If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Requests for reprints should be addressed to J. Antonio Aviña-Zubieta, MD, PhD, Arthritis Research Centre of Canada, 5591 No. 3 Road, Richmond, BC V6X 2C7, Canada.
Arthritis Research Centre of Canada, Vancouver, British Columbia, CanadaDivision of Rheumatology, Department of Medicine, University of British Columbia, Vancouver, Canada
Arthritis Research Centre of Canada, Vancouver, British Columbia, CanadaDepartment of Rheumatology, Division of Rheumatology, Allergy and Immunology, Harvard Medical School, Boston, Mass
The purpose of this study is to assess risk and time trends of newly recorded myocardial infarction and stroke in cases with systemic sclerosis.
Methods
We conducted a matched incident cohort study (1996-2010) among patients satisfying at least one of the following: 1) diagnosis of systemic sclerosis on at least 2 visits within a 2-year period by a nonrheumatologist physician; or 2) diagnosis of systemic sclerosis on at least one visit by a rheumatologist or from hospitalization; as well as receiving no prior systemic sclerosis diagnosis between 1990 and 1995. Ten controls were matched by birth year, sex, and calendar year of exposure from the general population for each case. Incident myocardial infarction, stroke, and myocardial infarction or stroke was recorded from hospital or death certificates. We estimated incidence rate ratios and hazard ratios (HRs) after adjusting for confounders.
Results
Among 1239 individuals with systemic sclerosis and no history of myocardial infarction (83% female, 56 years old), the incidence rate for myocardial infarction was 13.0/1000 person-years vs 4.1/1000 person-years in the comparison cohort. The incidence rate for stroke was 8.0/1000 person-years vs 3.7/1000 among controls. The adjusted HRs were 3.49 (95% confidence interval [CI], 2.52-4.83) and 2.35 (95% CI, 1.59-3.48) for myocardial infarction and stroke, respectively. For myocardial infarction and stroke, the risk was highest within the first year following diagnosis (HR 8.95; 95% CI, 5.43-14.74 and HR 5.25; 95% CI, 2.90-9.53, respectively).
Conclusion
This large general population-based study indicates an increased risk of myocardial infarction and stroke in patients with systemic sclerosis, especially within the first year of diagnosis.
Patients with new onset of systemic sclerosis have an increased risk of myocardial infarction and stroke compared with the general population.
•
The highest risk of myocardial infarction and stroke occurred within the first year after the diagnosis, suggesting that inflammation plays a pathogenic role in these complications.
•
Our findings support increased vigilance in cardiovascular prevention, surveillance, and risk modification in patients with systemic sclerosis.
Systemic sclerosis is a chronic disease of unknown etiology involving autoimmunity, and is characterized by vasculopathy and fibrosis of the dermis and visceral organs. The microvascular complications of systemic sclerosis, which are primarily a result of obliterative vasculopathy involving intimal proliferation in the small arterioles, manifest as Raynaud's phenomenon, pulmonary hypertension, and scleroderma renal crisis.
Increased risk of macrovascular disease, manifesting as higher rates of stroke and myocardial infarction, has been well established in other rheumatic diseases, such as rheumatoid arthritis
Many studies have examined subclinical radiographic evidence of atherosclerosis (ie, carotid intima-media thickness, brachial artery flow-mediated vasodilation) in systemic sclerosis.
Although the underlying causes are still unclear, 2 meta-analyses have concluded that patients with systemic sclerosis have more atherosclerosis than controls.
Our group was the first to demonstrate that patients with systemic sclerosis have an increased risk of myocardial infarction, stroke, and peripheral vascular disease using The Health Improvement Network (THIN), a primary care database of electronic medical records from the United Kingdom.
To address some of the limitations of previous studies and to assess, for the first time, the time-based risks after the diagnosis of systemic sclerosis, we evaluated the risk of myocardial infarction and stroke in an incident cohort of patients with systemic sclerosis compared with controls randomly selected from the general population.
