Monitoramento de tempos de sobrevivência com o gráfico CUSUM ajustado ao risco: Identificação do limite de controle e da magnitude na mudança
The statistical process control (SPC) provides several tools for monitoring some characteristic quality, especially in industry. However, there is also interest in the use of SPC for medical monitoring procedure, for example to assess performance hospitals or medical procedures. Unlike the industria...
Na minha lista:
| Autor principal: | |
|---|---|
| Outros Autores: | |
| Formato: | bachelorThesis |
| Idioma: | pt_BR |
| Publicado em: |
Universidade Federal do Rio Grande do Norte
|
| Assuntos: | |
| Endereço do item: | https://repositorio.ufrn.br/handle/123456789/34285 |
| Tags: |
Adicionar Tag
Sem tags, seja o primeiro a adicionar uma tag!
|
| Resumo: | The statistical process control (SPC) provides several tools for monitoring some characteristic quality, especially in industry. However, there is also interest in the use of SPC for medical monitoring procedure, for example to assess performance hospitals or medical procedures. Unlike the industrial data, which are usually composed of homogeneous observations (equipment, parts etc.), in the medical field observations generally come from people who have different characteristics, such as age, sex and risk factors. In this sense, some authors propose CUSUM control charts for this scenario, where heterogeneity is included by a regression structure incorporated into the model. These are called CUSUM graphs risk-adjusted (RA CUSUM). Sego et al (2009) proposed a risk-adjusted CUSUM chart for monitoring lifetimes based on accelerated failure time model, RAST calling this CUSUM chart (CUSUM charts set for the risk survival times). This proposal allows to detect, for example, changes in average life expectancy throughout the monitoring period. The use of these charts in practice dependent on obtaining adequate control limits. Moreover, when applied RAST CUSUM data from retrospective studies it is necessary to specify the change of intensity to be identified. From this perspective, this paper proposes a procedure for obtaining via bootstrap control limit for the survival time monitoring through RAST CUSUM chart. In addition, some procedures are studied to identify the real magnitude of change in a process in retrospective studies. Simulation studies are conducted to evaluate the proposals. To illustrate the results was considered an application with real data from the literature on hospitalized patients with myocardial infarction. |
|---|