Search

Polypharmacy and inappropriate medication use are very common in multimorbid older patients. This population has unfortunately been excluded from most large, randomized studies. In a recent multicenter randomized study (OPERAM), we included over 2000 multimorbid patients. We found that 86% of the patients aged 70 years and more had inappropriate medications and that these medications could be discontinued without negative impact on the health of these patients. This cohort of multimorbid patients will be followed for 10 years to evaluate their prognosis, life expectancy, treatments and quality of life, with numerous projects to better understand the inappropriate prescribing of individual drugs and their consequences on the health of this population.La polypharmacie et les médicaments inappropriés sont très fréquents chez les patients âgés multimorbides. Cette population a malheureusement été exclue de la plupart des grandes études randomisées. Dans une récente étude randomisée multicentrique (OPERAM), nous avons inclus plus de 2000 patients multimorbides. Celle-ci a montré que 86 % des patients âgés de 70 ans et plus avaient des médicaments inappropriés et qu’il était possible de stopper leur administration, sans répercussion négative sur leur santé. Ces patients multimorbides constituent une cohorte qui va être suivie sur 10 ans pour évaluer leurs pronostic, espérance de vie, traitements et qualité de vie. Cela permettra la réalisation de nombreux projets, notamment pour mieux comprendre les conséquences de la prescription inappropriée de médicaments.

International audience; Introduction: We aimed at identifying the associations of health issues managed in general practice encounters with patients aged 65 years or over, according to their sex. Methods: Ancillary study of the ECOGEN study on the consultation practice of French general practitioners. We have studied the 15 most frequent health issues managed in patients aged 65 years or over and identified the pairs most frequently and strongly associated. Results: The sample included 5,782 encounters corresponding to 17,182 health issues managed. Each encounter involved an average of 3.0 [2.9-3.0] health issues. The strongest associations mainly included cardio-metabolic conditions. The most frequent and strong pair (OR: 3.8 [3.3-4.4]) associated non-complicated arterial hypertension with lipid disorder. Some associations were specific to patient sex. Conclusion: The remarkable associations identified may enable to improve the management of elderly patients with multimorbidity by favouring synergic treatments.

The accurate characterization of health-state severity has been a central concern in most health-related research or policy fields. A common approach to take into account health-state severity is the use of summary measures, usually referred to as indices or scores, based on predictive modeling methods. With the increasing availability of routinely collected standardized data in medico-administrative databases, such summary measures are more widely used for research purposes or to inform policy making. The French national health data system (SNDS) is particularly adapted to predictive modeling as it allows morbidity measurement, longitudinal follow-up and the study of health-related outcomes. We propose two morbidity indices developed using SNDS data and illustrate their utilization in two application studies.We developed and validated two outcome-specific morbidity indices in a nationwide population of people aged 65 years or older: the Mortality-Related Morbidity Index (MRMI) predictive of 2-year mortality and the Expenditure-Related Morbidity Index (ERMI) predictive of healthcare expenditure over 2 years and reflecting the intensity of healthcare utilization. The MRMI and ERMI indices have better overall performance and better calibration than comparable Charlson indices.As an application of the morbidity indices to case-identification, we studied the risk of readmission in heart failure (HF) patients. To analyze readmission predictors, we distinguished HF severity from overall morbidity measured through the MRMI and ERMI indices and took into account the competing mortality risk. Risk-groups defined upon HF severity and overall morbidity, available at admission, have a 40% separation in HF