Patients and Methods
Data Source
Universal health coverage is available for all residents in British Columbia (BC), Canada (population ∼ 4.7 million). Population Data BC captures all provincially funded health care services since 1990, including: all outpatient medical visits,
British Columbia Ministry of Health [creator]. Medical Services Plan (MSP) payment information file. Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
British Columbia Ministry of Health [creator]. Discharge abstract database (hospital separations). Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
British Columbia Ministry of Health [creator]. Medical Services Plan (MSP) payment information file. Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
British Columbia Ministry of Health [creator]. Medical Services Plan (MSP) payment information file. Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
British Columbia Ministry of Health [creator]. Consolidation file (MSP registration & premium billing). Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
BC Vital Statistics Agency [creator]. (2012): Vital statistics deaths. Population Data BC [publisher]. Data extract. BC Vital Statistics Agency (2013). Available at: http://www.popdata.bc.ca/data.
BC Ministry of Health [creator]. (2013): PharmaNet. BC Ministry of Health [publisher]. Data extract. Data Stewardship Committee (2013). Available at: http://www.popdata.bc.ca/data.
which includes all dispensed medications for all BC residents since 1996. Numerous population-based studies have been successfully conducted using these databases.
Using Population Data BC, we conducted cohort analyses of incident myocardial infarction, stroke, and both outcomes combined among individuals with incident systemic sclerosis as compared with age-, sex-, and entry time-matched individuals without systemic sclerosis from a sample of 400,000 individuals randomly selected from the general population by Population Data BC. We created an incident systemic sclerosis cohort with cases diagnosed for the first time between January 1996 and December 2010 defined as individuals who received one of the following: 1) one International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code for systemic sclerosis by a rheumatologist (ICD-9-CM 710.1) or from hospitalization (ICD-9-CM 710.1 or ICD-10-CM M34.0); or 2) diagnosis of systemic sclerosis (>18 years) on at least 2 visits, at least 2 months apart, and within a 2-year period by a nonrheumatologist physician; as well as an absence of a prior systemic sclerosis diagnosis between January 1990 and December 1995 (to ensure incident systemic sclerosis cases). The use of ICD-9 codes to identify systemic sclerosis patients in administrative health databases has been successfully employed in a US study where the incidence rate of systemic sclerosis found was similar to other validated US estimates in the literature.
Moreover, a validation study of ICD-9 codes in a Canadian context found the specificity of systemic sclerosis diagnosis from administrative health data was 94.9% (95% confidence interval [CI], 93.0-96.2).
To further improve specificity, we excluded individuals with at least 2 visits ≥2 months apart subsequent to the systemic sclerosis diagnostic visit with a diagnosis of other systemic autoimmune rheumatic diseases (eg, rheumatoid arthritis, systemic lupus erythematosus). For each comparison cohort, we matched up to 10 individuals, randomly selected from the general population, without systemic sclerosis to each systemic sclerosis case based on age, sex, and calendar year of study entry. We excluded systemic sclerosis cases and controls with myocardial infarction or stroke before the index date.
Individuals with systemic sclerosis entered the case cohort after all inclusion and exclusion criteria had been met. Systemic sclerosis cases and controls were followed until they experienced an outcome, died (29% vs 10%, respectively), left the province (8% vs 13%, respectively), or the follow-up period ended (December 31, 2010), whichever occurred first.
Ascertainment of Myocardial Infarction and Stroke
The primary outcomes were the first myocardial infarction or stroke event during follow-up. We identified myocardial infarction (ICD-9 CM 410 or ICD-10 CM I21) and ischemic stroke events (ICD-9 CM 433 and 434 or ICD-10 CM I63-I66) from hospitalization data
British Columbia Ministry of Health [creator]. Discharge abstract database (hospital separations). Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
BC Vital Statistics Agency [creator]. (2012): Vital statistics deaths. Population Data BC [publisher]. Data extract. BC Vital Statistics Agency (2013). Available at: http://www.popdata.bc.ca/data.
including out-of-hospital deaths. Cardiovascular disease included myocardial infarction or stroke. The validity of ICD codes to identify myocardial infarction and stroke has been widely demonstrated, including in the Canadian context.
