Feasibility of using a computerized food frequency questionnaire (FFQ) to assess dietary patterns in patients with advanced gastrointestinal (GI) malignancies.

Abstract

465 Background: Identification of poor dietary patterns is important for the optimal management of patients with metastatic GI malignancies who suffer from high rates of malnutrition and sarcopenia which are associated with increased rates of treatment-related toxicity and worsened survival. Dietary records or recall methods and FFQs are available but have high patient burden and are seldom utilized in oncology settings outside of formal nutritional consultations. Thus, we evaluated the feasibility of conducting serial assessments of dietary patterns using a novel computerized FFQ technology. Methods: Assessments of diet using the computerized FFQ, nutritional status (NUTRISCORE), QOL and anorexia/cachexia burden (FAACT), anxiety/depression (HADS) and taste burden were done at the time of initiation of first-line chemotherapy (baseline) and 3 months later. The FFQ was done independently either at home on personal computer after visit or via iPad in infusion room. Dietary quality was defined by the Healthy Eating Index 2010 (HEI) score, which was automatically calculated by the computerized FFQ software. Feasibility was defined as at least 70% of patients completing baseline and 3-month assessments. Spearman’s correlations were used to determine associations between measures. Results: 29 patients with advanced (metastatic or unresectable) GI cancers were enrolled and 23 completed the baseline FFQ (8 colorectal, 5 pancreas, 5 gastroesophageal, 5 other). Median age was 58 (20-78) with M:F ratio of 10:13. The overall completion rate of baseline and 3-month assessments was 81.8%, meeting the pre-defined feasibility criteria. The mean baseline HEI score was 65 (range 38-88). Of patients who completed both baseline and 3-month assessments, the mean HEI score remained stable at 58 with changes in specific HEI components including a 20% decline in the seafood and plant protein score and a 13% decline in the whole grain score and small improvements in other areas (e.g. empty calories, dairy). There were no significant correlations between baseline HEI score and other assessments. Conclusions: It is feasible to use a computerized FFQ to assess for longitudinal changes in eating patterns and dietary quality in patients with advanced GI cancers, particularly when patients can complete the assessments during chemotherapy infusions. Preliminary findings also suggest that the FFQ can be helpful for highlighting areas in which diet quality could be improved. Additional analysis of other computerized FFQ data (e.g. macro/micronutrient consumption) is ongoing.