INTRODUÇÃO: As taxas de mortalidade para doença de Chagas variam, consideravelmente de 0,2 a 19,2% anualmente dependendo principalmente do estágio da doença. Considera-se que pacientes na fase indeterminada tem taxas baixas, semelhante a população em geral. Este estudo compara a taxa de mortalidade e as causas de morte de doadores de sangue da Fundação Pró-Sangue - Hemocentro de São Paulo (FPS) soropositivos para a doença de Chagas e de doadores soronegativos para todos os testes de triagem realizados no processo de doação. Doadores de sangue constituem uma população, na qual os indivíduos soropositivos são em sua maioria assintomáticos. MÉTODOS: Este é um estudo de coorte retrospectiva de doadores de sangue de 1996 a 2000: 2842 doadores soropositivos para doença de Chagas e 5684 doadores soronegativos para todos os testes de triagem. O status vital foi designado realizando-se o relacionamento probabilístico de dados (RPD) utilizando-se o Sistema de Informação sobre Mortalidade do Brasil (SIM). Utilizamos um método RDP validado preliminarmente com sensibilidade de 94% (IC 95%, 90% - 97%), especificidade de 100% (IC 95%, 98% - 100%). Apenas os dados dos anos 2001 a 2009 do SIM foram disponibilizados para análise. Análise univariada foi conduzida considerando desfechos e fatores de riscos potenciais. Os testes Chi-quadrado e Exato de Fisher foram utilizados conforme apropriado. A diferença de sobrevida dos dois grupos no tempo foi graficamente representada pela Curva de Kaplan-Meyer e testado pelo teste Logrank. Regressões múltiplas de Cox foram utilizadas para derivação de riscos relativos instântaneos (hazard ratios - HR) para os desfechos do estudo, ajustando-se por varíaveis de confusão. RESULTADOS: RPD encontrou 159 óbitos entre os 2842 doadores soropositivos (5,6%) e 103 óbitos entre os 5684 doadores soronegativos (1,8%). A sorologia de doadores foi associada com óbitos relacionados a doença de Chagas (HR 2,3 - IC 95%: 1,8 - 3,0). As curvas de sobrevida foram diferentes quando comparando doadores soropositivos e soronegativos, jovens e velhos (p < 0,001). Das 159 mortes entre doadores soropositivos, 26 foram devidas a doença de Chagas (CID-10: B57.0 - B57.5), 47 foram devidas a outras doenças cardíacas (CDI-10: I10 a I80.2), sem menção a doença de Chagas como uma causa associada, e apenas 1 com menção a doença de Chagas na Parte II da declaração de óbito. CONCLUSÕES: Doadores de sangue soropositovs seguidos retrospectivamente por 9 a 14 anos morrem ao menos 2,3 vezes mais que doadores soronegativos. As causas de morte no SIM podem estar mal classificadas para os doadores de sangue soropositivos assintomáticos / INTRODUCTION: Mortality rates for Chagas\' disease vary considerably between 0.2 and 19.2% annually depending mostly on the stage of the disease. Patients at the indeterminate phase are considered to have low rates, similar to the overall population. This study compares mortality rates and causes of death of blood donors seropositive for Chagas\' disease and seronegative for all screening tests at Fundação Pró-Sangue - Hemocentro de São Paulo (FPS). Blood donors constitute a particular population in which seropositive individuals are mostly asymptomatic. METHODS: This is a retrospective cohort study of blood donors from 1996 to 2000: 2842 seropositive for Chagas disease and 5684 seronegative. Death status was ascertained by performing probabilistic record linkage (RL) with the national mortality information system (SIM). RL was validated in a preliminary study and sensitivity and specificity of the method were determined to be 94% (CI 95%, 90% - 97%) and 100% (CI 95%, 98% - 100%) respectively. Only SIM records for 2001 to 2009 were available for analysis. Univariate analysis was conducted considering outcomes and potential risks factors. Chi-square test or Fisher exact test were used as appropriate. The different survival of the two groups over time was graphically represented by the Kaplan-Meier curve and tested with the Logrank test. Multiple Cox regressions were used to derive Hazard ratios (HR) for the outcomes studied, adjusting for confounders. RESULTS: Reidentified 159 deaths among the 2842 seropositive blood donors (5.6%) and 103 deaths among the 5684 seronegative (1.8%). Serology of blood donors was associated with Chagas disease related death (HR 2.3 - 95% IC: 1.8 - 3.0). The survival curves differed when comparing seropositive and seronegative donors, young and old (p < 0.001). Out of the 159 deaths among seropositive donors, 26 were due to Chagas disease (ICD10 B57.0 - B57.5), 47 were due to other heart diseases (ICD10 - I10 to I80.2), with no mention of Chagas disease as an associated cause, and only 1 with Chagas disease mentioned in Part II of death certificate. CONCLUSIONS: Seropositive blood donors die at least 2.3 times more than seronegative donors. Causes of death in SIM may be misclassified for asymptomatic seropositive individuals

A utilização de bases de dados para estudos epidemiológicos, avaliação da qualidade e quantidade dos serviços de saúde vem despertando a atenção dos pesquisadores no contexto da Saúde Pública. No Brasil, as bases de dados do Sistema Único de Saúde (SUS) são exemplos de repositórios importantes que reúnem informações fundamentais sobre a Saúde. Entretanto, apesar dos avanços em termos de coleta e de ferramentas públicas para a pesquisa nessas bases de dados, tais como o TABWIN e o TABNET, esses recursos ainda não fazem uso de técnicas mais avançadas para a produção de informação gerencial, como as disponíveis em ferramentas OLAP (On Line Analytical Processing) e de mineração de dados. A situação é extremamente agravada pelo fato dos dados da Saúde Pública, produzidos por vários sistemas isolados, não estarem integrados, impossibilitando pesquisas entre diferentes bases de dados. Consequentemente, a produção de informação gerencial torna-se uma tarefa extremamente difícil. Por outro lado, a integração dessas bases de dados pode constituir um recurso indispensável e fundamental para a