A study protocol (accepted for a presentation at the EGPRN conference in Riga 2017)

Objectives

1. To know prevalence of GF in GP consultations and its relation with patient characteristics (sociodemographic and clinical) and professional (gender, experience, personality and knowledge of the patient).
2. To assess the validity of gut feelings (sense of alarm and sense of reassurance) to predict severe disease and cancer in GP consultations.
3. To estimate the relationship of gut feelings (sense of alarm) with testing and referrals to specialist consultations.

Methods

• Prospective observational study of diagnostic validity.
• Sample size: 2951 patients attending consultation.
• Variables: patient (sociodemographic data, type of visit, date, time), professional (age, sex, training, environment, years of experience, rational-experiential Inventory, knew the patient?), consultation (GFQ, cancer related symptoms).
• Tracking variables 6 months: incident diagnosis of serious disease and cancer, testing, referrals, number of visits.

Analysis

• Objective 1: descriptive analysis of the variables and prevalence of both GFs. OR calculation and logistic regression with the main variables. Replicate multivariate subgroup analysis will be evaluated, possible interactions and application of random effects, considering a multilevel model.
• Objective 2: sensitivity, specificity, PPV and NPV, accuracy, LR + and LR- of the senses of alarm and reassurance.
• Objective 3: bivariate analysis of main variables of diagnostic test and referral, exploring relation to the sense of alarm and reassurance. Multivariate logistic regression analysis.

Authors: Bernardino Oliva, Sebastià March, Gaspar Tamborero, David Medina, María Martín-Rabadán, Erik Stolper, Magdalena Esteva.