Objetivo: A segurança no uso de medicamentos tem constituído importante indicador na assistência prestada em nível hospitalar. Para tanto, este estudo propõe avaliar o impacto clínico e econômico da implantação de um programa de farmacovigilância (PFV) intensiva em um hospital geral, através do monitoramento de reações adversas à vancomicina. Delineamento: Coorte. Amostra: 689 pacientes constituíram a amostra, sendo 151 acompanhados antes do PFV e 538 depois PFV. Principais Medidas: Ocorrência de reação adversa a medicamento (RAM) baseado no algoritmo de Naranjo e estimativa de custos hospitalares. Resultados: Houve redução na ocorrência de RAM após a implementação do PFV (19,9% vs. 13,4%; P < 0,05). O PFV foi um fator independente relacionado a uma diminuição da RAM, especialmente naqueles pacientes em uso de vancomicina inferior a 12 dias e em pacientes pediátricos, com um NNT de 15. Não houve diferença nos custos antes e após o programa, entretanto aqueles pacientes com RAM apresentaram um custo adicional de R$ 12.676,47, sendo que a atividade de vigilância representou um custo estimado de R$ 19,23 por caso acompanhado. Conclusão: Um PFV implementado ativamente, com busca e monitoramento de RAM, e com estratégias pró-ativas para minimizá-las, tem impacto positivo nos resultados clínicos e econômicos hospitalares. / Proper drug safety monitoring is a central aspect and an important quality indicator of hospital care. This study was carried to evaluate the clinical and economic impact of the implementation of an Intensive Phamacovigilance Program (IPP) in a general hospital. We selected the occurrence of Adverse Drug Reactions (ADR) related to the antimicrobial drug vancomycin to evaluate the program. Study design: Cohort. Methods: 689 patients were included, 151 of them were followed before the implementation of the IPP and 538 patients were followed after this. Main determinants: The occurrence of ADR was measured and classified based in the ‘Naranjo’ algorithm. All the hospital costs were estimated using the register of the financial department of the hospital. Results: There was a significant reduction in the ADR rates after the implementation of the IPP (19.9% versus 13.4%; P-value<0.05). The IPP was an independent factor for the reduction of the ADR rates; it was more important in those that used vancomycin for a period of time ≤ 12 days, and in pediatric patients, in this case the number needed to treat (NNT) estimated was 15. There was no difference between the costs comparing the period before and after the IPP implementation, but those patients that has ADR represented an additional cost of R$ 12,676.47. The cost of the IPP program was R$ 19.23 per individual case followed. Conclusion: The implementation of an IPP that monitored ADR and the strategies that were simultaneously developed to minimize them, had a positive impact in the clinical and economic aspects analized.

Objetivo: A segurança no uso de medicamentos tem constituído importante indicador na assistência prestada em nível hospitalar. Para tanto, este estudo propõe avaliar o impacto clínico e econômico da implantação de um programa de farmacovigilância (PFV) intensiva em um hospital geral, através do monitoramento de reações adversas à vancomicina. Delineamento: Coorte. Amostra: 689 pacientes constituíram a amostra, sendo 151 acompanhados antes do PFV e 538 depois PFV. Principais Medidas: Ocorrência de reação adversa a medicamento (RAM) baseado no algoritmo de Naranjo e estimativa de custos hospitalares. Resultados: Houve redução na ocorrência de RAM após a implementação do PFV (19,9% vs. 13,4%; P < 0,05). O PFV foi um fator independente relacionado a uma diminuição da RAM, especialmente naqueles pacientes em uso de vancomicina inferior a 12 dias e em pacientes pediátricos, com um NNT de 15. Não houve diferença nos custos antes e após o programa, entretanto aqueles pacientes com RAM apresentaram um custo adicional de R$ 12.676,47, sendo que a atividade de vigilância representou um custo estimado de R$ 19,23 por caso acompanhado. Conclusão: Um PFV implementado ativamente, com busca e monitoramento de RAM, e com estratégias pró-ativas para minimizá-las, tem impacto positivo nos resultados clínicos e econômicos hospitalares. / Proper drug safety monitoring is a central aspect and an important quality indicator of hospital care. This study was carried to evaluate the clinical and economic impact of the implementation of an Intensive Phamacovigilance Program (IPP) in a general hospital. We selected the occurrence of Adverse Drug Reactions (ADR) related to the antimicrobial drug vancomycin to evaluate the program. Study design: Cohort. Methods: 689 patients were included, 151 of them were followed before the implementation of the IPP and 538 patients were followed after this. Main determinants: The occurrence of ADR was measured and classified based in the ‘Naranjo’ algorithm. All the hospital costs were estimated using the register of the financial department of the hospital. Results: There was a significant reduction in the ADR rates after the implementation of the IPP (19.9% versus 13.4%; P-value<0.05). The IPP was an independent factor for the reduction of the ADR rates; it was more important in those that used vancomycin for a period of time ≤ 12 days, and in pediatric patients, in this case the number needed to treat (NNT) estimated was 15. There was no difference between the costs comparing the period before and after the IPP implementation, but those patients that has ADR represented an additional cost of R$ 12,676.47. The cost of the IPP program was R$ 19.23 per individual case followed. Conclusion: The implementation of an IPP that monitored ADR and the strategies that were simultaneously developed to minimize them, had a positive impact in the clinical and economic aspects analized.

