Open access Protocol Multimorbidity patterns in chronic older patients, potentially inappropriate prescribing and adverse drug reactions: protocol of the multicentre prospective cohort study MoPIM Marisa Baré ‍ ‍,1,2 Susana Herranz,3 Rosa Jordana,4 Maria Queralt Gorgas,5 Sara Ortonobes,5 Daniel Sevilla,6 Elisabet De Jaime,7 Olatz Ibarra,8 Candelaria Martín,9 on behalf of the MoPIM study group To cite: Baré M, Herranz S, Abstract Jordana R, et al. Multimorbidity Strengths and limitations of this study Introduction Multimorbidity is a major challenge for patterns in chronic older current healthcare systems and professionals. From patients, potentially ►► To our knowledge, this is the first published study the different approaches that have been proposed to inappropriate prescribing and that includes objectives related to chronic multimor- adverse drug reactions: protocol analyse this issue, the hypothesis of the existence of bidity, appropriateness of medication and adverse of the multicentre prospective association patterns of different chronic conditions drug reactions during hospitalisation at the same cohort study MoPIM. BMJ Open is gaining visibility. In addition, multimorbidity can time, in older patients. 2020;10:e033322. doi:10.1136/ be associated to polypharmacy, which can lead to a ►► Data of multimorbidity include also identification of bmjopen-2019-033322 higher risk of potentially inappropriate prescribing (PIP) chronic exacerbated diseases that cause hospital- ►► Prepublication history for and consequently to adverse drug reactions (ADRs). isation, geriatric syndromes and functional status. this paper is available online. The general objective of this novel study is to identify ►► The multicentre cohort design as well as the inno- To view these files, please visit the association between PIP, multimorbidity patterns, vative analytical approach of multimorbidity patterns the journal online (http://​dx.​doi.​ polypharmacy and the presence of ADRs in older patients can lead to novel results in this field. org/​10.​1136/​bmjopen-​2019-​ admitted for exacerbation of chronic diseases. ►► Adverse drug reactions rate could be subject to an 033322). Methods and analysis The MoPIM (morbidity, potentially infra-­estimation if active efforts to identify them are inappropriate medication) study is a multicentre not complete. Received 31 July 2019 Revised 22 November 2019 prospective cohort study of an estimated sample of Accepted 30 December 2019 800 older (≥65 years) patients admitted to five general hospitals in Spain due to an exacerbation of a chronic Universitario de Canarias. The results will be actively and disease. Patients referred to home hospitalisation, mainly disseminated through publication in peer-­reviewed admitted due to an acute process or with a fatal journals and communications in scientific conferences. outcome expected at the time of admission are excluded. Trial registration number NCT02830425. Sociodemographic data, chronic morbidities and geriatric syndromes, number of chronic prescribed medications, PIP at admission to hospital and on discharge, according Introduction to the newest screening tool of older screening tool of Multimorbidity and possible patterns older person's potentially inappropriate prescriptions/ In recent years, the healthcare landscape screening tool to alert doctors to right treatment criteria, has changed dramatically, with the consider- and ADRs during hospitalisation are being collected. able and gradual increase of older patients Multimorbidity patterns will be identified using cluster with multiple chronic health conditions.1 analyses techniques, and the frequency of polypharmacy, Multimorbidity has become an important © Author(s) (or their PIP and ADRs will be calculated. Finally, the possible challenge for the health system due to, on employer(s)) 2020. Re-­use relationship between those indicators will be identified the one hand, the population ageing (which permitted under CC BY-­NC. No through bivariate and multivariate analyses. commercial re-­use. See rights increases the percentage of people with multi- Ethics and dissemination The project has been and permissions. Published by approved by the clinical research ethics committees of morbidity) and, on the other hand, the diffi- BMJ. culty of clinical management of patients with each centre: Comité Ético de investigación Clínica del For numbered affiliations see Parc Taulí, Comitè Ètic d'Investigació Clínica Osona per multimorbidity.1 2 Far from treating diseases end of article. or isolated processes, in patients with several a la Recerca i Educació Sanitàries (FORES), Comité de Correspondence to Ètica de la Investigación con Medicamentos (CEIm)-­Parc chronic pathologies, and especially at times Dr Marisa Baré; de Salut MAR, Comité Ético de Investigación Clínica de of decompensation, the complexity inherent ​mbare@​tauli.