living. 51While these responsibilities can lead to caregiver burden with a significant impact on mental health and well-being of informal caregivers, 51 numerous studies have demonstrated that multicomponent interventions targeting informal caregivers can improve their health. 52The DemTab application offered a range of features designed to enhance communication and social participation.Although we employed numerous features specifically targeted Although some participating PCP used the DemTab application on a regular basis over the study period, overall, the general usage was low across PCP of the intervention group.Numerous reasons need to be considered for this finding.First, the application ran on a stand-alone tablet computer that could not be connected to the electronic clinical software used in the practice for data protection and compatibility reasons.This lack of interoperability was a major usage barrier and is a known challenge for the implementation of digital support systems in healthcare. 53Second, the COVID-19 pandemic occurred during the intervention period, which led to massive problems for the recruitment and implementation of the study.The pandemic as well as the measures to reduce the risk of infection caused severe limitations and burdens, both for the highly vulnerable study population (elderly and chronically ill people and their informal caregivers) and for the participating PCP.This may not only have impacted AGDG, but the pandemic itself may have also affected the secondary outcomes such as depression or quality of life.Finally, the GDG were not issued by the German College of General Practitioners and Family Physicians and thus may not be entirely suited for use in primary care settings.Accordingly, some of the PCP in our study reported that the current dementia guideline was little useful for their treatment of people living with dementia. 12Therefore, the perspective of PCP should be incorporated in future German dementia guidelines to meet their specific needs.
T A B L E 3 Results of the linear mixed models for the secondary outcomes for people living with dementia and their caregivers.

| Strengths and limitations
The DemTab study developed and tested a pioneer technology-based intervention, and was able to demonstrate the general feasibility of this approach.The implementation of novel apps for the purpose of supporting guideline-based treatment is still only in its early stages.
The DemTab study was conducted as pragmatic trial and implemented under real-world conditions in general practices and at people living with dementia's homes.Important barriers for implementation were identified, which can provide a basis for further practical research in this area.One major strength of the study was the rigorous accordance with quality guidelines for cRCT.A study protocol was published in a peer-reviewed journal to assure that our research is conducted in accordance with predefined hypotheses and procedures.However, while study protocols endorse the quality of a trial, unexpected and sometimes unavoidable events occur during the implementation of a trial, hindering the strict observance to a protocol.The COVID-19 pandemic affected the implementation of the DemTab Study and led to a number of protocol violations, in particular the recruitment of PCP and people living with dementia.
First, for participants living alone (n = 17) regular visits from informal caregivers were restricted, leading to violations of the protocol.Less frequent visits of informal caregivers may have affected the usage of the tablet-based intervention, especially for people living with dementia with major cognitive impairment.However, as most of included people living with dementia lived together with their informal caregivers, we believe this violation of the protocol is of small relevance.Second, in light of recruitment problems, we extended the inclusion criteria and included people living with dementia living in shared dementia homes and their caregivers in our study.As the main aim of the DemTab Study was to improve ambulatory primary care of dementia and in Germany, shared dementia homes fall under the ambulatory care sector, we believe the impact of this violation is marginal.Third, another impact of the pandemic was with regard to the assessments of cognitive functioning.As reported in the study protocol, the assessment of cognitive functioning was to be conducted with the MMSE.However, as previously described, the in-person assessment of data was changed to phone assessment due to the pandemic.The execution of the MMSE over phone was not feasible.However, this only affected 12 people living with dementia and the MMSE was not a primary outcome of the DemTab Study, thus we believe this violation did not have a major impact on the study results.In spite of various attempts to improve the recruitment of participants, including the adaptation of inclusion as exclusion criteria, the originally estimated sample size of N = 204 could not be reached.This represents a major limitation of the present study.9][60] Finally, Finally, the primary outcome was measured with a self-developed assessment and no large-scale validation of the primary outcome measurement was conducted.For example, as already discussed, the post hoc recoding of the category "not applicable" as missing data must be addressed.While no significant differences were found in comparisons of means and correlations across scoring methods, data labeled as "not applicable" can still offer valuable insights into responses.As such, it may serve as a meaningful contribution to the assessment and interpretation of AGDG.For example, adherence to a specific dementia guideline recommendation depends on the type and stage of the dementia and therefore not ever recommendation is applicable for all patients.Due to patient related (health) reasons such as the severity or current stage of dementia or the overall health status of a patient, PCP may still be adherent to the dementia guideline, although they are not following all the guideline recommendations.
The current AGDG score does not incorporate this aspect, and it is crucial to acknowledge this limitation when interpreting the AGDG score.In future research that utilizes the present checklist or any other tool to assess guideline adherence, it is advisable to define and incorporate the stage and severity of dementia in guideline recommendations.Further, the category "not applicable" should be included when measuring adherence to guidelines as it may enhance the score's validity and explanatory capacity.Further, the use of a self-report checklist for guideline adherence may have biased the results regarding the primary outcome.Objective assessment of treatment on the basis of patient records is very complex and was not possible within the scope of this research project.Future research should implement objective measures of guideline adherence in order to rule out self-report bias.

