PD999 - PATIENT AND MEAL RELATED FACTORS ASSOCIATED WITH PERCEIVED HOSPITAL FOOD QUALITY: AN EXPLORATORY DECISION-TREE ANALYSIS USING THE JAPANESE VERSION OF HOSPITAL FOOD EXPERIENCE QUESTIONNAIRE (HFEQ-J)
PD999
PATIENT AND MEAL RELATED FACTORS ASSOCIATED WITH PERCEIVED HOSPITAL FOOD QUALITY: AN EXPLORATORY DECISION-TREE ANALYSIS USING THE JAPANESE VERSION OF HOSPITAL FOOD EXPERIENCE QUESTIONNAIRE (HFEQ-J)
G. Okada1,*, K. Amano2, M. Hachijin2, C. Kambara1, Y. Matsumoto3
1Department of Health Sciences, Prefectural University of Hiroshima, 2Division of Nutrition Management, Hiroshima University Hospital, Hiroshima, 3Department of Nutrition, Graduate School of Human Life and Ecology, Osaka Metropolitan University, Osaka, Japan
Rationale: Identifying modifiable factors influencing perceived hospital food (HF) quality is essential for improving in‑hospital nutritional care, yet these factors have rarely been examined beyond immediate sensory experience. We explored patient‑ and meal‑related factors associated with perceived HF quality using the Japanese version of the Hospital Food Experience Questionnaire (HFEQ‑J).
Methods:
In this cross‑sectional single‑institution survey, perceived overall HF quality (5‑point Likert scale) was dichotomized as ≥4 vs <4. Candidate predictors included patient characteristics, eating behavior (complete HF intake), and patient‑reported HF priorities and evaluations (HFEQ‑J items). A χ² automatic interaction detection (CHAID) decision‑tree model was applied. In a sub‑analysis, sensory and behavioral factors (e.g., taste, smell, appetite, intake) were excluded to identify upstream service‑level contributors.
Results:
Among 201 inpatients, 62.2% rated overall HF quality ≥4. In the primary analysis, taste of HF consumed was the first split, followed by combination, smell, importance placed on HF smell, and complete intake, showing performance (misclassification 0.164; accuracy 88%). In the sub‑analysis, service‑level and cognitive attributes—including importance placed on HF being healthy, serving temperature, food familiarity, and container usability—emerged as contributors, although predictive accuracy was lower (misclassification 0.353; accuracy 73%).
Conclusion:
Perceived HF quality was strongly driven by sensory attributes. When these were excluded, patients’ cognitive evaluations and service‑level aspects became relevant. Because placing importance on HF being “healthy” remained an important factor, dietitian‑led explanations may support understanding of therapeutic diets and HF quality perceptions.
Disclosure of Interest: G. Okada: None declared, K. Amano: None declared, M. Hachijin: None declared, C. Kambara: None declared, Y. Matsumoto Consultant for: Original Nutrition Co. Ltd.