Methods

Study design

This retrospective cohort study used the Taiwan National Health Insurance Research (NHIR) database from January 1, 1997 to December 31, 2013. Subjects are those RA patients who use biologics after 18 years old. The medication before and after 1 year of biologics will be recorded. We defined early use of biologics as biologics prescribed within 2.24 years after the RA diagnosis, and the late use of biologics was defined as those prescribed after 2.24 years of the RA diagnosis. These definitions are based on previous studies defining early arthritis as the onset of symptoms within 2 years of diagnosis [14-17], while we needed another 3 months for application biologics here in Taiwan, which equals a total of 2.24 years. We further defined the cut-off value of a 50% reduction in days of using DMARDs, steroids, or NSAIDs as the clinically meaningful tapering of medication [18], a protocol found in other studies.

Data source

Taiwan’s National Health Insurance (NHI) Program began to be implemented on March 1, 1995. This program provides broad health insurance, and more than 99% of Taiwan’s 23 million citizens have been included and received various healthcare services under this program, including physical therapy, inpatient and outpatient care, dental care, childbirth, Chinese medicine, etc. This NHIR provides information regarding hospitalization, epidemiological research, information on prescribed medication, diagnostic information, etc., all of which is considered high quality [19]. The NHIR randomly sampled a database of 1,000,000 subjects from all of its beneficiaries and database of subjects with major illnesses and has been releasing the data set to the public for studies since 1997.
Each person has been assigned a distinct identity number in the NHIR database, and identification data of the beneficiaries has been randomized to protect their privacy. This current study used the database of subjects with major illnesses and was financially supported by Kaohsiung Chang Gung Memorial Hospital, Taiwan (CMRP: CFRPG8H0231; IRB: 201801196B0).

Study cohort

The International Classification of Diseases, 9thversion (ICD-9) code was used for encoding diseases of interest. Patients aged at least 16 years old who were diagnosed with RA (ICD-9 code 714.0) in the NHI database at least three times in an outpatient department or at least one time in an inpatient department within 12 months were defined as RA patients in this study. RA patients who used one biological agent at least three times within 6 months to treat RA were defined as biologics users and have been included in this study starting from March 1997. We calculated the total days of prescribing NSAIDs, oral steroids, intra-articular steroid, MTX, and DMARDs by physicians. Furthermore, the overall medication prescribed days within 12 months before the initiation of biologics and 12 months following a one-month washout after discontinuation of a biological agent were recorded and analyzed. Exclusion criteria were as follows: the use of a biological agent prior to diagnosis of RA; a diagnosis of ulcerative colitis (ICD-9 code 556.9, 556.8, 556) , Crohn’s disease (ICD-9 code 555, 555.0, 555.1, 555.2, 555.9), psoriasis, and/or psoriatic arthritis (ICD-9 code 696.0, 696.1, 696.2, 6961, 696) within 5 years before the use of a biological agent [20]; RA patients who had never used any biological agent; and a follow-up period less than 12 months.

Statistical Analysis

We used t-tests and chi-square tests to compare baseline characteristics between these two groups. Logistic regression was used to estimate crude and adjusted odds ratios (OR) and 95% confidence intervals. All statistical analyses were performed using commercial software (SAS 9.4, SAS Institute, Cary, NC).