In database studies, the true value of the outcome of interest is measured with error through indicators implemented by appropriate algorithms. Differential sensitivity of the study outcome across exposure groups may severely bias measures of association in both directions. We propose a statistical test for differential sensitivity across exposure groups based on estimates of positive predictive value (PPV) obtained by validating a sample of cases.
A statistical test for differential sensitivity of study outcome / Limoncella, G; Grilli, L; Dreassi, E; Rampichini, C; Platt, RW; Gini, R. - In: PHARMACOEPIDEMIOLOGY AND DRUG SAFETY. - ISSN 1053-8569. - ELETTRONICO. - 32:(2023), pp. 574-575. (Intervento presentato al convegno ICPE 2023 tenutosi a Halifax nel August 23-27).
A statistical test for differential sensitivity of study outcome
Limoncella, G
;Grilli, L;Dreassi, E;Rampichini, C;
2023
Abstract
In database studies, the true value of the outcome of interest is measured with error through indicators implemented by appropriate algorithms. Differential sensitivity of the study outcome across exposure groups may severely bias measures of association in both directions. We propose a statistical test for differential sensitivity across exposure groups based on estimates of positive predictive value (PPV) obtained by validating a sample of cases.
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
