Anaphylaxis is a severe and potentially life threatening allergic reaction which can affect the airway, breathing and/or circulation. This reaction can be triggered by a number of different allergens but the most common are food, medications, insect venom and latex. Because these reactions occur suddenly and are potentially very serious, the best management lies in the correct use of the prescribed emergency medication. Epinephrine, or adrenaline as it is more commonly known, is the recommended drug for the treatment of anaphylaxis. Injecting the epinephrine into the outer thigh muscle is the preferred route of administration. For health professionals, patients and carers, this rapid administration is facilitated by the manufacture of preloaded syringes and autoinjectors. Despite the availability of these devices, a review of studies shows poor knowledge and skills amongst both health professionals and patients with regards to using autoinjectable epinephrine devices correctly. These studies found that as well as poor knowledge in using the autoinjectors, there was a lack of confidence amongst patients and an unwillingness to carry the device with them at all times.

In other disease groups like asthma and diabetes, psychological models which involve asking people how they think about their illness and their related behaviours, have been found to help in the understanding of why some people follow or adhere to health professionals advice and why others do not. Based on these findings, this study will look at two appropriate psychological models and their ability to predict variation in adherence to self-care regimens in adolescents and young adults with food allergy related anaphylaxis.

Study Design: Cross-sectional comparison of the Self Regulatory Model (Leventhal, 1984) and The Health Belief Model (Rosenstock, 1966) by which the models will be used to predict and understand adherence to selfcare regimens of adolescents and young adults with food allergy.

Proposed Sample Size / Power Calculations:

The power calculation was based on multiple regression analysis for all predictor variables, 25 in total. To predict a minimum of 10% of the variance in adherence with an alpha=0.05 and power=0.90, a minimum proposed sample size of 275 participants will be needed. A 40% attrition rate has been anticipated due to the use of postal questionnaires and so a minimum of 459 participants will need to be recruited.

Analyses Nonparametric item response theory analysis, univariate correlations and multiregressional analysis for predictor models will be conducted. Structural equation modelling will also be applied to assess the goodness of fit of the two social cognition models.

Setting This study will be based in an outpatients setting and in the community. Participants will be identified and recruited from their hospital allergy clinic either whilst attending an appointment or from a hospital mail out. The third method of recruitment will involve participants from a convenience sample of the universities of Sussex and Brighton.

Procedure Phase 1 Recruitment and consent The recruitment of study participants will involve three methods. The first method will involve recruiting participants in person from their hospital allergy clinic. Eligible individuals identified by clinic staff will be given a participant information leaflet and asked if they are interested in taking part in the study. Once verbal consent is given, participants will be directed to the chief investigator who will be present in clinic. Interested individuals would then have the opportunity to discuss the research with the chief investigator and/or the clinic staff. If still interested, individuals (aged 16-21 or parents of those aged 13-15) will be asked to sign two consent forms (one copy to keep for themselves) and be given a questionnaire. The participants will be given the opportunity to complete the questionnaire within the clinic, or take it home to complete. If the latter option is chosen, a freepost envelope addressed to the research team will be provided. Participants will also be given a correspondence slip on which to write their contact details to return separately to the research team if they are interested in receiving a summary of findings.

The second method of recruitment will involve a patient database search by clinic staff and subsequent mail out to eligible individuals. Clinic staff will search their patient records for individuals who fit the inclusion criteria. The consultant in charge will confirm their eligibility and sign the invitation letter prepared on hospital headed paper. This will be put in an envelope together with the participant information leaflet, two consent forms (one copy for the participant to keep and the other to be returned), the questionnaire, future correspondence slip and freepost envelope addressed to the research team.

The third method of recruitment will be a convenience sample through the Universities of Sussex and Brighton. A generic email will be sent to staff and students consisting of the participant information leaflet and detailing the main inclusion criteria. Individuals who are eligible and interested in taking part in the research will be asked to reply to the email with their contact details. Similarly to those recruited from the clinic mail out, these participants will be sent a pack containing another participant information leaflet, two consent forms, a questionnaire, a correspondence slip and a freepost envelope addressed to the research team.

Phase 2 - Questionnaire In order to measure the models, carefully constructed and piloted questionnaires have been developed. Each participant will be asked to complete one questionnaire on one occasion. The questionnaire will collect demographic details, assess knowledge of how and when to use medication, confidence in the individual's own ability to use the medication, management of food allergy, the Revised Illness Perception Questionnaire (Moss-Morris et al., 2002) which is the validated questionnaire to measure items from the Self-Regulatory Model (Leventhal, 1984), items which relate to factors found within the Health Belief Model (Rosenstock, 1966) and finally the Revised Life Orientation Test (Scheier et al., 1994). The questionnaire completion time has been confirmed with individuals within the targeted age range and is expected to take no longer than 40 minutes to complete. In order to improve the response rate of questionnaires, a second mail out will be sent to all eligible participants one month after the initial mail out.

Phase 3 - Feedback to participants Those participants who expressed an interest in receiving a summary of findings and provided contact details will be sent a summary document upon completion of the study.

Measures Demographic/background Participants will be asked their age, gender, ethnicity, size of family, history of food allergy and anaphylaxis, use of the emergency medication autoinjectable epinephrine, membership of an allergy support group, possession of a management plan for their food allergy.

Knowledge Knowledge about how and when to use an autoinjectable epinephrine device will be measured and the individuals confidence in their ability to do so correctly will also be looked at. This will enable us to explore whether underuse is as a result of poor knowledge or whether this can be better explained by other factors.

Adherence Adherence will be measured by asking participants about their current methods of managing their food allergy. This will include questions on allergen avoidance and carrying emergency medication (auto-injectable epinephrine and additional medicines they may use to treat their food allergy).

Revised Illness Perception Questionnaire (Moss-Morris et al., 2002), Health Belief Model (Rosenstock, 1966) items and the Revised Life Orientation Test (Scheier et al., 1994) The Revised Illness Perception Questionnaire (Moss-Morris et al., 2002) measures constructs from the Self-Regulatory Model (Leventhal, 1984) which include symptoms, the duration of food allergy, the consequences of having food allergy, the personal control an individual feels they have over their food allergy, the amount the individual feels the treatment can control their food allergy, emotions they associate with their food allergy, how much they comprehend their food allergy and what they believe are the causes. The Health Belief Model (Rosenstock, 1966) similarly looks at attitudes but in this case how susceptible and serious individuals feel suffering an allergic reaction would be, the barriers and benefits to following recommended advice and what prompts these individuals to follow this advice. The Revised Life Orientation Test (Scheier et al., 1994) will also be measured to determine levels of optimism which has been found in other studies to impact on the way participants respond to items on questionnaires.

Inclusion Criteria:

• Aged between 13-21 years
• Diagnosis of food allergy
• Prescription of an epinephrine auto-injector

Exclusion Criteria:

• If participants are unable to write or understand English

Hospital and community sample