​cat Euskadi, Comité de Ética de Investigación del Hospital to this type of patients becomes more evident. Baré M, et al. BMJ Open 2020;10:e033322. doi:10.1136/bmjopen-2019-033322 1 Open access In addition, in older patients, a process of exacerbation or alert doctors to right treatment) are the most used and decompensation of chronic problems may coincide with validated in European older people. In fact, they were the existence of geriatric syndromes such as acute confu- developed under the aegis of the European Union Geri- sional syndrome or frailty. In these cases, therapeutic atric Medicine Society.19 STOPP-­START criteria include decisions may require great knowledge and skills for drug interactions and therapeutic duplicity.19 The 2008 sharing decision making to achieve an adequate balance version consisted of a list of 84 medication indications, between risks and benefits for the patient.3 However, clin- developed using a Delphi method applied to experts ical practice guidelines do not usually consider or resolve from different disciplines, who carried out a literature the uncertainties related to these situations. In fact, review. The criteria are directed to prevalent diseases in randomised controlled trials often exclude patients with older patients and are ordered by physiological systems. concomitant pathologies.4 These criteria are easy to relate to active diagnoses and Currently, basic knowledge about the interrelationship to the patients’ medication lists that appear in their elec- of diseases is very limited, partly because of the existing tronic health records. The systematic review published by confidence in a scientific method that maximises internal Hill-­Taylor et al in 201316 concluded that these criteria validity but excludes patients with comorbidity in both were more sensitive than those of Beers to detect PIP. In observational studies and clinical trials. Attempts to study 2015, O’Mahony et al20 published an updated version that multimorbidity clash with the absence of an internation- includes 114 criteria, which was subsequently translated ally agreed definition and quantification.5 A systematic into Spanish.21 review of Violan et al6 identified important variations It should be noted that although the criteria for evalu- in both the definition of multimorbidity and its preva- ating PIP can be helpful in making decisions, the clinical lence in different contexts and ages. Despite these vari- and social context of the patient has to be always taken ations, multimorbidity is much more frequent in older into account. Hence there is need to share and assess patients.7 8 Among the different definitions, the one that the findings or PIP detected with the prescribing profes- advocates for the existence of association patterns of some sional. Fortunately, it is increasingly common to use these of these chronic diseases, either because they coincide, or or other criteria to identify a possible treatment inade- because they share some physiopathological mechanisms, quacy and assess a possible deprescription or prescrip- is recently gaining strength.9 10 tion. However, its systematic use could be far from being real. Polypharmacy, PIP and ADRs In recent years, many studies have been published Polypharmacy is a remarkable feature of patients with using the STOPP-­START criteria to assess the adequacy of multiple chronic pathologies. For this same reason, the medications in the community, socio-­health centres and/ prevalence of polypharmacy is usually high in the older or nursing homes, and the hospital setting.14 18 22–28 At the population.11 Although there is no consensus on the time of initiating this study, no published analyses were minimum number of prescribed drugs to consider a available with the objective to describe the relationships patient as polymedicated, the chronic daily consumption between multimorbidity, PIP and ADRs. of five or more drugs is a commonly accepted figure.12 Polypharmacy, in turn, is associated with a higher proba- Multimorbidity, polypharmacy, PIP and ADRs bility of adverse drug reactions (ADRs). In older people, Given the importance of this phenomenon, it is necessary this probability is even higher as a result of physiological to deepen our knowledge to offer the best clinical care changes associated with ageing, changes in the pharma- and optimise health outcomes. cokinetic and pharmacodynamic behaviour of drugs, Taking into account all the previous considerations, the functional problems and social aspects. It has been esti- hypotheses of the study were, first, that the morbidity of mated that ADRs are responsible for up to 30% of hospital older patients admitted to hospital due to exacerbation admissions in older patients.13 Potentially