Discussion

Such factors as data collection interval, race, provider type (general physician vs. specialist), and type of drug coverage are associated with the use of DMARDs or biological agents among RA patients [10], and financial burden of certain expensive biological agents, usually leads to insufficient treatment among RA patients [10]. The data collection interval in this study was between 1998 and 2012, and during which period, the most available biologics agents for RA were etanercept (etanercept was available in Taiwan was since May 12, 2005) and adalimumab (adalimumab was available in Taiwan was since Aug 19, 2008). All the other currently available biologic agents for RA were neither available nor reimbursed by health insurance during the interval. Considering the study method, the cumulative dosage of analgesics for treating lower back pain has been reported in a previous study [21]. This similar method of cumulative days of administering a certain drug was applied in this present study to appropriately represent the severity of RA, since the medications were entirely reimbursed by Taiwan’s health insurance, and all the medications prescribed are recorded and could be processed in the future, as in this study. Therefore, this study focused on the changes of cumulative days of the three aforementioned types of oral medication within 12 months before and after biological agents in the same RA patient. As a result, we were able to evaluate whether the use of a biological agent could taper the subsequent cumulative days of the aforementioned medications.
The most frequent causes of death in RA patients are cardiovascular disease, neoplasms, and sepsis [22], but none of these were considered as a covariate in this study because treatment of these diseases is irrelevant to the aforementioned medication, and we excluded patients that had passed away during the follow-up period. We focused on comparing cumulative days of oral medication in the same individual.
Previous studies have suggested that both smoking [23] and genes [24] may be involved in increased RA severity. However, due to the limitation of the NIHR database, we could not include these two variables in this current study. Furthermore, temperature and humidity are also claimed to influence RA severity, with both sunny conditions and less humid conditions significantly lowering RA activity [25]. We believe that these factors have a limited influence in this study due to the similar climate cycle throughout Taiwan. One interesting finding that we did not show in our result is that some patients started use biologics agent before his adulthood, which is before 18 years old, which by definition was juvenile RA.
The footnote in Table 1, we mark the early use of biologics and late use of biologics as either before or after 2.24 years (equals to 27 months) diagnosis of RA, which we combine the idea of two-year treatment window of opportunity from previous recommendation and evidence [16, 17], and the real-world situation in Taiwan that all the reimbursement cases of biologics are required to be authorized first before the prescription of biologics, and the average processing period is around 3 months (0.24 years). These patients who use biologics agent within 2.24 years of diagnosis of RA are representatives those patients within 2-year treatment window of opportunity. Otherwise, if we pick up those patients treated with biologics with exact 2 years within diagnosis of RA, we might pick up those patients with only 1.76 years (21 months) of RA duration, which could exaggerate the results in Table 2, and make the comparison of oral medication 1 year before and after the use of biologics unreliable.
In Table 3, we demonstrate the advantages between giving biological treatment in the first 2.24 years, compared to those who receive it later, which shows that the number of patients using glucocorticoids could be reduced significantly compared to the other group. (p=0.047) Those patients with delayed use of biologics have a tendency to increase use glucocorticoids. Despites of the statistics of difference of NSAIDs dose not reach significance, we still can see there is a trend of using more NSAIDs in those patients with delayed treatment with biologics (p=0.06). In delayed treatment subgroup, the NSAIDs tend to be prescribed more; 236 patients (57.42% of overall 411 patients) were having more than 75% of NSAIDs prescription days even after treated with biologics. It gives us the hint that delayed treatment with biologics might hinder the process of tapering glucocorticoids and NSAIDs.[16, 26, 27]
Indication bias, comorbidities, and adherence rates (differences between oral prescribed agents and how much patients actually took) are listed as our study limitations. The reimbursement of biologics other than the Etanercept and the Adalimumab as first-line biologics treatment was not available in Rituximab and Tocilizumab by 2013, and the Golimumab was not available until the end of 2012, which could limit the case numbers in this research. Even though early treatment with an immune modulation agent has been proven to be beneficial in rheumatic patients, the adherence rates and comorbidities could be biased. However, due to increased risk of infectious diseases [28] to those with TB, have active or suspected infections, or easily get infected are not recommended to receive early full-dose DMARD agents and glucocorticoids treatment unless infections are under control.
Although biological agents have been considered appropriate pharmaceutical treatment for RA, immunological tolerance, which results in long-term remission, has not yet been established [29], despites several choices of biologics currently available. The search for alternative cures is still needed, and our study has provided some hints that early treatment with biologics may be a better choice than conventional oral medication, and this was the main purpose of our study.
We set the study period to 12 months prior to and 12 months following a one-month washout period after a biological agent in a bid to avoid such time varying covariates as adverse drug effects due to long-term use or progression of disease severity. Residence type (city/country), age, gender, community/nursing home, type of healthcare, and comorbidities [30] are commonly considered, and the current study focused on the efficacy of biologics and treatment timing by comparing changes of days of related drug administration in the same individual. Only gender and age were found to be independent factors that could decrease days of prescription of both steroids and NSAIDs in early use of biologics agents, preferentially Etanercept and Adalimumab after three months’ treatment. This study indicates that younger patients only need short-term (2.53±1.92 years, p=0.03) and early treatment with biologics (within 2.24 years of diagnosis of RA) in order to taper steroids and NSAIDs to less than 50% than before biologics treatment. This result has an important clinical implication that reflects updated treatment guidelines to use steroids at the lowest dose possible [31].