manipulação do enorme volume de dados disponível nesses ambientes e, assim, possibilitar a produção de informação e conhecimento relevantes, que contribuam para a melhoria da gestão em Saúde Pública. Acompanhar o seguimento de pacientes e comparar diferentes populações são outras importantes limitações das atuais bases de dados, uma vez que não há um identificador unívoco do paciente que possibilite executar tais tarefas. Esta Tese teve como objetivo a construção de um armazém de dados (data warehouse), a partir da análise de dez anos (período de 2000 a 2009) das principais bases de dados do SUS. Os métodos propostos para coleta, limpeza, padronização das estruturas dos bancos de dados, associação de registros ao paciente e integração dos sistemas de informação do SUS permitiram a identificação e o seguimento do paciente com sensibilidade de 99,68% e a especificidade de 97,94%. / The use of databases for epidemiologic studies, quality and quantity evaluation of health services have attracted the attention of researchers in the context of Public Health. In Brazil, the databases of the Sistema Único de Saúde (SUS) are examples of important repositories, which store fundamental information about health. However, despite of the advances in terms of load and public tools for research in those databases, such as TABWIN and TABNET, these resources do not use advanced techniques to produce management information as available in OLAP (On Line Analytical Processing) and data mining tools. The situation is drastically increased for the fact that data in public health, produced for different systems, are not integrated. This makes impossible to do research between different databases. As a consequence, the production of management information is a very difficult task. On the other hand, the integration of these databases can offer an important and fundamental resource to manipulate the enormous volume of data available in those environments and, in this way, to permit the production of relevant information and knowledge to improve the management of public health. The patient follow up and the comparison of different populations are other important limitations of the available databases, due to the absence of a common patient identifier. The objective of this Thesis was the construction of a data warehouse to analyze ten years (period from 2000 to 2009) of the principal databases of SUS. The proposed methods to load, clean, database structure standardization, patient record linkage and SUS information systems integration have been permitted patient identification and follow up with sensitivity of 99.6% and specificity of 97.94%.

As consequências do uso de substâncias psicoativas na saúde da população mundial são questões de preocupação - a maioria dos problemas dos usuários continua sendo a falta de acesso ao tratamento. Estudos longitudinais buscam identificar desfechos de recaída, mas são de alto custo. Como alternativa, estudos epidemiológicos com bases de dados secundários têm sido implementados em todo o mundo utilizando técnicas de record linkage. No Brasil, tem havido aumento do uso de registros médicos; porém, há pouca literatura sobre seu uso para seguimento de pacientes psiquiátricos, especialmente para estudos sobre usuários de drogas. Há muitos sistemas de informação na área de saúde pública sem um identificador único que possa ser utilizado para localizar um paciente em múltiplas bases de dados, condição de aplicação prática para o uso da técnica. O objetivo desse estudo foi o de produzir informações com base em dados secundários para mapear a trajetória dos usuários de crack na rede de atenção à saúde, utilizando metodologia de record linkage para o seguimento dos pacientes após sua alta hospitalar. Para isso, foi realizada a análise dos dados de atendimento disponibilizados pelos sistemas de informação em saúde pública para identificar a viabilidade de produzir informação para o seguimento dos usuários de crack na rede de atenção à saúde após a sua alta hospitalar. Com a impossibilidade de realizar esse seguimento com os dados disponíveis, a metodologia de record linkage probabilístico foi utilizada para rastrear as hospitalizações de usuários de crack e a continuidade de seu tratamento ambulatorial para estudo de seguimento desses usuários. A parcela da informação pública disponível pelos Sistemas de Informação em Saúde, não permitiu o acompanhamento de usuários entre os dispositivos de atendimento da rede de atenção. Utilizando uma amostra de 293 pacientes em tratamento para o uso de crack em duas instituições foi utilizada a metodologia e localizados 217 pacientes nos dados de internação hospitalar e 180 na base de atendimento ambulatorial; 55% foram identificados como pares verdadeiros na primeira base, enquanto apenas 12% na base de atendimento ambulatorial. Entre os dados hospitalares e ambulatoriais, demonstrou-se que dentre os usuários que realizam tratamento hospitalar, apenas 10 foram atendidos na rede ambulatorial, nesse período. Produzir informação para mapear a trajetória de usuários de crack utilizando as bases de dados dos sistemas de informação em saúde é possível por meio da metodologia de record linkage, como alternativa aos estudos longitudinais desta população de difícil acesso. Este estudo tem especial importância, pois pode contribuir também para a avaliação de programas de tratamento, por meio de indicadores de reinternação, tempos de permanência, curva de sobrevida e outros. A informação é fundamental para a implantação dos modelos de gestão que garantam as intervenções necessárias aos usuários com transtornos por uso de substâncias, especialmente no caso dos usuários de crack. / The consequences of use of psychotropic substances on the world population’s health are a matter of concern – most of the problems faced by users is still related to the lack of access to treatment. Longitudinal studies seek to identify these outcomes, but they are expensive. Alternatively, epidemiological studies based on secondary data have been