[Skip to Navigation]
September 1990

Calculation of Sensitivity and Specificity

Author Affiliations

Faculty of Rehabilitation Medicine Department of Pediatrics University of Alberta Edmonton, Alberta Canada T6G 2G4

Am J Dis Child. 1990;144(9):958-959. doi:10.1001/archpedi.1990.02150330016010

Sir.—The November 1989 issue of AJDC included an article by Susan Harris.1 We believe that Dr Harris used an inappropriate method in the calculation of the sensitivity and specificity rates for the Movement Assessment of Infants and Bayley Mental and Motor Scales, resulting in invalid rates. Harris had 3-year outcome data on 229 infants, including 36 with cerebral palsy, 75 who were developmentally delayed, and 118 who were normal. The 75 infants who were delayed at 3 years (32%) were not included in the analyses. Although these 75 infants presented diagnostic problems with their variable outcomes,2,3 they cannot be excluded from the analyses for the sake of simplicity.

Sensitivity and specificity were accurately defined as, respectively, the percentage of abnormal cases correctly identified and the percentage of normal cases correctly identified. Obviously, sensitivity and specificity will vary depending on the value at which the screening test is

Add or change institution