An Epidemiological Study of the Incidence of Inflammatory Bowel Disease in Spain

Study design This is a prospective, observational, multicenter, population-based incidence cohort study which will enroll cases of IBD (CD, UC, or indeterminate colitis) diagnosed in adults over 1 year in Spain.

In addition, each incident case that gave his/her informed consent, will be followed up for 12 months to determine changes in phenotype or disease location, the need for immunosuppressive and biologic treatments, and the need for hospital admissions and surgery during the first year after diagnosis.

Study population Incident cases diagnosed with IBD over 12 months in the Spanish territory, which has a population of 46,439,864 inhabitants (1 January 2015) according to data from the National Statistics Institute (INE). According to recent data, only 3% of incident cases are diagnosed in the pediatric population, so the study will focus on adult patients (over 15 years of age).

Case detection In Spain, health care is performed mostly by the public health services. According to recent data, approximately 15% of the Spanish population has private health insurance (Informe Sanidad Privada: Aportando Valor. Análisis de situación 2014 (Report on Private Health Care: Providing Value. 2014 Situation Analysis); available at: https://www.fundacionidis.com/wp-content/informes/informe_analisis situac_2014_0.pdf). In addition, of those persons having private health insurance, only about 15% make exclusive use of it. For these reasons and taking into account the specific characteristics of IBD, the risk of underestimating the incidence of IBD considering only cases seen in public health centers would be of little relevance. Therefore, this study will be conducted at centers providing public health care within the National Health System.

Furthermore, as previously mentioned, according to recent data only 3% of incident cases are diagnosed in the pediatric population, so the study will focus on adult patients (over 16 years of age)4.

To conduct this research project, an IBD specialist who is a member of GETECCU has been selected from each Autonomous Community, who will act as the coordinator in their region. Gastroenterologists in charge of IBD Units in Spain will be contacted, or if unavailable, a gastroenterologist from each of the hospitals included in the National Catalog of Hospitals as of 31 December 2014 at http://www.msssi.gob.es/ciudadanos/prestaciones/centrosServiciosSNS/hospitales/ home.htm (last accessed 13 August 2015). The most appropriate strategy for identification of cases will be planned on a case by case basis in each Autonomous Community, according to the specific characteristics of this Community, each province and health area, with the aim of detecting all incident cases. The databases of the endoscopy units and pathology departments will be reviewed in all sites participating in the study during the year of enrollment, in order to detect any possible incident cases of IBD that not might not have been identified initially.

Each participating investigator will confirm the diagnosis at the time of entering patients in the study and 3 months later to assure this diagnosis and the phenotypic characteristics of the disease, and thereby have greater diagnostic accuracy. External monitoring of incident cases included in the registry will also be performed by review of cases selected at random by the research team of Hospital Universitario de la Princesa.

Health areas in which detection and enrollment of all incident cases cannot finally be assured will be excluded from the study.

Definitions

• Disease location and phenotype: IBD location and phenotype will be defined according to the Montreal classification.
• Time to diagnosis: It will be defined as the time from the first medical consultation made by the patient after onset of symptoms to the diagnosis of IBD.
• Population center: The type of population center at the patient's birth and at diagnosis of IBD will be recorded. Whether the population center of origin of the patient is considered rural or urban will be based on the classification of the National Statistics Institute (INE) of each municipality.
• Socioeconomic level: Socioeconomic level will be assessed through different variables, such as the patient's educational level (primary education or lower, secondary education, higher education or equivalent), occupational status (self-employed, employee, unemployed, retired), professional status (nonsalaried or salaried) and type of working hours (full time or part time).
• Number of cohabitants: The number of cohabitants in the patient's home during childhood (up to 16 years) and at diagnosis of IBD will be recorded.
• Smoking: Smoking status will be categorized as "nonsmoker", "smoker", or "ex-smoker", and will be considered at the time of diagnosis of IBD. Patients will be considered "smokers" if they have a smoked more than 7 cigarettes per week for at least 6 months or smoked at least 1 cigarette in the 6 months prior to diagnosis. Patients will be considered "ex-smokers" l if they quit smoking at least 6 months before diagnosis. Patients will be considered "nonsmokers" if they never smoked or did so in a very small amount or occasionally.
• Treatments: Treatments received by the patient in the 12 months since diagnosis of the disease will be included, provided they were received for IBD. Only the first prescription of each therapeutic group will be recorded.
• Changes in phenotype: Changes in phenotype will be considered as the appearance of new lesions not present at diagnosis subsequent to the initial tests performed to determine disease extent and severity. In these cases, the phenotype, the complication leading to classification of the patient in a different phenotype and date of occurrence of the complication will be recorded.
• Hospital admission Hospital admission occurring during the first year from diagnosis of the disease will be included. The date of admission, date of discharge, if related or not to IBD and the cause of admission.
• Surgical procedures: The surgical procedures performed on the patient since diagnosis of IBD (including those performed before knowing the patient had IBD and which led to its diagnosis), the indication for surgery and the date of surgery will be recorded. An emergency surgical procedure will be considered as any surgery performed within 24 from admission of the patient to the emergency department. An elective surgical procedure will be considered as any procedure performed subsequent to the first day of admission and by the usual surgical team.