applied worldwide, using record linkage methods. In Brazil, there has been an increase in the use of medical records. However, literature on its use for the follow-up of psychiatric patients, especially for studies on drug users is scarce. In the public health area, there are several information systems without an identification field that enable the location of a patient in multiple databases – which is one of the practical applications of the record linkage technique. The aim of this study was to produce information based on secondary data for mapping crack users pathway in the public healthcare network based on data linkage method, to follow them up after hospital discharge. For this, analysis of public health information systems was conducted to identify the feasibility of producing information for the follow-up of crack users in the network of health care. With the inability to conduct follow-up with the available data, the probabilistic record linkage methodology was used for tracing out crack users hospitalizations and the continuity of outpatient treatment after their discharge. The available public information from National Information Systems does not allow follow-up of patients of the health system across healthcare services, neither the monitoring of the continuation of treatment within the healthcare network. In a sample of 293 patients in treatment for crack use in two hospitals of Porto Alegre/RS; 217 patients were located in hospital admission data and 180 in the outpatient care database; 55% were identified as exact matches in the first database, whereas the outpatient database provided only 12%. Data from both hospital and outpatient care revealed that, among patients who received hospital treatment, only 10 attended outpatient care during the studied period. To produce information to track patient´s pathway is possible by record linkage method as an alternative to longitudinal studies of hard to reach populations. This study is particularly relevant, because it can also contribute to the evaluation of treatment programs, by means of indicators of rehospitalization, length of stay, survival rate etc. The formulation of public policies requires evidences based on information that, up to now, has not been adequately used, particularly that produced by existing Health Information Systems. Information is crucial for the implementation of administration models able to guarantee the necessary care to individuals with disorders resulting from drug use, especially in cases of crack users.

L’objectif de cette thèse était d’investiguer différents choix méthodologiques, en particulier le choix du délai et la prise en compte des causes médicales de décès, dans l’élaboration des indicateurs de mortalité post-hospitalière visant à refléter la qualité des soins.Dans une première phase, les données médico-administratives hospitalières des bénéficiaires du Régime Général (RG) de l’Assurance Maladie décédés dans l'année suivant une hospitalisation en 2008 ou 2009 ont été appariées aux causes de décès (base du CépiDc). Le taux d’appariement était de 96,4%.Dans une deuxième phase les séjours pour lesquels la cause initiale de décès pouvait être qualifiée d'indépendante du diagnostic principal du séjour ont été repérés à l'aide d'un algorithme et d'un logiciel s'appuyant sur des standards internationaux. Dans une troisième phase, le modèle le plus souvent utilisé à l'international pour évaluer la mortalité intra-hospitalière (modèle « de Jarman ») a été reproduit et utilisé pour construire des indicateurs de mortalité par établissement à 30, 60, 90, 180 et 365 jours post-admission, pour l'année 2009 (12 322 831 séjours PMSI-MCO des bénéficiaires du RG).L’indicateur de mortalité intra-hospitalière s’est révélé biaisé par les pratiques de sortie des établissements (caractérisées par la durée moyenne de séjour et le taux de transfert vers d’autres établissements). Les indicateurs à 60 ou 90 jours post-admission doivent être préférés à l’indicateur à 30 jours car ils ont l’avantage d’inclure presque tous les décès intra-hospitaliers, limitant notamment les incitations à maintenir les patients en vie jusqu’à la fin de la période de suivi et/ou à cesser de leur dédier des ressources une fois ce terme atteint. L’utilisation des causes de décès en supprimant les décès indépendants change de façon négligeable les indicateurs de mortalité globale par établissement, toutefois elle pourrait être utile pour des indicateurs spécifiques, limités à certaines pathologies ou procédures.Des réserves quant à la pertinence de ces indicateurs ont été décrites (limites du modèle et des variables d'ajustement, hétérogénéité de la qualité du codage entre les établissements), mettant en évidence la nécessité de recherches complémentaires, en particulier sur leur capacité à refléter la qualité des soins et sur l’impact de leur diffusion publique. A ce jour, l’interprétation des indicateurs de mortalité par établissement nécessite la plus grande prudence. / The main objective of this PhD work was to investigate different methodological options for the elaboration of post hospital mortality indicators aiming at reflecting quality of care, in particular to identify the most relevant timeframes and to assess the contribution of the causes of death information.In a first phase, the hospital discharge data of the French General health insurance scheme beneficiaries who died during the year following an hospital stay in 2008 or 2009 were linked to the cause of death register. The matching rate was 96.4%.In a second phase, the hospital stays for which the underlying cause of death could be qualified as independent from the main diagnosis were identified with an algorithm and a software relying on international standards.In a third phase, the method most widely used to assess in-hospital mortality (Dr Foster Unit method) was reproduced and used to construct hospital mortality indicators at 30, 60, 90, 180 et 365 days post-admission, on year 2009 (12 322 831 acute-care stays)..As in