inappropriate of their chronic pathology may show certain patterns or prescribing (PIP) of drugs and poor monitoring of profiles of association. Second, that geriatric syndromes prescribed treatments are predisposing factors for the may be relevant in these patterns, and that these patterns appearance of ADRs.13 Fernandez et al14 observed that may be associated with greater polypharmacy and PIP. 69% of the ADRs detected in hospitalised polymedicated And third, that potentially inappropriate medications, or older patients were due to treatments that were identi- some of them, may, in turn, be associated with a greater fied as inappropriate. In 2011, Hamilton and colleagues presence of ADRs, already detected at the time of hospital also identified an association between overtreatment and admission or developed during the same hospitalisa- ADRs in patients.15 tion, compared with those patients with no medication inappropriateness. The STOPP-START criteria There are several tools to evaluate potentially inade- Objectives quate prescribing.16–18 Among all of them, the explicit The general objective of the study is to identify the asso- criteria STOPP-­START (screening tool of older person's ciation between multimorbidity, polypharmacy, PIP and potentially inappropriate prescriptions/screening tool to the presence of ADRs in older patients admitted for 2 Baré M, et al. BMJ Open 2020;10:e033322. doi:10.1136/bmjopen-2019-033322 Open access exacerbation of chronic diseases. The specific objectives Only patients attended by the physician participating in are to: the study are included. 1. Identify and describe the PIP of chronic medication During recruitment, which lasted from September according to STOPP-­START criteria at the time of ad- 2016 until the end of 2018, each patient was included mission and on discharge. only once. 2. Estimate the prevalence of multimorbidity and, togeth- er with geriatric syndromes and risk factors, identify Data acquisition possible morbidity association patterns in older pa- Each patient is monitored until hospital discharge (or tients admitted as a result of an exacerbation of their death during hospitalisation), with data collection during chronic disease. the first days of admission to hospital ward and at the time 3. Describe the possible polypharmacy (≥5 chronic medi- of discharge. cations) of those patients. As part of the usual patient care routine for older 4. Assess the difference in PIP on admission and at dis- patients, the pharmacist of the clinical committee reviews charge. the prescribed chronic medication at the time of admis- 5. Evaluate if any of the possible multimorbidity patterns sion. STOPP/START (SS) criteria are checked for each are associated with greater PIP or with specific types patient. These consist of 80 STOPP criteria (which detect of PIP. medication that would not meet criteria for indication 6. Analyse the possible association between the amount to a patient or a specific clinical situation or medications or the types of PIP and the number or types of ADRs. prescribed included drug-­ drug and drug-­ disease inter- actions) and 34 START criteria (detect medication that would be recommended to incorporate, including some Methods and analysis vaccines).20 For each patient, possible PIP are recorded at Design and setting the time of admission and, following the usual practice, A multicentre prospective cohort study has been the clinical committee evaluates the PIP together with designed, including patients admitted to the internal the possible modification of the medication according to medicine or geriatric services of five general hospitals the SS criteria. For the purposes of the study, the criteria in three different regions of Spain: Parc Taulí Hospital are re-­applied and recorded by the pharmacist to the Universitari, Consorci Hospitalari de Vic, Hospital del prescribed medication on discharge, in this case without Mar, Hospital de Galdakao and Complejo Hospitalario evaluation by the clinical committee. Universitario de Canarias. All of them have an internal In addition, ADRs notified to the pharmacy department medicine and/or geriatric service and are part of, or by the clinical team or registered in the clinical course collaborate with, the Spanish Health Services Research are identified by the pharmacist, and the consequences on Chronic Patients Network (REDISSEC). In addition, a derived from them are classified as well as if they have process of medication review is established in all of them, been resolved at the time of discharge. According to the using criteria for older patients. WHO and the European Medicine Agencies, an ADR is Inclusion criteria any undesirable event that has happened to the patient Patients older than 64 years admitted to the internal while using a medication and there is a