Avaliação da confiabilidade interobservadores e validação de construto do formulário POLST (physician orders for life-sustaining treatment) para documentação de preferências de cuidados no fim da vida
Carregando...
Arquivos
Data
2020-12-08
Autores
Lovadini, Gustavo Bigaton [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
No Brasil a maior parte dos profissionais e instituições de saúde ainda se encontram longe de constituírem uma rotina de discussão sobre preferências de cuidados no fim da vida junto a pacientes com prognóstico reservado. Esta corresponde a uma grande lacuna na atenção à saúde em nosso país, a qual frequentemente se associa a sofrimento evitável de pacientes e familiares, bem como ao mau uso dos recursos disponíveis no Sistema Único de Saúde (SUS). Em 1991 nos Estados Unidos da América, foi iniciado um programa de discussão de preferências de cuidados no fim da vida denominado POLST (Physician Orders for Life-Sustaining Treatment). Trata-se de um sistema coordenado para evocar, documentar e comunicar as preferências de pacientes/familiares quanto a tratamentos prolongadores da vida para enfermos com expectativa de vida reduzida, de forma a orientar as condutas médicas em diferentes cenários de cuidados. Em nossa revisão de literatura identificamos a inexistência de estudos de avaliação da validade de construto ou da confiabilidade interobservadores desse formulário. Portanto, com o intuito de produzir dados que subsidiem a implantação do paradigma POLST no Brasil, a presente pesquisa teve como objetivo avaliar essas propriedades psicométricas desse instrumento, já adaptado para o Brasil, de acordo com o referencial do COSMIN (Consensus-based Standards for the selection of health Measurement Instruments). Dessa forma obtivemos como resultado dois artigos. O primeiro artigo avaliou a confiabilidade interobservadores do instrumento, ou seja, a estabilidade das prescrições de tratamentos médicos no fim da vida documentadas em formulários POLST quando pesquisadores diferentes entrevistavam em dias distintos um mesmo participante da pesquisa seguindo uma abordagem padronizada. De um total de 64 participantes entrevistados, houve 5 (8%) casos de discordância na comparação entre avaliações envolvendo ao menos uma das prescrições médicas de cuidados no fim da vida. A estatística Kappa revelou níveis elevados de concordância interobservadores em todos os itens do instrumento. O segundo artigo analisou a validade de construto do formulário POLST, ou seja, se as prescrições médicas documentadas por meio do instrumento refletiam as preferências de cuidados dos pacientes estudados conforme registaradas seguindo uma outra abordagem. A hipótese testada foi que a prescrição de cuidados no final da vida contida nos formulários POLST apresentaria elevado nível de concordância com a documentação em texto livre de entrevistas realizada por um médico experiente em cuidados paliativos seguindo a proposta de Sudore & Fried para discussões de planejamento antecipado de cuidados. Dos 62 pacientes entrevistados, houve 7 (11%) casos de discordância relativas a ao menos uma das prescrições de tratamentos no fim da vida, sendo que a análise dos dados utilizando da estatística Kappa novamente evidenciou elevada concordância entre as duas abordagens. Em conclusão, por um lado os resultados dos estudos que compõem essa tese fornecem evidências acerca da confiabilidade interobservadores e da validade de construto do formulário POLST, o que certamente contribuirá para a futura implantação desse paradigma no Brasil. Por outro lado, o fato de terem sido observados casos de discordâncias em ambas as pesquisas ressalta a complexidade envolvida em qualquer processo de planejamento antecipado de cuidados e aponta para a necessidade de que prescrições de tratamentos médicos no fim da vida registradas em formulários POLST sejam revisadas com frequência de modo a confirmar que as mesmas são consistentes com as preferências atuais dos pacientes.