other countries, in-hospital mortality revealed biased by discharge patterns in the French data: hospitals : short length-of-stay or high transfer-out rates for comparable casemix tend to have lower in-hospital mortality. The 60-day and 90-day indicators should be preferred to the 30-day indicator, because they reflect a larger part of in-hospital mortality, and are less subject to the incentives either to maintain patients alive until the end of the follow-up window or to shift resources away when this length of stay is reached. The contribution of the causes of death seems negligible in the context of hospital-wide indicators, but it could prove its utility in future health services research about specific indicators limited to selected conditions or procedures.However, reservations about the relevance of hospital-wide mortality indicators aiming at assessing quality of care are described (limits of the statistical model and adjustment variables available, heterogeneity of the coding quality between hospitals). Further research is needed, in particular on the capacity of these indicators to reflect quality of care and on the impact of their public reporting. To date, the use of hospital-wide mortality indicators needs to be extremely cautious.

INTRODUÇÃO: As taxas de mortalidade para doença de Chagas variam, consideravelmente de 0,2 a 19,2% anualmente dependendo principalmente do estágio da doença. Considera-se que pacientes na fase indeterminada tem taxas baixas, semelhante a população em geral. Este estudo compara a taxa de mortalidade e as causas de morte de doadores de sangue da Fundação Pró-Sangue - Hemocentro de São Paulo (FPS) soropositivos para a doença de Chagas e de doadores soronegativos para todos os testes de triagem realizados no processo de doação. Doadores de sangue constituem uma população, na qual os indivíduos soropositivos são em sua maioria assintomáticos. MÉTODOS: Este é um estudo de coorte retrospectiva de doadores de sangue de 1996 a 2000: 2842 doadores soropositivos para doença de Chagas e 5684 doadores soronegativos para todos os testes de triagem. O status vital foi designado realizando-se o relacionamento probabilístico de dados (RPD) utilizando-se o Sistema de Informação sobre Mortalidade do Brasil (SIM). Utilizamos um método RDP validado preliminarmente com sensibilidade de 94% (IC 95%, 90% - 97%), especificidade de 100% (IC 95%, 98% - 100%). Apenas os dados dos anos 2001 a 2009 do SIM foram disponibilizados para análise. Análise univariada foi conduzida considerando desfechos e fatores de riscos potenciais. Os testes Chi-quadrado e Exato de Fisher foram utilizados conforme apropriado. A diferença de sobrevida dos dois grupos no tempo foi graficamente representada pela Curva de Kaplan-Meyer e testado pelo teste Logrank. Regressões múltiplas de Cox foram utilizadas para derivação de riscos relativos instântaneos (hazard ratios - HR) para os desfechos do estudo, ajustando-se por varíaveis de confusão. RESULTADOS: RPD encontrou 159 óbitos entre os 2842 doadores soropositivos (5,6%) e 103 óbitos entre os 5684 doadores soronegativos (1,8%). A sorologia de doadores foi associada com óbitos relacionados a doença de Chagas (HR 2,3 - IC 95%: 1,8 - 3,0). As curvas de sobrevida foram diferentes quando comparando doadores soropositivos e soronegativos, jovens e velhos (p < 0,001). Das 159 mortes entre doadores soropositivos, 26 foram devidas a doença de Chagas (CID-10: B57.0 - B57.5), 47 foram devidas a outras doenças cardíacas (CDI-10: I10 a I80.2), sem menção a doença de Chagas como uma causa associada, e apenas 1 com menção a doença de Chagas na Parte II da declaração de óbito. CONCLUSÕES: Doadores de sangue soropositovs seguidos retrospectivamente por 9 a 14 anos morrem ao menos 2,3 vezes mais que doadores soronegativos. As causas de morte no SIM podem estar mal classificadas para os doadores de sangue soropositivos assintomáticos / INTRODUCTION: Mortality rates for Chagas\' disease vary considerably between 0.2 and 19.2% annually depending mostly on the stage of the disease. Patients at the indeterminate phase are considered to have low rates, similar to the overall population. This study compares mortality rates and causes of death of blood donors seropositive for Chagas\' disease and seronegative for all screening tests at Fundação Pró-Sangue - Hemocentro de São Paulo (FPS). Blood donors constitute a particular population in which seropositive individuals are mostly asymptomatic. METHODS: This is a retrospective cohort study of blood donors from 1996 to 2000: 2842 seropositive for Chagas disease and 5684 seronegative. Death status was ascertained by performing probabilistic record linkage (RL) with the national mortality information system (SIM). RL was validated in a preliminary study and sensitivity and specificity of the method were determined to be 94% (CI 95%, 90% - 97%) and 100% (CI 95%, 98% - 100%) respectively. Only SIM records for 2001 to 2009 were available for analysis. Univariate analysis was conducted considering outcomes and potential risks factors. Chi-square test or Fisher exact test were used as appropriate. The different survival of the two groups over time was graphically represented by the Kaplan-Meier curve and tested with the Logrank test. Multiple Cox regressions were used to derive Hazard ratios (HR) for the outcomes studied, adjusting for confounders. RESULTS: Reidentified 159 deaths among the 2842 seropositive blood donors (5.6%) and 103 deaths among the 5684 seronegative (1.8%). Serology of blood donors was associated with Chagas disease related death (HR 2.3 - 95% IC: 1.8 - 3.0). The survival curves differed when comparing seropositive and seronegative donors, young and old (p < 0.001). Out of the 159 deaths among seropositive donors, 26 were due to Chagas disease (ICD10 B57.0 - B57.5), 47 were due to other heart diseases (ICD10 - I10 to I80.2), with no mention of Chagas disease as an associated cause, and only 1 with Chagas disease mentioned in Part II of death certificate. CONCLUSIONS: Seropositive blood donors die at least 2.3 times more than seronegative donors. Causes of death in SIM may be misclassified for asymptomatic seropositive individuals