readmission incidence and specific patterns of risk over the 1-year follow-up period.As an application to payment models, we compared the performance of the MRMI and ERMI indices, among other predictors, for expenditure prediction. Models were applied to prediction of overall, inpatient and outpatient individual expenditure, for people aged 65 or older and for HF patients aged 65 or older. To illustrate the use of predictive models for prospective resource allocation, we compared predicted with observed mean individual expenditure at the area of residence level. Health-state severity measured through the MRMI and ERMI was the most important predictor, both at the individual and area of residence level.The MRMI and ERMI indices are performant tools for outcome-specific severity adjustment in studies using SNDS data. They can be used as stratification or adjustment variables, or among other predictors in predictive models.; La mesure pronostique de l’état de santé constitue un enjeu important dans de nombreux domaines liés aux soins, à la recherche ou à la décision publique en santé. Une approche commune à ces différents domaines est de synthétiser l’information disponible sur la morbidité d’une population sous la forme d’indices, appelés parfois scores, en faisant appel à des méthodes de modélisation prédictive. La disponibilité croissante de données médico-administratives et l’essor de leur utilisation à des fins de recherche ou d’aide à la décision ont souligné l’importance de ce type de mesures pronostiques. En France, les données du SNDS et en particulier la « cartographie des pathologies et des dépenses » développée par l’Assurance Maladie, permettent la mesure de la morbidité, un suivi individuel longitudinal sur plusieurs années et l’étude de différents résultats de santé. Nous proposons deux indices de morbidité élaborés en appliquant des méthodes de modélisation prédictive aux données médico-administratives françaises et nous illustrons leur apport par deux études d’application.Les deux indices proposés sont élaborés et validés en appliquant un cadre méthodologique commun à une population nationale de personnes âgées de 65 ans ou plus. L’indice MRMI (Mortality-Related Morbidity Index) est prédictif de la mortalité à deux ans et l’indice ERMI (Expenditure-Related Morbidity Index) est prédictif des dépenses de soins remboursées sur deux ans et reflète l’intensité du recours au système de santé. Leur performance prédictive est supérieure aux indices comparables les plus communément utilisés, indices de Charlson et mesures d’Elixhauser.Dans une première étude d’application, nous étudions le risque de réhospitalisation pour les patients atteints d’insuffisance cardiaque (IC), en utilisant des méthodes adaptées à la prise en compte du risque compétitif de décès. Nous distinguons la stabilité de l’IC de la sévérité globale de l’état de santé, mesurée à travers les deux indices proposés. Ces deux informations, disponibles à l’admission d’un séjour pour IC, permettent de segmenter la population en groupes de risque avec un écart de 40% d’incidence cumulée de réhospitalisation pour IC au bout d’un an de suivi.Pour la deuxième étude d’application, nous comparons différents modèles prédictifs afin de quantifier l’apport des indices de morbidité dans la prédiction des dépenses individuelles. Nous étudions trois périmètres de dépenses : totales, hospitalières et ambulatoires, parmi deux populations différentes : l’ensemble des personnes âgées de 65 ans ou plus et les personnes âgées de 65 ans ou plus et atteintes d’IC. Nous illustrons les enjeux de la définition de paiements populationnels prospectifs de type capitation, en comparant les dépenses observées aux dépenses prédites par ces modèles à l’échelle des départements métropolitains. La sévérité de l’état de santé, mesurée à travers les deux indices proposés, est le déterminant le plus important de la performance prédictive des dépenses, aussi bien au niveau individuel que départemental.Les indices MRMI et ERMI sont des outils performants pour prendre en compte la sévérité de l’état de santé dans les travaux basés sur des données du SNDS et de manière adaptée au résultat étudié. Ils peuvent servir de variables de stratification ou d’ajustement, ou être inclus parmi d’autres variables dans des modèles prédictifs.