Covariates consisted of potential risk factors for cardiovascular disease assessed during the year before the index date, including: health care utilization (outpatient and hospital visits), medication use (glucocorticoids, hormone replacement therapy, contraceptives, COX-2 inhibitors, nonsteroidal anti-inflammatory drugs (NSAIDs), and cardiovascular medications [antihypertensives, cardiac glycosides, diuretics, antiarrhythmic, nitrates, fibrates, and statins]). In addition, the modified Charlson Comorbidity Index for administrative data was calculated in the year before the index date.
We identified incident cases of myocardial infarction and stroke during follow-up and calculated incidence rates of myocardial infarction and stroke, both individually and in combination (ie, cardiovascular disease), per 1000 person-years. We estimated the cumulative incidence of each event for myocardial infarction and stroke accounting for the competing risk of death by using the SAS macro CUMINCID (SAS Institute, Inc., Cary, NC) and graphically presented these results.
We used Cox proportional hazard regression models to assess the risk of myocardial infarction and stroke associated with systemic sclerosis after adjusting for confounders listed above (ie, preexisting comorbidities, cardiovascular medications, nonsteroidal anti-inflammatory drugs, COX-2 inhibitors, and health care resource use). We entered confounders one at a time into the Cox models in a forward selection according to each confounder's impact on the hazard ratio of the systemic sclerosis, relative to the hazard ratio (HR) in the model selected in the previous step. Cut-off for the minimum important relative effect at each step was set at 5%.
In order to evaluate the impact of duration of systemic sclerosis (ie, follow-up time after systemic sclerosis diagnosis), we estimated the HR yearly for the first 5 years.
We used SAS version 9.3 for all analyses. For all HRs, we calculated 95% CIs. All P-values were 2 sided. Ethics approval was obtained from the University of British Columbia.
Sensitivity Analyses
We performed 2 sensitivity analyses to test the robustness of our results. First, we estimated the cumulative incidence of each event, accounting for the competing risk of death according to Lau et al,
and expressed the results as sub-distribution HRs with a 95% CI. Second, to assesses how a hypothetical unmeasured confounder (eg, smoking) might have affected our estimates of the association between systemic sclerosis and the risk of myocardial infarction and stroke,
we simulated unmeasured confounders with their prevalence ranging from 10% to 20% in the systemic sclerosis and control cohorts, and odds ratios for the associations between the unmeasured confounder and myocardial infarction and stroke ranging from 1.3 to 3.0.
Results
We identified 1239 new cases with systemic sclerosis free of myocardial infarction and stroke. Table 1 summarizes the baseline characteristics of this cohort and its comparison cohort. As expected, the systemic sclerosis cases had significantly more prescriptions for cardiovascular drugs, glucocorticoids, hormone replacement therapy, traditional NSAIDs, and COX-2 inhibitors, but not for contraceptives or lipid-lowering medications. In addition, our systemic sclerosis cases had a higher Charlson Comorbidity score and more health care resource utilization than the controls.