With the inevitable adoption of information technology into all areas of human pursuit, the potential benefits for health care should not be overlooked. In Australia, details of most health care encounters are currently recorded for administrative purposes. This results in an impressive electronic data-bank that could provide a national resource for health service evaluation. ¶ Evaluation of health services has become increasingly important to provide indicators of the benefits, risks and cost-effectiveness of treatments. However, if administrative data are to be used for this purpose, several questions must first be addressed: Are the current data collections accessible? What outcome measures can be derived from these data? Can privacy issues be managed? Could the quality of the data be improved? Is the existing infrastructure adequate to supply data for evaluation purposes? Could the existing system provide a basis for the development of an integrated health information system? ¶ The aims of the project were: · To examine the potential for using administrative data to generate outcome measures and surveillance indicators. · To investigate the logistics of gaining access to these data for the purpose of research. This to be achieved within the current ethical, political and financial framework. · To compare the Australian health-service data system with the current international state-of-the-art. · To develop suggestions for expansion of the present system as part of an integrated health record and information system. This system to manage patient records and provide data for quality management, treatment surveillance and cost-effectiveness evaluation as a routine activity. ¶ The thesis is presented in two parts. In the first part, a historical cohort study is described that involved patients with implantable medical devices. The potential to evaluate outcomes was investigated using all national health-service information currently available in electronic form. Record linkage techniques were used to combine and augment the existing data collections. Australia’s national health databases are to varying degrees, amenable to such linkage and cover doctor visits, pharmaceuticals, hospital admissions and deaths. The study focused on medical devices as an illustrative case but the results are applicable to the routine assessment of all medical and surgical interventions. ¶ For the Australian ‘Medical Devices study’, the records of 5,316 patients who had medical device implants in 1993-94 were selected from the archives of a major private health insurer. Five groups of medical implants were studied: heart valves, pacemakers, hips, vascular grafts and intra-optic lenses. Outcomes for these patients, including death, re-operation and health service utilisation, were compared and analysed. ¶ A comparison study was performed using data from the Manitoba Health database in Winnipeg, Canada. Manitoba provides a very similar demographic group to that found in Australia and is an example of a prototype integrated-health-information system. One of the principal advantages for research is that personally identified data about medical and hospital services are collected for all patients. Selection bias is eliminated because individual consent is not required for this type of research and all selected patients could be included in the study. ¶ The two studies revealed many barriers to the use of administrative data for health outcomes research. Service event data for the Australian cohort could be collected but only after long delays and hospital morbidity data were not available for the entire cohort. In contrast to the situation in Australia, the Manitoba data were both accessible and complete, but were lacking in detail in some areas. ¶ Analysis of the collected data demonstrated that without the addition of clinical data only general indications of trends could be deduced. However, with minimal supplementary clinical data, it was possible to examine differences in performance between brands of medical devices thus indicating one of the uses for this type of data collection. ¶ In the second part of the thesis, conclusions are presented about the potential uses and limitations of the existing system and its use as a basis for the development of a national Integrated Health Record and Information System (IHRIS). The need for the establishment of a systemic quality management system for health care is discussed. ¶ The study shows that linked administrative data can provide information about health outcomes which is not readily available from other sources. If expanded and integrated, the system that is currently used to collect and manage administrative data, could provide the basis for a national health information system. This system would provide many benefits for health care. Benefits would include the monitoring, surveillance and cost-effectiveness analysis of new and existing treatments involving medical devices, drugs and surgical procedures. An integrated health information system could thus provide for both clinical and administrative needs, while in addition providing data for research. ¶ Unfortunately, in Australia, the use of administrative data for this purpose is not currently feasible. The principal barrier is the existence of a culture within the Australian health care system which is not supportive of research and is deficient in quality and safety measures. ¶ Recent initiatives by both the Commonwealth and state governments have supported the introduction of measures to improve quality and safety in health care. It is argued here that an Integrated Health Record and Information System (IHRIS) would provide an essential component of any such scheme. The results of this study have important policy implications for health care management in both the administrative and clinical domains.

