Scandinavian Journal of Primary Health Care ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/ipri20 Unburdening dementia – a basic social process grounded theory based on a primary care physician survey from 25 countries Ferdinando Petrazzuoli, Shlomo Vinker, Sebastian Palmqvist, Patrik Midlöv, Jan De Lepeleire, Alessandro Pirani, Thomas Frese, Nicola Buono, Jette Ahrensberg, Radost Asenova, Quintí Foguet Boreu, Gülsen Ceyhun Peker, Claire Collins, Miro Hanževački, Kathryn Hoffmann, Claudia Iftode, Tuomas H. Koskela, Donata Kurpas, Jean Yves Le Reste, Bjørn Lichtwarck, Davorina Petek, Diego Schrans, Jean Karl Soler, Sven Streit, Athina Tatsioni, Péter Torzsa, Pemra C. Unalan, Harm van Marwijk & Hans Thulesius To cite this article: Ferdinando Petrazzuoli, Shlomo Vinker, Sebastian Palmqvist, Patrik Midlöv, Jan De Lepeleire, Alessandro Pirani, Thomas Frese, Nicola Buono, Jette Ahrensberg, Radost Asenova, Quintí Foguet Boreu, Gülsen Ceyhun Peker, Claire Collins, Miro Hanževački, Kathryn Hoffmann, Claudia Iftode, Tuomas H. Koskela, Donata Kurpas, Jean Yves Le Reste, Bjørn Lichtwarck, Davorina Petek, Diego Schrans, Jean Karl Soler, Sven Streit, Athina Tatsioni, Péter Torzsa, Pemra C. Unalan, Harm van Marwijk & Hans Thulesius (2020) Unburdening dementia – a basic social process grounded theory based on a primary care physician survey from 25 countries, Scandinavian Journal of Primary Health Care, 38:3, 253-264, DOI: 10.1080/02813432.2020.1794166 To link to this article: https://doi.org/10.1080/02813432.2020.1794166 © 2020 The Author(s). Published by Informa Published online: 28 Jul 2020. UK Limited, trading as Taylor & Francis Group. Submit your article to this journal Article views: 2054 View related articles View Crossmark data Citing articles: 8 View citing articles Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=ipri20 SCANDINAVIAN JOURNAL OF PRIMARY HEALTH CARE 2020, VOL. 38, NO. 3, 253–264 https://doi.org/10.1080/02813432.2020.1794166 RESEARCH ARTICLE Unburdening dementia – a basic social process grounded theory based on a primary care physician survey from 25 countries Ferdinando Petrazzuolia, Shlomo Vinkerb, Sebastian Palmqvistc, Patrik Midlo €va, Jan De Lepeleired, Alessandro Pirani , Thomas Frese , Nicola Buono , Jette Ahrensberg , Radost Asenovai, e f g h €lsen Ceyhun Pekerk, Claire Collinsl, Miro Hanzevackim, Kathryn Hoffmannn, Quintı Foguet Boreuj, Gu Claudia Iftode , Tuomas H. Koskelap, Donata Kurpasq, Jean Yves Le Rester, Bjørn Lichtwarcks, Davorina o Petekt, Diego Schransu, Jean Karl Solerv, Sven Streitw, Athina Tatsionix, Peter Torzsay, Pemra C. Unalanz, Harm van Marwijkaa and Hans Thulesiusa,ab a Center for Primary Health Care Research, Department of Clinical Sciences in Malmo€, Lund University, Malm€o, Sweden; bDepartment of Family Medicine, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel; cClinical Memory Research Unit, Department of Clinical Sciences in Malm€o, Lund University, Sweden; dDepartment of Public Health and Primary Care, General Practice, University of Leuven, Leuven, Belgium; eFamily and Nursing Home Practice – Memory Clinic, Alzheimer’s Association “Francesco Mazzucca” Onlus, Ferrara, Italy; fInstitute of General Practice and Family Medicine, Medical Faculty, Martin-Luther-University Halle-Wittenberg, Halle/ Saale, Germany; gSNAMID (National Society of Medical Education in General Practice), Caserta, Italy; hResearch Center for Emergency Medicine, Aarhus University, Aarhus, Denmark; iDepartment of Urology and General Medicine, Medical University of Plovdiv, Plovdiv, Bulgaria; jInstitut Universitari d’Investigacio en Atencio Primaria- IDIAP Jordi Gol, Universitat Autonoma de Barcelona, Catalonia, Spain; kDepartment of Family Medicine, Ankara University School of Medicine, Ankara, Turkey; lIrish College of General Practitioners, Dublin, Ireland; mDirector Health Care Center of Zagreb, Zagreb, Croatia; nDepartment of General Practice and Family Medicine, Center for Public Health, Medical University of Vienna, Vienna, Austria; oTimis Society of Family Medicine, Timisoara, Romania; p Department of General Practice, University of Tampere, Finland; qFamily Medicine Department, Wroclaw Medical University, Wroclaw, Poland; rEA 7479 SPURBO. Department of General Practice, Universite de Bretagne Occidentale, Brest, France; sThe Research Centre for Age-related Functional Decline and Disease, Innlandet Hospital Trust, Ottestad, Norway; tDepartment of Family medicine, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia; uDepartment of Family Medicine and Primary, Health Care Ghent University, Ghent, Belgium; vMediterranean Institute of Primary Care, Attard, Malta; wInstitute of Primary Health Care (BIHAM), University of Bern, Bern, Switzerland; xDepartment of Internal Medicine, General Medicine, Faculty of Medicine, University of Ioannina School of Health Sciences, Ioannina, Greece; yDepartment of Family Medicine, Faculty of Medicine, Semmelweis University, Budapest, Hungary; zDepartment of Family Medicine, Marmara University Medical Faculty, Istanbul, Turkey; aaBrighton and Sussex Medical School, University of Brighton, UK; abDepartment of Medicine and Optometry, Linnaeus University, Kalmar, Sweden ABSTRACT ARTICLE HISTORY Objective: To explore dementia management from a primary care physician perspective. Received 27 October 2019 Design: One-page seven-item multiple choice questionnaire; free text space for every item; final Accepted 7 June 2020 narrative question of a dementia case story. Inductive explorative grounded theory analysis. KEYWORDS Derived results in cluster analyses. Appropriateness of dementia drugs assessed by tertiary care Dementia; drug specialist. prescription; primary care; Setting: Twenty-five European General Practice Research Network member countries. unburdening; elderly Subjects: Four hundred and forty-five key informant primary care physician respondents of people; grounded theory which 106 presented 155 case stories. Main outcome measures: Processes and typologies of dementia management. Proportion of case stories with drug treatment and treatment according to guidelines. Results: Unburdening dementia – a basic social process – explained physicians’ dementia man- agement according to a grounded theory analysis using both qualitative and quantitative data. Unburdening starts with Recognizing the dementia burden by Burden Identification and Burden Assessment followed by Burden Relief. Drugs to relieve the dementia burden were reported for 130 of 155 patients; acetylcholinesterase inhibitors or memantine treatment in 89 of 155 patients – 60% appropriate according to guidelines and 