Os eventos adversos, devido à grande ocorrência, são considerados um importante problema de saúde pública além de elevarem os custos com a saúde. Dessa forma, existe a necessidade de um sistema eficiente de farmacovigilância para prevení-los sendo a monitorização intensiva uma excelente estratégia para se alcançar este objetivo, pois permite a identificação e o conhecimento dos eventos adversos, além de medir sua incidência e melhorar as notificações voluntárias. Por esta razão, o presente estudo propôs a avaliação deste método em eventos adversos relacionados ao antimicrobiano vancomicina. As informações coletadas, provenientes dos bancos de dados e dos prontuários do Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto (HCFMRP-USP), são referentes aos pacientes internados no hospital em tratamento com vancomicina no período de 01/09/2008 a 31/08/2009. Identificou-se 1719 tratamentos de 1302 usuários. Destes tratamentos, 521 foram excluídos por serem profiláticos. A idade média dos 1198 pacientes monitorados foi de 39,33 (± 27,66) anos e 58,68% pertenciam ao gênero masculino. Os diagnósticos infecciosos prevalentes para o tratamento com vancomicina foram pneumonia (18,95%), sepse (11,19%) e infecção relacionada a cateter (10,27%) com período médio de duração do tratamento de 11,47 (±7,79) dias geralmente na dose de 2000 mg ao dia (35,98%). Foram identificados indicadores de eventos adversos em 85,81% dos tratamentos, sendo a incidência de eventos adversos de 7,18%. A incidência de reações adversas (RAM) foi de 5,93% e de erros de medicação 2,57%. A causalidade das RAM foi prevalentemente provável (49,29%) e a gravidade moderada (54,93%). As RAM foram essencialmente do tipo B (81,69%) e os erros de medicação relativos à prescrição (87,10%). A comparação entre os grupos controle e estudo não apresentou diferenças significativas, demonstrando a necessidade de monitorização de todos os parâmetros relacionados a eventos adversos com a vancomicina. Dessa forma, fica evidente a importância da monitorização intensiva e da presença dos farmacêuticos clínicos na equipe de saúde para melhorar a eficiência da farmacovigilância, como demonstrado por este estudo. Diante dos resultados, conclui-se que o método de monitorização intensiva utilizando base de dados e prontuários, além da participação de um farmacêutico clínico, promove o conhecimento, a avaliação e a compreensão dos eventos adversos presentes em uma instituição, permitindo a esquematização de estratégias para a sua prevenção. / Adverse events, due to the high occurrence, are considered an important public health problem in addition to driving up health care costs. There is therefore a need for an efficient system of pharmacovigilance to prevent them being the intensive monitoring an excellent strategy for achieving this goal because it allows the identification and knowledge of the adverse events, in addition to measuring its incidence and improve spontaneous reports. For this reason, the present study evaluated this method in adverse events related to antibiotic vancomycin. The informations collected from databases and records of the Hospital of the Medical School of Ribeirão Preto-SP (USP-HCFMRP), refers to hospital patients treated with vancomycin for the period 01/09/2008 to 31/08/2009. We evaluated 1719 treatments in 1302 users, of which 521 were excluded because they were prophylactic. The average age of 1198 patients was 39.33 years (±27.66) and 58.68% were males. Prevalent infectious diagnoses for treatment with vancomycin were pneumonia (18.95%), sepsis (11.19%) and catheter-related infection (10.27%) with mean duration of treatment of 11.47 days (± 7.79) usually at a dose of 2000 mg daily (35.98%). We identified indicators of adverse events in 85.81% of treatments, and the incidence of adverse events was 7.18%. The adverse drug reactions (ADR) incidence was 5.93% and 2.57% of medication errors. The causality of ADR was probable (49.29%) and severity was moderate (54.93%). The ADR were mainly of type B (81.69%) and medication errors related to prescription (87.10%). Comparisons between the study and control groups showed no significant differences, demonstrating the need for monitoring of all parameters related to adverse events with vancomycin. Thus, it is clear the importance of the intensive monitoring and the presence of clinical pharmacists in the health team to improve the efficiency of pharmacovigilance, as demonstrated by this study. Before the results, we conclude that the method of intensive monitoring using the database and records, plus the involvement of a clinical pharmacist, promotes knowledge, assessment and understanding of adverse events presents in an institution, enabling the layout of strategies to prevention.