Diss. (sammanfattning) Linköping : Linköpings universitet, 2009. / Härtill 5 uppsatser.

As consequências do uso de substâncias psicoativas na saúde da população mundial são questões de preocupação - a maioria dos problemas dos usuários continua sendo a falta de acesso ao tratamento. Estudos longitudinais buscam identificar desfechos de recaída, mas são de alto custo. Como alternativa, estudos epidemiológicos com bases de dados secundários têm sido implementados em todo o mundo utilizando técnicas de record linkage. No Brasil, tem havido aumento do uso de registros médicos; porém, há pouca literatura sobre seu uso para seguimento de pacientes psiquiátricos, especialmente para estudos sobre usuários de drogas. Há muitos sistemas de informação na área de saúde pública sem um identificador único que possa ser utilizado para localizar um paciente em múltiplas bases de dados, condição de aplicação prática para o uso da técnica. O objetivo desse estudo foi o de produzir informações com base em dados secundários para mapear a trajetória dos usuários de crack na rede de atenção à saúde, utilizando metodologia de record linkage para o seguimento dos pacientes após sua alta hospitalar. Para isso, foi realizada a análise dos dados de atendimento disponibilizados pelos sistemas de informação em saúde pública para identificar a viabilidade de produzir informação para o seguimento dos usuários de crack na rede de atenção à saúde após a sua alta hospitalar. Com a impossibilidade de realizar esse seguimento com os dados disponíveis, a metodologia de record linkage probabilístico foi utilizada para rastrear as hospitalizações de usuários de crack e a continuidade de seu tratamento ambulatorial para estudo de seguimento desses usuários. A parcela da informação pública disponível pelos Sistemas de Informação em Saúde, não permitiu o acompanhamento de usuários entre os dispositivos de atendimento da rede de atenção. Utilizando uma amostra de 293 pacientes em tratamento para o uso de crack em duas instituições foi utilizada a metodologia e localizados 217 pacientes nos dados de internação hospitalar e 180 na base de atendimento ambulatorial; 55% foram identificados como pares verdadeiros na primeira base, enquanto apenas 12% na base de atendimento ambulatorial. Entre os dados hospitalares e ambulatoriais, demonstrou-se que dentre os usuários que realizam tratamento hospitalar, apenas 10 foram atendidos na rede ambulatorial, nesse período. Produzir informação para mapear a trajetória de usuários de crack utilizando as bases de dados dos sistemas de informação em saúde é possível por meio da metodologia de record linkage, como alternativa aos estudos longitudinais desta população de difícil acesso. Este estudo tem especial importância, pois pode contribuir também para a avaliação de programas de tratamento, por meio de indicadores de reinternação, tempos de permanência, curva de sobrevida e outros. A informação é fundamental para a implantação dos modelos de gestão que garantam as intervenções necessárias aos usuários com transtornos por uso de substâncias, especialmente no caso dos usuários de crack. / The consequences of use of psychotropic substances on the world population’s health are a matter of concern – most of the problems faced by users is still related to the lack of access to treatment. Longitudinal studies seek to identify these outcomes, but they are expensive. Alternatively, epidemiological studies based on secondary data have been applied worldwide, using record linkage methods. In Brazil, there has been an increase in the use of medical records. However, literature on its use for the follow-up of psychiatric patients, especially for studies on drug users is scarce. In the public health area, there are several information systems without an identification field that enable the location of a patient in multiple databases – which is one of the practical applications of the record linkage technique. The aim of this study was to produce information based on secondary data for mapping crack users pathway in the public healthcare network based on data linkage method, to follow them up after hospital discharge. For this, analysis of public health information systems was conducted to identify the feasibility of producing information for the follow-up of crack users in the network of health care. With the inability to conduct follow-up with the available data, the probabilistic record linkage methodology was used for tracing out crack users hospitalizations and the continuity of outpatient treatment after their discharge. The available public information from National Information Systems does not allow follow-up of patients of the health system across healthcare services, neither the monitoring of the continuation of treatment within the healthcare network. In a sample of 293 patients in treatment for crack use in two hospitals of Porto Alegre/RS; 217 patients were located in hospital admission data and 180 in the outpatient care database; 55% were identified as exact matches in the first database, whereas the outpatient database provided only 12%. Data from both hospital and outpatient care revealed that, among patients who received hospital treatment, only 10 attended outpatient care during the studied period. To produce information to track patient´s pathway is possible by record linkage method as an alternative to longitudinal studies of hard to reach populations. This study is particularly relevant, because it can also contribute to the evaluation of treatment programs, by means of indicators of rehospitalization, length of stay, survival rate etc. The formulation of public policies requires evidences based on information that, up to now, has not been adequately used, particularly that produced by existing Health Information Systems. Information is crucial for the implementation of administration models able to guarantee the necessary care to individuals with disorders resulting from drug use, especially in cases of crack users.