40% outside of guidelines. More Central and Northern primary care physicians were allowed to prescribe, and more were engaged in dementia management than Eastern and Mediterranean physicians according to cluster analyses. Physicians typically identified and assessed the dementia burden and then tried to relieve it, commonly by drug prescriptions, but also by community health and home help services, men- tioned in more than half of the case stories. CONTACT Ferdinando Petrazzuoli ferdinando.petrazzuoli@med.lu.se Center for Primary Health Care Research, Clinical Research Center, Lund University, Box 50332, Malm€ o 202 13, Sweden This article has been republished with minor changes. These changes do not impact the academic content of the article. ß 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 254 F. PETRAZZUOLI ET AL. Conclusions: Primary care physician dementia management was explained by an Unburdening process with the goal to relieve the dementia burden, mainly by drugs often prescribed outside of guideline indications. Implications: Unique data about dementia management by European primary care physicians to inform appropriate stakeholders. KEY POINTS Dementia as a syndrome of cognitive and functional decline and behavioural and psychological symptoms causes a tremendous burden on patients, their families, and society.  We found that a basic social process of Unburdening dementia explained dementia manage- ment according to case stories and survey comments from primary care physicians in 25 countries.  First, Burden Recognition by Identification and Assessment and then Burden Relief – often by drugs.  Prescribing physicians repeatedly broadened guideline indications for dementia drugs. The more physicians were allowed to prescribe dementia drugs, the more they were responsible for the dementia work-up. Our study provides unique data about dementia management in European primary care for the benefit of national and international stakeholders. Introduction explanatory theory of dementia management in pri- mary care. Dementia arises together with the accumulation of harms and disease burdens over the life course, and clusters with other endemic conditions exacerbating each other synergistically [1,2]. Multimorbidity and car- Methods diovascular comorbidity show prevalences of 65–98% Study design for those more than 65 years of age [3] with accompa- nying polypharmacy and potentially inappropiate For this survey study directed to primary care physi- medications adding to the burdens [4]. Dementia can cians in 25 countries of the European General Practice be seen as a triad of impairments and burdens: cogni- Research Network (EGPRN), we developed a seven- tive, emotional, and physical [2] in a ‘frail brain’ [5]. item multiple choice questionnaire (MCQ). The design Dementia eventually results in disabled instrumental of the survey has been presented elsewhere [9,10]. At and basic activities of daily life, neuropsychiatric the end of every MCQ, optional free text comments issues, personality changes, social impairment and could be added [10]. In the eighth and last survey stigma [6]. These disease burdens are often long last- item, we invited respondents to contribute anony- ing and indeed affect family caregivers who often mised case stories of dementia from their own develop depression and physical illnesses while caring practice. for their loved ones [7]. A systematic review domi- We applied an analytic classic grounded theory nated by qualitative studies of dementia management approach characterised by inductive reasoning with in primary care showed lack of support for patients, no a priori hypotheses allowing both qualitative and caregivers and primary care physicians and an attitude quantitative data analysis [11–13]. of ‘therapeutic nihilism’ [8]. There was also limited access to and knowledge about community services and resources and absence of interdisciplinary teams Setting to enhance management [8]. Other main themes from the qualitative studies were ‘time constraints, financial In the 25 countries, national key informants were iden- constraints, stigma, diagnostic uncertainty, and disclos- tified and contacted face to face by author FP during ing the diagnosis’ [8]. With this gloomy review as meetings of the EGPRN and WONCA Europe conferen- background, we wanted to explore dementia manage- ces in 2013–2015. National key informants organised ment using an inductive approach from classic translation of the survey into their own languages and grounded theory where the basic research question is back translation to English. They also selected further ‘what is going on?’ We wanted to know how primary key informants from different geographical and socio- care physicians in different countries treat their economic regions within their country by convenience patients suffering from dementia, and to generate an or snowball sampling. SCANDINAVIAN JOURNAL OF PRIMARY HEALTH CARE 255 Participants and data We also ran numerous statistical analyses based on both grounded theory codes and results of MCQ ques- 445 key informant primary care physicians from 25 tions. These analyses were included in memos to countries, of which 106 physicians from 23 countries deductively fit the emerging concepts by applying the presented 155 case stories. ‘constant comparative method’ of grounded theory Data coded for this study: and indeed using its dictum ‘all is data’ [14,16]. i. Secondary analysis of MCQ data provided by The grounded theory analysis was mainly done by 445 key informants and 8,000 words of free text the first and last author in collaboration and analytic data from the same survey [9,10]. consensus was reached during the write-up process ii. Data from 106 key informants who provided 155 lasting several years and ending with the submission case stories consisting of 40,000 words. of this manuscript. Good grounded theories should indeed be modifiable when new data enters such as relevant critique from reviewers [14,16]. Grounded theory analysis Classic grounded theory is the world’s most cited behavioral research method with 124,055 Google Scholar citations (15 July 2020) [11]. In classic Dementia analysis grounded theory, hypotheses are inductively gener- Proxies for treatment appropriateness were aligned ated to explain how participants resolve their main with the guidelines for dementia drug therapy and concern abstract of time, place and people [11,13,14]. established by a tertiary care physician researcher Classic grounded theory differs from other methods expert on neurocognitive disorders (author SP) [17,18]. using only qualitative data by