suspicion that it medicine or geriatrics service as a result of the exacerba- is caused by the medication, including: (a) any noxious, tion or decompensation of their chronic pathology are unintended and undesired effect of a drug after doses included. A stay of >48 hours in hospital is considered an used in humans for prophylaxis, diagnosis or therapy and admission. (b) noxious and unintended effects resulting not only No written informed consent was deemed necessary for from the authorised use of a medicinal product at normal this study. doses but also from medication errors and uses outside the terms of the marketing authorisation, including the Exclusion criteria misuse and abuse of the medicinal product.29 30 As a Patients referred to home hospitalisation (patient at routine standard practice, a physician, a nurse, a clinical home but the clinical team provides punctual attention by assistant or the pharmacist can notify any ADRs to the telephone or home visits), those admitted due to an acute pharmacy department or to the electronic notification process not related to the chronic diseases (for example, system for adverse reactions to medication, held by the an infection) according to clinical judgement of the Spanish Ministry of Health. corresponding physician or those with a fatal outcome expected at the time of admission are not included. Variables In the case that various patients met the inclusion The different types of information obtained according to criteria during the same day, and so that the clinical team the moment of monitoring are shown in figure 1. All of could take the tasks of registering all variables of the study them are described below: together with their care tasks, they were sorted alphabet- a. Sociodemographic: patient’s code, centre, date of ically and selected the first of all of them to be included birth, sex, type of cohabitation (alone, with relatives or in this study. other people, in a nursing home). Baré M, et al. BMJ Open 2020;10:e033322. doi:10.1136/bmjopen-2019-033322 3 Open access Table 1 Chronic conditions and geriatric syndromes recorded Geriatric syndromes Chronic conditions and risk factors Charlson Index 1. AIDS/HIV Acute confusional syndrome/delirium 2. Any malignancy (excluding skin) Chronic pain 3. Cerebrovascular disease Cognitive/intellectual impairment 4. Chronic pulmonary disease Constipation 5. Congestive heart failure Depression or Anxiety 6. Dementia Dysphagia 7. Diabetes with complication Frailty 8. Diabetes without complication Immobility Figure 1 Data registered for each patient included in the morbidity, potentially inappropriate medication study. 9. Hemiplegia Incontinence (Urinary/ ADR, adverse drug reactions; PIP, potentially inappropriate faecal) prescribing. 10.Leukaemia Instability/falls 11.Lymphoma Malnutrition b. Clinical: date of arrival at the emergency room, date of 12.Metastatic solid tumour Polypharmacy admission, date of discharge, destination at discharge 13.Mild liver disease Pressure ulcers (home, transfer to another hospital, transfer to anoth- 14.Moderate or severe liver disease Sensorial deficit er service of the hospital itself, transfer to a nursing 15.Moderate or severe renal disease Sleep disorders/ home), cause of the death (chronic illness, complica- Insomnia tion or others), functional status just before entering 16.Myocardial infarction the hospital (Barthel index), existence of hospitalisa- tion and/or visits to healthcare services in the three 17.Peptic ulcer disease previous months due to exacerbation of any chronic 18.Peripheral vascular disease pathology, existence of pharmacological allergies and 19.Rheumatologic disease active principles involved. Other conditions Chronic active diseases of the patient are recorded. For 20.Amputation this purpose, the physicians of the project have defined, on a consensual basis, a limited list of 64 chronic problems or 21.Anaemia conditions, coming from the 114 groups defined by Salis- 22.Asthma bury and colleagues,8 and including the 19 categories of the 23.Cardiac arrhythmia Charlson Index.31 In order to define the list of conditions to 24.Cataract be recorded, the physicians that participated in the project 25.Chronic hepatitis (B or C) considered the subjective estimated frequency of the condi- tion in older hospitalised patients in their departments. 26.Chronic pancreatic disease Following the same criteria as Salisbury, a condition is 27.Degenerative arthropathy considered to be chronic when it lasts for at least 6 months, 28.Dermatitis or eczema including past conditions that require ongoing disease or 29.Diverticular disease of the colon risk management, important conditions with a significant 30.Drug-­related conditions risk of recurrence or past conditions that have continuing implications for patient management.8 31.Dyslipidaemia For each of the chronic condition (see table 1), it is 32.Fibromyalgia also recorded if they have required attention or clinical 33.Gallstones (previous hepatic colic) management throughout the hospitalisation, and the 34.Chronic gastritis or gastro-­oesophageal reflux attending physician of the clinical team assigns a (subjec- tive) correlative score (1, 2, 3…) to each one, according 35.Glaucoma to the importance or weight during the attention process. 