Mortality in Patients with rheumatoid Arthritis: A Retrospec-

Background and Objectives: Mortality rates among patients with rheumatoid arthritis (RA) have been reported to be higher than in the general population. The long-term prognosis of RA has improved in recent years due to early diagnosis as well as effective pharmacological treatment and may be able to diminish the excess mortality risk. This study was designed to investigate mortality (a) in patients with RA in a retrospectively defined national RA cohort in comparison with the general Lithuanian population and (b) to conduct a systematic review of the literature from different countries and meta-analysis. Materials and Methods: In this national retrospective cohort study patients with the first-time diagnosis of RA during the period between 1 January 2013 and 31 December 2017 were identified from the Lithuanian Compulsory Health Insurance Information System database SVEIDRA. All cases were cross-checked with Health Information center at the Institute of Hygiene, for the vital status of these patients and date of death if the fact of death was documented. The standardized mortality ratios (SMRs) with 95% confidence intervals (CI) obtained for all-cause mortality in patients with RA adjusted for age, sex, and calendar year, were calculated. The search for published studies by a combination of keywords “rheumatoid arthritis AND standardized mortality ratio” was performed in MEDLINE (via PubMed, OVID and EBSCO), Science Direct, Tylor & Francis, Springer databases. Studies were selected according to described inclusion and exclusion criteria listed in the paper and a meta-analysis was conducted. A random-effect meta-analysis model was used to compute the pooled standardized mortality ratios (meta-SMRs). Results: Overall, 4623 patients with newly diagnosed RA during the 2013–2017 period were identified and enrolled in the Lithuanian population-based cohort. The mean age of patients at the time of RA diagnosis was 58.7 (standard deviation (SD) 15.1) years, and 77.1% of the patients were women. The estimated SMR for all-cause mortality was 1.15 (95% CI 1.02, 1.29). The SMR for men (SMR 1.14, 95% CI 0.94, 1.39) was higher than for women (SMR 1.03, 95% CI 0.89, 1.19). A systematic literature search revealed 12 studies meeting the inclusion criteria, starting from 2010 to 2020, representing 50 072 patients. The meta-SMR in patients with RA for all-cause mortality was 1.41 (95% CI 1.29, 1.55). Allcause mortality risk was higher for men (meta-SMR 1.53, 95% CI 1.31, 1.78) than for women (metaSMR 1.46, 95% CI 1.2, 1.77). Conclusions: In a retrospectively defined population-based national RA cohort a 15% excess risk of death was observed among patients with RA compared to the general Lithuanian population. Patients with RA have a higher mortality risk than the general population. Published data indicate that the risk of mortality is increased by 41% in patients with RA compared to the general population. Excessive all-cause mortality risk is higher in males than in women. National data showed lower standardized mortality if compared to literature data.


Introduction
Rheumatoid arthritis (RA) is a chronic, systemic, autoimmune disorder that limits a person's performance of physical functions and has a negative impact on quality of life [1,2]. Higher mortality rates in patients with RA compared to the general population were first described in a longitudinal observational study over 65 years ago [3]. The long-term prognosis of RA has improved in recent years due to early diagnosis and effective pharmacological treatment [4], resulting in most RA patients achieving stable clinical remission or experiencing lower disease activity [5]. Nevertheless, most studies have found that RA is still associated with higher mortality risk than the general population [6][7][8][9][10]. The main causes of death identified among RA patients are increased incidence of circulatory system diseases, cancer and respiratory conditions [11]. However, some studies report that the mortality in patients with RA was similar or even lower to that of the general population [12][13][14]. Reason for the contradictory results could be explained by different types of cohorts, length of the follow-up and the geographical location of the study [7]. Given these conflicting results, a better estimation of the mortality risk among RA patients over time is essential to understand the prognosis of this disease.
Hence, the objectives of this study were to investigate mortality (a) in patients with RA in a retrospectively defined national RA cohort in comparison with the general Lithuanian population and (b) to conduct a systematic review of the literature from different countries and meta-analysis.

Data Sources
This national retrospective cohort study was performed using the data of Lithuanian Compulsory Health Insurance Information System database SVEIDRA. It is a populationbased database with data that captures all physician visits, procedures, hospitalizations, diagnoses, and prescribed reimbursed medications to all residents of Lithuania since 1995. The main information sources of this database are health care institutions and medication prescriptions released by pharmacies.
We were allowed to use the information from SVEIDRA on all patients who had a first-time diagnosis of RA (diagnosis codes M05 and M06 according to International Classification of Diseases 10th version (ICD-10) during the period between 1 January 2013 and 31 December 2017. All participants were classified as cases if they had records of at least one prescription of the medications for RA reimbursed by the state, including glucocorticoids, conventional synthetic (methotrexate, azathioprine, leflunomide, sulfasalazine, and hydrochloroquine), or biological disease modifying drugs. Cases of children (<18 years old at the time of diagnosis) were excluded, as well as cases with unidentifiable identification code. The final 4623 cases were cross-checked with Health Information center at the Institute of Hygiene, for the vital status of these patients and date of death if the fact of death was documented.