emphasising explana- SP also assessed the specific dementia type based on tory concepts rather than descriptions. the information given in the case stories in accordance Theoretical memos, in the shapes of text, diagrams, with the DSM-5 [19]. and figures, were written, typed, or drawn to create a 200-page memo bank from which this paper was writ- ten. ‘Memos are the theorizing write-up of ideas about substantive codes and their theoretically coded rela- Statistical analysis tionships as they emerge during coding, collecting and analysing data’ [14]. These memos were analysed IBM SPSS Statistics for Windows, Version 22.0. (IBM, based on grounded theory quality principles of fit, Armonk, NY USA, 2013) was used for descriptive and relevance, workability and modifiability [14,16]. Then exploratory statistical analyses. Cluster analysis is an followed selective coding where the core concept exploratory method used to identify structures within guided the analysis during theoretical memoing. We the data such as homogenous groups of cases if coded case stories and survey text comments asking grouping and differences between dependent and for each incident in the data ‘What is the participants’ independent variables is previously unknown [10,20]. main concern and how is it being resolved?’ and We did two-step cluster analyses with variables emerg- ‘What are the participants doing to resolve their main ing from memos. Overall goodness-of-fit of clusters concern’ and ‘What is this a study about?’. Answers to had silhouette coefficients with measures of less than these coding questions yielded several hundred indi- 0.2 classified as poor, between 0.2 and 0.5 fair, and cators of preliminary dementia discovery, diagnosis, more than 0.5 as a good solution quality. Fair or work up and treatment. Theoretical memos included higher was considered acceptable clustering with a codes related to care and caregivers – spouse, chil- cut-off score at 0.3 [20]. More advanced statistical dren, institutional and non-institutional care etc. We methods such as Directed Acyclical Graphs were further analysed survey free text including descriptions applied to the data in another version of this study of formal and informal rules of dementia work up and available online [10]. diagnostic and treatment structures of the jurisdiction In summary, this study involved a mix of data and in which the case story took place [9]. methods emphasising an inductive approach of classic The core concept was theoretically coded using a grounded theory with the goal of exploring and gen- basic social process [11,14–16] to explain the ongoing erating a hypothesis of what was going on in the field resolution of the main concern in the case stories. of primary care physician dementia management. 256 F. PETRAZZUOLI ET AL. Results More than two-thirds of patients in our study that were prescribed acetylcholinesterase inhibitors and/or A multitude of burdens was the main concern sur- memantine had too many exclusion criteria to be rounding the care of patients with dementia and their accepted into dementia drug trials [10]. caregivers. Patients and caregivers were overwhelmed In five out of 25 countries, primary care physicians not only by the burden of dementia but often by mul- were entitled to officially prescribe reimbursable timorbidity – present in 74% of the case stories – and dementia drugs. In many other countries, primary care by behavioural and psychological symptoms of physicians could diagnose dementia unofficially and dementia. prescribe specific dementia drugs to patients that The core concept, Unburdening dementia, was would not be reimbursed by their health insur- deduced as a label explaining the resolution of the ance [9,10]. main concern. The participating countries were split into types Unburdening dementia is seen as a basic social pro- according to the ability of primary care physicians to cess starting with Identifying the burden, mostly done prescribe a dementia drug which could subsequently by family members of the patient, then Assessing the be reimbursed by the health insurance system: burden by cognitive tests, especially the Mini Mental ‘permissive’ ¼ primary care physicians always able to State Examination (MMSE), the most used cognitive prescribe, ‘partially permissive’ ¼ need for first pre- psychometric test internationally [21]. The dementia scription by secondary care specialist, and ‘non-per- burden is then Recognised by formal caregivers and missive’ ¼ mandatory secondary care prescribing and eventually pursued by Burden Relief, often by prescrip- follow-up. Some primary care physicians from ‘non- tion drugs in 130 of 155 (84%) of the cases but also permissive’ countries prescribed dementia drugs to by community health and home help services, men- their patients despite the restrictions for the purpose tioned in more than half of the case stories. The basic of mercy and equity as an unburdening action. social process of Unburdening dementia is presented The MCQ analysis showed that the more primary in Table 1 with sub-categories and different care physicians were officially allowed to prescribe properties. dementia drugs, the more they felt responsible for the Unburdening explains what many physicians – both diagnostic work-up. But they also then increasingly primary care physicians and secondary care specialists seemed to engage multiprofessional community – were doing to help patients and their families. health and home help services in unburdening tasks, Unburdening often resulted in Burden Relief consisting thus providing psychosocial support. of drug prescriptions – the most available therapeutic Demographics, dementia prevalence, grades of tool. Acetylcholinesterase inhibitors or memantine dementia drug prescribing permissiveness in the par- treatment was reported in 89 of 155 patients (57%): ticipating countries and share of dementia drugs pre- 60% of the prescriptions were appropriate according scribed in the case stories per region are shown in to guidelines and 40% prescribed outside of guide- Table 2. lines. Antidepressants were mentioned in 27% of the Statistical analytical procedures also showed the patients, and to a lesser degree other drugs such as relevance of Unburdening. Thus, statistical data were antipsychotics. Only 22 out of 155 case stories did not used in memos to test the basic social process core mention any drug to relieve the dementia burden. variable. This means that the statistics supported Table 1. The basic social process of Unburdening Dementia. Unburdening Dementia - a two-step Basic Social Process BASIC SOCIAL PROCESS STAGE: Recognising burden after burden identification and Burden Relief assessment. APPLIED TO DEMENTIA CARE : Identifying cognitive and mental burden by family caregiver. Cognitive, mental and social Burden Relief Assessing burden by psychometric tools and then Recognising burden by formal caregiver. CASE STORY ILLUSTRATION: i. Consultation together: spouse/child report Burden Relief by community health services, memory loss. home help services and drugs ii. Physician notices lack of collaboration during consultation. iii. Assessment with Mini Mental State Examination. iv. Physician diagnoses the patient with dementia. The table illustrates the basic social process of Unburdening Dementia that emerged to explain the action in the data provided by primary care physi- cians across EGPRN countries. Basic social processes are grounded theory core variables that are ‘processural’ meaning that they have two or more clear emergent stages [13,14]. Descriptive incidents of community health and home help services (such as nursing home care and dementia services) appeared in more than half of the case stories as indicators of a property of Burden Relief. SCANDINAVIAN JOURNAL OF PRIMARY HEALTH CARE 257 Table 2. Demographics, dementia prevalence, grades of dementia drug prescribing permissiveness in the participating countries and share of dementia drugs prescribed in the case stories per region. Primary care Proportion of Population 65 physician allowed typical dementia Population years old or Dementia to prescribe Number of case drug treatment per Country million people over (%) prevalence (%) dementia drugs? stories (% of total) case per region Denmark 5.8 18.2 1.5 PARTIAL 3 (2) Finland 5.5 19.4 1.7 NO 4 (3) Norway 5.2 15.9 1.6 YES 6 (4) Sweden 9.7 19.4 1.8 YES 8 (5) North Countries, total 21 57% Austria 8.6 18.3 1.7 NO 3 (2) Belgium 11.3 17.8 1.8 PARTIAL 2 (1) France 66.4 18.0 1.8 PARTIAL 5 (3) Germany 81.2 20.8 1.9 YES 5 (3) Ireland 4.6 12.6 1.1 YES 35 (23) Switzerland 8.2 17.6 1.7 YES 5 (3) The Netherlands 16.9 17.3 1.5 PARTIAL 5 (3) West Countries, total 60 67% Bulgaria 7.2 19.6 1.5 NO 5 (3) Croatia 4.2 18.4 1.5 NO 3 (2) Hungary 9.8 17.5 1.5 NO 3 (2) Poland 38.0 14.9 1.3 NO 20 (13) Romania 19.9 16.5 1.3 NO 10 (6) Slovenia 2.1 17.5 1.6 PARTIAL 4 (3) East Countries, total 45 40% Greece 10.8 20.5 1.8 NO 4 (3) Israel 8.5 10.3 1.1 NO 7 (5) Italy 60.6 21.4 2.1 NO 8 (5) Malta 0.4 17.9 1.3 NO 1 (1) Spain 46.4 18.1 1.8 NO 6 (4) Turkey 77.7 7.7 0.4 NO 3 (2) Mediterranean Countries, total 29 66% Total 155 57% Legend: Demography, dementia prevalence, prescription rules for primary care physicians, dementia cases stories and proportion of typical dementia drug treatment from 25 countries in the EGPRN. Data on dementia prevalence by Prince et al. 2013 and on dementia drug prescribing rules by the Alzheimer Europe Association, 2012. theory generation by making it less dependent on SP) are shown in Table 3. Alzheimer’s disease was the qualitative data only, due to the large amount of data most common dementia subtype followed by vascular readily coded quantitatively. After basic descriptive dementia. In a third of the patients, no subtype could statistics, we employed the analytic procedures below be specified due to lack of diagnostic information, during the selective coding phase. which resulted in the unspecified diagnoses of mild or major neurocognitive disorders. (a) Cluster analyses A two-step cluster analysis divided 151 case stories (b) Cross-sectional associations by chi-square into two groups (four cases had missing data). Group testing 1 had 76 stories and group 2 had 75 stories. ‘Permissiveness to prescribe dementia drugs’ and There was a significant positive association between ‘country’ predicted group affiliation (Figure 1). permissiveness to prescribe dementia drugs and pri- Group 1 had high involvement of primary care mary care physicians’ diagnosis of dementia. MMSE physicians in the dementia work up and treatment. testing was reported in 128 out of 155 cases (83%) Patients were from Belgium, Denmark, Finland, France, and was significantly associated with permissiveness Germany, Ireland, the Netherlands, Norway, Sweden, to prescribe dementia drugs but not with effective and Switzerland. prescription of dementia drugs. We found no signifi- Group 2 had low involvement of primary care cant association between permissiveness to prescribe physicians in the dementia work up and treatment. dementia drugs and the mentioning of dementia Patients were from Austria, Bulgaria, Croatia, Greece, drugs or appropriateness of treatment. Neither did we Hungary, Israel, Italy, Malta, Poland, Romania, Slovenia, find any significant association between appropriate- Spain, and Turkey. ness of treatment and early specialist involvement or The different types of dementia (neurocognitive dis- referral. Statistical measures of the testing above are orders) as assessed by a tertiary care physician (author available online together with results of more 258 F. PETRAZZUOLI ET AL. Figure 1. Importance of predictors of primary care physician involvement in dementia work-up and treatment. Degree of involvement is dichotomised into two groups by a two-step cluster analysis. Cluster analysis is an exploratory method to identify structures within the data such as homogenous groups of cases if grouping is not previously known. ‘Exploratory’ means that it makes no distinction between dependent and independent variables. We used the SPSS Version 22 two-step cluster analysis. The most important predicting variables to allocate the informants in the two groups were ‘permissiveness to prescribe dementia drugs’ (Predictor importance PI ¼ 1) and ‘country’ (PI ¼ 0.61). All the other variables had a PI of < 0.02. Permissiveness: official permissive- ness to prescribe typical dementia drug reimbursed by the local health care system. Country: country of the case and his/her primary care physician. Appropriate treatment: appropriate treatment according to the dementia guidelines and the label indication. Age: age of the case in years. Gender: gender of the case. MMSE: mini mental state examination. Number of words in case: word count in the case description. Table 3. Dementia drug treatment stratified by diagnostic group. Dementia Drug No regular Regular dementia drug (acetylcholinesterase Diagnosis dementia drug inhibitors/memantine) Total number of patients No diagnostic information available 4 0 4 Alzheimer’s Disease/Major Neurocognitive Disorder () 9 31 40 Vascular Dementia /Major Neurocognitive Disorder (†) 12 4 16 Major Neurocognitive Disorder unspecified (b) 19 31 50 Normal Pressure Hydrocephalus 