36.Gout Thus, chronic problems that have not had any signifi- 37.Haemorrhoids cance during hospitalisation do not represent any score. 38.Haematologic disorders (myelodysplastic syndrome, Drug-­related conditions of this list refer to poor manage- gammapathy, polycythaemia) ment of medication for a chronic disease that have clinical Continued 4 Baré M, et al. BMJ Open 2020;10:e033322. doi:10.1136/bmjopen-2019-033322 Open access considered chronic if it has been prescribed, at least, Table 1 Continued 3 months ago. Active principles were only considered Geriatric syndromes separately, regardless of the combinations. Creams, Chronic conditions and risk factors ointments and healing material were not considered. 39.Hypertension d. ADRs identified both at the time of admission and 40.Inflammatory osteoarticular disease during the course of admission: drug involved, type of ADR according to the Wills and Brown classification33 41.Irritable bowel syndrome (predictable, unpredictable, continuous or prolonged, 42.Ischaemic heart disease without infarction with carcinogenic effect, rebound effect), start date if 43.Migraine it appeared along stay in the hospital, consequenc- 44.Neurologic disorder of the central nervous system es in terms of health (causes death, threatens life, 45.Non-­congestive heart failure lengthens the time of hospitalisation, other important consequences under medical criteria) and if they are 46.Non-­ischaemic heart disease (miocardiopatie, valvulopatie) resolved during admission or at discharge. 47.Non-­schizophrenic mental disorders (excluding Pilot study depression and anxiety) At the beginning of the recruitment, the first 10 admis- 48.Obesity sions per centre were selected to validate the data collec- 49.Osteoporosis tion instruments and identify the problems that could 50.Other neurological pathologies (essential tremor) arise during this stage. Then, relevant changes were made in the protocol and in the questionnaires. 51.Other vascular diseases (ischaemia, aneurism) 52.Parkinson’s disease Sampling and analysis 53.Peripheral neuropathy or neuritis An estimated consecutive sample of 800 patients meeting 54.Post-­traumatic stress disorder the inclusion criteria are being selected, proportionally 55.Previous fractures (not hip) distributed to the volume of income of the medicine 56.Previous hip fracture and/or geriatric services of each centre. The sample size calculation was based on the indicator 57.Prostatic benign hypertrophy ‘Prevalence of patients with PIP’. Taking into consider- 58.Schizophrenia ation the variability of the prevalence for older patients in 59.Sleep apnoea a published review,16 that ranged between 21% and 79% 60.Chronic thyroid disease of cases fulfilling any STOPP/START criteria, estimation 61.Tuberculosis was based on a prevalence of PIP of 50% (which would provide the highest number of patients to include), an 62.Urinary tract stones (nephritic colic) alpha risk of 5% and a power of 80%. Then, a minimum of 63.Varicose veins of lower extremities 400 patients would be required for estimation of the prev- 64.Vertigo alence of PIP in the whole sample. In order to increase the statistical power of the bivariate and multivariate anal- yses, and specifically the proposed cluster analysis, for implications in that hospitalisation (as, for example, any which there is no sample calculation formula, and taking drug intolerance or an excess drug poisoning). also into account the feasibility and possible difficulties Specific geriatric syndromes and risk factors (acute of each centre and the funding of the project, the initial confusional syndrome/delirium, chronic pain, cogni- sample size was increased by 100%. tive/intellectual impairment, constipation, depression According to the objectives of the study, the following or anxiety, dysphagia, frailty, immobility, incontinence steps of data analysis will be performed: (urinary/faecal), instability/falls, malnutrition, polyphar- The main descriptive indicators will be: macy, pressure ulcers, sensorial deficit, sleep disorders/ ►► Number and percentage of patients with PIP, both on insomnia) are also recorded, as usual. Two of the depart- admission and at discharge. ments systematically apply a recently developed scale for ►► Number and percentage of the STOPP-­START criteria frailty,32 while the others consider clinical