The aim of our work was to appreciate the importance of comorbidities of heart failure individually and globally in patients hospitalized at the Cardiology Institute of Abidjan.This was a prospective cohort study of adult heart failure patients hospitalized from January to December 2015, and followed up over 12 months. Co-morbidities were analysed through their prevalence, their relationship with the etiologies, and their impact on the prognosis.Three hundred and two patients (mean age: 55.5±16.9 years, 61.6 % male) were recruited. High blood pressure, anaemia and kidney dysfunction were the most common co-morbidities (48 %, 43.7 % and 41.3 % respectively). There was an average of 3.4±1.8 comorbidities per patient with an increase in the number of comorbidities with age (P0.05) and a more frequent association with hypertensive and ischemic heart disease (P0.001). During the one-year follow-up, 96 patients died. Apart from hepatic dysfunction (RR=1.97, 95 % CI [1,19-3.25], P=0.008, a high score of Charlson index appeared as a risk factor of death as much in univariate analysis (RR=4.15 95 % CI [2.32-7.41], P0.001), as in multivariate analysis according to the Cox model (RR=2.48. 95 % CI [1.08-5.09], P=0.03) confirmed by Kaplan Meier curves (P0.001).Comorbidities are common in our heart failure patients and significantly affect their prognosis.

We have been studying multimorbidity in epidemiology and clinical practice since the 1970s. Growing experience shows that the concept of multimorbidity is helpful for epidemiologists to describe population-level changes, but of little use to patients and clinicians who 'diagnose' multimorbidity. Several large, rigorous clinical trials show that organizational changes in primary care can improve the care of these complex patients, but do not improve directly health-related quality of life. This article describes the history of multimorbidity and results of recent randomized trials to prompt a discussion of new approaches to improve the experience of people living with multiple chronic diseases.Depuis plusieurs décennies, la multimorbidité fait l’objet d’études, non seulement en épidémiologie, mais également en médecine générale. Il en ressort que la multimorbidité est essentiellement une entité épidémiologique et les récentes études publiées dans ce domaine tendent à montrer qu’il est difficile d’en faire un concept ayant une réelle consistance clinique. En revanche, la littérature nous apprend également qu’il est essentiel de modifier l’organisation des structures de soins afin d’améliorer la prise en charge de ces patient·e·s particulièrement complexes. Cet article se veut une réflexion autour du concept de multimorbidité dans une perspective historique, ainsi que les besoins de transformation permettant la prise en charge de ces patient·e·s.

Thèse présentée sous la forme d'une "thèse article"; Importance: The postoperative patient care in neurosurgery has grown in recent decades due to an aging population and chronic diseases. In some countries, the co-management responds to this demand. Hospitalists are assuming an increasing role in the postoperative care of surgical patients but the impact of the model on postoperative outcomes is not clearly assessed.Objective: Which role can play the hospitalist in optimizing the medical postoperative patient care in a neurosurgery department?Design: Prospective and comparative six-month cohort studySetting: Academic medical center (La Timone, Marseille, France), adult neurosurgery department of Professor Dufour, from September 1, 2017 to February 28, 2018Participants: 234 patients divided into 2 branches over 2 months in 2 different units of the department (Unit A or Unit B), according to the allocation of the hospitalist (in September within unit A, and in October within unit B).Measurements: Inpatient postoperative complications and complications in the mid-term (re-admissions), length of stay, medical consultations.Intervention: A co-management hospitalist-neurosurgeon team in postoperative care compared with standard postoperative care by neurosurgeons with medical consultation.Material and methods: All cranial or spine cases operated or not in the framework of an urgent or regulated surgery were included and followed over a period of 1 to 4 months.Results: The patients were comparable in number, age, gender, reasons for hospitalization, means of admission and scales of multimorbidity (Asa and Charlson).The average length of stay was statistically reduced by 10.13 days (standard deviation 7.76, CI 8.77 to 11.49) to 7.07 days (standard deviation 3.94, CI 6.31 to 7.82, p = 0.0087), without and with the presence of a hospitalist. The medical consultation rate was