1 0 1 Mild Neurocognitive Disorder unspecified 5 0 5 Major Neurocognitive Disorders, mixed aetiologies 10 22 32 Fronto-Temporal Dementia (b) 3 1 4 Dementia of Lewy Body 1 0 1 Alzheimer’s Disease/ Mild Neurocognitive Disorders 2 0 2 Total 66 89 155 Legend: Memantine and acetylcholinesterase inhibitors were considered appropriately used if they were prescribed to patients with Alzheimer’s disease (AD), mixed dementia (if AD was considered one of the components), dementia with Lewy bodies or Parkinson’s disease dementia. The indication for the treatment was considered correct regardless of dementia stage (i.e. memantine was not restricted to moderate to severe AD and acetylcholinesterase inhibitors was not restricted to mild to moderate AD) [21]. () ¼ appropriate prescription of acetylcholinesterase inhibitors/memantine. (b) ¼ questionable prescription of acetylcholinesterase inhibitors/memantine. (†) ¼ inappropriate prescription of acetylcholinesterase inhibitors/memantine. advanced statistical methods applied to the data in a with dementia and their caregivers after analysing previous version of this study [10]. both qualitative and quantitative data. Unburdening dementia is, according to grounded theory taxonomy, a core variable and was coded as a basic social pro- Discussion cess [14–16] starting with Recognising the burden by Unburdening dementia emerged as an explanation of Burden Identification and Assessment followed by what physicians from 25 countries did to help patients Burden Relief. As a core variable, Unburdening is SCANDINAVIAN JOURNAL OF PRIMARY HEALTH CARE 259 suggested as a cohesive explanation for the resolution difficulty of establishing a diagnosis. The discrepancy of a main concern – the comprehensive burden of further applies to patients being included in disease dementia – and as such a tentative explanatory theory modifying anti-amyloid trials aimed at relieving the of dementia management in primary care. dementia burden at the micro level [24,25]. Our study We analysed case histories and survey responses suggests that since many patients with dementia in from primary care physicians in 25 countries and primary care are treated outside of the traditional pro- found that only in 14% of 155 case stories was there tocols for the drug trials that were designed to prove no mention of any drug to relieve dementia problems. their effects, it would seem logical to expand the drug In more than half of the case stories, acetylcholinester- trial populations to include typical primary care ase inhibitors and/or memantine were mentioned as patients. being prescribed and in one quarter of case stories, guideline indications of dementia drug treatment Limitations were broadened [17,18]. The physicians also used other drugs to unburden dementia such as Day care rehabilitation, respite care and caregiver sup- antidepressants. port [26] etc. are much less mentioned as Burden Free text data and MCQ responses, confirmed by a Relief procedures in our study as compared to pre- two-step cluster analysis, showed that the more the scribing dementia drugs of sometimes marginal thera- primary care physicians were officially allowed to pre- peutic effect [17,18]. Why Unburdening by drugs scribe dementia drugs the more they were responsible emerged as the most common Burden Relief strategy for the dementia work-up [9,10] also engaging com- used by the primary care physicians could perhaps be munity health and home help services in unburdening explained by a ‘prescription reflex’ response against tasks. Pro-active dementia management by primary therapeutic nihilism [27]. The big picture of the case care physicians was related to geographical region stories analysed for this study resembles data from a with Central and Northern European physicians show- systematic review of barriers to dementia manage- ing high permissiveness to prescribe dementia drugs ment which showed limited access to and knowledge in comparison with a low prescribing permissiveness about community services and absence of interdiscip- in Eastern European and Mediterranean countries. This linary teams as well as attitudes of therapeutic nihilism finding of regional differences within Europe con- [8]. The respondents in our study could have been firmed a pattern seen previously [9,10]. biased for the eighth case story survey item by ques- ‘Label indications’ of drugs are often limited and tions 3 and 4 in the MCQ survey that asked about narrow [6,18]. Yet, both primary and secondary care drug treatment. physicians do not always apply a strict indication per- Our recruitment and sampling of key respondents spective. They have a patient with multiple severe was strategic, which limits generalizability. Also, the conditions in front of them and try to do their best to proportion of case story data from Ireland and Poland help that patient [22]. The most common tool that was higher than from other countries and differences many physicians use to help their patients is the pre- within countries are not considered since the survey scription of drugs. Physicians inevitably tend to was anonymous. Yet, we used the same data collec- broaden the indications for drugs which could be tion approach for all countries. Moreover, with hun- beneficial even though the patient does not fulfill the dreds of pages of case story text and free text eligibility criteria for prescription [23]. We call this pat- responses from several hundred primary care physi- tern of physician behaviour Unburdening by drugs. cians we reached conceptual saturation according to A majority of patients in our study that were pre- grounded theory principles of fit, relevance, workabil- scribed specific dementia drugs would not have been ity and modifiability [14,16]. These are qualities that in accepted in dementia drug trials that sit as the base traditional qualitative data analysis often are labeled for the alleged evidence of their effects [10]. Patients trustworthiness and generalizability. Hence, we sug- in our study were indeed on average 80 years old and gest that Unburdening as a basic social process is had significant co-morbidity. This suggests a discrep- enough abstract of time, place and people to be con- ancy between real world primary care dementia sidered useful for a general understanding of practice patients and patients from acetylcholinesterase inhibi- habits of primary care physicians. tor trials where the efficacy of these drugs was deter- Our inductive method was based on classic mined. Patients in our study were thus old, with high grounded theory that mostly analyses qualitative data. co-morbidity and