As consequências do uso de substâncias psicoativas na saúde da população mundial são questões de preocupação - a maioria dos problemas dos usuários continua sendo a falta de acesso ao tratamento. Estudos longitudinais buscam identificar desfechos de recaída, mas são de alto custo. Como alternativa, estudos epidemiológicos com bases de dados secundários têm sido implementados em todo o mundo utilizando técnicas de record linkage. No Brasil, tem havido aumento do uso de registros médicos; porém, há pouca literatura sobre seu uso para seguimento de pacientes psiquiátricos, especialmente para estudos sobre usuários de drogas. Há muitos sistemas de informação na área de saúde pública sem um identificador único que possa ser utilizado para localizar um paciente em múltiplas bases de dados, condição de aplicação prática para o uso da técnica. O objetivo desse estudo foi o de produzir informações com base em dados secundários para mapear a trajetória dos usuários de crack na rede de atenção à saúde, utilizando metodologia de record linkage para o seguimento dos pacientes após sua alta hospitalar. Para isso, foi realizada a análise dos dados de atendimento disponibilizados pelos sistemas de informação em saúde pública para identificar a viabilidade de produzir informação para o seguimento dos usuários de crack na rede de atenção à saúde após a sua alta hospitalar. Com a impossibilidade de realizar esse seguimento com os dados disponíveis, a metodologia de record linkage probabilístico foi utilizada para rastrear as hospitalizações de usuários de crack e a continuidade de seu tratamento ambulatorial para estudo de seguimento desses usuários. A parcela da informação pública disponível pelos Sistemas de Informação em Saúde, não permitiu o acompanhamento de usuários entre os dispositivos de atendimento da rede de atenção. Utilizando uma amostra de 293 pacientes em tratamento para o uso de crack em duas instituições foi utilizada a metodologia e localizados 217 pacientes nos dados de internação hospitalar e 180 na base de atendimento ambulatorial; 55% foram identificados como pares verdadeiros na primeira base, enquanto apenas 12% na base de atendimento ambulatorial. Entre os dados hospitalares e ambulatoriais, demonstrou-se que dentre os usuários que realizam tratamento hospitalar, apenas 10 foram atendidos na rede ambulatorial, nesse período. Produzir informação para mapear a trajetória de usuários de crack utilizando as bases de dados dos sistemas de informação em saúde é possível por meio da metodologia de record linkage, como alternativa aos estudos longitudinais desta população de difícil acesso. Este estudo tem especial importância, pois pode contribuir também para a avaliação de programas de tratamento, por meio de indicadores de reinternação, tempos de permanência, curva de sobrevida e outros. A informação é fundamental para a implantação dos modelos de gestão que garantam as intervenções necessárias aos usuários com transtornos por uso de substâncias, especialmente no caso dos usuários de crack. / The consequences of use of psychotropic substances on the world population’s health are a matter of concern – most of the problems faced by users is still related to the lack of access to treatment. Longitudinal studies seek to identify these outcomes, but they are expensive. Alternatively, epidemiological studies based on secondary data have been applied worldwide, using record linkage methods. In Brazil, there has been an increase in the use of medical records. However, literature on its use for the follow-up of psychiatric patients, especially for studies on drug users is scarce. In the public health area, there are several information systems without an identification field that enable the location of a patient in multiple databases – which is one of the practical applications of the record linkage technique. The aim of this study was to produce information based on secondary data for mapping crack users pathway in the public healthcare network based on data linkage method, to follow them up after hospital discharge. For this, analysis of public health information systems was conducted to identify the feasibility of producing information for the follow-up of crack users in the network of health care. With the inability to conduct follow-up with the available data, the probabilistic record linkage methodology was used for tracing out crack users hospitalizations and the continuity of outpatient treatment after their discharge. The available public information from National Information Systems does not allow follow-up of patients of the health system across healthcare services, neither the monitoring of the continuation of treatment within the healthcare network. In a sample of 293 patients in treatment for crack use in two hospitals of Porto Alegre/RS; 217 patients were located in hospital admission data and 180 in the outpatient care database; 55% were identified as exact matches in the first database, whereas the outpatient database provided only 12%. Data from both hospital and outpatient care revealed that, among patients who received hospital treatment, only 10 attended outpatient care during the studied period. To produce information to track patient´s pathway is possible by record linkage method as an alternative to longitudinal studies of hard to reach populations. This study is particularly relevant, because it can also contribute to the evaluation of treatment programs, by means of indicators of rehospitalization, length of stay, survival rate etc. The formulation of public policies requires evidences based on information that, up to now, has not been adequately used, particularly that produced by existing Health Information Systems. Information is crucial for the implementation of administration models able to guarantee the necessary care to individuals with disorders resulting from drug use, especially in cases of crack users.