also statistically reduced. The one-third decrease in medical complications, was not statistically significant. However, there was a significant association between age, multimorbidity and the occurrence of inpatient complications.Conclusion: The co-management hospitalist-neurosurgeon reduced medical postoperative consultations and the length of stay with no statistically difference in complication rates. Additional research on the job description as well as the economic impact of this hospitalist model is warranted.; Introduction : la prise en charge (PEC) des patients en post-opératoire en service de neurochirurgie ne fait l’objet d’aucun consensus en terme d’intervenant et de responsabilité associée. Or cette PEC s’est alourdie lors des dernières décennies suite au vieillissement de la population et à l’émergence de pathologies chroniques, complexifiant la gestion de patients de plus en plus âgés et/ou porteurs de plusieurs pathologies. L’émergence du co-management entre chirurgiens et médecins, notamment dans les pays anglo-saxons ainsi que dans le secteur privé, pour les services de chirurgie orthopédique, vasculaire et de neurochirurgie ainsi que la demande récurrente de renfort médical dans ces services nous amène à l’objet de cette étude : quelle optimisation est possible dans la PEC médicale des patients, en péri-opératoire, par le médecin généraliste (MG) dans un service de neurochirurgie ? Nous nous sommes intéressés au détail des complications post-opératoires, chirurgicales ou médicales, au nombre de demandes d’avis spécialisés ainsi qu’à la durée moyenne de séjour (DMS), en fonction de la présence, ou non du MG.Matériel et méthodes : nous avons conduit une étude observationnelle prospective et comparative de 6 mois au sein du service de neurochirurgie adultes du Pr Dufour au CHU de La Timone à Marseille du 01 septembre 2017 au 28 février 2018. Tous les cas neurochirurgicaux crâniens ou rachidiens, opérés ou non, dans le cadre d’une intervention urgente ou réglée ont été inclus et suivis sur une durée de 1 à 4 mois. Les patients hospitalisés en neurochirurgie fonctionnelle, neurochirurgie pédiatrique ainsi que les patients de neurochirurgie hébergés dans d’autres services n’ont pas été inclus dans cette étude. Nous avons constitué une cohorte de 234 patients, répartis en 2 bras sur 2 mois dans les unités 3 et 4 du service, selon l’affectation du MG (septembre, unité 4 et octobre, unité 3). Nous les avons ensuite suivis jusqu’à leur consultation de contrôle.Résultats : la DMS a statistiquement diminué, de 10,13 jours à 7,07 jours (p < à 0,05), sans puis avec la présence d’un MG, pour des populations comparables en nombre, âge, sexe, motifs d’hospitalisation, modes d’entrée (intervention urgente ou programmée) et scores de multimorbidité (Asa et Charlson). La diminution du nombre de demandes d’avis spécialisés liée à l’intervention du MG, selon les unités concernées, est aussi statistiquement significative, représentant 14,29% et 8% en présence du MG versus 22,81% et 19,15% en son absence. La baisse de 30% des complications médicales, qui représentent 45% de l’ensemble des complications, n’apparaît pas statistiquement significative. L’observation des complications confirme cependant une association significative entre l’âge, la multimorbidité (selon le score de Charlson) et la survenue de celles-ci en post-opératoire immédiat mais pas à distance de l’opération.Discussion : la PEC des patients en post-opératoire en service de neurochirurgie peut être optimisée par la présence d’un MG, grâce à la réduction du nombre de complications, de demandes d’avis et de la DMS. Le profil, le temps et l’organisation du travail ainsi que le financement de celui-ci restent à définir. L’étude d’autres organisations, en complément du benchmarking réalisé, permettra de trouver un nouvel équilibre économique, entre réduction des délais et renforcement de l’équipe médico-chirurgicale d’une part, amélioration de la cotation en fonction des multimorbidités et optimisation de la qualité des soins d’autre part.

En réponse aux défis de la fragmentation des soins pour les personnes atteintes de maladies chroniques, l’état fédéral belge a proposé le projet Integreo, un plan conjoint en faveur des maladies chroniques. Il s’agit d’une quête à la poursuite du « triple/quadruple aim ». Après l’exposé de la genèse et du contenu du plan Integreo, nous présenterons le devis de recherche mixte. L’exposé finira par l’introduction d’une discussion sur l’ambition de vouloir inclure les acteurs des différents niveaux de pouvoir aux différentes étapes de la recherche, et sur la conciliation des agendas de recherche et politique, tout en gardant une indépendance intellectuelle.

