Atualização do modelo FRAX para o Brasil
Data
2024-08-29
Autores
Albergaria, Ben-Hur 
Orientadores
Lazaretti-Castro, Marise
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Introdução: O manejo clínico da osteoporose foi revolucionado pela mudança nos critérios de indicação para tratamento, inicialmente baseados na medida de risco relativo dada pelo T-score definido pela DXA (Absorciometria por duplo feixe de raios-X , do inglês, Dual energy X-ray absorptiometry), para critérios atuais baseados na medida de risco absoluto com base em ferramentas de predição clínica. Dentre as ferramentas disponíveis e validadas, o FRAX (Fracture Risk Assessment Tool) se destaca em vários aspectos, incluindo sua capacidade de personalizar predições de risco de fraturas osteoporóticas para uma população-alvo específica. Este processo, conhecido como calibração tem, portanto, importância fundamental e demanda atualização periódica com a aquisição de dados de alta qualidade relativos às taxas de incidência de fratura de quadril daquela população. O objetivo da primeira etapa desta pesquisa foi realizar um estudo epidemiológico multicêntrico descritivo , o Estudo Brasileiro de Validação em Osteoporose- (BRAVOS), para estimar a incidência de fraturas de quadril conduzido especificamente para atualizar os dados do FRAX Brasil. Em uma segunda etapa, recalibrar o modelo FRAX Brasil utilizando os dados do estudo BRAVOS e comparar este novo modelo com o modelo original. Materiais e métodos: Estudo retrospectivo e observacional incluindo todos pacientes com idade ≥ 50 anos internados em hospitais devido uma fratura de quadril residentes em três cidades (Belém, Joinville e Vitória) de áreas geográficas representativas do Brasil de 2010 a 2012 a partir de prontuários médicos. Analisamos taxas de incidência (brutas e padronizadas por idade e sexo) para fraturas de quadril. A seguir, para a atualização do modelo FRAX Brasil, as taxas de fratura de quadril das três cidades foram combinadas, ponderadas pela população de cada região. Para outras fraturas maiores, as taxas de incidência para o Brasil foram estimadas usando proporções suecas para quadril e outras fraturas osteoporóticas maiores. As estimativas de mortalidade foram obtidas da base de dados da Organização das Nações Unidas (ONU). Resultados:1.025 (310 em homens e 715 em mulheres) fraturas de quadril na população com mais de 50 anos de idade foram identificadas nas três cidades. A taxa bruta de incidência de fratura de quadril foi de 103,3/100.000 (intervalo de confiança de 95% [IC = 97,0; 109,7), sendo 77,4/100.000 (IC 95% = 68,8; 86,0) em homens e 125,2/100.000 (IC 95% = 116,0; 134,4) em mulheres. Incidência padronizada para a idade e o sexo foi de 105,9 casos por 100.000 pessoas por ano (IC 95% = 99,4; 112,4); 78,5 casos por 100.000 (IC 95% = 69,8;87,3) em homens e 130,6 casos por 100.000 em mulheres (IC 95% = 121,0; 140,2) por ano. Belém, localizada na região equatorial (latitude 1° 27′ S), teve incidência bruta e ajustada por idade significativamente menor que Joinville (latitude 26° 18′ S) e Vitória (latitude 20° 19′ S), que não eram diferentes entre si. A incidência de fraturas aumentou exponencialmente com a idade, e as mulheres tiveram cerca de duas vezes mais risco de fraturas do que os homens. Comparado ao modelo FRAX original, o modelo atualizado apresentou menores probabilidades de fratura em 10 anos em homens e mulheres em todas as idades. Entretanto, houve também uma correlação muito próxima nas probabilidades de fratura entre o modelo original e o atualizado (r > 0,99). Conclusões: A fratura de quadril acomete principalmente mulheres idosas e apresenta grande variabilidade de incidência entre as diferentes regiões do Brasil. A incidência de fraturas de quadril no Brasil diferiu acentuadamente daquela relatada anteriormente, de modo que as diretrizes nacionais e o modelo FRAX para o Brasil devem ser revisados. As disparidades entre os modelos FRAX original e atualizado indicam a importância de atualizar os modelos FRAX específicos do país com o advento de mudanças significativas na epidemiologia das fraturas.
