Comparative Effectiveness of Different Drugs Used to tr€at Patients in Rheumatoid Arthritis Saudi Database (RASD)"

Rheumatoid arthritis (RA) is a chronic progressive inflammatory disease that causes irreversible joint deformities which can have debilitating effects on a patient's overall wellbeing. RA has a global prevalence rate of 0.5-1.1% with an annual incidence rate of 20-50 cases per 100 000 of the American and North European population.

Registries are widely mentioned in literature as a method that studied a given population in a way that is more applicable to clinical practice. Although there are numerous registries that previously set place in Europe and north America, the middle east and Africa remain scarcely studied with a few published small, scaled hospital reports. Registries are imperative to study a disease's unique course in each population, which may or may not align with its counterparts in western populations. Therefore, we established a Saudi registry for rheumatoid arthritis called Rheumatoid Arthritis Saudi Database (RASD).

We have published the results of RASD from one center recently. Our study was conducted in a single center in Saudi Arabia using a pool of 433 patients. The aim was to describe the population of rheumatoid arthritis and to compare these findings to previously published data registries.

Currently, RASD has expanded to include several centers in Saudi Arabia. It is essential to conduct a study that address the comparative effectiveness of different drugs used to treat RA in Saudi population. There are many reasons that make management of RA in our country different than other places in the world. This study should help us to compare our comparative effectiveness to other studies with similar methodologies that reported on effectiveness in different population than ours. We may include centers who are even not part of our RASD as long as they can provide a minimum of one year data on their RA patients with at least 3 visits describing the outcome measures that we are interested in.

By the end of this research project the following information will be available for health care leaders for strategic planning: The effectiveness of different biological and targeted synthetic DMARDs in Saudi population, showing the most effective and the least effective drug. This information can help in the following:

• Guide the development of a local guidelines focusing on the use of the most effective drugs in our population.
• Prioritizing the use of different drugs to treat RA in Saudi population.
• The drug survival rates and the discontinuation rates of different drugs in Saudi population.

This is a multicenter study. The data will be obtained primarily from Rheumatoid Arthritis Saudi Database (RASD). We will also include RA patients from other centers who are not part of the registry if they are providing minimum of 12 months follow up with at least 3 documented visits of these patients to health care facilities following "treat to target approach" or preferably "treat to work approach".

For the comparison of drug efficacy the best level of evidence is obtained by blinded randomized control design, however use of data of registry is also an acceptable option with known limitation. This is a retrospective study design. We will observe all the cohorts of registry as well as those meeting the criteria, to find out the most effective drug for the treatment of rheumatoid arthritis as well as to find out the association of various risk factors with the treatment outcome.

We will use non probability, convenience sampling technique to enroll the patients.

Sample size: Sample size was calculated as 798 based on a confidence interval of 95% and power of 80% with a p value of less than 0.05

All the collected data will be entered in SPSS version 22, edited, cleaned and analyzed. Categorical data will be measured by proportion and percentages while continuous data will be measured by mean and standard deviated where normality assumption is met. Homogeneity of variables will be measured by Levene's test for equality of variance. Student t test will be used to compare between two groups and ANOVA for the more than two groups in case of continuous data. For the categorical data Chi-Square test will be used for large data and Fisher Exact test for smaller data. Logistic regression model will be used in order to predict the outcome.

Exposure and outcome variables included: Age of patient, Age of onset, sex, urban vs rural areas, level of education, comorbidities, number of drugs used before using biological and targeted synthetic DMARDs, duration of methotrexate use before starting biological and targeted synthetic DMARDs, effectiveness of drugs, lag time, drug survival rate.