International audience; Introduction: The high prevalence of chronic diseases is accompanied by an underestimation of multimorbidity. This reality is also found in primary care. The therapeutic patient education (TPE) program for multimorbidity, based on a model developed by the Education and Health Practice Laboratory (LEPS, EA 3412), has been proposed by the Mutualité Sociale Agricole (MSA), the french health insurance body for farmers, to patients with several chronic diseases attending multi-professional primary care centers (MSPs). It was tested in 2015–2016.Objectives: To design, validate and experiment an operational model of TPE for multimorbidity patients with cardiovascular disease, diabetes, and/or chronic obstructive pulmonary disease, educated in two MSPs. Method: Assessment of patients by questionnaire (before and immediately after education) focused on their knowledge of the causes and means for managing their illnesses, their feelings of trust, mastery and efficiency. Six months after the TPE program, patients were interviewed on the competencies they had acquired.Results: 47 patients were evaluated (35 patients had 2 pathologies, 11 had 3 and 1 had 4). After education, an increase in their useful knowledge has been observed: explanation of the diseases and their causes (T1: 38,5%; T2: 88,4%, p

Inappropriate use of medication which occurs from the care of the persons with multimorbidity is recognized as a public health issue for the northern countries. This is stressed by the ageing of their population. In sub-Saharan Africa and more specifically in Burkina Faso, the care of the people with multimorbidity, in particular at the older adults, still seems little taken into account in the health care systems organization. This system is still health problems-centered and acute diseases- centered. The aim of this thesis was to analysis the health care system by a tracer «Inappropriate use of medication in older adults with multimorbidity». The three questions of research were 1) what are the characteristics of the older adults with multimorbidity which could influence the inappropriate use of medication? 2) What are the characteristics of the health care system which influence the care of older adults with multimorbidity? 3) How the characteristics of health care system influence the care of older adults with multimorbidity? The research was carried out at Bobo-Dioulasso in Burkina Faso. We conducted an analysis based-system centered on the care of older adults with multimorbidity. For their care, the older adults are in interaction with the care providers, the organization of health facilities and the social-family network. This system was analyzed through different methods using various sources of data. This analysis had for starting point the older adults at home. This first step was carried out in 2012. This first step consisted of a household survey to understand health problems (chronic diseases) and the problems of the older adults care (potentially inappropriate medication). In a second step (2013-2014), we carried out, in a logical way, sub-studies in the health facilities 1) the health care system dysfunctions, 2) the factors of heath care organization associated with the inappropriate use of medication according to the older adult’s perspective and 3) the reasons of the inappropriate use of medication during the care of older adults. Our sample was constituted by older adults with a mean age of 69 years. Among the older adults, 55.3% were men, 54.1% illiteracy 58.7%, married, 58% without retirement incomes and 59.1% having already used a health facility during the last six months. We found older adults with polypharmacy (39%), multimorbidity (65 %), having used at least one inappropriate medication (59%), having used the formal and informal care (self-medication by informal market, the traditional medication) for the same episode of disease. These older adults were affected by the following main chronic diseases: arterial hypertension (82.3%), malnutrition (39.2%), visual impairment (28.3%), and diabetes mellitus (26.5%). About their care, our research showed an acute management of the chronic diseases, a lack of coordination with others care providers in the private and traditional sector, a weak implementation of the activities of information, educational and communication to strengthen patients empowerment, a weakness in the in-house training of the care providers, an insufficiency of the teamwork, a lack of standards and protocols for the care of the older adults with multimorbidity. The main prescribers of inappropriate medication were the paramedical in the first line. The characteristics of care providers who can contribute to explain the use of the inappropriate