Magister Public Health - MPH / Background: Adverse Drug Reactions (ADRs) contribute to potentially expensive hospital admissions and are regarded as a major public health priority. ADRs in South Africa are mainly detected by a spontaneous reporting system but it is plagued by under-reporting. Previous records indicated under-reporting of ADRs in the Cape Winelands District amongst healthcare workers. Pharmacists, in particular, did not report ADRs compared to other healthcare cadres whilst they are generally considered to be the custodians of medicines. Study Aim: This study aimed to explore and describe the perceptions and experiences of rural public sector pharmacists’ reporting of ADRs and to understand why pharmacists in this rural health district under-reported ADRs. Study Design: A qualitative study design was appropriate for this research question as the researcher wanted to gain an in-depth understanding of human behavior related to the phenomena of under-reporting. Study Population and Sampling: The primary study population consisted of 24 public sector pharmacists in the Cape Winelands District. A purposive sampling strategy enabled the selection of 16 pharmacists ranging in gender, age, experience and rank. Eight pharmacists were supervisor pharmacists while the rest were production pharmacists, including a community service pharmacist and an intern pharmacist. Supervisor pharmacists are more involved with managerial tasks and the attendance of meetings compared to production pharmacists that focus on patient care and dispensing of medication. Two key informants involved in the Western Cape Pharmacovigilance System were included in the study. One key stakeholder was a policy specialist pharmacist working at Directorate: Pharmacy Services and primarily involved with the Provincial Pharmacy and Therapeutics Committee. The other key policy stakeholder, at the time of the study, was the manager of the Medicines Information Centre which forms part of the University of Cape Town’s (UCT) Pharmacology Division. Both were highly experienced pharmacists familiar with the pharmacovigilance system. Data Collection: In-depth interviews were conducted using a semi-structured interview guide consisting of open- ended questions. The semi-structured interview guide was tested on a participant outside the primary study population. Interviews were conducted in English and Afrikaans. Interviews were tape-recorded and the interviewers made field notes to supplement the data recorded. Two researchers with experience in qualitative data collection, briefed by the investigator, interviewed the pharmacists who worked in the district and the investigator interviewed the two key stakeholders. Data Analysis: The tape recordings were translated, where applicable, and all were transcribed verbatim by the investigator. The transcribed recordings were analyzed by the investigator by assigning codes to material on an Excel spreadsheet. This approach enabled the identification of themes which aided the understanding of the research phenomena. Ethics: Ethical approval was obtained from the University of the Western Cape Senate Research Committee and permission from the Western Cape Department of Health Research Committee. Written informed consent (See Appendix 1, page 73) was obtained from each participant prior to conducting the interviews and interviewees were assured of confidentiality throughout the research. Key Results and Discussion: Pharmacists in the study strongly acknowledged the importance of ADR reporting which is linked with pharmacists seeing themselves as the custodians of medication. Pharmacists in the study associated the reporting of ADRs with medication safety and felt responsible for ensuring it. In spite of this acknowledgement of the importance of ADR reporting, pharmacists rarely reported an ADR themselves. This finding was in line with previous research conducted and linked with barriers pharmacists faced in practice. The study revealed that pharmacists identified ADR reporting opportunities during their normal clinical work and enabled other health care professionals (HCPs) to confirm the occurrence of an ADR and report it. Pharmacists primarily identified ADRs when they scanned patient folders for clues that could indicate that an ADR had occurred. Other research conducted confirmed that the use of patient records could be used in the identification of ADRs. This finding was important to inform future training workshops to promote reporting of ADRs. Some pharmacists in the study associated an ADR