judgement identified on admission and at discharge. (although based on the same variables). ►► Number and percentage of polymedicated patients c. Pharmacological: Number of chronic medications of (≥5 chronic active principles). the electronic prescribing at the time of admission ►► Number and percentage distribution of chronic and discharge, SS criteria detected on admission, ac- pathologies. tive principle involved, clinical decisions to modify ►► Number and percentage of geriatric syndromes and the prescription associated with the PIP detected, SS risk factors. criteria detected on discharge and the active princi- ►► Number and percentage of the chronic pathologies ple involved in the PIP detected. Medication is only that have motivated admission due to exacerbation. Baré M, et al. BMJ Open 2020;10:e033322. doi:10.1136/bmjopen-2019-033322 5 Open access ►► Number and percentage of patients with any ADR The dissemination plan includes publication in peer-­ detected on or during admission. reviewed journals of several disciplines such as internal b) The possible multimorbidity patterns will be iden- medicine, geriatrics or public health, as well as research tified using a cluster analysis algorithm, similar to that communications in some scientific conferences and mass used by Marengoni et al.34 With that intention, diseases or media. Other kind of dissemination activities to clini- syndromes with a very low prevalence will be excluded to cians, managers and policy makers will be done in the avoid statistical noise and therefore spurious findings in proper format. the cluster solutions. To characterise the multimorbidity patterns corresponding to each cluster of individuals, the Author affiliations 1 frequency of chronic diseases or syndromes in each cluster Department of Epidemiology and Cancer Screening, Consorci Corporació Sanitària Parc Taulí, Sabadell, Catalonia, Spain will be calculated. Observed/expected ratios will be calcu- 2 Department of Pediatrics, Obstetrics and Gynecology, Preventive Medicine and lated by dividing the prevalence of a given disease within Public Health, Faculty of Medicine, Universitat Autònoma de Barcelona, Bellaterra, a cluster by its prevalence in the overall population. The Catalonia, Spain 3 exclusivity of different diseases, defined as the fraction of Internal Medicine Department, Acute Care Geriatric Unit, Consorci Corporació participants with the disease included in the cluster over Sanitària Parc Taulí, Sabadell, Catalonia, Spain 4 Internal Medicine Department, Consorci Corporació Sanitària Parc Taulí, Sabadell, the total number of participants with the disease, will be Catalonia, Spain also calculated. 5 Pharmacy Department, Consorci Corporació Sanitària Parc Taulí, Sabadell, c) To evaluate the relationship between multimorbidity Catalonia, Spain 6 and PIP, and to analyse the possible association between PIP Pharmacy Department, Consorci Hospitalari de Vic, Vic, Catalonia, Spain 7 and ADRs, a bivariate analysis will be carried out, where the Geriatrics Department, Consorci Parc de Salut MAR de Barcelona, Barcelona, Catalonia, Spain variables will be treated quantitatively and/or categorically 8 Pharmacy Department, Hospital Galdakao-­Usansolo, Galdacano, País Vasco, Spain (number of morbidities, number of PIP or ADRs, patterns 9 Internal Medicine Department, Hospital Universitario de Canarias, La Laguna, of morbidity, existence of PIP or ADRs, types of PIP or Canarias, Spain ADRs). Parametric or non-­parametric tests will be applied according to the normal distribution of the PIP or ADRs Twitter Marisa Baré @BareMarisa number. In addition, ORs and their CI will be estimated Acknowledgements The authors acknowledge the dedication and support of the at 95% among multimorbidity patterns, existence of PIP entire morbidity, potentially inappropriate medication research group. as well as between PIP and ADRs detected at the time of Collaborators Morbidity, potentially inappropriate medication study group: Parc admission. Taulí University Hospital: Marisa Baré (Clinical Epidemiology and Cancer Screening; REDISSEC), Susana Herranz (Acute Care Geriatric Unit), Rosa Jordana (Department d) Finally, multiple regression and multilevel analysis of Internal Medicine), Maria Queralt Gorgas (Pharmacy Department, REDISSEC), techniques (where the levels would be the patient and Mònica Gómez and Sara Ortonobes (Pharmacy Department), Marina Lleal (Fundació the hospital) will be applied to adjust for other possible Parc Taulí, REDISSEC); University Hospital of Vic: Pere Roura (Epidemiology Unit; factors (such as age, sex, Barthel index or previous visits) REDISSEC), Daniel Sevilla, Núria Solà and Javier González (Pharmacy Department), Núria Molist, Mariona Espauella (Department of Geriatrics); Oscar Mascaró statistically significant in the