Introduction: The clinical management of osteoporosis was revolutionized by the change in indication criteria for treatment, initially based on the measure of relative risk given by the T-score defined by DXA, to current criteria based on the measure of absolute risk based on clinical prediction tools. Among the available and validated tools, FRAX (Fracture Risk Assessment Tool) stands out in several aspects, including its ability to personalize osteoporotic fracture risk predictions for a specific target population. This process, known as calibration, is therefore of fundamental importance and requires periodic updating with the acquisition of high-quality data relating to the incidence rates of hip fractures in that population. The objective of the first stage of this research was to carry out a multicentric descriptive epidemiological study, the Brazilian Osteoporosis Validation Study (BRAVOS), to estimate the incidence of hip fractures, conducted specifically to update the FRAX Brazil data. In a second step, recalibrate the FRAX Brazil model using data from the BRAVOS study and compare this new model with the original model. Materials and methods: Retrospective and observational study including all patients aged ≥ 50 years admitted to hospitals due to a hip fracture residing in three cities (Belem, Joinville and Vitória) in representative geographic areas of Brazil from 2010 to 2012 based on medical records. We analyzed incidence rates (crude and standardized by age and sex) for hip fractures. Next, to update the FRAX Brazil model, the hip fracture rates of the three cities were combined, weighted by the population of each region. For other major fractures, incidence rates for Brazil were estimated using Swedish proportions for hip and other major osteoporotic fractures. Mortality estimates were obtained from the United Nations (UN) database. Results: 1,025 (310 in men and 715 in women) hip fractures in the population over 50 years of age were identified in the three cities. The crude incidence rate of hip fracture was 103.3/100,000 (95% confidence interval [CI = 97.0; 109.7), being 77.4/100,000 (95% CI = 68.8; 86.0) in men and 125.2/100,000 (95% CI = 116.0; 134.4) in women. Age- and sex-standardized incidence was 105.9 cases per 100,000 people per year (95% CI = 99.4; 112.4); 78.5 cases per 100,000 (95% CI = 69.8; 87.3) in men and 130.6 cases per 100,000 in women (95% CI = 121.0; 140.2) per year. Belem, located in the equatorial region (latitude 1° 27′ S), had a crude and age-adjusted incidence significantly lower than Joinville (latitude 26° 18′ S) and Vitória (latitude 20° 19′ S), which were not different between yes. The incidence of fractures increased exponentially with age, and women had about twice the risk of fractures as men. Compared to the original FRAX model, the updated model showed lower 10-year fracture odds in men and women at all ages. Nevertheless, there was also a very close match in fracture probabilities between the original and updated models (r > 0.99). Conclusions: Hip fracture mainly affects elderly women and presents great variability in incidence between the different regions in Brazil. The incidence of hip fractures in Brazil differed markedly from that reported previously, so that national guidelines and the FRAX model for Brazil should be revised. The disparities between the original and updated FRAX models indicate the importance of updating country-specific FRAX models with the advent of significant changes in fracture epidemiology.
Introduction: The clinical management of osteoporosis was revolutionized by the change in indication criteria for treatment, initially based on the measure of relative risk given by the T-score defined by DXA, to current criteria based on the measure of absolute risk based on clinical prediction tools. Among the available and validated tools, FRAX (Fracture Risk Assessment Tool) stands out in several aspects, including its ability to personalize osteoporotic fracture risk predictions for a specific target population. This process, known as calibration, is therefore of fundamental importance and requires periodic updating with the acquisition of high-quality data relating to the incidence rates of hip fractures in that population. The objective of the first stage of this research was to carry out a multicentric descriptive epidemiological study, the Brazilian Osteoporosis Validation Study (BRAVOS), to estimate the incidence of hip fractures, conducted specifically to update the FRAX Brazil data. In a second step, recalibrate the FRAX Brazil model using data from the BRAVOS study and compare this new model with the original model. Materials and methods: Retrospective and observational study including all patients aged ≥ 50 years admitted to hospitals due to a hip fracture residing in three cities (Belem, Joinville and Vitória) in representative geographic areas of Brazil from 2010 to 2012 based on medical records. We analyzed incidence rates (crude and standardized by age and sex) for hip fractures. Next, to update the FRAX Brazil model, the hip fracture rates of the three cities were combined, weighted by the population of each region. For other major fractures, incidence rates for Brazil were estimated using Swedish proportions for hip and other major osteoporotic fractures. Mortality estimates were obtained from the United Nations (UN) database. Results: 1,025 (310 in men and 715 in women) hip fractures in the population over 50 years of age were identified in the three cities. The crude incidence rate of hip fracture was 103.3/100,000 (95% confidence interval [CI = 97.0; 109.7), being 77.4/100,000 (95% CI = 68.8; 86.0) in men and 125.2/100,000 (95% CI = 116.0; 134.4) in women. Age- and sex-standardized incidence was 105.9 cases per 100,000 people per year (95% CI = 99.4; 112.4); 78.5 cases per 100,000 (95% CI = 69.8; 87.3) in men and 130.6 cases per 100,000 in women (95% CI = 121.0; 140.2) per year. Belem, located in the equatorial region (latitude 1° 27′ S), had a crude and age-adjusted incidence significantly lower than Joinville (latitude 26° 18′ S) and Vitória (latitude 20° 19′ S), which were not different between yes. The incidence of fractures increased exponentially with age, and women had about twice the risk of fractures as men. Compared to the original FRAX model, the updated model showed lower 10-year fracture odds in men and women at all ages. Nevertheless, there was also a very close match in fracture probabilities between the original and updated models (r > 0.99). Conclusions: Hip fracture mainly affects elderly women and presents great variability in incidence between the different regions in Brazil. The incidence of hip fractures in Brazil differed markedly from that reported previously, so that national guidelines and the FRAX model for Brazil should be revised. The disparities between the original and updated FRAX models indicate the importance of updating country-specific FRAX models with the advent of significant changes in fracture epidemiology.
Descrição
Citação
ALBERGARIA, Ben-Hur. Um novo modelo FRAX para o Brasil. 2024. 90 f. Tese (Doutorado em Endocrinologia Clínica) - Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, 2024