Elaboração e validação de questionário para análise do uso de serviços de saúde da atenção básica e relação com fatores de risco à saúde
Carregando...
Arquivos
Data
2018-02-22
Autores
Ferro, Izabela dos Santos [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Investigações referentes à avaliação dos gastos com saúde são importantes, entretanto os métodos existentes apresentam limitações em suas aplicações, como subestimação dos valores e divergência, erros de preenchimento, limitações de ordem burocrática e falta de validação científica em relação a questionários. Objetivo: Elaborar e verificar a validade e a confiabilidade de um questionário para a avaliação do uso de serviços de saúde na atenção primária do Sistema Único de Saúde (SUS) em adultos com idade ≥ 50 anos. Métodos: Amostra composta por 64 usuários do SUS da cidade de Presidente Prudente – SP. O questionário proposto avaliou a utilização de serviços de saúde da atenção básica considerando as variáveis exames, consultas e medicamentos, nos períodos de 30 e 60 dias retroativos à data da coleta, e sua validade foi testada segundo informações contidas nos prontuários dos participantes. Para a confiabilidade foi utilizado o método “teste-reteste”, considerando um intervalo de 7 dias após a primeira avaliação. Posteriormente, os resultados foram convertidos em valores monetários e comparados segundo indicadores de saúde: idade, sexo, atividade física, composição corporal, condição econômica e escolaridade, ademais as formas de conversão foram apresentadas. Quanto à análise estatística, para observar a validade, confiabilidade, concordância numérica e consistência interna entre os dados, bem como comparação com indicadores de saúde foram utilizados os seguintes testes: Coeficiente de Correlação Intraclasse (CCI), técnica de alfa (α) de Cronbach, correlação de Spearman (rho) e teste de Wilcoxon. Para a relação entre subitens do questionário proposto e produto final foi utilizada a correlação de Pearson (r). Sendo os valores de significância estatística considerados significativos quando inferiores a 5% e o software utilizado foi o BioEstat (versão 5.0). Resultados: Quando analisada a correlação entre a quantidade total de gastos reportados através do questionário com as variáveis analisadas, consultas, exames e medicamentos, foram observados valores de correlação fraca para consultas (r=0,305) quando retroagidos 30 dias e correlação perfeita para medicamentos (r=1,00) para os dois períodos retroativos. Para a correlação entre questionário e prontuário foi observado para o período de 30 dias e 60 dias valores de CCI de fraco a moderado para consultas (CCI=0,429; CCI=0,576) exames (CCI=0,181; CCI= 0,020) e medicamentos (CCI=0,505; CCI= 0,642). Para a correlação da diferença entre valores reportados nos questionários e valores de prontuários com as variáveis descritivas, foi observada diferença estatisticamente significante para “Medicamentos 60 dias” e idade (rho=0,267) e condição econômica (rho=-0,386). Foi observada correlação fraca entre atividade física ocupacional e exames (r=0,259) somente xi para questionário, para o período retroativo de 60 dias. Para diferença entre os valores brutos de gastos reais do questionário proposto e do prontuário, houve diferença estatisticamente significativa para a variável medicamentos, quando retroagidos 30 e 60 dias (p-valor=0,016; p-valor=0,001). Para a confiabilidade intra-avaliador, retroagindo 30 dias foi observado valores de CCI excelente para consultas (CCI=0,805) e fraco (CCI=0,257) para medicamentos, e para 60 dias, foi observado CCI moderado para consultas (CCI=0,543), e fraco para exames e medicamentos (CCI=-0,104 e CCI=0,338). Para a confiabilidade interavaliadores retroagindo 30 ou 60 dias foram observados valores de CCI fraco para todas as variáveis analisadas. Conclusão: O questionário proposto apresentou validade moderada, porém não obteve resultados consistentes para a análise de confiabilidade. Mesmo existindo limitações no seu processo de elaboração, há uma relevância prática no instrumento no que diz respeito a sua facilidade de aplicação no âmbito da saúde e ainda, uma maior efetividade na análise da variável “consultas”.