| CONCLUSION
Considering rising numbers of people living with dementia in modern societies, PCP play an essential role in planning adequate treatment and care for both patients and their informal caregivers.While S3 guidelines represent the highest quality guidelines and should be the gold standard in dementia care, barriers to guideline adherence compromise their translation into real-world clinical practice.Within our study, the use of a multi-component technology-based intervention did not improve self-reported guideline adherence in PCP.
However, important implementation barriers such as lack of interoperability and low applicability of existing German S3 guidelines in the primary care setting were identified.Interventions specifically tailored to the needs of PCP, people living with dementia and informal caregivers are needed to improve adherence to dementia guidelines in primary care.To reduce the burden on PCP, future interventions should also explore the role of multi-professional approaches and emphasize the inclusion of informal caregivers.In addition, greater involvement of PCP in modifying the new German dementia guideline, which is currently being developed, could increase their acceptance.
LECH ET AL.

AFigure 1 .of 14 -
Figure 1.Randomization was conducted on PCP Level (cluster) to avoid contamination effects across groups.All study participants (PCP, people living with dementia, and informal caregivers) received comprehensive information material and detailed verbal information on the DemTab study and the trial.Written informed consent was

F32. 3 )
, (2) a planned hospital or rehabilitation stay longer than four weeks, and (3) a planned relocation to an inpatient care-facility or nursing home within the study period.Inclusion criteria for informal caregivers included (1) living with or regularly visiting the person living with dementia, and (2) signed informed consent.Originally, informal caregivers such as partners, children, other family members as well as friends were addressed.However, as many PCP treated F I G U R E 1 Overview of the DemTab study's design.
After randomization, both PCP and the person living with dementia of the intervention group received a tablet, a training on the usage of the tablet and a handbook on the DemTab application (DemTab App) prior to the intervention's beginning.During the training people living with dementia and informal caregivers were invited to use the DemTab App solely or engage with informal caregiver and other family members or friends.Participants were instructed to use the DemTab App as often as possible for 9 months.A 9-month timeframe was primarily selected because of the structure of the German healthcare system, which operates on a quarterly billing system.

32 2. 4 |
linked to another function, the development of a care plan where PCP and people living with dementia could establish specific goals and care plans.Another function assisted PCP, if necessary, in the prescription of suitable antidementia drugs.Based on preset filters (e.g., dementia type and severity), tailored individual and guidelinebased prescription recommendations for antidementia drugs and dosages are delivered via tablet, and automatic reminders for correct medication intake and adjustment are provided to both PCP and people living with dementia.Other functions included direct (via a messenger) and indirect (via people living with dementia's health information such as blood pressure or mood reported on the tablet) communication and information exchange between PCP, people living with dementia, and informal caregivers.Finally, a full electronic version of the German Dementia Guideline20 as well as further information about outpatient dementia care was provided on the tablet (see Figure2).For people living with dementia and informal caregivers the main functions of the DemTab App include serious games and programs developed to engage, stimulate, and motivate the people living with dementia, 31 including a quiz for cognitive stimulation, games emphasizing activities of daily living or a picture gallery for biography work.In addition, the DemTab App provides a variety of information on dementia and dementia care, including a location service provided by the Federal Ministry for Family Affairs, Senior Citizens, Women and Youth (2019) pointing out social and healthcare services in the vicinity.Finally, a guided audio-relaxation technique for both people living with dementia and their informal caregivers can be found in the DemTab App.Measurement Data was collected at baseline and after 9 months (follow up) by a trained study nurse.Baseline data were collected from July 2019 to July 2020.Most baseline data was collected in the participant's home using a self-report questionnaire.We aimed at obtaining answers directly from the individual living with dementia.However, if a people living with dementia was no longer able to provide answers or the validity of answers was questionable, answers/information were verified or obtained via the informal caregiver.In addition, due to the COVID-19 pandemic curfews and contact regulations, data collection was changed from in-person assessment to phone assessment in March 2020.Except for the Mini Mental State Examination (MMSE), data collection via phone proved to be feasible.This affected n = 12 people living with dementia for baseline and all participants for follow-up data collection.The study nurse as well as all participants management program.Adherence to dementia guidelines was assessed at baseline and follow-up based on medical records and a caregiver's survey.The power calculation for the DemTab study was conducted with G*Power 3.1.In a first step, an estimated minimum sample size of n = 71 for each group at a type I error rate of alpha = 0.05

