BACKGROUND: Quality control methods are key components of dietary assessment, but have rarely been evaluated. METHODS: One hundred forty-four 8-10-year-old African-American girls at three field centers completed two 24-h dietary recalls at baseline before a pilot weight gain prevention intervention (one recall collected in-person and one by telephone). The dietary recall data were initially reviewed by the dietary interviewer (Phase 1), then by a local lead nutritionist at the field center (Phase 2), and then by the Nutrition Coordinating Center (NCC) (Phase 3); any differences identified by NCC were reconciled (Phase 4). Bland-Altman and generalizability theory methods were used to assess agreement of consumption for selected food variables and nutrients between phases. RESULTS: Only small differences occurred. Quality control procedures primarily reduced the variances of nutrients rather than caused the means to shift. Most of the variability among phases was due to individual level variability in dietary intake. CONCLUSIONS: Decisions to review dietary recall data beyond local review should be based on the level of precision and accuracy required for the study outcomes and the availability of financial resources.
BACKGROUND: Quality control methods are key components of dietary assessment, but have rarely been evaluated. METHODS: One hundred forty-four 8-10-year-old African-American girls at three field centers completed two 24-h dietary recalls at baseline before a pilot weight gain prevention intervention (one recall collected in-person and one by telephone). The dietary recall data were initially reviewed by the dietary interviewer (Phase 1), then by a local lead nutritionist at the field center (Phase 2), and then by the Nutrition Coordinating Center (NCC) (Phase 3); any differences identified by NCC were reconciled (Phase 4). Bland-Altman and generalizability theory methods were used to assess agreement of consumption for selected food variables and nutrients between phases. RESULTS: Only small differences occurred. Quality control procedures primarily reduced the variances of nutrients rather than caused the means to shift. Most of the variability among phases was due to individual level variability in dietary intake. CONCLUSIONS: Decisions to review dietary recall data beyond local review should be based on the level of precision and accuracy required for the study outcomes and the availability of financial resources.
Authors: Robert C Klesges; Eva Obarzanek; Shiriki Kumanyika; David M Murray; Lisa M Klesges; George E Relyea; Michelle B Stockton; Jennifer Q Lanctot; Bettina M Beech; Barbara S McClanahan; Deborah Sherrill-Mittleman; Deborah L Slawson Journal: Arch Pediatr Adolesc Med Date: 2010-11
Authors: Thomas N Robinson; Donna M Matheson; Helena C Kraemer; Darrell M Wilson; Eva Obarzanek; Nikko S Thompson; Sofiya Alhassan; Tirzah R Spencer; K Farish Haydel; Michelle Fujimoto; Ann Varady; Joel D Killen Journal: Arch Pediatr Adolesc Med Date: 2010-11
Authors: E Whitney Evans; Paul F Jacques; Gerard E Dallal; Jennifer Sacheck; Aviva Must Journal: Public Health Nutr Date: 2014-04-29 Impact factor: 4.022
Authors: Thomas N Robinson; Helena C Kraemer; Donna M Matheson; Eva Obarzanek; Darrell M Wilson; William L Haskell; Leslie A Pruitt; Nikko S Thompson; K Farish Haydel; Michelle Fujimoto; Ann Varady; Sally McCarthy; Connie Watanabe; Joel D Killen Journal: Contemp Clin Trials Date: 2007-05-25 Impact factor: 2.226