In both groups, all self-reported taste alterations were first assessed using the Chemotherapy Induced Taste Alteration Scale (CITAS). All patients in both groups then received usual care (the gold standard), which meant that they ate regular meals that they were used to at home and accounted for any dietary advice if needed.
A chef gastro-engineering (CGE) [16] from the research team carried out an on-site taste assessment for each consecutive patient in the intervention group using an O-box [14] (Fig. 1).
The O-box test is used to assess human gustatory and olfactory sensitivity together with somatosensory functions of taste. In a sense, it is comparable to the well-known Sniffin sticks test [17], but instead of smelling different scents, an individual has to taste different foods. To this end, the O-box contains 21 carefully selected natural foods: some foods are provided with a pasta and others are provided in a liquid form, all in well-defined and reproducible concentrations [14]. These food products can be used in a multitude of concentrations and combinations on a taste stick to determine an individual’s minimum and maximum threshold values within which each taste is experienced as tasty or not. To avoid interrater variations, all taste assessments in this study were executed by one CGE, a trained member of the research staff. The individual taste assessment results, together with data about documented food allergies and patient-reported food likes and dislikes as well as any prescriptions from the treating oncologist, dietician and/or speech therapist, are used in algorithms by the CGE to generate personalized recipes. In this case, STSwas used to personalize recipes for bread and soup meals that could easily be prepared at home by the patients or their family caregivers. All the recipes were provided to the included patients online within 48 hours. To address any practical difficulties in applying these recipes at home, a helpdesk was available to assist the patients or informal caregivers if needed.
The data from both cohorts were collected from March 2022–December 2023 at three different times—at baseline (t0), after one month (t1) and after three months (t2) of follow-up by using validated tools, including the MNA-SF [18], the CITAS [19], and the EORTC QLQ-C30 (version 3.0) [20]. For the patients in the intervention group, an additional questionnaire was used to monitor the application of STS and to assess food intake at t1 and t2.
The MNA-SF comprises six questions and simple measurements: anthropometric measurements (body mass index, weight loss); a global assessment (mobility); and a dietary questionnaire and subjective assessment (food intake, neuropsychological problems, acute disease). A total MNA-SF score < 8, 8–11, and > 11 points indicates malnutrition, risk of malnutrition, and no malnutrition, respectively [21, 22].
The CITAS enables valid, reliable measurement of specific symptoms of chemotherapy-induced taste alterations. The CITAS, which comprises 18 items and 4 subscales scored using a 5-point Likert-type scale, was first developed by Kano and Kanda [19]. The first subscale, ‘Decline in Basic Taste’, concerns the condition of sensing bitter, sweet, salty, sour, and umami. The second subscale, ‘Discomfort’, concerns the relationships among taste alterations and nausea and vomiting, alterations in the sense of smell, difficulty eating spicy/oily/meaty meals, and reduced appetite. The third subscale concerns individuals’ experiences of phantogeusia and parageusia, and the fourth subscale, ‘General taste alterations’, concerns individuals’ experiences of ageusia, cacogeusia, and hypogeusia. For assessments using the CITAS, the scores for each subscale are evaluated rather than the total scale score [19]. The subscale scores are obtained by dividing the number of items by the sum of the scores of those items. The maximum possible score is 5 points, whereas the minimum possible score is 1 point. An increase in the score indicates that the intensity of taste alterations and discomfort has also increased.
The EORTC QLQ-C30 (version 3.0) [20] is composed of both multi-item scales and single-item measures, including five functional scales, three symptom scales, a global health status/quality of life scale, and six single items. Each of the multi-item scales includes a different set of items, and no item is present in more than one scale. The data are analysed using the scoring procedures in the EORTC QLQ-C30 Scoring Manual [23]. All of the scales and single-item measures have scores ranging from 0 to 100 points. A high score for a functional scale represents a high/healthy level of functioning. A high score for the global health status/quality of life scale represents a high quality of life. Finally, a high score for a symptom scale/item represents a high level of symptomatology/problems.
The palatability and daily intake of soups and bread were assessed for all patients in the intervention group. Both factors were evaluated for seven days: during the week before t1 and during the week before t2. To this end, a Likert-type scale ranging from 1 to 7 was used. A palatability score of 1 corresponds to an 'extremely bad taste', and a score of 7 corresponds to an 'excellent taste'. A daily intake score of 1 corresponds to 'I ate much less than normal', and a score of 7 corresponds to 'I ate much more than normal'. The average scores for palatability and daily intake are used to measure the effect of STS on these two factors.
Statistical analyses
All the statistical analyses were performed using SPSS version 29 (IBM Corp, Armonk, New York). A p value less than 0.05 was considered to indicate statistical significance. Descriptive statistics are provided for sociodemographic and clinical factors, as well as the main study variables, and are reported as the means and standard deviations for continuous variables and as numbers and proportions for dichotomous variables. Finally, a hierarchical regression analysis, controlling for age, sex and nutritional status at baseline, was performed to identify multivariate relationships between the interventions and outcomes.