a late diagnosis often due to the Yet, we used quantitative measures too, rare for classic 260 F. PETRAZZUOLI ET AL. grounded theory but they fit our data [11–13]. Thus, counter therapeutic nihilism [8,29]. Alas, ‘prescription Unburdening dementia emerged as a relevant basic reflex’ doctoring is especially risky for geriatric patients social process. Since our data was based on what [27]. In this study, physicians thus seemed to rely informants chose to report the results should, as in all more on drug treatment to relieve dementia burdens classic grounded theory research, not be seen as facts than what is proposed by today’s updated prescription but as tentative hypotheses. This is a limitation in guidelines where deprescribing is actually emphasised deductive research but fits the tenets of the inductive in unspecified dementia (but not in Alzheimer’s dis- property of the classic grounded theory method. ease) [17]. In many case stories, specification of the type of It is almost needless to say that Unburdening by dementia was not given by the respondents and this drugs has general implications for physicians concern- is similar to other studies in primary care where the ing how they deal with many different health issues. subtype of dementia was not considered [28]. Hence, We thus propose that Unburdening by drugs is what a tertiary care physician researcher carefully examined most practicing physicians do every day with or with- all case stories and tried to establish the specific out evidence-based support for the drugs they are dementia type in accordance with DSM-5 [19]. prescribing. So why should it be different for patients with dementia? Contribution to policy and practice The conceptual framework of Unburdening could be further developed in order to understand practice High permissiveness to prescribe dementia drugs was habits for different diseases and hypotheses based on related to extensive dementia management by pri- Unburdening may be tested deductively. mary care physicians when compared with low per- missiveness. This could be of interest for dementia care stakeholders as a way to incentivize dementia Conclusions management [29]. Since studies exploring contempor- ary dementia management across different countries In this grounded theory analysis of an international from a primary care perspective are rare Unburdening primary care physician survey, a basic social process dementia could be a key for developing policies of that we call Unburdening dementia emerged as a reso- dementia management not limited only to health care lution for a core concern of the physicians. professionals. Unburdening is suggested as a tentative explanatory To our knowledge, Unburdening presented as a theory of dementia management in primary care start- basic social process [15] has not been published ing with Recognising a dementia burden by before for any health issue, let alone for dementia. Identification and Assessment of dementia followed by Recognising dementia – the first stage of dementia Burden Relief. Unburdening – has indeed been much explored [29] Physicians typically identify the burden of both the and the burden concept often appears in qualitative patient and family members and then assess the cog- research on chronic disease management in general nitive impairment with MMSE in order to relieve the and in dementia care in particular [7,8]. Hence, there dementia burden, commonly by drug prescriptions, is a vast literature about caregiver burden in dementia but also by community health and home help services care resulting in emotional distress and burn-out [7]. in unburdening tasks. Dementia drug therapy seemed In a grounded theory dissertation, unburdening was a inappropriate in 40% of patients where primary care subcore concept explaining caregivers coping by physicians broadened the guidelines with the purpose ‘telling their stories’ to the researcher [30]. of Unburdening by drugs. Our findings have implica- Unburdening by drugs was accompanied by a tions with regard to how dementia is managed in real broadening of guideline indications for dementia world primary care across many jurisdictions. drugs. This was in some cases mentioned by physi- cians as motivated by mercy and equity. In a system- atic review of psychosocial factors of patients’ and Acknowledgments carers’ experiences of dementia diagnosis and treat- The authors are grateful to Mariella Stamato, nurse in ment, it was found that dementia drugs gave hope Uithoorn (the Netherlands), for help with the collection of and were considered as ‘worth a try’ although the Dutch cases and to Dr Barney G Glaser for encouraging benefits were not clear [31]. This view of expectations remarks regarding theoretical coding of the data using a should not be neglected in dementia management to basic social process. SCANDINAVIAN JOURNAL OF PRIMARY HEALTH CARE 261 Ethics approval and consent to participate [11] Glaser B, Strauss AL. The discovery of grounded the- ory: strategies for qualitative research. New The participation of the physicians was voluntary. Except for Brunswick, NJ: Aldine; 1967. in Ireland, the country that supplied the highest number of [12] Glaser BG. Doing quantitative grounded theory. Mill cases and where research ethics approval was obtained Valley, CA: Sociology Press; 2008. (ICGP-Bradley 2014/10/01), no formal research ethics [13] Holton JA, Walsh I. Classic grounded theory: applica- approval was needed and therefore requested for the survey tions with qualitative and quantitative data. Newbury study in the remaining countries. Park, CA: Sage; 2017. [14] Glaser BG. Doing grounded theory: issues and discus- sions. Mill Valley, CA: Sociology Press; 1998. Disclosure statement [15] Glaser BG, Holton J. Basic social processes. Grounded The authors declare that they have no conflict of interest. theory review. 2005;4:1–29. [16] Glaser BG. Theoretical sensitivity: advances in the methodology of grounded theory. Mill Valley, CA: Data availability statement Sociology Press; 1978. [17] Reeve E, Farrell B, Thompson W, et al. Deprescribing All data generated or analysed during this study are cholinesterase inhibitors and memantine in dementia: included in this published article and its supplementary guideline summary. Med J Aust. 2019;210(4):174–179. information files as well as in reference [10]. The datasets [18] O’Brien JT, Holmes C, Jones M, et al. Clinical practice analysed during this study will be available from the corre- with anti-dementia drugs: A revised (third) consensus sponding author on reasonable request. statement from the British Association for Psychopharmacology. J Psychopharmacol (Oxford)). 