6 of 14 - 3 | FINDINGS 3 . 1 |
Descriptive analyses (mean, standard deviation (SD) and range for continuous variables, frequencies for nominal and ordinal variables) of sociodemographic and clinical characteristics for PCP, people living with dementia and informal caregivers were calculated.Primary and secondary outcomes of the DemTab study at 9 months were analyzed applying intention-to-treat analysis and Linear Mixed Models (LMM).LMM allow to account for the nested structure of the data (PCP, level 2).The ID of PCP was used as a clustering variable.For the primary outcomes, LMM were adjusted for the baseline values of the respective score and the MMSE.For the secondary outcomes at people living with dementia level, LMM were adjusted for the respective score and the MMSE at baseline; at informal caregiver's level, LMM were adjusted for the respective score and age at baseline.An intraclass correlation coefficient (ICC) representing the ratio of the between-GP variance to the total variance was calculated.Missing values in the present study were replaced with multiple imputation in (MI) 10 replicates under the assumption that the mechanism of the missing values was Missing at Random (MAR).For this purpose, the standard method in SPSS was used, which automatically decides on the optimal imputation method based on the data, either for monotonic missingness patterns or an iterative Markov chain Monte Carlo (MCMC) LECH ET AL. method for arbitrary missingness patterns.For LMM, pooled results of the 10 MI data set are shown.Statistical analyses were performed using IBM SPSS Statistics version 27.The significance levels for tests were based on α = 0.05 without adjustment for multiple testing for the secondary outcomes.Participant flow A total of N = 28 PCP, N = 102 people living with dementia and N = 99 informal caregivers were recruited for the DemTab study; n = 3 people living with dementia were recruited without an informal caregiver.Further, out of the n = 88 caregivers participating, n = 10 were professional caregivers.Details can be obtained from Figure 3.
50 years old (range: 38-67 years).Less than half of PCP (n = 12, 42.9%) worked in a single-handed practice and PCP reported a mean of about 12 years of experience as practitioners (range: 1-29 years).During the last three months, PCP treated on average 1489 patients (range: 700-2990 patients) and 61 people living with dementia (range: 9-200 people living with dementia).People living with dementia main characteristics at baseline can be obtained from Table 2. Overall, more than half of people living with dementia (n = 54) were female.People living with dementia were on average 80 years old (range: 63-94 years) and reported an average of 10.2 years of education (range: 8-17 years).The mean MMSE score was 19.4 (range: 0-30).Informal caregiver's main characteristics at baseline can be obtained from Table 2. Overall, two third of informal caregivers (n = 63) were female, on average 68.5 years old (range: 42-90 years) and reported an average of 10.4 years of education (range: 9-12 years).
analyses were conducted.First, to account for possible dose of intervention effects, per protocol analyses were conducted.During the implementation of the intervention study, large differences across PCP usage of the tablet-based application were observed.The research team had continuously communicated with PCP to facilitate the implementation of the intervention.Based on observations (only few or no usage data were recorded on the tablets) and communications with PCP, we found strong reasons to believe that the intervention was not or only very rarely conducted by two of the PCP in the intervention group.As a sensitivity analysis, we conducted a per protocol analysis, where we allocated those two PCP (and the associated seven people living with dementia) to the control group, resulting in n = 45 patients in the control and n = 46 patients in the intervention group.No impact on guideline adherence of the intervention compared to the control group were found (people living with dementia level: β(95%CI) = 1.73 (−045; 3.90), p = 0.12; informal caregiver's level: β(95%CI) = 1.45 (−0.34; 3.23), p = 0.11).Since the present study failed to reach the originally estimated necessary found that barriers to guideline implementation can be differentiated into personal factors, guidelinerelated factors and external factors, and that structured implementation can improve guideline adherence.Accordingly, we expected the structured, tablet-based conversation-guide would F I G U R E 3 Flow chart of the DemTab Study.
disparity.For example, a national dementia education and research initiative named Primary Care Education, Pathways and Research of Dementia project (PREPARED) focused on the development, delivery and evaluation of dementia educational resources and trainings for PCP and other relevant community healthcare professionals.The DemTab app has made efforts to meet these requirements and has strived to integrate educational elements during the development of the DemTab application.Nevertheless, future research is likely to place greater emphasis on enhancing educational resources and training in primary care.

1 Baseline data of all people living with dementia. Characteristic n All participants (N = 91) IG (n = 56) CG (n = 35)
Abbreviations: ADL = Activities of Daily Living; AGDG, Adherence to German Dementia Guideline; CG, Control group; DIAS, Depression in Old Age; EQ5DL, Health Status; IG, Intervention group; M, Mean; MMSE, Mini Mental State Examination; NPI, Neuropsychiatric Inventory; QOL-AD, Quality of Life in Alzheimer's Disease questionnaire; SD, Standard Deviation.