with a therapeutic or clinical intervention. In general, therapeutic interventions usually involved a clinical action more closely associated with medical officers and were viewed by pharmacists in the study as being outside their legal and clinical scope of practice. A clinical intervention could include a change of medication, change of dose, and other prescription changes or might involve the medical officer referring the patient to a higher level of care depending on the severity of the suspected ADR experienced. A clinical intervention could include performing complex diagnostic tests, observations and laboratory investigations. Pharmacists’ association of an ADR experience with a clinical intervention was an important factor limiting their reporting of ADRs. The implication of this belief is that patients were referred from the pharmacy back to medical officers for the clinical intervention. In this way, although pharmacists do not directly report an ADR, their referral to medical officers would help improve reporting of ADRs. An unexpected and contrasting finding compared to previous research was the strong belief of some pharmacists in this study that common ADRs should be reported. Pharmacists believed that by reporting common ADRs en masse, authorities might decide to remove the problematic medication from the approved public sector formulary. This was in contrast to previous research where pharmacists either acknowledged that authorities only want novel or serious ADRs from newly marketed medication or believed that reporting well-known ADRs was a waste of time. Pharmacists reported that they faced several barriers in reporting ADRs. The main barriers that were mentioned were a lack of adequate feedback, heavy workload and time constraints, uncertainty in identifying the cause of an ADR and issues pharmacists had with the reporting process. These barriers were consistent with previous research conducted. Finally, pharmacists suggested various means of facilitating ADR reporting including use of electronic reporting aids, creating increased awareness amongst healthcare professionals, conducting continuous training and making amendments to the reporting form, some of which were in line with previous research conducted. Conclusion: Exploring the perceptions and experiences of pharmacists with respect to the under-reporting of ADRs revealed key knowledge about the spontaneous reporting system that could be applied to strengthen the current reporting system and enable more reporting. Whilst it was clear that pharmacists play an important role as the gatekeepers and drivers of the reporting process enabling other HCPs to report ADRs, more should be done to empower pharmacists in managing ADR reporting opportunities. This could benefit the healthcare system in ensuring that more ADRs are reported, as well as decrease the waiting time of patients and the workload of medical officers. In addition, engaging with pharmacists and HCPs to overcome barriers to reporting would facilitate increased ADR reporting. Recommendations: Several recommendations emerged from the study. Future circulars, training workshops and awareness posters about the ADR reporting process should inform all HCPs to report any medication suspected of being the cause of an ADR and not waste time in trying to identify the medication that caused it. A training workshop should be conducted with pharmacists to improve their skills in terms of identifying ADRs, how and what to report and of the appropriate referral of patients to the medical officers. An annual assessment on the availability of reporting forms in all health facilities should be conducted. In addition, the MIC should conduct a survey on the user-friendliness of the reporting form and enable HCPs to provide recommendations to help improve the reporting form template. Pharmacovigilance should be a standing item on the agendas of sub-district PTC meetings at which supervisor pharmacists should give quarterly updates to sub-district management on ADRs reported. As this study focused primarily on the experiences and perceptions of pharmacists in a rural health district, a follow-up study should explore perceptions and knowledge of medical officers and nurses of ADR reporting, specifically on the availability and complexity of the reporting form. Finally, the MIC should explore the development of a basic ADR causality assessment tool that could assist pharmacists and other HCPs in identifying a possible ADR and improve confidence amongst pharmacists and HCPs in reporting ADRs.

