| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 5.63 MB | Adobe PDF |
Advisor(s)
Abstract(s)
A validação farmacêutica de prescrições medicamentosas é um trabalho essencial para a segurança e bem-estar de todos os doentes internados nos hospitais nacionais e internacionais. O aumento da segurança para os doentes, assim como o ganho de tempo para a realização das outras tarefas, são uns dos principais desafios do atual sistema de validação farmacêutica dos hospitais públicos nacionais. O problema que trata a dissertação deriva da sobrecarga dos colaboradores da farmácia hospitalar do CHSJ que, em cada dia, em poucas horas, têm de validar centenas/milhares de prescrições de doentes internados, o que potencia a ocorrência de falhas neste processo de validação, devido à escassez de tempo. Esta sobrecarga leva a um consumo indevido de medicamentos gerando problemas para o doente, associados à utilização incorreta e massiva destes, podendo provocar aumento dos tempos de internamento e consequente aumento das despesas dos hospitais com doentes que poderiam ter alta médica, caso não ocorressem estes erros. O presente trabalho tem como principal objetivo modelar e desenvolver um protótipo de um Sistema de Apoio à Decisão Clínica (SADC) para a área da Farmácia. Inicialmente foram analisados alguns SADC orientados para a área da saúde. Os quatro sistemas analisados são orientados para a área de validação farmacêutica/ fármaco-vigilância. Da análise resulta que estes sistemas são um importante mais valia, no entanto identificou-se a possibilidade de acrescentar novas funcionalidades. No desenvolvimento do projeto foi seguido uma metodologia de prototipagem onde foi utilizado o Modelo de Desenvolvimento Evolucionário. Na realização da solução, são criadas regras baseadas em conhecimento, que foi obtido através de pesquisas de informação em literatura relativas ao tema e através da consulta de um especialista na área de validação farmacêutica. Estas regras serão implementadas utilizando a ferramenta Power BI, dado que esta se mostrou, ao nível de integração com SQL Server e ao nível de integração do conhecimento, a mais completa e melhor das estudadas. Foi desenvolvida uma plataforma que reúne informação proveniente da consulta de prescrição médica, do estado clínico do doente, informação sobre o percurso hospitalar do doente, tais como serviço no qual o doente está internado e se este possui cirurgias agendadas. Com base nestas informações a plataforma vai emitir alertas, avisos e sugerir possíveis erros na prescrição, bem como prescrições que não necessitem de se validadas pelo farmacêutico. Por fim foi realizado um inquérito com o objetivo de verificar e validar a qualidade das decisões sugeridas pelo sistema. O inquérito foi respondido por cinco especialistas em Farmacologia com mais de dez anos de experiência. Desse inquérito obtiveram-se resultados positivos. Os utilizadores indicaram elevados graus de satisfação relativamente à facilidade de utilização, ao tempo de resposta para cada tarefa e à terminologia. Relativamente à emissão de avisos relativos aos riscos e auto validados, identificados pelo sistema, os farmacêuticos consideraram que estas funcionalidade foram ao encontro das suas expectativas. Foi realizado também uma análise estatística correspondente ao impacto de quatro indicadores de risco da plataforma em 30 dias de prescrições. Os resultados obtidos foram significativos uma vez que um dos objetivos é prevenir erros na prescrição. Com base nos resultados obtidos foram escolhidos propostas de melhorias futuras na área de integração da plataforma com o sistema atual e desenvolvimento de um sistema de registos de intervenções complementar.