bivariate analysis or those (Department of Internal Medicine); Hospital del Mar Medical Research Institute-­ clinically relevant, where the dependent variable, PIP IMIM: Elisabet de Jaime, Miguel Ángel Márquez, Marta Arellano, Carlos Clemente or ADRs, according to objective, may be dichotomous and Olga Sabartés (Department of Geriatrics), Olivia Ferrandez, Núria Carballo or quantitative. The adjusted β coefficients and their and Marta de Antonio (Pharmacy Department), Priscila Giraldo and Maria Sala (Department of Epidemiology and Evaluation, REDISSEC); Hospital de Galdakao: 95% CI, and the exponentials of β (OR) will be estimated Rafael Estrada (Department of Internal Medicine), Maria Olatz Ibarra (Pharmacy according to whether it is a multiple linear regression or Department); Complejo Hospitalario Universitario de Canarias: Candelaria Martin a generalised estimating equation model. (Department of Internal Medicine), Gloria Julia Nazco (Pharmacy Department), All analyses will be performed with R (R Foundation for Rubén Hernández (Department of Internal Medicine). Statistical Computing, Vienna, V.3.6.0) and SPSS (IBM Contributors MB, the principal investigator, conceived the study and wrote the Corp, V.25.0). First results are expected to be available by protocol, in collaboration with SH, RJ, and MQG. SO, DS, EDJ, OI and CM reviewed and made amendments to its preliminary version. They also contributed to the the end of 2019. design of the final questionnaires and were responsible of the organisation of the fieldwork in each centre. All of the members of the study group participated in the Patient and public involvement inclusion of patients, as well as in the collection and quality assessment of clinical Since this is an observational study with variables and data for each patient. outcomes related to the healthcare process, this research Funding This work was supported by grants from Instituto de Salud Carlos III-­ is being developed without patient involvement. Patients FEDER, [PI15/00552] and by the Network for Research into Healthcare in Chronic Diseases, REDISSEC (RD16/0001/0002). are not invited to comment on the study design and are not consulted to develop patient relevant outcomes or Competing interests None declared. interpret the results. Patient consent for publication Not required. Provenance and peer review Not commissioned; externally peer reviewed. Open access This is an open access article distributed in accordance with the Ethics and dissemination Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-­commercially, All the personal data will be anonymised, assigning a code and license their derivative works on different terms, provided the original work is to each patient, exclusive of the study, so that they no properly cited, appropriate credit is given, any changes made indicated, and the use longer relate to identifiable people. is non-­commercial. See: http://​creativecommons.​org/​licenses/​by-​nc/​4.​0/. 6 Baré M, et al. BMJ Open 2020;10:e033322. doi:10.1136/bmjopen-2019-033322 Open access ORCID iD 17 Patterson SM, Cadogan CA, Kerse N, et al. Interventions to improve Marisa Baré http://​orcid.​org/​0000-​0001-​8085-​2833 the appropriate use of polypharmacy for older people. Cochrane Database Syst Rev 2014:CD008165. 18 Motter FR, Fritzen JS, Hilmer SN, et al. Potentially inappropriate medication in the elderly: a systematic review of validated explicit criteria. Eur J Clin Pharmacol 2018;74:679–700. References 19 Gallagher P, Ryan C, Byrne S, et al. STOPP (screening tool of older 1 Palladino R, Pennino F, Finbarr M, et al. Multimorbidity and health Person’s prescriptions) and start (screening tool to alert doctors to outcomes in older adults in ten European health systems, 2006–15. right treatment). consensus validation. Int J Clin Pharmacol Ther Health Aff 2019;38:613–23. 2008;46:72–83. 2 The Academy of Medical Sciences. Multimorbidity: a priority for 20 O'Mahony D, O'Sullivan D, Byrne S, et al. STOPP/START criteria for global health research, 2015. Available: https://​acmedsci.​ac.​uk/​file-​ potentially inappropriate prescribing in older people: version 2. Age download/​82222577 Ageing 2014;44:213–8. 3 Zulman DM, Asch SM, Martins SB, et al. Quality of care for 21 Delgado Silveira E, Montero Errasquín B, Muñoz García M, et al. patients with multiple chronic conditions: the role of comorbidity Mejorando La prescripción de medicamentos en Las personas interrelatedness. J Gen Intern Med 2014;29:529–37. mayores: Una nueva edición de Los criterios STOPP-­START. Revista 4 Buffel du Vaure C, Dechartres A, Battin C, et al. Exclusion of patients Española de Geriatría y Gerontología 2015;50:89–96. with concomitant chronic conditions in ongoing randomised 22 García-­Gollarte F, Baleriola-­Júlvez J, Ferrero-­López I, et al. controlled trials targeting 10 common chronic conditions and Inappropriate drug prescription at nursing home admission. J Am registered at C ​ linicalTrials.