COMPARISON OF TWO METHODS TO ASSESS CAUSALITY OF ADVERSE DRUG REACTIONS
Comparación de dos métodos de evaluación de causalidad de sospechas de reacciones adversas a medicamentos (RAM). 2003-2009.
Author
Kyonen, Mónica; particular
Folatre, Isabel; 2Médico Onco-Hematólogo Infantil, Instituto Pediátrico, Universidad Austral de Chile. Valdivia, Chile
Lagos, Ximena; Químico Farmacéutico, Docente Escuela de Química y Farmacia, Universidad Austral de Chile.
vargas, silvia; Químico Farmacéutico, Clínica Alemana, Valdivia, Chile
Abstract
Background: Assessment of causality is an essential part of pharmacological surveillance. Aim: To compare the causality assessment of suspected adverse drug reactions (ADRs) using algorithms proposed by Karch-Lasagna (K & L) and the World Health Organization (WHO). Material and methods: All spontaneous reports of suspected ADRs in a pediatric ward of a regional hospital were included. Causality was categorized as definitive, probable, possible, conditional or unlikely. The agreement between K & L and WHO algorithms was assessed using binomial test proportions and kappa coefficients. Results: One hundred thirty reports of ADRs in 126 patients aged 2 to 11 years were analyzed. The suspected medications were antineoplastic drugs in 59% of cases and antimicrobials in 23%. The most common affected system was the skin and appendages in 35%. Using K & L algorithm, causality was categorized as definitive in 10% of cases, probable in 28.5%, possible in 35.4%, conditional in 23.1% and unlikely in 3.0%. Using WHO algorithm, the figures were 2.3, 34.6, 59.2, 2.3 and 1.5%, respectively. The degree of agreement between K & L and WHO algorithms was 32.3% (kappa = 0.004). Conclusions: K & L attributed a higher level of ADR causality than WHO algorithm. Background: Assessment of causality is an essential part of pharmacological surveillance. Aim: To compare the causality assessment of suspected adverse drug reactions (ADRs) using algorithms proposed by Karch-Lasagna (K & L) and the World Health Organization (WHO). Material and methods: All spontaneous reports of suspected ADRs in a pediatric ward of a regional hospital were included. Causality was categorized as definitive, probable, possible, conditional or unlikely. The agreement between K & L and WHO algorithms was assessed using binomial test proportions and kappa coefficients. Results: One hundred thirty reports of ADRs in 126 patients aged 2 to 11 years were analyzed. The suspected medications were antineoplastic drugs in 59% of cases and antimicrobials in 23%. The most common affected system was the skin and appendages in 35%. Using K & L algorithm, causality was categorized as definitive in 10% of cases, probable in 28.5%, possible in 35.4%, conditional in 23.1% and unlikely in 3.0%. Using WHO algorithm, the figures were 2.3, 34.6, 59.2, 2.3 and 1.5%, respectively. The degree of agreement between K & L and WHO algorithms was 32.3% (kappa = 0.004). Conclusions: K & L attributed a higher level of ADR causality than WHO algorithm.