medication are: their missing work, their initial training, the lack of their work follow-up (assessment), their relation with the older adult during the consultation (paternalist), their low feeling of non-accountability to the patients and to their peers. To improve the use of inappropriate medication at the older adults with multimorbidity, we propose track at various levels and by priority, which should bring a profit beyond this target group. At macro level, the current dynamic in Burkina Faso about the development of an universal health insurance (UHI) and as well as the discussions about the financing based-result (FBR) could be taken as opportunities. In the first place, the identification of a basket of care for the UHI is an opportunity to include the necessary services for the rational care of the older adults with multimorbidity. Secondly, the inclusion of the private care providers in the system of UHI could allow the improvement of the health care system compartmentalization. Thirdly, the opportunity offered by the UHI and the FBR to include private and public care providers in the care is a track to introduce a process of accreditation for these providers in the rational prescription. Fourthly, to support this rational care, the implementation of consensus experts' committee could allow the producing of pharmaceutical and therapeutic directives centered on the care of the older adults with multimorbidity. This committee can be constituted by the experts of the learned societies and by the general practitioners of hospitals. At meso level: in the first place, a committee of regional experts constituted by specialists and general practitioners of hospitals can design therapeutic and pharmaceutical protocols and insure their application for the care of older adults with multimorbidity. This committee offers an opportunity to work on the conception of a medical record in hospitals. Secondly, the implementation of this committee is an opportunity to ensure the in-house training of the care providers. This in-house training could join the logic of the accreditation of the care providers in the rational prescription. At micro level: in the first place, the implementation of the pharmaceutical and therapeutic committees in the district hospitals could allow a rational prescription for the older adults with multimorbidity. These committees can consist of doctors, pharmacists, nurses and of social workers. Secondly, these committees can promote the use of the therapeutic and pharmaceutical protocols, the test of a medical shared tool (medical record). Thirdly these committees can be an opportunity to develop a local multidisciplinary team of care, to think about forms of information, educational and communication to strengthen patient’s empowerment and to define the role of a home-care coordinator. (SP - Sciences de la santé publique) -- UCL, 2015

Contexte : Les médecins spécialistes peuvent participer aux soins ambulatoires des personnes atteintes de maladies chroniques (MCs) et comorbidité comme co-gestionnaire ou consultant selon qu’ils sont responsables ou non du suivi du patient. Il y a un manque d’évidences sur les déterminants et l’impact du type d’implication du médecin spécialiste, ainsi que sur la façon optimale de mesurer la comorbidité pour recueillir ces évidences. Objectifs : 1) déterminer chez les patients atteints de MCs les facteurs associés à la cogestion en spécialité, dont les caractéristiques des organisations de première ligne et la comorbidité; 2) évaluer si le type d’implication du spécialiste influence le recours à l’urgence; 3) identifier et critiquer les méthodes de sélection d’un indice de comorbidité pour la recherche sur l’implication des spécialistes dans le suivi des patients. Méthodologie : 709 adultes (65 +/- 11 ans) atteints de diabète, d’arthrite, de maladie pulmonaire obstructive chronique ou d’insuffisance cardiaque furent recrutés dans 33 cliniques de première ligne. Des enquêtes standardisées ont permis de mesurer les caractéristiques des patients (sociodémographiques, comorbidité et qualité de vie) et des cliniques (modèle, ressources). L’utilisation des services de spécialistes et de l’urgence fut mesurée avec une base de données médico-administratives. Des régressions logistiques multivariées furent utilisées pour modéliser les variables associées à la cogestion et comparer le recours à l’urgence selon le type d’implication du spécialiste. Une revue systématique des études sur l’utilisation des services de spécialistes, ainsi que des revues sur les indices de comorbidité fut réalisée pour identifier les méthodes de sélection d’un indice de comorbidité utilisées et recommandées. Résultats : Le tiers des sujets a utilisé les services de spécialistes, dont 