L’objectif de ce travail est de présenter et évaluer une nouvelle méthode, PGRx,systématique cas-référents en pharmaco-épidémiologie avec son application à l’étudede l’infarctus du myocarde (IDM). Elle se distingue par la collecte systématique etcontinue de cas d’événements dans des centres spécialisés et d’un pool de référenceen médecine générale dont sont tirés les témoins appariés aux cas. L’évaluation durisque d’IDM associé au diclofénac (OR 1.5) et celle du bénéfice associé aux statines(OR 0.75) montrent des résultats similaires à ceux publiés (respectivement OR 1.4 etOR 0.74). Nous montrons que les référents sont un échantillon représentatif de lapopulation française en termes de motif de consultation et valide en termesd’évaluation de facteurs de risque. La concordance entre la mesure de l’exposition parla déclaration du patient et par les prescriptions médicales est excellente pour lesmédicaments cardiovasculaires (95%). Notre travail sur une série d’études montre quela collecte systématique de cas et d’un pool de référence, selon la méthode PGRx, estfaisable, reproductible et valide en termes de résultats et d’indicateurs de qualité. / The objective of this work is to present and assess PGRx, a new systematic case-referentsmethod in pharmacoepidemiology and its application on the study of themyocardial infarction (MI). The originality of PGRx is the systematic and on-goingcollect of cases of events in a network of specialized centres and of a pool of referentsin general practice (GP), from which controls are selected by matching to the cases.The assessment of the risk of MI associated with diclofenac (OR 1.5) and of thebenefit on MI associated with statins (OR 0.75) displays similar results than theliterature (respectively OR 1.4 and OR 0.74).We show that the pool of referents is arepresentative sample of the French population in terms of reasons of consulting a GPand valid in terms of risk factors’assessment. Agreement between the measure ofexposure from patients’ self-report and from physician’s report of their prescriptionsis excellent on cardiovascular drugs (95%). Our work, through several studies, showsthat the systematic collect of cases and of a reference pool by the PGRx method isfeasible, reproducible and valid in terms of results and quality indicators.

Analysis of spontaneous adverse events reports of hypolipidemics Author: Anežka Škabradová Supervisor: PharmDr. Kateřina Malá, Ph.D. Consultant: PharmDr. Josef Malý, Ph.D. Department of Social and Clinical Pharmacy, Faculty of Pharmacy in Hradec Králové, Charles University Introduction: Hypolipidemic drugs are widely used in today's medicine and their consumption is still rising. Besides all benefits, their usage includes also some risks that needs to be reconsidered regularly. Analysis of spontaneous adverse effects (AEs) reports contributes to a safer therapy by detecting especially serious and unexpected adverse drug reactions. Objective: The aim of this thesis was to analyse spontaneous adverse events reports of hypolipidemic drugs, which were sent to the State Institute for Drug Control (SÚKL) database in the period from June 2004 to October 2017. Methods: The data obtained from SÚKL during the above mentioned period was analysed using descriptive statistics. The frequency of occurrence of certain AEs, which were categorized according to MedDRA and distinguished by hypolipidemics, was analysed, as well as their severity and expectability. Part of the analysis was for example also by whom was the report submitted and which way they used to submit it. Results: Overall 290 reports were obtained,...

Charles University, Faculty of Pharmacy in Hradci Králové Department of: Social and Clinical Pharmacy Author: Roman Vorel Supervisor: PharmDr. Jan Kostřiba, Ph.D. Title of Diploma Thesis: Pharmacovigilance of active substance zolpidem Introduction: Drug safety became very important topic in the second half of 20th century, even beyond pharmacy. It relates to increasing impact on testing of drugs before launch onto the market, but also following monitoring, so the benefit and risk ratio of drug can be determined. Pharmacovigilance takes care of monitoring of drug use risk. Zolpidem is active substance from class hypnotics and it is used for short-term treatment of insomnia. Most professionals recommend zolpidem as first line treatment of acute insomnia. That is why it can be considered as one of the most important hypnotic drugs. Objectives: The aim of this diploma thesis is to analyze spontaneous reports for adverse drug reaction for zolpidem from years 2004 to 2017. The thesis focus on reports from the Czech republic but also from the whole world. The information in SPC of medicinal products containing zolpidem which were distributed from 2004 to 2017 are compared. Methods: Following data from years 2004 to 2017 were used: 1) List of spontaneous reports for adverse reaction from the Czech republic...