Data collection and follow-up:

Demographic data (age, sex, smoking), family history of IBD, socioeconomic characteristics, IBD type, pattern, and location and presence of extraintestinal manifestations at diagnosis will collected from each patient. The occurrence of complications (fistulas, stenosis, abscesses), changes in disease location, treatments for IBD, surgeries for IBD, and hospital admission during the first year since diagnosis will also be recorded.

After the patient is included, two other visits will be recorded during the 12 months, coinciding with the routine visits of the patient for follow-up of his/her disease. Thus, over the course of the study, 3 visits will be recorded, as described below:

• Visit 0 (baseline): inclusion of patient in the study and collection of socioeconomic data and on diagnosis of IBD.
• Visit 1 (month 3): confirmation of IBD diagnosis and updating of data related to treatment, changes in phenotype, hospital admissions, and surgery.
• Visit 2 (month 12): confirmation of IBD diagnosis and updating of data related to treatment, changes in phenotype, hospital admissions, and surgery. End of study.

In the case of patients not included at the time of diagnosis or later because they were referred from other centers, the corresponding visits will be performed retrospectively, completing the data retrospectively in the other visits (the previous ones). In addition, the inclusion period will be extended for 6 months in which patients diagnosed during the year of the study belonging to a participating area and not previously identified by the responsible investigator may be enrolled.

In the case of sites participating in ENEIDA, to avoid duplication of recording of the data by the responsible physician, the possibility of modifying the database will be proposed to the ENEIDA committee, including, in such a way that they are only visible to the investigators of this research project, the necessary variables that are not included in ENEIDA. In the event that this modification is not accepted by the ENEIDA committee, the study data will be recorded in an electronic database created specifically for this purpose.

In any case, the sites not currently participating in ENEIDA will record the data in an electronic database of REDCap created for this purpose. Subsequently, the two databases, both ENEIDA and that created specifically for the study, can be exported and combined for their analysis.

Statistical analysis The reference population for the study will be made up of the reference population of the areas of the public centers where the participating study physicians work, which will be based on the estimates of the National Statistics Institute (INE). The incidence rate (number of incident cases per 100,000 inhabitants) during 1 year will be calculated. Age-standardized incidence rates adjusted to the European population will be calculated6.

For qualitative variables, percentages will be calculated (with their 95% confidence intervals) and for quantitative variables, the arithmetic mean and standard deviation will be determined. In the univariate analysis, categorical variables will be compared using a Chi-squared (X2) test and quantitative variables will be compared using the appropriate test (Student's t-test, Wilcoxon, etc. depending on whether their values follow a normal distribution or not). To examine in detail the time course of the use of treatments, hospital admissions, surgery or occurrence of complications, the Kaplan-Meier method will be used, and the differences between the curves will be compared with the log-rank test. A Cox regression model will be used to study which variables are associated with the probability of occurrence of complications, progression to more severe phenotypes, use of immunosuppressive or biologic drugs, hospital admission, and performance of surgery.