Search Strategy, Study Selection and Data Extraction
We conducted a systematic review and meta-analysis from 1 September 2020 to 15 December according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [15]. The search for published studies by a combination of keywords "rheumatoid arthritis AND standardized mortality ratio" was performed in MEDLINE (via PubMed, OVID and EBSCO), Science Direct, Tylor & Francis, Springer databases. Eligibility criteria were (1) study population with rheumatoid arthritis aged 15 and over; (2) retrospective or prospective cohort studies published in the period 2010-2020; and (3) outcome of interest mortality, reported as a standardized mortality ratio (SMR), or easily calculated from reported data for the entire study period. If data were duplicated in more than one article, only the most recent one was included. Studies on cancers, cardiovascular diseases or infections related to the musculoskeletal system, as well as studies dealing solely with predictors of mortality but not reporting rates or reporting only hypothetical empirical data were excluded for this review.
Two authors independently reviewed the title and abstract of studies identified in the electronic search to exclude studies that did not address the main research question of interest, based on described inclusion and exclusion criteria (see above). The full texts of the remaining articles were examined to determine whether they contained relevant information; discrepancy in article selection was resolved by consensus.
Data collected were the general characteristics of each study and the outcomes measured: primary author, year of publication, geographic location of the population studied, study design, number of participants, age at baseline of the patients, time period of study, SMR and confidence intervals (CI), observed and expected death.
Assessment of risk of bias relied on the New Castle-Ottawa scale for cohorts and only moderate to high-quality studies were included in the review [16]. According to the reported scale, Studies were evaluated across three domains: selection (four questions) and comparability (two questions) of study groups and determination of the outcome of interest (three questions), with all questions with a score of 1, except for comparability of study groups, where separate points were awarded for controlling age and/or sex (maximum 2 points). Hence, a score with a range of 0-9 was allocated to each study, and those with a score of 8 or more were considered to be high-quality studies; with a score from 6 to 7 were categorized as moderate-quality studies.

Statistical Methods
The SMRs were identified in order to compare the mortality in the retrospectively defined national RA cohort with the general population. The SMRs were computed as the ratio of the number of observed deaths in a study population divided by the number of expected deaths if the study population had the same age, sex, and calendar year specific rates as the general population, with 95% CI. Indirect standardization was used to calculate the expected number of deaths for the study population. Person-years were estimated from the date of RA diagnosis to the first of the death or end of the follow-up (31 December 2017). Expected numbers were computed as multiplication of the exact person-years under observation in the cohort by sex-, calendar year-, and 5-year-age-groups-specific national deaths rates. National mortality data were obtained from Statistics Lithuania for the years 2013-2017 [17]. We calculated the SMR for all-cause mortality.
Only studies reporting SMR were included in the systematic review and meta-analysis. If an SMR or its 95% CI were not directly provided, then they were estimated from the reported observed (O) and expected (E) deaths, as SMR = O/E and its 95% CI = SMR ± 1.96√ O/E. A random-effect meta-analysis model described by DerSimonian and Laird was used to compute the pooled standardized mortality ratios (meta-SMRs) with 95% CI [18]. The meta-SMR shows a summary estimate of the increased risk of death in patients with RA compared with the general population, weighted by the inverse of the variance the log of the SMRs of each study. We assessed heterogeneity between study-specific estimates using the inconsistency index (I 2 statistic). All analyses were performed using WINPEPI [19].