62% pour de la cogestion. La cogestion était associée avec une augmentation de la gravité de la maladie, du niveau d’éducation et du revenu. La cogestion diminuait avec l’âge et la réception de soins dans les cliniques avec infirmière ayant un rôle innovateur. Le recours à l’urgence n’était pas influencé par l’implication du spécialiste, en tant que co-gestionnaire (OR ajusté = 1.06, 95%CI = 0.61-1.85) ou consultant (OR ajusté = 0.97, 95%CI = 0.63-1.50). Le nombre de comorbidités n’était pas associé avec la cogestion, ni l’impact du spécialiste sur le recours à l’urgence. Les revues systématiques ont révélé qu’il n’y avait pas standardisation des procédures recommandées pour sélectionner un indice de comorbidité, mais que 10 critères concernant principalement la justesse et l’applicabilité des instruments de mesure pouvaient être utilisés. Les études sur l’utilisation des services de spécialistes utilisent majoritairement l’indice de Charlson, mais n’en expliquent pas les raisons. Conclusion : L’implication du spécialiste dans le suivi des patients atteints de MCs et de comorbidité pourrait se faire essentiellement à titre de consultant plutôt que de co-gestionnaire. Les organisations avec infirmières ayant un rôle innovateur pourraient réduire le besoin pour la cogestion en spécialité. Une méthode structurée, basée sur des critères standardisés devrait être utilisée pour sélectionner l’indice de comorbidité le plus approprié en recherche sur les services de spécialistes. Les indices incluant la gravité des comorbidités seraient les plus pertinents à utiliser., Background: Medical specialist physicians can be involved either as comanagers (responsible for follow-up of patients) or consultants (provide advice/specialized interventions) in the care of patients with chronic diseases (CDs) managed in a primary health care (PHC) setting. Evidences concerning determinants and impact of type of specialist involvement are currently lacking, in particular the influence of comorbidity and how best to measure this factor. Objectives: The objectives were 1) to determine clinical, patient and PHC organizational characteristics associated with type of specialist involvement in patients with CDs; 2) to assess whether type of specialist involvement is associated with emergency department (ED) use and; 3) to identify methods for selecting a comorbidity index for specialist services research. Methods: 709 adults (65 +/- 11 years) with diabetes, heart failure, arthritis, or chronic obstructive pulmonary disease were recruited in 33 PHC practices. Standardized surveys were used to measure patient (gender, age, education, income, comorbidity, quality of life) and practice characteristics (model, remuneration mode, resources, role of nurse). Information on specialist services and ED use was procured from the Quebec physician claims database. We used multivariate logistic regression to 1) model variables associated with being comanaged and 2) compare ED use among persons with different types of specialist involvement. We conducted two systematic reviews: 1) review articles on comorbidity indices to identify proposed selection procedures and 2) studies on specialist services utilization to identify selection processes actually used. Results: One third of our sample saw a specialist; the majority (62%) was as a comanager. Comanagement was associated with higher disease severity, younger age, higher education level and income and primary care management in practices without a nurse in advanced practice role. There was no difference in rates of ED use over one year between patients with or without specialist involvement, either as a comanager (adjusted OR = 1.06, 95%CI = 0.61-1.85) or as a consultant (adjusted OR = 0.97, 95%CI = 0.63-1.50). Quantity of comorbidity was not associated with either comanagement or impact of specialist involvement on ED use. Our systematic review revealed no standardized selection process of a comorbidity index. However, 10 distinct criteria related to accuracy and applicability of a measurement scale or validity of reported studies were compiled. Studies on specialist services utilization mostly used the Charlson comorbidity index, but none justified their choice. Conclusion: Specialist support in the management of patients with CDs and comorbidity should be provided on a consultant basis. The PHC practice model with a nurse in an advanced practice role may reduce the need for specialist comanagement. When adjusting for comorbidity, researchers should use a structured process to select the appropriate index based on standard criteria such as validity and applicability. Indices considering severity of comorbidities may be more useful than sole disease count in specialist services research.