Submitted by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-10-17T12:42:28Z No. of bitstreams: 1 TES_PEDRO_HENRIQUE_SOARES_LEIVAS_COMPLETO.pdf: 1132784 bytes, checksum: 11c61ab6cd63f1363810d559ca598a23 (MD5) / Approved for entry into archive by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-10-17T12:42:37Z (GMT) No. of bitstreams: 1 TES_PEDRO_HENRIQUE_SOARES_LEIVAS_COMPLETO.pdf: 1132784 bytes, checksum: 11c61ab6cd63f1363810d559ca598a23 (MD5) / Made available in DSpace on 2017-10-17T12:42:46Z (GMT). No. of bitstreams: 1 TES_PEDRO_HENRIQUE_SOARES_LEIVAS_COMPLETO.pdf: 1132784 bytes, checksum: 11c61ab6cd63f1363810d559ca598a23 (MD5) Previous issue date: 2017-08-11 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior - CAPES / This thesis presents three essays related to the quality of hospital care in Brazil, made using data from 2008 to 2014 from the Hospital Information System of the Unified Health System (SIH-SUS) and the National Registry of Health Establishments (CNES). In the first essay, the objective was to verify if the health outcome of patients undergoing hip replacement are better as the hospital performs more of this procedure. In other words, we tried to verify the existence of the so-called volume-outcome effect in the Brazilian hospital sector. Evidences pointed to the existence of a volume-outcome effect in Brazil. Exploring the geographical distribution of patient residences and hospitals as a source of exogenous variation for volume, we could identify that the effect is a result of the practice-makes-perfect hypothesis, occurring through static scale economies. In the second essay, the objective was to verify if the probability of death of patients submitted to percutaneous coronary intervention is lower if it was treated in a specialized hospital in cardiovascular diseases instead of general hospital. The use of the instrumental variables model allowed us to evaluate the relationship from a causal perspective. The findings pointed out that the health outcomes of patients treated in specialized hospitals are better than the results of patients treated in general hospitals. Finally, the third essay aimed to investigate the existence of a weekend effect in the Brazilian hospital sector. Considering the patients admitted in an emergency with a diagnosis of acute myocardial infarction (AMI), odds ratios for hospital death were estimated using logistic regression. The results were consistent with the existence of the weekend effect, even controlling a series of factors, including hospital effect. / Esta tese apresenta tr?s ensaios relacionados a qualidade do cuidado hospitalar no Brasil, os quais foram confeccionados utilizando dados de 2008 a 2014, provenientes do Sistema de Informa??es Hospitalares do Sistema ?nico de Sa?de (SIH-SUS) e do Cadastro Nacional de Estabelecimentos de Sa?de (CNES). No primeiro ensaio, o objetivo foi verificar se o resultado de sa?de dos pacientes submetidos a artroplastia de quadril s?o melhores ? medida que o hospital realiza mais desse procedimento. Em outras palavras, buscou-se verificar a exist?ncia do chamado volume-outcome effect no setor hospitalar brasileiro. As evid?ncias apontaram a exist?ncia do volume-outcome effect no Brasil. Explorando a distribui??o geogr?fica das resid?ncias dos pacientes e dos hospitais como fonte de varia??o ex?gena para o volume, fomos capazes de identificar que o referido efeito ? resultado da hip?tese de practice-makes-perfect, ocorrendo atrav?s de ganhos est?ticos de escala. No segundo ensaio, o objetivo foi verificar se a probabilidade de ?bito de pacientes submetidos a Interven??o Coron?ria Percut?nea ? menor se o mesmo foi tratado em hospital especializado em doen?as cardiovasculares ao inv?s de hospital geral. A utiliza??o do modelo de vari?veis instrumentais permitiu avaliar a rela??o sob uma perspectiva causal. Os achados apontaram que os resultados de sa?de dos pacientes tratados em hospitais especializados s?o melhores do que os resultados dos pacientes tratados em hospitais gerais. Por fim, o terceiro ensaio teve por objetivo investigar a exist?ncia de um weekend effect no setor hospitalar brasileiro. Considerando os pacientes internados em car?ter de urg?ncia com diagn?stico de Infarto Agudo do Mio?rdio (IAM), foram estimadas raz?es de chance para o ?bito hospitalar por meio de regress?o log?stica. Os resultados foram coerentes com a exist?ncia do weekend effect, mesmo controlando uma s?rie de fatores, inclusive efeito fixo de hospital.