Investigations related to the evaluation of health spending are important, however, existing methods present limitations in their applications, such as underestimation of values and divergence, errors of filling, bureaucratic limitations and lack of scientific validation regarding questionnaires. Aim: To elaborate and verify the validity and reliability of a questionnaire for the evaluation of the use of health services in primary care of the Unified Health System (UHS) in adults aged ≥ 50 years. Methods: Sample composed of 64 UHS users from the city of Presidente Prudente - SP. The questionnaire proposed evaluated the use of basic health care services considering the variables exams, consultations and medications, in the periods of 30 and 60 days retroactive to the date of collection, and their validity was tested according to information contained in the participants' medical records. For reliability, the "test-retest" method was used, considering an interval of 7 days after the first evaluation. Subsequently, the results were converted into monetary values and compared according to health indicators: age, sex, physical activity, body composition, economic condition and schooling, in addition the forms of conversion were presented. In order to observe the validity, reliability, numerical concordance and internal consistency between the data, as well as comparison with health indicators, the following tests were used: Intraclass Correlation Coefficient (ICC), Cronbach's alpha (α) technique, Spearman correlation (rho) and Wilcoxon's test. For the relationship between sub items of the proposed questionnaire and final product, the Pearson correlation (r) was used. Statistical significance’s values were considered significant when less than 5% and the software used was BioEstat (version 5.0). Results: When analyzing the correlation between the total amount of expenses reported through the questionnaire with the analyzed variables, consultations, exams and medications, low correlation values were observed for consultations (r=0.305) when retroactive 30 days and perfect correlation for medications (r=1.00) for the two retroactive periods. For the correlation between the questionnaire and the medical record, it was observed for the period of 30 days and 60 days the values of low to moderate ICC for consultations (ICC=0,429; ICC=0.576), exams (ICC=0,181; CCI=0,020) and medications (ICC=0.505, ICC=0.642). For the correlation of the difference between the values reported in the questionnaires and the values of the medical records with the descriptive variables, a statistically significant difference was observed for "Medications 60 days" and age (rho=0.267) and economic condition (rho=-0.386). There was a low correlation between occupational physical activity and exams (r=0.259) for questionnaire only, for the retroactive period of 60 days. For the xiii difference between the raw data of the actual spending of the proposed questionnaire and the medical record, there was a statistically significant difference for the variable drugs, when 30 and 60 days retrograde (p-value=0.016, p-value=0.001). For the intra-rater reliability, 30 days were observed values of ICC excellent for consultations (ICC=0.805) and low for medications (ICC=0.257), and for 60 days, moderate ICC was observed for consultations (ICC=0.543), and low for exams and medications (ICC=-0.104 and ICC=0.338). For the interrater reliability of 30 or 60 days, low ICC values were observed for all variables analyzed. Conclusion: The proposed questionnaire had a moderate validity, but did not obtain consistent results for the reliability analysis. Although there are limitations in its elaboration process, there is a practical relevance in the instrument regarding its ease of application in the health area and also, a greater effectiveness in the analysis of the variable "consultations".
Investigations related to the evaluation of health spending are important, however, existing methods present limitations in their applications, such as underestimation of values and divergence, errors of filling, bureaucratic limitations and lack of scientific validation regarding questionnaires. Aim: To elaborate and verify the validity and reliability of a questionnaire for the evaluation of the use of health services in primary care of the Unified Health System (UHS) in adults aged ≥ 50 years. Methods: Sample composed of 64 UHS users from the city of Presidente Prudente - SP. The questionnaire proposed evaluated the use of basic health care services considering the variables exams, consultations and medications, in the periods of 30 and 60 days retroactive to the date of collection, and their validity was tested according to information contained in the participants' medical records. For reliability, the "test-retest" method was used, considering an interval of 7 days after the first evaluation. Subsequently, the results were converted into monetary values and compared according to health indicators: age, sex, physical activity, body composition, economic condition and schooling, in addition the forms of conversion were presented. In order to observe the validity, reliability, numerical concordance and internal consistency between the data, as well as comparison with health indicators, the following tests were used: Intraclass Correlation Coefficient (ICC), Cronbach's alpha (α) technique, Spearman correlation (rho) and Wilcoxon's test. For the relationship between sub items of the proposed questionnaire and final product, the Pearson correlation (r) was used. Statistical significance’s values were considered significant when less than 5% and the software used was BioEstat (version 5.0). Results: When analyzing the correlation between the total amount of expenses reported through the questionnaire with the analyzed variables, consultations, exams and medications, low correlation values were observed for consultations (r=0.305) when retroactive 30 days and perfect correlation for medications (r=1.00) for the two retroactive periods. For the correlation between the questionnaire and the medical record, it was observed for the period of 30 days and 60 days the values of low to moderate ICC for consultations (ICC=0,429; ICC=0.576), exams (ICC=0,181; CCI=0,020) and medications (ICC=0.505, ICC=0.642). For the correlation of the difference between the values reported in the questionnaires and the values of the medical records with the descriptive variables, a statistically significant difference was observed for "Medications 60 days" and age (rho=0.267) and economic condition (rho=-0.386). There was a low correlation between occupational physical activity and exams (r=0.259) for questionnaire only, for the retroactive period of 60 days. For the xiii difference between the raw data of the actual spending of the proposed questionnaire and the medical record, there was a statistically significant difference for the variable drugs, when 30 and 60 days retrograde (p-value=0.016, p-value=0.001). For the intra-rater reliability, 30 days were observed values of ICC excellent for consultations (ICC=0.805) and low for medications (ICC=0.257), and for 60 days, moderate ICC was observed for consultations (ICC=0.543), and low for exams and medications (ICC=-0.104 and ICC=0.338). For the interrater reliability of 30 or 60 days, low ICC values were observed for all variables analyzed. Conclusion: The proposed questionnaire had a moderate validity, but did not obtain consistent results for the reliability analysis. Although there are limitations in its elaboration process, there is a practical relevance in the instrument regarding its ease of application in the health area and also, a greater effectiveness in the analysis of the variable "consultations".
Descrição
Palavras-chave
Validação, Gastos com saúde, Indicadores de saúde, Validation, Healthcare expenditures, Health indicators