Challenges experienced with early introduction and sustained consumption of allergenic foods in the Enquiring About Tolerance (EAT) study: A qualitative analysis

Abstract

Background: The early introduction group participants of the Enquiring About Tolerance study were asked to undertake a proscriptive regimen of early introduction and sustained consumption of 6 allergenic foods. It was envisaged that this might be challenging, and early introduction group families were presented with an open-text question to express any problems they were experiencing with the regimen in recurring online questionnaires.

Objective: We sought to analyze these open-text questionnaire responses with the aim of identifying challenges associated with the introduction and regular consumption of allergenic foods.

Methods: Three combinations of interim questionnaire responses were selected for analysis, representing the early period (4, 5, and 6 months), middle period (8 and 12 months), and late period (24 and 36 months) of participation in the Enquiring About Tolerance study. Responses were assigned a code to describe their content and subsequently grouped into themes to portray key messages. A thematic content analysis allowed for conversion of qualitative codes into quantitative summaries.

Results: Three main challenges to allergenic food consumption were identified. First, some children refused the allergenic food, causing a sense of defeat among caregivers. Second, caregivers were concerned that allergenic foods might be causing a reaction, triggering a need for reassurance. Third, practical problems associated with the regimen compromised caregivers' capacity to persist.

Conclusion: Understanding the challenges experienced with allergenic food introduction and sustained consumption is the necessary precursor to developing specific communication and support strategies that could be used by caregivers, practitioners, policymakers, and key stakeholders to address these problems.

Keywords: Food allergy; adherence; allergens; breastfeeding; diet; infancy; qualitative; randomized controlled trial.

Graphical abstract

Fig 1

Problem question response types across all interim questionnaires. Relative distribution of the 3 possible response types at each interim questionnaire time point is demonstrated. For all EIG participants (A), the proportion giving a text response diminished over time (green bar). Over time, the proportion of EIG families not completing the questionnaire increased significantly (blue bar), except for the 12- and 36-month interim questionnaires, which were completed in conjunction with a clinical visit. There was a routing issue within the 7-month interim questionnaire, which meant that the majority of EIG families who completed the questionnaire were inadvertently not offered the problem question. If EIG families who did not complete the interim questionnaire are excluded, it can be seen (B) that there is a steady increase over time in EIG families leaving the problem question blank (red bar) with a concomitant decrease in those entering a response (green bar).

Fig 2

A word cloud of the most frequent words used in the open-text responses in all interim questionnaires analyzed. Depicted in this word cloud are the top 200 words that were most frequently used, excluding common words and combining like words (eg, sesame and tahini combined as sesame). The size of the words corresponds to the frequency with which they were used in the responses analyzed. Words that depict the main themes identified within the qualitative analysis of the open-text responses are color coded according to theme (blue, infant refusal; orange, reaction concerns; green, practical problems; yellow, issues with starting; and black, no problems). The 6 early introduction foods are identified in pink, and their relative size gives a clear indication of the frequency with which the specific food featured in the open-text responses.

Fig 3

A bubble chart displaying the proportion of text responses coded into a thematic category within each interim questionnaire grouped by period. In each interim questionnaire (columns), all text responses were coded to represent a thematic category (rows). Sizes of bubbles are proportional to percentages of responses in any individual interim questionnaire that were coded to a particular theme. The variation in bubble size within and across months shows how the dominance of thematic categories can change in relation to the other thematic categories and in relation to themselves over time.

Fig 4

Distribution of thematic responses across time periods by EIG adherence status. Among adherent, nonadherent, and nonevaluable EIG participants, respectively, all qualitative codes recorded by these families in each time period (early period, middle period, and late period) were counted and totaled. Bars depict the relative distribution of thematic responses for each EIG adherence group by each time period.