Reliability between Cardiovascular Risk Assessment Tools: A Pilot Study

Abstract Background The prevention of cardiovascular disease (CVD) is important in clinical practice due to its high morbidity and mortality. Different guidelines have recommended the use of different cardiovascular risk assessment tools, which may have implications on therapeutic decisions. Objective To evaluate the agreement rate between the Framingham risk score (FRS) and the Systematic Coronary Risk Evaluation (SCORE) tool on CVD risk assessment in disease-free subjects. Methods Cross-sectional study with a sample of 51 subjects treated at the outpatient clinic of a university hospital in Brazil between January 2014 and January 2015. The FRS and two versions of the European SCORE (SCORE-High and SCORE-Low) were used to assess CVD risk; patients were classified as low/moderate risk (< 20% and <5%, respectively) or high risk (≥ 20% and ≥5%, respectively). The agreement rate was evaluated using kappa statistics, a test for interrater reliability that ranges from -1 to 1, and results above 0.6 represent a high agreement rate. Results The FRS classified a higher proportion of subjects as high risk for CVD (35.3% [18/51] vs. 23.5% [12/51] with the SCORE-High and 13.7% [7/51] with SCORE-Low). However, there was a high agreement rate between FRS and SCORE-High (k=0.628). The agreement between FRS and SCORE-Low was poor (k=0.352). Conclusions There was a high agreement rate between FRS and SCORE-High in cardiovascular risk assessment in the study sample. (Int J Cardiovasc Sci. 2020; [online].ahead print, PP.0-0)

Degree of Agreement between Cardiovascular Risk Stratification Tools / Grau de Concordância entre Instrumentos de Estratificação de Risco Cardiovascular

Abstract Background: Cardiovascular disease (CVD) is the leading cause of morbidity and mortality in Brazil, and primary prevention care may be guided by risk stratification tools. The Framingham (FRS) and QRISK-2 (QRS) risk scores estimate 10-year overall cardiovascular risk in asymptomatic individuals, but the instrument of choice may lead to different therapeutic strategies. Objective: To evaluate the degree of agreement between FRS and QRS in 10-year overall cardiovascular risk stratification in disease-free individuals. Methods: Cross-sectional, observational, descriptive and analytical study in a convenience sample of 74 individuals attending the outpatient care service of a university hospital in Brazil between January 2014 and January 2015. After application of FRS and QRS, patients were classified in low/moderate risk (< 20%) or high risk (≥ 20%). Results: The proportion of individuals classified as at high risk was higher in FRS than in QRS (33.7% vs 21.6%). A synergic effect of male gender with systemic arterial hypertension was observed in both tools, and with for geriatric age group in QRS (p < 0.05) in high-risk stratum. The Kappa index was 0.519 (95%CI = 0.386-0.652; p < 0.001) between both instruments. Conclusion: There was a moderate agreement between FRS and QRS in estimating 10-year overall cardiovascular risk. The risk scores used in this study can identify synergism between variables, and their behavior is influenced by the population in which it was derived. It is important to recognize the need for calibrating risk scores for the Brazilian population.

Resumo Fundamento: A doença cardiovascular (DCV) é a principal causa de morbimortalidade no Brasil, e a prevenção primária pode ser direcionada com ferramentas que estratificam o risco. Os escores de Framingham (ERF) e QRISK-2 (ERQ) estimam o risco cardiovascular (RCV) global em 10 anos em indivíduos assintomáticos, mas a escolha do instrumento pode implicar em terapêuticas distintas. Objetivo: Observar o grau de concordância entre o ERF e o ERQ, na estratificação do risco cardiovascular global em 10 anos, nos indivíduos livres da doença. Métodos: Estudo transversal, observacional, descritivo e analítico, com uma amostra de conveniência de 74 indivíduos, atendidos em um ambulatório de ensino de um hospital universitário brasileiro, no sul do país, de janeiro de 2014 a janeiro de 2015. O ERF e o ERQ foram aplicados nos pacientes, que foram classificados em baixo/moderado (< 20%) ou alto risco (≥ 20%). Resultados: A proporção de indivíduos classificados no estrato de alto risco foi superior no ERF que no ERQ (33,7% vs 21,6%), sendo identificado efeito sinérgico do gênero masculino com hipertensão arterial sistêmica nas duas ferramentas, e com faixa etária geriátrica no ERQ (p < 0,05) nesse estrato de risco. O índice de concordância Kappa entre os dois escores foi igual a 0,519 (IC95% = 0,386-0,652; p < 0,001). Conclusão: Houve concordância moderada entre o ERF e o ERQ, na estimativa de RCV global em 10 anos. Os escores utilizados podem identificar sinergismo entre as variáveis, e têm comportamento influenciado pela população na qual foram originados. É importante reconhecer a necessidade de escores calibrados para a população brasileira.