Results
During the period between 2013 and 2017, we have identified 4 623 patients with RA. The mean age of these patients at the time of RA diagnosis was 58.7 (standard deviation (SD) 15.1) years. The cohort consisted mainly of women (77.1%), and the mean duration of follow-up was 2.78 years. A total number of 278 patients died during the study period: 98 men and 180 women. The overall mean age at death was 74.8 (SD 10.6) years.
The estimated SMR for all-cause mortality was 1.15 (95% CI 1.02, 1.29) over the period between 2013 and 2017 (Table 1). This indicates a 15% higher risk of mortality compared with the general Lithuanian population. In men, SMR was 1.14 (95% CI 0.94, 1.39), and in women, it was 1.03 (95% CI 0.89, 1.19).  (Figure 1), in all 394 after removal of duplicates, incomplete text and articles older than 2010. The selection by title and abstract left 23 articles. Finally, only 12 studies met the complete set of selection criteria and were included in the review. Articles were excluded mainly because of no SMR was provided or could not be calculated from the reported data for the entire study period (n = 5), unspecified age of participants or study sample consisted of patients under 15 years of age (n = 5) and hypothetical empirical data were reported (n = 1).
Revealed 12 studies represented 50 072 patients, and 6 060 deaths occurred during follow-up. Eight of the studies were performed in Europe and four in other parts of the world. Inclusion start-up ranged from 1985 to 2010, and RA diagnosis was mostly based on American College of Rheumatology or American Rheumatism Association classification criteria. Table 2 represents the main characteristics of all included studies, and Table  3 shows the results on mortality by study.   The results of the meta-analysis are shown in Figure 2. The meta-SMR in patients with RA for all-cause mortality was 1.41 (95% CI 1.29, 1.55). Literature data indicate that the risk of mortality is increased by 41% in patients with RA compared to the general population. Excessive all-cause mortality risk was higher for men (meta-SMR 1.53, 95% CI 1.31, 1.78) than for women (meta-SMR 1.46, 95% CI 1.2, 1.77) ( Table 4).

Discussion
Using national registries data from official state-run sources, we have assessed patients with RA in Lithuania and found a 15% excess risk of death in this cohort compared with the general population. However, identified national data showed lower standardized mortality if compared to literature data. In our performed systematic review and meta-analysis patients with RA had a 41% higher risk of mortality compared with the general population. Our findings are in line with the previously two published meta-analyses on the issue of mortality in RA. In both studies, revealed meta-SMR was similar to that identified in our study; reported meta-SMR in meta-analysis performed by Toledano et al. was 1.44 (95% CI 1.23, 1.69) [6], in another meta-analysis conducted by Dadoun et al. estimated meta-SMR was 1.47 (95% CI 1.19, 1.83) [7]. These similarities to other studies reported results reveal that excess mortality in RA still occurs and that the overall mortality rate is still as high as it was in the past decades [7]. According to the review of T. Sokka et al. the leading causes of death in RA patients are similar to the general population, with cardiovascular diseases being the most common cause of death, and with more infection, pulmonary and renal disease in RA than in the general population [29]. Therefore, the reasons for the indicated lower mortality among RA patients in Lithuania might be the same reasons that could be applied to the general population, improvements in rheumatology care and new treatment strategies such as conventional synthetic and biological disease modifying drugs introduced since 2003 may have had an impact on mortality of patients with RA.
Additionally, mortality risk was slightly higher for men than for women in our retrospectively defined national RA cohort. Data supporting this finding were reported in studies of patients living in Sweden and Denmark [24,27]. Nonetheless, our observation is contrary to the studies performed in countries such as Germany and Spain [9,20]. Evidence suggests that difference between the mortality in men and women with RA can be due to innate and adaptive immune responses and environmental, dietary and lifestyle factors [30,31]. Notably, our identified meta-SMR in a conducted meta-analysis was also higher among men than in women. However, this result was in contrast to meta-analysis performed by Toledano et al., which reported higher mortality risk among women compared to men [6].
Some limitations of this study should be mentioned. One of the inclusion criteria in this national retrospective cohort study was information about at least one prescription of the medications for RA reimbursed by the state, therefore, some cases of RA might be omitted in a case when patient is not treated with state-reimbursed medications. The major limitation of this study was a short follow-up period being in average 2.78 years. Furthermore, some studies evaluating mortality in RA patients were excluded because the data needed in order to calculate the SMR were not available in the articles.
Based on our established results and on the studies included in a meta-analysis, we can conclude that patients with RA have an increased risk of death compared to the general population. Despite new treatment strategies, RA remains a serious disease posing an increased risk of mortality and other studies are required to identify factors that may decrease this risk to patients with RA.

Conclusions
Our findings support the hypothesis that patients with RA have higher risks of mortality than the general population. In a retrospectively defined population-based national RA cohort a 15% excess risk of death was observed among patients with RA compared to the general Lithuanian population. National data showed lower standardized mortality if compared to literature data, which reveal that the risk of mortality is increased by 41% in patients with RA compared with the general population. Excessive all-cause mortality risk was higher in males than in women.

Institutional Review Board Statement:
The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Lithuanian Ethics Committee: approval number 158200-17-958-462, November 7, 2017. The document with permission to conduct this study may be added upon request.

Informed Consent Statement:
This study received the waiver form from Informed consent being signed by participants. The document with permission to conduct the study without informed consent may be added upon request.

Conflicts of Interest:
The authors declare no conflict of interest.