2017;31(2):147–168. References [19] Association AP. Diagnostic and statistical manual of [1] Singer M, Bulled N, Ostrach B, et al. Syndemics and mental disorders (DSM-5V R ). Washington DC: American the biosocial conception of health. Lancet. 2017; Psychiatric Pub; 2013. 389(10072):941–950. [20] Sarstedt M, Mooi E. Cluster analysis. A concise guide to [2] Iliffe S, Manthorpe J. Dementia: is the biopsychosocial market research. NYC, NY: Springer; 2014. p. 273–324. model vindicated? Br J Gen Pract. 2017;67(661):344–345. [21] Creavin ST, Wisniewski S, Noel-Storr AH, et al. Mini- [3] Glynn LG, Valderas JM, Healy P, et al. The prevalence of Mental State Examination (MMSE) for the detection of multimorbidity in primary care and its effect on health dementia in clinically unevaluated people aged 65 and over in community and primary care populations. care utilization and cost. Fam Pract. 2011;28(5):516–523. €m J, Wettermark B, et al. Cochrane Database Syst Rev. 2016;1:CD011145. [4] Schmidt-Mende K, Hasselstro [22] Grimsmo A, Lohre A, Rosstad T, et al. Disease-specific General practitioners’ and nurses’ views on medica- clinical pathways - are they feasible in primary care? tion reviews and potentially inappropriate medicines A mixed-methods study. Scand J Prim Health Care. in elderly patients - a qualitative study of reports by 2018;36(2):152–160. educating pharmacists. Scand J Prim Health Care. [23] Gupta SK, Nayak RP. Off-label use of medicine: 2018;36(3):329–341. Perspective of physicians, patients, pharmaceutical [5] Clegg A, Young J, Iliffe S, et al. Frailty in elderly peo- companies and regulatory authorities. J Pharmacol ple. Lancet. 2013;381(9868):752–762. Pharmacother. 2014;5(2):88–92. [6] Livingston G, Sommerlad A, Orgeta V, et al. Dementia [24] van Dyck CH. Anti-amyloid-b monoclonal antibodies prevention, intervention, and care. Lancet. 2017; for Alzheimer’s disease: pitfalls and promise. Biol 390(10113):2673–2734. Psychiatry. 2018;83(4):311–319. [7] Cheng ST. Dementia caregiver burden: a research [25] Roostaei T, Nazeri A, Felsky D, et al.; Alzheimer’s update and critical analysis. Curr Psychiatry Rep. 2017; Disease Neuroimaging Initiative (ADNI). Genome-wide 19(9):64. interaction study of brain beta-amyloid burden and [8] Koch T, Iliffe S, Project E-E, EVIDEM-ED project. Rapid cognitive impairment in Alzheimer’s disease. Mol appraisal of barriers to the diagnosis and manage- Psychiatry. 2017;22(2):287–295. ment of patients with dementia in primary care: a [26] de Waard CS, Poot AJ, den Elzen WPJ, et al. Perceived systematic review. BMC Fam Pract. 2010;11:52. doctor-patient relationship and satisfaction with gen- [9] Petrazzuoli F, Vinker S, Koskela TH, et al. Exploring eral practitioner care in older persons in residential dementia management attitudes in primary care: a homes. Scand J Prim Health Care. 2018;36(2):189–197. key informant survey to primary care physicians in 25 [27] O’Malley K, Laher M, Cusack B, et al. Clinical pharma- European countries. Int Psychogeriatr. 2017;29(9): cology and the elderly patient. In: Denham M.J., edi- 1413–1423. tor. The treatment of medical problems in the elderly. [10] Petrazzuoli F. Dementia management in European pri- Current status of modern therapy., vol 3. Springer: mary care. [Dissertation]. Lund: Lund University: Dordrecht, 1980. Faculty of Medicine, 2019. 107 p. https://portal. [28] Ko€hler S, Buntinx F, Palmer K, et al. Depression, vascu- research.lu.se/portal/files/64009821/Ferdinando_ lar factors, and risk of dementia in primary care: a Petrazzuoli_WEBB_1_.pdf. (accessed 22 February retrospective cohort study. J Am Geriatr Soc. 2015; 2020) 63(4):692–698. 262 F. PETRAZZUOLI ET AL. [29] Koch T, Iliffe S, Evidem Programme. The role of pri- No mary care in the recognition of and response to dementia. J Nutr Health Aging. 2010;14(2):107–109. Comments: … … … … . … … … … … … … … … … … [30] Keady J. The dynamics of dementia: a modified ………………………………………………… grounded theory study [Dissertation].University of …………… Wales: Bangor, 1999. [31] Bunn F, Goodman C, Sworn K, et al. Psychosocial fac- 4. Is continued dementia drug treatment reimbursed if tors that shape patient and carer experiences of prescribed by GPs in your country? dementia diagnosis and treatment: a systematic  Yes review of qualitative studies. PLoS Med. 2012;9(10): No e1001331. Comments: … … … … . … … … … … … … … … … … ………………………………………………… … … … … … … … … … … … … …. Appendix 1 5. Do GPs try to establish a diagnosis of dementia on their Exploratory survey on dementia diagnosis and own? treatment in your country (jurisdiction)  never Country … … … … … … … . … … .  rarely Name: … … … … … … … … … … .  often Email: … … … … … … @ … … … … … … … … …  always These questions concern the current situation in your own Comments: … … … … . … … … … … … … … … … … country or jurisdiction (region or health care district) ………………………………………………… Please respond with as many comments as possible since …………………………… multiple choice questions do not cover all that is important in an exploratory study! If you write your e-mail address we 6. Do GPs refer a suspected case of dementia to a second- will write you back! ary care specialist?  never 1. Which healthcare professionals are officially responsible rarely for the diagnosis of dementia? often  Secondary care specialists (neurologist, geriatrician, always psychiatrist, neuropsychiatrist) 2. GPs Comments: … … … … . … … … … … … … … … … … 3. Both ………………………………………………… …………………………… Comments: … … … … . … … … … … … … … … … … ………………………………………………… 7. What would GPs need to be able to detect dementia …………… better? More than one response option possible.  More time in the consultation 2. Which are the most popular dementia screening tests Incentives used? Please comment if Easy and short dementia tools  MMSE Other … … … … … … … … … … … … … … … Clock Drawing Test Other. Please specify … … … … … … … … … … … … … Comments: … … … … . … … … … … … … … … … … …………………………… ………………………………………………… …………… Comments: … … … … . … … … … … … … … … … … ………………………………………………… 8. Please tell as much as possible about a case of demen- …………… tia in your practice! (or tell us in a later e-mail) 3. Are GPs allowed to start prescribing drug treatment for … … … ….