Introduction : L’administration d’opioïdes par voie sous-cutanée peut occasionner des épisodes de dépression respiratoire menant à des décès évitables. Grâce à leur efficacité, les opioïdes demeurent la thérapie de premier choix pour le soulagement de la douleur modérée à sévère. Cependant, de plus en plus d’usagers présentent un ou plusieurs facteurs de risque de dépression respiratoire qui justifient une surveillance clinique adaptée à leur condition. Pour assurer une surveillance sécuritaire des usagers, des lignes directrices professionnelles doublées de mesures organisationnelles sous forme de directives et protocoles de surveillance infirmière ont été instaurées. La surveillance clinique demeure toutefois inadéquate, nous permettant de supposer qu’il existe, au-delà des éléments organisationnels, des déterminants individuels influençant la surveillance infirmière. Objectif : Identifier les déterminants de l’intention des infirmières d’exercer la surveillance clinique lors du pic d’action des opioïdes sous-cutanés auprès d’adultes hospitalisés. Méthodes : Étude corrélationnelle prédictive auprès d’infirmières d’unités de chirurgie et médecine (n = 104) d’un établissement de santé québécois par le biais d’un questionnaire auto-rapporté. Des régressions multiples ont été réalisées pour identifier les déterminants associés à l’intention d’adopter le comportement souhaité. Résultats : La perception de contrôle et la norme professionnelle ont été identifiés comme étant les principaux déterminants de l’intention des infirmières à exercer une surveillance lors du pic d’action des opioïdes. Discussion et conclusion : Par l’identification des déterminants individuels les plus susceptibles d’expliquer l’intention d’exercer une surveillance adaptée lors du pic d’action des opioïdes, cette étude permet de cibler des pistes d’intervention en matière de sécurité clinique. / Introduction: Subcutaneous administration of opioids may cause episodes of respiratory depression leading to preventable deaths. Due to their efficacy, opioids remain the first-choice therapy for the relief of moderate to acute pain. However, more and more inpatients have one or more risk factors for respiratory depression that justifies clinical surveillance adapted to their condition. To ensure safe monitoring of inpatients, professional guidelines, coupled with organizational measures in the form of guidelines and nurse monitoring protocols, have been introduced. Despite the action taken, clinical surveillance remains inadequate, allowing us to assume that there are, beyond the organizational elements, individual determinants influencing nurse supervision. Objective: Identify the determinants of the intention of nurses to conduct clinical surveillance of adults in hospital during the peak action of subcutaneous opioids. Methods: Predictive correlational study among nurses of surgical and medicine units of a Quebec health facility (n = 104) through a self-reported questionnaire. Multiple regressions were carried out to identify the determinants associated with the intention to perform the desired behaviour. Results: Perceived control and professional standards have been identified as the key determinants of nurses' intention to perform clinical surveillance associated with opioid at the time of peak effect. Discussion and conclusion: By identifying the individual determinants most likely to explain the intention to exercise appropriate surveillance at the peak of opioid action, this study helps to target clinical safety interventions.

En France, en 2012, les sujets âgés d’au moins 65 ans représentaient 71 % de l’incidence du Cancer ColoRectal (CCR). Depuis 2005, des thérapies ciblées ont été autorisées dans le CCR métastatique (CCRm) et sont recommandées en 1ère ligne en association à une chimiothérapie. Face à un manque d’évaluation de ces médicaments chez le sujet âgé, l’objectif de ce travail était l’étude, en situation réelle de soins, de l’utilisation, des bénéfices et de la sécurité d'emploi des médicaments anticancéreux, dont les thérapies ciblées, chez le sujet âgé et/ou fragile atteint de CCRm. Une revue systématique de la littérature a confirmé la faible inclusion des sujets âgés et fragiles dans les essais cliniques évaluant les thérapies ciblées dans le CCRm. D’après une étude sur des données mondiales de pharmacovigilance et la réunion de deux cohortes de terrain de patients traités par thérapies ciblées, nous avons montré que leur effectivité et sécurité d’emploi chez le sujet âgé étaient équivalentes à celles du sujet plus jeune. Dans la cohorte, la fragilité vis-à-vis des effets indésirables graves et du décès était liée aux caractéristiques du CCRm. Enfin, d’après une étude de terrain pilote et une cohorte dans les données de l’Assurance Maladie française, nous avons montré que, chez le sujet âgé, le traitement par médicaments anticancéreux, dont les thérapies ciblées, n’était pas optimal et encore très lié à l’âge des patients. Le rapport bénéfices/risques des thérapies ciblées