Table 1Characteristics of Systemic Sclerosis and Comparison Cohort at Baseline
Variable
SSc (n = 1223)
Non-SSc (n = 12,433)
P-Value
Age, mean (SD)
56.1 (±14.9)
56.1 ± 14.8
NS
Female
1018 (83.2)
10,328 (83.1)
NS
Hospitalized
384 (31.4)
1916 (15.4)
<.001
Number of outpatient visits, mean (SD)
20.35 (±17.20)
8.81 ± 10.81
<.001
Charlson comorbidity index, mean (SD)
1.12 (±1.42)
0.32 ± 1.03
<.001
Angina
72 (5.9)
335 (2.7)
<.001
COPD
182 (14.9)
904 (7.3)
<.001
Obesity
9 (0.7)
27 (0.2)
.004
Glucocorticoids
321 (26.2)
493 (4)
<.001
Cardiovascular drugs
473 (38.7)
3153 (25.4)
<.001
Dipyridamole
5 (0.4)
7 (0.1)
.003
Antidiabetes medication
48 (3.9)
688 (5.5)
.017
HRT
136 (11.1)
952 (7.7)
<.001
Contraceptives
45 (3.7)
474 (3.8)
NS
Fibrates
7 (0.6)
79 (0.6)
NS
Statin
124 (10.1)
1178 (9.5)
NS
Traditional NSAIDs
312 (25.5)
1806 (14.5)
<.001
Cox-2 inhibitors
108 (8.8)
383 (3.1)
<.001
Values are n (percentages) unless otherwise noted.
Overall, systemic sclerosis was associated with a substantial increased risk for myocardial infarction and stroke events (Table 2 and Figure). The age-, sex-, and entry time-matched HRs among systemic sclerosis patients compared with the comparison cohort were 4.26 (95% CI, 3.17-5.73) for myocardial infarction, 3.08 (95% CI, 2.14-4.44) for stroke, and 3.54 (95% CI, 2.77-4.53) for cardiovascular disease. After adjusting for confounders, the corresponding HRs were 3.49 (95% CI, 2.52-4.83), 2.35 (95% CI, 1.59-3.48), and 2.70 (95% CI, 2.07-3.51), respectively (Table 2).
Table 2Risk of MI, Stroke, or Both (CVD) According to Systemic Sclerosis Status
Fully adjusted models include the following selected covariates: MI: number of outpatient visits, glucocorticoids, and diabetes medication; Stroke: number of outpatient visits and glucocorticoids; CVD: number of outpatient visits and glucocorticoids.
Fully adjusted models include the following selected covariates: MI: number of outpatient visits, glucocorticoids, and diabetes medication; Stroke: number of outpatient visits and glucocorticoids; CVD: number of outpatient visits and glucocorticoids.
Fully adjusted models include the following selected covariates: MI: number of outpatient visits, glucocorticoids, and diabetes medication; Stroke: number of outpatient visits and glucocorticoids; CVD: number of outpatient visits and glucocorticoids.
(95% CI)
2.70 (2.07-3.51)
1.0
CI = confidence interval; CVD = cardiovascular disease; HR = hazard ratio; MI = myocardial infarction; SSc = systemic sclerosis.
∗ Fully adjusted models include the following selected covariates: MI: number of outpatient visits, glucocorticoids, and diabetes medication; Stroke: number of outpatient visits and glucocorticoids; CVD: number of outpatient visits and glucocorticoids.
Overall, when we evaluated the impact of follow-up time after systemic sclerosis diagnosis, the HRs for myocardial infarction, stroke, and cardiovascular disease were greatest in the first year after diagnosis (Table 3). The risks attenuated in the following years, but remained statistically significant for over 5 years following the initial diagnosis for all outcomes.
Table 3Age- and Sex-adjusted Cox HRs for MI, Stroke, or Both (CVD) in SSc According to Follow-up Period
Time after Diagnosis
MI HR (95% CI)
Stroke HR (95% CI)
CVD HR (95% CI)
<1 y
8.95 (5.43-14.74)
5.25 (2.90-9.53)
6.81 (4.55-10.20)
<2 y
6.20 (4.10-9.38)
4.57 (2.75-7.59)
5.00 (3.55-7.05)
<3 y
6.35 (4.45-9.07)
3.60 (2.26-5.74)
4.61 (3.41-6.22)
<4 y
5.55 (3.98-7.74)
3.21 (2.08-4.95)
4.18 (3.16-5.53)
<5 y
5.20 (3.77-7.18)
3.29 (2.18-4.95)
4.00 (3.05-5.24)
CI = confidence interval; CVD = cardiovascular disease; HR = Hazard ratio; MI = myocardial infarction; SSc = systemic sclerosis.