À ce jour, les mesures de risque des cancers ou d’efficacité de leur suivi, se font à partir de recueils de données médicales spécifiques initiés par les médecins épidémiologistes. Ces recueils disposent néanmoins de certaines limites : perte d’information, biais de déclaration, absence de données pour un risque non connu, biais de mesure (par exemple pour les données de nature médico-économiques). Le partage sécurisé de données médicales entre différentes structures médicales publiques et/ou privées est à ce jour en pleine mutation technologique. Les technologies proposées doivent rendre possible un partage électronique et sécurisé de ces données de manière à les rendre disponible à tout instant dans le cadre de l’observation sanitaire à l’évaluation de prises en charge ou de politiques de santé. Pour répondre à ces besoins, l’infrastructure GINSENG se base sur des informations produites dans le cadre des soins, sans nouvelles modalités de recueil, permettant à la fois une vitesse d’accès à l’information et une exhaustivité accrue. Ce recueil se fait par ailleurs avec de meilleures garanties d’anonymat et un chaînage de l’information médicale pour chaque patient. Une autorisation de la CNIL a été octroyée à l’infrastructure informatique du projet ainsi qu’à son utilisation pour le suivi des cancers en octobre 2013. Depuis le portail web e-ginseng.com, les médecins habilités s’authentifient grâce à leur Carte de Professionnel de Santé (CPS). Chaque patient, dont les données médicales sont réparties dans les établissements de santé, est identifié avec son accord, par les attributs suivants : nom, prénom, année et mois de naissance ainsi que son code postal de résidence avant d’être assigné à un numéro d’identification unique et anonyme. La mise à jour des données médicales de chaque patient est réalisée une fois par semaine ; chaque médecin peut alors consulter toutes les informations médicales relatives à chaque patient par une simple connexion au réseau. Ces informations lui apparaissent sous forme d’une arborescence d’évènements médicaux. Par exemple, un médecin chargé du suivi des patients dans le cadre du dépistage organisé pourra accéder directement depuis le portail web aux informations médicales dont il aura besoin pour établir une fiche médicale exhaustive du parcours du patient pour lequel un cancer aurait été détecté ou bien une suspicion de cancer qui se serait avérée négative suite à plusieurs examens médicaux. Un médecin épidémiologiste peut également réaliser des requêtes statistiques d’envergure sur les données médicales afin de répondre à des questions d’intérêt en santé publique. Pour aller plus loin, les requêtes épidémiologiques lancées sur les données médicales peuvent être couplées à des informations d’utilité publique recueillies sur d’autres bases de données en accès libre sur internet. L’infrastructure informatique GINSENG est actuellement déployée pour le suivi des cancers en région Auvergne entre les structures de gestion du dépistage organisé du cancer (SGDO) et le cabinet d’anatomie et cytologie pathologiques (ACP) Sipath-Unilabs. Le recours à un hébergeur de données de santé (HADS), nommé Informatique de sécurité (IDS), est également proposé pour le stockage des informations confidentielles des patients. Cette infrastructure permet actuellement de collecter toutes les informations médicales d’intérêt pour le suivi des cancers et l’évaluation des pratiques médicales. Les équipes de bio-statistiques et de santé publique du CHU de Clermont-Ferrand établissent actuellement les analyses épidémiologiques d’intérêt à partir des données collectées par le réseau. / The implementation of a grid network to support large-scale epidemiology analysis (based on distributed medical data sources) and medical data sharing require medical data integration and semantic alignment. In this thesis, we present the GINSENG (Global Initiative for Sentinel eHealth Network on Grid) network that federates existing Electronic Health Records through a rich metamodel (FedEHR), a semantic data model (SemEHR) and distributed query toolkits. A query interface based on the VIP platform, and available through the e-ginseng.com web portal helps medical end-users in the design of epidemiological studies and the retrieval of relevant medical data sets.