… … … … … … … … … … … … … … … dementia? …………………………………………………  Yes ………………… SCANDINAVIAN JOURNAL OF PRIMARY HEALTH CARE 263 Appendix 2 Galantamine Exclusion criteria for anti-amyloid trials Some stable and well-controlled concomitant medical disorders were not reason for exclusion (hypertension, heart failure, diabetes Anti-amyloid trials and hypothyroidism). The list of reasons for exclusion was quite Key Exclusion Criteria: extensive and consistent across studies, other neurodegenerative illness, cardiovascular disease, or active cerebrovascular disease,  Any medical or neurological condition (other than Alzheimer’s clinically significant infarct dementia, psychiatric, hepatic, renal, D.) that might be a contributing cause of cognitive pulmonary, metabolic, endocrine, active peptic ulcer, history of impairment. epilepsy, drug or alcohol abuse.  Stroke or Transient Ischaemic Attack (TIA) or unexplained loss of consciousness in the past year.  Clinically significant psychiatric illness in past six months.  Seizure in the past three years.  Poorly controlled diabetes mellitus.  Unstable angina, myocardial infarction, chronic heart failure, or clinical significant conduction abnormalities past year.  Impaired renal or liver function.  Human immunodeficiency virus (HIV) infection. Rivastigmine  Significant systematic illness or infection in past 30 days. The list of exclusions was not extensive. Patients with severe and  Brain MRI showing acute or sub-acute micro or macrohae- unstable illnesses (cardiovascular or pulmonary disease, unstable morrhage, >4 microhaemorrhages, cortical infarct or > 1 diabetes mellitus, peptic ulceration within the preceding five lunar infarct. years, evidence of alcohol or substance abuse) were excluded, as  Any contraindications to brain MRI or PET scans. were subjects taking medications such as anticholinergic drugs, acetyl-choline precursor health food supplements, memory  Negative PET scan with any amyloid-targeting ligand last 48 enhancers, insulin and psychotropic drugs. weeks.  Clinically significant 12-lead electrocardiogram (ECG) abnormalities.  Alcohol or substance abuse in past 1 year.  Taking blood thinners (except for aspirin at prophylactic dose or less)  Changes in medications or doses of medication in past 4 weeks. FACT BOX Adapted from Sevigny et al, Nature 2016 (The antibody aducanu- Donepezil studies mab reduces Ab plaques in Alzheimer’s disease) Patients were excluded if they had: insulin-dependent diabetes mellitus, other endocrine disorder, asthma, obstructive pulmonary disease or clinically significant uncontrolled gastrointestinal hep- atic or cardiovascular diseases. Medications such as anticholiner- gics, anticonvulsants, antidepressants and antipsychotics. Galantamine studies Patients were excluded if they had: other neurodegenerative ill- ness, cardiovascular disease, or active cerebrovascular disease, Donepezil clinically significant infarct dementia, psychiatric, hepatic, renal, Patients were excluded from the donepezil studies if they had pulmonary, metabolic, endocrine, active peptic ulcer, history of insulin-dependent diabetes mellitus or other endocrine disorder, epilepsy, drug or alcohol abuse. asthma, obstructive pulmonary disease or clinically significant Rivastigmine studies uncontrolled gastrointestinal hepatic or cardiovascular diseases. Patients were excluded if they had: severe and unstable illnesses Patients known to be hypersensitive to cholinesterase inhibitors (cardiovascular or pulmonary disease, unstable diabetes mellitus, or who had taken tacrine or other investigational medicines peptic ulceration within five years, alcohol or substance abuse), within one month of baseline were excluded. Concomitant medi- medications such as anticholinergic drugs, acetylcholine precursor cations such as anticholinergics, anticonvulsants, antidepressants health food supplements, memory enhancers, insulin and psycho- and antipsychotics were not allowed. Drugs with central nervous tropic drugs. system (CNS) activity were prohibited or partially restricted. Adapted from: AChEI, Cholinesterase inhibitors for Alzheimer’s disease, Birks J, The Cochrane Library 2009, Issue 1 264 F. PETRAZZUOLI ET AL. . Donepezil . Rivastigmine Patients were excluded from the donepezil studies if they had The list of exclusions was not extensive. Patients with severe and insulin-dependent diabetes mellitus or other endocrine disorder, unstable illnesses (cardiovascular or pulmonary disease, unstable asthma, obstructive pulmonary disease or clinically significant diabetes mellitus, peptic ulceration within the preceding five uncontrolled gastrointestinal hepatic or cardiovascular diseases. years, evidence of alcohol or substance abuse) were excluded, as Patients known to be hypersensitive to cholinesterase inhibitors were subjects taking medications such as anticholinergic drugs, or who had taken tacrine or other investigational medicines acetyl-choline precursor health food supplements, memory within one month of baseline were excluded. Concomitant medi- enhancers, insulin and psychotropic drugs. cations such as anticholinergics, anticonvulsants, antidepressants and antipsychotics were not allowed. Drugs with central nervous system (CNS) activity were prohibited or partially restricted. . FACT BOX Donepezil studies Appendix 3 Patients were excluded if they had: insulin-dependent diabetes mellitus, other endocrine disorder, asthma, obstructive pulmonary Exclusion criteria for dementia drug RCTs disease or clinically significant uncontrolled gastrointestinal hep- atic or cardiovascular diseases. Medications such as anticholiner- . Galantamine gics, anticonvulsants, antidepressants and antipsychotics. Galantamine studies Some stable and well-controlled concomitant medical disorders Patients were excluded if they had: other neurodegenerative ill- were not reason for exclusion (hypertension, heart failure, dia- ness, cardiovascular disease, or active cerebrovascular disease, betes and hypothyroidism). The list of reasons for exclusion was clinically significant infarct dementia, psychiatric, hepatic, renal, quite extensive and consistent across studies, other neurodege- pulmonary, metabolic, endocrine, active peptic ulcer, history of nerative illness, cardiovascular disease, or active cerebrovascular epilepsy, drug or alcohol abuse. disease, clinically significant infarct dementia, psychiatric, hepatic, Rivastigmine studies renal, pulmonary, metabolic, endocrine, active peptic ulcer, his- Patients were excluded if they had: severe and unstable illnesses tory of epilepsy, drug or alcohol abuse. (cardiovascular or pulmonary disease, unstable diabetes mellitus, peptic ulceration within five years, alcohol or substance abuse), medications such as anticholinergic drugs, acetylcholine precursor health food supplements, memory enhancers, insulin and psycho- tropic drugs. Adapted from: AChEI, Cholinesterase inhibitors for Alzheimer’s dis- ease, Birks J, The Cochrane Library 2009, Issue 1