HRs from analyses of the association between systemic sclerosis and myocardial infarction or stroke that accounted for the competing event of death gave slightly attenuated estimates, but remained significant (Table 4). Moreover, when assessing for the potential impact of an unmeasured confounder between systemic sclerosis and all outcomes, the adjusted HRs remained statistically significant (Table 4).
Hypothetical level of association between the unmeasured confounder and the outcome.
= 3.0
MI
3.49 (2.52-4.83)
2.58 (1.87-3.56)
3.44 (2.49-4.77)
2.77 (1.98-3.87)
3.48 (2.51-4.82)
2.89 (2.08-4.01)
Stroke
2.35 (1.59-3.48)
1.72 (1.16-2.54)
2.31 (1.56-3.42)
1.78 (1.19-2.66)
2.30 (1.55-3.41)
1.79 (1.19-2.69)
CVD
2.70 (2.07-3.51)
2.06 (1.58-2.69)
2.67 (2.05-3.49)
2.35 (1.79-3.07)
2.62 (2.01-3.42)
2.04 (1.55-2.69)
Primary analysis is the fully adjusted model; subdistribution models are adjusted for the same covariates as the primary analysis in addition to age and sex; simulated confounder models include additional covariates per the selection algorithm.
CI = confidence interval; CVD = cardiovascular disease; HR = hazard ratio; MI = myocardial infarction; OR = odds ratio.
∗ Hypothetical prevalence of unmeasured confounder in systemic sclerosis cases.
† Hypothetical level of association between the unmeasured confounder and the outcome.
In this large general population-based study, we found that systemic sclerosis was associated with a substantially increased risk of myocardial infarction and stroke. This increased risk was independent of relevant risk factors. The increased risk was the highest during the first year after the diagnosis of systemic sclerosis, and remained significantly high even after 5 years for both outcomes. These findings add a new dimension to our current understanding of the cardiovascular risk profile in systemic sclerosis in a truly general population-based context.
For example, 2 previous studies reported an increased prevalence of coronary heart disease, and ischemic heart disease in systemic sclerosis patients; however, they reported no difference in the prevalence of stroke.
Additionally, in a study using 865 systemic sclerosis patients from THIN, our group reported a significantly increased risk of myocardial infarction (relative risk [RR] 1.80; 95% CI, 1.07-3.05) and stroke (RR 2.61; 95% CI, 1.54-4.44).
In comparison with previous studies, our current study has shown an even higher risk of both myocardial infarction and stroke in systemic sclerosis patients. The current study used a case definition that employed systemic sclerosis diagnostic codes from a rheumatologist or hospitalization. Conversely, THIN relies on cases identified from Read codes in the records from primary care physicians. Thus, compared with the THIN database, our study likely captured all myocardial infarctions and strokes because we utilized hospitalization data
British Columbia Ministry of Health [creator]. Discharge abstract database (hospital separations). Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
BC Vital Statistics Agency [creator]. (2012): Vital statistics deaths. Population Data BC [publisher]. Data extract. BC Vital Statistics Agency (2013). Available at: http://www.popdata.bc.ca/data.
Moreover, we included cases recorded by both rheumatologists and nonrheumatologist physicians, whereas other studies, such as the one utilizing the THIN database, may have missed severe cases that would have only been captured by specialists, which may or may not have been at higher risk for cardiovascular disease. Furthermore, a unique feature from our data is that we have all medications dispensed for all individuals.
likely included nearly all systemic sclerosis cases, and their results are closer to the estimates of the current study. However, controls in this study were matched by comorbidity level; therefore, having a control group with a higher baseline risk than the general population attenuates the risk estimates.