semble donc positif dans la population âgée atteinte de CCRm actuellement traitée. Cependant, la population âgée traitée ne semble pas encore totalement correspondre à celle qui pourrait bénéficier du traitement. / In France, in 2012, patients aged at least 65 years accounted for 71% of the incidence of ColoRectal Cancer (CCR). Since 2005, targeted therapies have been authorized in metastatic CRC (mCRC) and are recommended in first-line in combination with conventional chemotherapy. Given this lack of evaluation of these drugs in the elderly, the objective of this work was the study, in real-life setting, of the use, the benefits and the safety of anticancer drugs, including targeted therapies in elderly and/or frail mCRC patients. A systematic review of the literature confirmed the low inclusion of elderly and frail patients in clinical trials evaluating targeted therapies in the mCRC. According to a study on international pharmacovigilance data and the pooling of two field cohorts of patients treated by targeted therapies, we showed that their effectiveness and safety in the elderly were equivalent to those of the younger. In the cohort, frailty regarding serious adverse events and death was related to the characteristics of the mCRC. Finally, according to a pilot field study and a cohort in the French health insurance data, we have shown that, in the elderly, the treatment with anticancer drugs, including targeted therapies, was not optimal and still very related to the age of the patients. The benefit/risk ratio of targeted therapies therefore seems positive in the elderly population with mCRC currently treated. However, the elderly population treated does not seem to be fully matching with those who would benefit from treatment.

Le phénomène de massification des données de santé constitue une opportunité de répondre aux questions des vigilances et de qualité des soins. Dans les travaux effectués au cours de cette thèse, nous présenterons des approches permettant d’exploiter la richesse et le volume des données intra hospitalières pour des cas d’usage de pharmacovigilance et de surveillance de bon usage du médicament. Cette approche reposera sur la modélisation de trajectoires de soins intra hospitalières adaptées aux besoins spécifiques de la pharmacovigilance. Il s’agira, à partir des données d’un entrepôt hospitalier de caractériser les événements d’intérêt et d’identifier un lien entre l’administration de ces produits de santé et l’apparition des effets indésirables, ou encore de rechercher les cas de mésusage du médicament. L’hypothèse posée dans cette thèse est qu’une approche visuelle interactive serait adaptée pour l’exploitation de ces données biomédicales hétérogènes et multi-domaines dans le champ de la pharmacovigilance. Nous avons développé deux prototypes permettant la visualisation et l’analyse des trajectoires de soins. Le premier prototype est un outil de visualisation du dossier patient sous forme de frise chronologique. La deuxième application est un outil de visualisation et fouille d’une cohorte de séquences d’événements. Ce dernier outil repose sur la mise en œuvre d’algorithme d’analyse de séquences (Smith-Waterman, Apriori, GSP) pour la recherche de similarité ou de motifs d’événements récurrents. Ces interfaces homme-machine ont fait l’objet d’études d’utilisabilité sur des cas d’usage tirées de la pratique réelle qui ont prouvé leur potentiel pour un usage en routine. / The massification of health data is an opportunity to answer questions about vigilance and quality of care. The emergence of big data in health is an opportunity to answer questions about vigilance and quality of care. In this thesis work, we will present approaches to exploit the diversity and volume of intra-hospital data for pharmacovigilance use and monitoring the proper use of drugs. This approach will be based on the modelling of intra-hospital care trajectories adapted to the specific needs of pharmacovigilance. Using data from a hospital warehouse, it will be necessary to characterize events of interest and identify a link between the administration of these health products and the occurrence of adverse reactions, or to look for cases of misuse of the drug. The hypothesis put forward in this thesis is that an interactive visual approach would be suitable for the exploitation of these heterogeneous and multi-domain biomedical data in the field of pharmacovigilance. We have developed two prototypes allowing the visualization and analysis of care trajectories. The first prototype is a tool for visualizing the patient file in the form of a timeline. The second application is a tool for visualizing and searching a cohort of event sequences The latter tool is based on the implementation of sequence analysis algorithms (Smith-Waterman, Apriori, GSP) for the search for similarity or patterns of recurring events. These human-machine interfaces have been the subject of usability studies on use cases from actual practice that have proven their potential for routine use.