JCPSLP Vol 20 No 3 November 2018

styles of food to choose from within a texture category including sweet and savoury. Other considerations included ingredient-related factors, such as cost, availability and variety of flavours; as well as practicalities such as cooking time and requirements for special equipment/utensils. Once the recipes were considered for these characteristics, a process of cooking and auditing each of the recipes for consistency in texture compliance could begin. At this point, ten recipes were chosen for each diet texture (30 recipes in total) with five sweet and five savoury dishes to be tested for each texture. Do and Study – cooking and compliance testing The re-cooking and compliance testing process was undertaken by a group of student volunteers (n = 10) in their own homes. Each volunteer cooked two to three recipes developed by peers. Home recipe testing by volunteers was an essential component of the dysphagia cookbook preparation process. This allowed the opportunity to see how recipes would “hold up” in a real-world cooking scenario with no special equipment, or expert knowledge of how to make the recipe work. Through this process not only were the meals tested, but the instructions given via the methods section were scrutinised and corrected to ensure all recipes as clear and easy to follow as possible. As per Phase 1, the Do and Study components of Phase 2 was also iterative in nature, with each recipe. Copies of each recipe were saved onto Google Drive and shared with volunteers. Recipe names and volunteer allocations were entered into a spreadsheet that was made available to all volunteers via Google Docs. This spreadsheet served as a visual tracking method as to the outcome of the compliance trials. Each volunteer would input the results of their compliance test into the spreadsheet using a colour coding system – green for a pass, red for a fail, and yellow for recipes that required further modification before being deemed compliant. Comments were also provided within the spreadsheet, or as track-changes on the digital file of the corresponding recipe. After the first round of cooking was completed and feedback was received, volunteers were then allocated a second different recipe to cook and audit. Recipes that received two “green” ratings were deemed to be consistently replicable at the level of compliance being sought (e.g., soft, minced moist or smooth pureed). Any recipes that received a “yellow” or “red” response in one of the two trials were given a third trial before a final decision was made about their inclusion in the cookbook. Recipes that received two or more “yellow” or “red” ratings with multiple modifications were discarded, and another recipe from the wider bank of recipes was substituted into the review process. The compliance testing resulted in a total of 30 recipes approved for inclusion in the cookbook (see Table 1 for a full listing of the recipes included in the Cookbook). Act – cookbook design and editing Following the home testing and auditing process, the recipes were ready for inclusion in the dysphagia cookbook. Preparing recipes ready for publication involved rewriting many of the recipes so the style of writing was consistent and important information such as storage instructions, dietary information (e.g., if the recipe was dairy and gluten free) and total cooking time was listed for each recipe. The cookbook team then consulted with the university’s graphic design studio, LiveWorm, to assist in turning the

of preparing ingredients to minimise preparation time, or substituting ingredients that could not be adapted to be compliant with the Australian Standards for Texture Modified Foods and Fluids (Atherton et al., 2007). In this way, the Do and Study components of the PDSA cycle were an iterative process with each recipe design decision being reviewed carefully before moving to the next step (e.g., what went right, what went wrong, and what needed to be changed in the next test cycle [Varkey et al., 2007]). During this phase some students created recipes that could be adapted to fit the texture requirements for more than one texture. For example, if a student’s task was to create a minced moist recipe, they experimented with recipe instructions to create a recipe that could also be made to be compliant with a soft or puree diet consistency. From 2017, this became a requirement of the task for all students. Act – recipe making and auditing Once the recipes had been developed at home, students attended class to present their final recipes and meals to their peers. It was during this class that an auditing process was undertaken by peers and academic staff. Each cohort group from 2014 to 2017 followed this process in a purpose-built teaching kitchen; this classroom experience became known as the “dysphagia kitchen”. The auditing process sought to determine compliance with the Australian Standards for Texture Modified Foods and Fluids (Atherton et al., 2007) criteria for each diet texture. Students and staff sampled and evaluated each recipe for compliance as part of the auditing process, using a diet audit tool adapted from Atherton et al., 2007. Students provided comments on their peers’ work, praising particularly innovative or creative methods of creating, and making suggestions for improvements where needed. The dysphagia kitchen provided the opportunity for the students to sample a variety of texturally compliant meals, and gain practical knowledge about the look, consistency, and mouth-feel of each of the modified textures. After the dysphagia kitchen experience, students were given 48 hours to make any necessary changes to their recipes based on received feedback, prior to electronic submission of recipes to the academic staff member overseeing the activity. From this process, a bank of 50 recipes from the 2014–2017 cohorts was established. Phase 2: Cookbook development Plan – recipe review With a considerable volume of recipes, further auditing processes were ready to begin. Lead by an academic member of staff, a small group of student volunteers began the process of further auditing the recipes for potential inclusion in the cookbook. The students started by reviewing all of the recipes, and identifying those that may or may not be appropriate for inclusion. Several factors helped in deciding which recipes could progress to the next stage of compliance trials including duplication and variety. For example, as the dysphagia kitchen experience took place across different cohorts, and as students were not aware of the recipes created by their peers from previous cohorts, there were a number of instances where recipes were similar and choices needed to be made about which ones to include. In addition, choosing a variety of recipes representing a range of main ingredients was also important in the recipe selection process. This process was designed to ensure consumers would have a range of flavours and

152

JCPSLP Volume 20, Number 3 2018

Journal of Clinical Practice in Speech-Language Pathology