There are a number of potential mechanisms by which systemic sclerosis may lead to an increased risk of myocardial infarction and stroke events. The etiology can be ascribed to microvascular and macrovascular abnormalities,
however, which one is the primary cause remains a matter of ongoing debate. It has been thought that enhanced atherosclerosis, a macrovascular pathology, underlies the association between myocardial infarction and systemic sclerosis.
Systemic sclerosis-specific vasculopathy involving chronic thrombotic microangiopathic processes may result in small vessel thrombi, microvascular impairment, and intimal fibrosis, causing ischemia in both the heart and the brain. Indeed, an autopsy study of 8 systemic sclerosis patients showed narrowing and platelet-fibrin clots in the small coronary arteries.
This led to the theory of a “myocardial Raynaud's phenomenon,” according to which, damage to the heart, and possibly to the brain as well, is caused by repeated ischemia-reperfusion events.
It remains unknown how the antibody profiles associated with systemic sclerosis subsets contribute to the increased risks of myocardial infarction and stroke. Nordin et al
found that a greater number of ischemic arterial events (defined as ischemic heart disease, cerebrovascular, and peripheral vascular diseases) and more plaques were observed in anticentromere antibody-positive systemic sclerosis patients as compared with both anticentromere antibody-negative patients and controls.
In contrast, ischemic events were rare in the antitopoisomerase 1-positive group. Interestingly, it was previously reported that carotid wall stiffness was more pronounced in diffuse cutaneous systemic sclerosis than in limited cutaneous systemic sclerosis.
Unfortunately, we did not have laboratory data in our databases and were subsequently unable to assess the risks among the different autoantibodies associated with systemic sclerosis, or to distinguish between diffuse and limited systemic sclerosis.
Regardless of the microvascular or macrovascular manifestations of systemic sclerosis, acute and chronic inflammation from the disease itself likely contributes to both myocardial infarction and stroke.
Inflammation in systemic autoimmune diseases tends to peak at the time of diagnosis, which is when patients are likely the most symptomatic. This may account for our observed increased risk of myocardial infarction and stroke events in systemic sclerosis patients within the first year following initial diagnosis.
Similar patterns of increased risk of deep vein thrombosis and pulmonary embolism within the first year of diagnosis have also been observed in systemic sclerosis and other autoimmune diseases.
Although the risk of myocardial infarction and stroke slightly trended down over 5 years, it remained significantly elevated in comparison with the control cohort, suggesting the chronic inflammation may continue to contribute to the development of vascular complications. Indeed, it has been well established that chronic inflammation is a risk factor for atherosclerosis.
We found our systemic sclerosis cases used significantly more NSAIDs, COX-2 inhibitors, and glucocorticoids when compared with controls from the general population, all of which have been associated with increased myocardial infarction and stroke events.
Coxib and traditional NSAID Traialists (CNT) Collaboration Vascular and upper gastrointestinal effects of non-steroidal anti-inflammatory drugs: meta-analyses of individual participant data from randomised trials.
However, only glucocorticoids were selected in the multivariable model, and after adjusting for that, the results remained statistically significant. Similar results were observed in the study by Man et al,
lending further evidence to the fact that our observed increased risk of stroke events is primarily due to the presence of underlying systemic sclerosis.
The findings of increased myocardial infarction and stroke among patients with systemic sclerosis could have important implications for clinical care, both immediately after a diagnosis of systemic sclerosis and in the long-term follow-up. A diagnosis of systemic sclerosis could prompt clinicians to conduct a careful history and examination for symptoms and signs of cardiovascular disease, as well as to obtain cardiovascular testing when appropriate. Consideration could also be given to avoid medications and habits that increase the risk of ischemic arterial events, such as encouraging the cessation of smoking and the cautious use of NSAIDs.
The strengths and limitations of our study deserve comment. Our current study has the advantage of the utilization of data covering nearly the entire population of BC, making our results generalizable to the general population. In addition, we were able to adjust for preexisting comorbidities, health resource utilization, and medication use for all cases and controls, unmeasured confounders, and competing risk of death. We also used one of the strictest case definitions for administrative databases.