​gov: a systematic review of registration Med Dir Assoc 2012;13:83.e9–83.e15. details. BMJ Open 2016;6:e012265. 23 Sevilla-­Sánchez D, Molist-­Brunet N, Amblàs-­Novellas J, et al. 5 Calderón-­Larrañaga A, Vetrano DL, Ferrucci L, et al. Multimorbidity and functional impairment-­bidirectional interplay, synergistic effects Potentially inappropriate medication at hospital admission in patients and common pathways. J Intern Med 2019;285:255–71. with palliative care needs. Int J Clin Pharm 2017;39:1018–30. 6 Violan C, Foguet-­Boreu Q, Flores-­Mateo G, et al. Prevalence, 24 San-­José A, Agustí A, Vidal X, et al. Inappropriate prescribing determinants and patterns of multimorbidity in primary care: to older patients admitted to hospital: a comparison of different a systematic review of observational studies. PLoS One tools of misprescribing and underprescribing. Eur J Intern Med 2014;9:e102149. 2014;25:710–6. 7 Fortin M, Stewart M, Poitras M-­E, et al. A systematic review of 25 Salgueiro E, Elizarde BC, Elola AI, et al. [The most common STOPP/ prevalence studies on multimorbidity: toward a more uniform START criteria in Spain. A review of the literature]. Rev Esp Geriatr methodology. Ann Fam Med 2012;10:142–51. Gerontol 2018;53:274–8. 8 Salisbury C, Johnson L, Purdy S, et al. Epidemiology and impact of 26 Moriarty F, Bennett K, Cahir C, et al. Potentially inappropriate multimorbidity in primary care: a retrospective cohort study. Br J Gen prescribing according to STOPP and START and adverse outcomes Pract 2011;61:e12–21. in community‐dwelling older people: a prospective cohort study. Br J 9 Poblador-­Plou B, van den Akker M, Vos R, et al. Similar Clin Pharmacol 2016;82:849–57. multimorbidity patterns in primary care patients from two European 27 Gibert P, Cabaret M, Moulis M, et al. Optimizing medication use regions: results of a factor analysis. PLoS One 2014;9:e100375. in elderly people in primary care: impact of STOPP criteria on 10 Prados-­Torres A, Calderón-­Larrañaga A, Hancco-­Saavedra J, et al. inappropriate prescriptions. Arch Gerontol Geriatr 2018;75:16–19. Multimorbidity patterns: a systematic review. J Clin Epidemiol 28 Thomas RE, Thomas BC. A systematic review of studies of the 2014;67:254–66. STOPP/START 2015 and American geriatric Society beers 2015 11 Gutiérrez-­Valencia M, Aldaz Herce P, Lacalle-­Fabo E, et al. criteria. Curr Aging Sci 2019;26:187–92. Prevalence of polypharmacy and associated factors in older adults 29 European Medicines Agency and Heads of Medicines Agencies. in Spain: data from the National health survey 2017. Medicina Clínica Guideline on good pharmacovigilance practices (GVP) module 2019;153:141–50. VI – management and reporting of adverse reactions to medicinal 12 Masnoon N, Shakib S, Kalisch-­Ellett L, et al. What is polypharmacy? products (Rev 1, 2014. A systematic review of definitions. BMC Geriatr 2017;17. 30 Mascolo A, Scavone C, Sessa M, et al. Can causality assessment 13 Cabré M, Elias L, Garcia M, et al. Avoidable hospitalizations due to fulfill the new European definition of adverse drug reaction? A adverse drug reactions in an acute geriatric unit. Analysis of 3,292 review of methods used in spontaneous reporting. Pharmacol Res patients. Medicina Clínica 2018;150:209–14. 2017;123:122–9. 14 Fernández-­Regueiro R, Fonseca-­Aizpuru E, López-­Colina G, et al. 31 Charlson ME, Pompei P, Ales KL, et al. A new method of classifying Prescripción inadecuada Y efectos adversos a medicamentos en prognostic comorbidity in longitudinal studies: development and pacientes de edad avanzada. Rev Clin Esp 2011;221:400–6. validation. J Chronic Dis 1987;40:373–83. 15 Hamilton H, Gallagher P, Ryan C, et al. Potentially inappropriate 32 Amblàs-­Novellas J, Martori JC, Espaulella J, et al. Frail-­VIG index: medications defined by STOPP criteria and the risk of adverse a concise frailty evaluation tool for rapid geriatric assessment. BMC drug events in older hospitalized patients. Arch Intern Med Geriatr 2018;18:29. 2011;171:1013–9. 33 Wills S, Brown D. A proposed new means of classifying adverse 16 Hill-­Taylor B, Sketris I, Hayden J, et al. Application of the reactions to medicines. Pharm J 1999;262:163–5. STOPP/START criteria: a systematic review of the prevalence of 34 Marengoni A, Roso-­Llorach A, Vetrano DL, et al. Patterns of potentially inappropriate prescribing in older adults, and evidence multimorbidity in a population-­based cohort of older people: of clinical, humanistic and economic impact. J Clin Pharm Ther sociodemographic, lifestyle, clinical, and functional differences. J 2013;38:360–72. Gerontol A Biol Sci Med Sci 2019;72. Baré M, et al. BMJ Open 2020;10:e033322. doi:10.1136/bmjopen-2019-033322 7