Pharmaceutical validation of drug prescriptions is a necessary and essential task for the safety and well-being of all patients admitted to national and international hospitals. Increasing patient safety as well as gaining time to perform other tasks that are required to be performed by pharmacists are one of the main challenges of the current pharmaceutical validation system in our national public hospitals. The current problem comes from the overload of employees of the pharmacy at Hospital São João who every day, in a few hours, has to validate hundreds / thousands of inpatient prescriptions, which leads to mistake being made in this validation process, due to scarcity of time, leading to improper consumption of drugs, generating problems for the patient associated with the incorrect and massive use of these, which can cause the increase in hospital stay and treatment expenses for patients who might have already been discharged if these mistakes did not occur. The present work has as main objective to model and develop a prototype of a Clinical Decision Support System (CDSS) for the Pharmacy area. Initially, a couple of health-oriented DSS were analyzed. The four systems analyzed were oriented towards the area of pharmaceutical / drug-surveillance validation. From this study it was found that these systems are an important asset, however the possibility of adding new features was identified. In the development of the project, a prototyping methodology was used, using the Evolutionary Development Model. To realize the solution, knowledge-based rules were created, which were obtained by searching for information in the literature related to the topic and by consulting a specialist in the field of pharmaceutical validation. These rules were implemented using the Power BI tool, as it has been proved to be the most complete and it has one of the best level of integration with SQL Server and of knowledge integration, from the technologies that were studied. A platform was developed that gathers information from the medical prescription, the patient's clinical status, information about the patient's hospital path, such as the service in which the patient is hospitalized and if he has scheduled surgeries. Based on the information, the platform will issue alerts, warnings and suggestions for, both possible errors in the prescription and prescriptions that do not need to be validated by the pharmacist. Finally, a survey was carried out in order to verify and validate the quality of the decisions suggested by the system. The survey was answered by five pharmacology experts with more than ten years of experience. From this survey, positive results were obtained. Users indicated a high level of satisfaction regarding ease of use, the response time for each task and terminology. Regarding the issuance of warnings related to risks and self-validations, identified by the system, pharmacists considered that these features met their expectations. A statistical analysis was also carried out corresponding to the impact of four platform risk indicators in 30 days of medical prescriptions. The results obtained were significant since one of the main objectives is to prevent errors in the prescription. Based on the results obtained, proposals for future improvements were made in the area of integration of the platform with the current system and development of a complementary intervention registration system.
Pharmaceutical validation of drug prescriptions is a necessary and essential task for the safety and well-being of all patients admitted to national and international hospitals. Increasing patient safety as well as gaining time to perform other tasks that are required to be performed by pharmacists are one of the main challenges of the current pharmaceutical validation system in our national public hospitals. The current problem comes from the overload of employees of the pharmacy at Hospital São João who every day, in a few hours, has to validate hundreds / thousands of inpatient prescriptions, which leads to mistake being made in this validation process, due to scarcity of time, leading to improper consumption of drugs, generating problems for the patient associated with the incorrect and massive use of these, which can cause the increase in hospital stay and treatment expenses for patients who might have already been discharged if these mistakes did not occur. The present work has as main objective to model and develop a prototype of a Clinical Decision Support System (CDSS) for the Pharmacy area. Initially, a couple of health-oriented DSS were analyzed. The four systems analyzed were oriented towards the area of pharmaceutical / drug-surveillance validation. From this study it was found that these systems are an important asset, however the possibility of adding new features was identified. In the development of the project, a prototyping methodology was used, using the Evolutionary Development Model. To realize the solution, knowledge-based rules were created, which were obtained by searching for information in the literature related to the topic and by consulting a specialist in the field of pharmaceutical validation. These rules were implemented using the Power BI tool, as it has been proved to be the most complete and it has one of the best level of integration with SQL Server and of knowledge integration, from the technologies that were studied. A platform was developed that gathers information from the medical prescription, the patient's clinical status, information about the patient's hospital path, such as the service in which the patient is hospitalized and if he has scheduled surgeries. Based on the information, the platform will issue alerts, warnings and suggestions for, both possible errors in the prescription and prescriptions that do not need to be validated by the pharmacist. Finally, a survey was carried out in order to verify and validate the quality of the decisions suggested by the system. The survey was answered by five pharmacology experts with more than ten years of experience. From this survey, positive results were obtained. Users indicated a high level of satisfaction regarding ease of use, the response time for each task and terminology. Regarding the issuance of warnings related to risks and self-validations, identified by the system, pharmacists considered that these features met their expectations. A statistical analysis was also carried out corresponding to the impact of four platform risk indicators in 30 days of medical prescriptions. The results obtained were significant since one of the main objectives is to prevent errors in the prescription. Based on the results obtained, proposals for future improvements were made in the area of integration of the platform with the current system and development of a complementary intervention registration system.
Description
Keywords
Validação de Prescrições Farmácia Hospitalar Sistemas de Informação Saúde Sistemas de Apoio à Decisão Clínica Sistema Informático Hospitalar Power BI Sql Server DAX Excel Prescription Validation Hospital Pharmacy Health Information Systems Clinical Decision Support Systems Hospital Computer System