The potential limitations of our study are inherent in all observational studies utilizing administrative data. For example, both the diagnosis of systemic sclerosis and outcomes of myocardial infarction and stroke were based solely on ICD coding data, which allows for the possibility of misclassification. However, ICD coding for systemic sclerosis in Canadian administrative databases has been found to have a high specificity (94.9%).
Moreover, 80% of our cases met the case definition from hospitalization or from rheumatologist diagnosis, making our identified cases more likely to be true cases.
Furthermore, “false positive cases” would make the point estimates closer to the null hypothesis, thus resulting only in a conservative bias. A number of studies have also examined the accuracy of ICD-9 and ICD-10 coding for both myocardial infarction
Accuracy of Medicare claims-based diagnosis of acute myocardial infarction: estimating positive predictive value on the basis of review of hospital records.
Identification of acute myocardial infarction from electronic healthcare records using different disease coding systems: a validation study in three European countries.
and found the coding to be valid and accurate for both outcomes.
An additional limitation of our study was that we did not have access to specific disease subtypes, level of disease activity, autoantibody profiles, or the extent of organ involvement in our systemic sclerosis cases. Lastly, we did not have access to body mass index or smoking data, which are relevant potential confounders. However, in our previous study (using the THIN database), in which we did have access to data on these risk factors, the difference in our results was minimal. Nevertheless, to account for this we conducted a sensitivity analysis involving a simulated unmeasured confounder, and all our results remained statistically significant, suggesting that our results are robust.
Conclusion
In summary, we found that patients with systemic sclerosis have over a fourfold increased risk of myocardial infarction and nearly a threefold increased risk of stroke when compared with the general population. The risk was greatest in the first year following diagnosis. Although decreasing with time, the risk remained significantly elevated for at least 5 years after diagnosis. Our findings support increased vigilance in cardiovascular prevention, surveillance, and risk modification in patients with systemic sclerosis.
Acknowledgments
We want to thank Kathryn Reimer for her editorial assistance.
British Columbia Ministry of Health [creator]. Medical Services Plan (MSP) payment information file. Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
British Columbia Ministry of Health [creator]. Discharge abstract database (hospital separations). Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
British Columbia Ministry of Health [creator]. Consolidation file (MSP registration & premium billing). Population Data BC [publisher]. Data extract. MOH (2013). Available at: http://www.popdata.bc.ca/data.
BC Vital Statistics Agency [creator]. (2012): Vital statistics deaths. Population Data BC [publisher]. Data extract. BC Vital Statistics Agency (2013). Available at: http://www.popdata.bc.ca/data.
BC Ministry of Health [creator]. (2013): PharmaNet. BC Ministry of Health [publisher]. Data extract. Data Stewardship Committee (2013). Available at: http://www.popdata.bc.ca/data.
Coxib and traditional NSAID Traialists (CNT) Collaboration
Vascular and upper gastrointestinal effects of non-steroidal anti-inflammatory drugs: meta-analyses of individual participant data from randomised trials.
Accuracy of Medicare claims-based diagnosis of acute myocardial infarction: estimating positive predictive value on the basis of review of hospital records.
Identification of acute myocardial infarction from electronic healthcare records using different disease coding systems: a validation study in three European countries.
Funding: The study was funded by the Canadian Arthritis Network, the Arthritis Society of Canada, the British Columbia Lupus Society (Grant 10-SRP-IJD-01 ) and the Canadian Institutes for Health Research (Grant THC 13523 ). JAA-Z is the British Columbia Lupus Society Research Scholar and a Michael Smith Foundation Research Scholar.
Conflict of Interest: We certify there is no conflict of interest with any financial organization regarding the material discussed in this manuscript.
Authorship: Drs Avina-Zubieta and Sayre had access to the data and all authors had a role in writing the manuscript.
01051-7&id=gr1.jpg){kind=link}