In Brazil, most healthcare professionals and institutions still have not included discussions about preferences of care at the end of life with patients with decreased life expectancy as part of their daily routine. This represents a major gap in the care at the close of life in our country, which is often associated with the avoidable suffering of patients and their loved ones, as well as the misuse of Public Health resources. In 1991 the Physician Orders for Life-Sustaining Treatment (POLST) was started in the United States. It is a coordinated system to elicit, document and communicate patients’ and families’ preferences regarding life-sustaining treatments for individuals with reduced life expectancy across a variety of different healthcare settings. In our literature review we identified that there were no studies evaluating the construct validity or the inter-rater reliability of the POLST form. Therefore, with the intent of producing data to support the implementation of POLST in Brazil, we aimed to evaluate the inter-rater reliability and the construct validity of the POLST form that we had previously adapted to Brazil. We followed the COSMIN (Consensus-based Standards for the selection of health Measurement Instruments) guidelines when designing our methods and reporting our findings. As a result, we produced two articles. The first article evaluated the interrater reliability of the POLST form, i.e., the stability of medical orders for treatments at the end of life recorded in those forms when two different researchers interviewed the same participant on different days following a standardized approach. Out of 64 participants that we interviewed, there were 5 (8%) cases of disagreement between the interviews regarding at least one medical order documented in the POLST forms, which yielded Kappa statistics with high levels of agreement for all items of the instrument. The second article evaluated the construct validity of the POLST form, i.e., whether the medical orders recorded in POLST forms reflected preferences of care documented following another approach. We tested the hypothesis that medical orders in POLST forms would show high levels of agreement with the free-form text documentation of an advance care planning conversation performed following the proposal of Sudore & Fried by and experienced palliative care physician. Out of 62 patients that we interviewed, there were 7 (11%) cases of disagreement between at least one medical order recorded in POLST forms and the free-form text documentation of the other advance care planning conversation, which again yielded Kappa statistics with high levels of agreement for all items of the POLST form. In conclusion, on the one hand our results provide evidence on the interrater reliability and construct validity of the POLST form, which will be instrumental for the future implementation of the POLST paradigm in Brazil. On the other hand, the finding of cases of discordances regarding medical orders in POLST forms highlights the complexity involved in any advance care planning process and points towards the need to revise medical orders in POLST forms frequently to confirm that they accurately reflect patients’ current preferences for care.
In Brazil, most healthcare professionals and institutions still have not included discussions about preferences of care at the end of life with patients with decreased life expectancy as part of their daily routine. This represents a major gap in the care at the close of life in our country, which is often associated with the avoidable suffering of patients and their loved ones, as well as the misuse of Public Health resources. In 1991 the Physician Orders for Life-Sustaining Treatment (POLST) was started in the United States. It is a coordinated system to elicit, document and communicate patients’ and families’ preferences regarding life-sustaining treatments for individuals with reduced life expectancy across a variety of different healthcare settings. In our literature review we identified that there were no studies evaluating the construct validity or the inter-rater reliability of the POLST form. Therefore, with the intent of producing data to support the implementation of POLST in Brazil, we aimed to evaluate the inter-rater reliability and the construct validity of the POLST form that we had previously adapted to Brazil. We followed the COSMIN (Consensus-based Standards for the selection of health Measurement Instruments) guidelines when designing our methods and reporting our findings. As a result, we produced two articles. The first article evaluated the interrater reliability of the POLST form, i.e., the stability of medical orders for treatments at the end of life recorded in those forms when two different researchers interviewed the same participant on different days following a standardized approach. Out of 64 participants that we interviewed, there were 5 (8%) cases of disagreement between the interviews regarding at least one medical order documented in the POLST forms, which yielded Kappa statistics with high levels of agreement for all items of the instrument. The second article evaluated the construct validity of the POLST form, i.e., whether the medical orders recorded in POLST forms reflected preferences of care documented following another approach. We tested the hypothesis that medical orders in POLST forms would show high levels of agreement with the free-form text documentation of an advance care planning conversation performed following the proposal of Sudore & Fried by and experienced palliative care physician. Out of 62 patients that we interviewed, there were 7 (11%) cases of disagreement between at least one medical order recorded in POLST forms and the free-form text documentation of the other advance care planning conversation, which again yielded Kappa statistics with high levels of agreement for all items of the POLST form. In conclusion, on the one hand our results provide evidence on the interrater reliability and construct validity of the POLST form, which will be instrumental for the future implementation of the POLST paradigm in Brazil. On the other hand, the finding of cases of discordances regarding medical orders in POLST forms highlights the complexity involved in any advance care planning process and points towards the need to revise medical orders in POLST forms frequently to confirm that they accurately reflect patients’ current preferences for care.
Descrição
Palavras-chave
POLST, Reprodutibilidade dos testes, Estudos de validação, Planejamento antecipado de cuidados, Diretivas antecipadas, Reproducibility of tests, Validation studies, Advance Care Planning, Advance Directives.