Rheumatology Advance Access originally published online on November 6, 2007
Rheumatology 2007 46(12):1808-1813; doi:10.1093/rheumatology/kem273
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Published by Oxford University Press 2007.
Hospitalizations and mortality in systemic sclerosis: results from the Nationwide Inpatient Sample
1Division of Immunology and Rheumatology, Stanford University, Stanford, CA, 2Palo Alto Veteran Affairs Health Care System, Palo Alto, CA and 3Department of Rheumatology, University of Pittsburgh, Pittsburgh, PA, USA.
Correspondence to: L. Chung, 3801 Miranda Ave, Palo Alto VA Health Care System, Palo Alto, CA 94304, USA. E-mail: shauwei{at}stanford.edu
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Objective. To study the causes of hospitalizations and predictors of subsequent adverse outcomes for contemporary cohorts of patients with systemic sclerosis (SSc) in the USA.
Methods. The data source was the 2002 and 2003 Healthcare Cost and Utilization Project-Nationwide Inpatient Sample (HCUP-NIS) databases. We identified all discharges with an International Classification of Diseases-Clinical Modification (ICD9-CM) code of 710.1 (limited and diffuse SSc), then excluded those with concomitant diagnoses for lupus or rheumatoid arthritis. We calculated hospitalization rates, in-hospital mortality rates and mean length of stay (LOS). Multivariate logistic and linear regression models for in-hospital death and LOS were performed adjusting for sociodemographic and comorbidity covariates.
Results. The overall in-hospital mortality rate was 6.3% and the mean LOS was 6.6 days. Hospitalization rates were 4.5 times higher in women than in men, but in-hospital mortality was 
25% lower (P = 0.005). SSc was the most common principal diagnosis for all SSc hospitalizations, with the most common secondary diagnosis (24%) being pulmonary fibrosis. After SSc, respiratory failure was the second most common principal diagnosis in patients who died. Pulmonary fibrosis increased the odds of in-hospital death by 2.63 [95% confidence interval (CI) 1.98–3.49] fold and increased LOS by 7.25% (95% CI 0.90–13.60).
Conclusions. Women with SSc had higher rates of hospitalization but lower in-hospital mortality than men. Pulmonary fibrosis was the major predictor of poor hospitalization outcomes in SSc patients in recent years, emphasizing the importance of continuing to develop more effective therapies for this fatal complication of the disease.
KEY WORDS: Systemic sclerosis, Hospitalizations, Mortality, Epidemiology, Pulmonary fibrosis
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Systemic sclerosis (SSc) is an autoimmune disease characterized by microvascular abnormalities and fibrosis of the skin and internal organs. Patients are classified as limited or diffuse based on the extent of skin thickening. Patients with limited SSc have cutaneous sclerosis of the face and distal extremities alone, while those with diffuse disease have cutaneous involvement of the trunk, abdomen and proximal extremities as well. Internal organ involvement tends to occur earlier in the course of disease in patients with diffuse compared with limited disease [1]. Females are affected 4.6 times more than males and the prevalence of disease is 1.15 times higher in blacks than in whites [2].
SSc is associated with significant morbidity and mortality, with an overall mortality rate in the USA of 3.9 per million person-years [3]. Although multiple studies have evaluated mortality rates and causes of death in patients with SSc [4–9], few studies have analysed hospitalization patterns. A previous study using the 1995 Healthcare Cost and Utilization Project (HCUP)-Nationwide Inpatient Sample (NIS) estimated an overall in-hospital mortality rate of 7.1% and mean length of stay (LOS) of 7.5 days [10]. This study included data from only 19 states to determine national estimates. In addition, indications for hospitalization were not examined.
Over the past decade, several advances in the management of SSc have occurred that may affect hospitalization patterns. Early recognition and treatment of scleroderma renal crisis with angiotensin-converting enzyme (ACE) inhibitors has improved the 5-yr survival from 10 to 50–65% [11, 12]. Early screening for alveolitis and treatment with cyclophosphamide has resulted in beneficial effects on lung function [13–15]. Lastly, since 2002, additional medications have become available for the treatment of pulmonary arterial hypertension (PAH), including inhaled and subcutaneous prostanoids, endothelin receptor antagonists and phosphodiesterase-5 inhibitors [16, 17]. This has led to an improvement in 2-yr survival of patients with SSc-associated PAH from 47 to 71% [17]. Given the advances in management of SSc in the last decade, we sought to evaluate hospitalization patterns in the USA in more recent years by determining hospitalization rates, indications for hospitalization and factors associated with in-hospital death and LOS during the years 2002–03.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Study design and population
The Stanford University Institutional Review Board approved our study. We used the HCUP-NIS database to perform a cross-sectional analysis of all discharges from a stratified sample of USA hospitals for the years 2002 and 2003. The NIS is a nationally representative sample of 20% of all non-federal hospitals in the USA with information from
1000 hospitals in 35 states in 2002 and 37 in 2003. This database contains clinical and demographic information from discharge abstracts. Hospital and discharge weights that account for missing states are provided for calculating national estimates (http://hcup-us.ahrq.gov/db/nation/nis/reports/NIS_2004_Design_Report). Because HCUP does not release patient identifiers for confidentiality reasons, we used hospital discharge as the unit of measure for our analyses. Discharge abstracts included the following variables: principal diagnosis (the primary reason for hospitalization), up to 14 secondary diagnoses, discharge status (died in the hospital or not), LOS and insurance status. We identified all discharges for which an International Classification of Diseases-Clinical Modification, Ninth Revision (ICD9-CM) code of 710.1 was included in any of the discharge diagnoses, thereby identifying patients with limited and diffuse SSc, but excluding localized scleroderma/morphea. We excluded all patients with ICD9-CM codes for systemic lupus erythematosus (SLE; 710.0) or rheumatoid arthritis (RA; 714.0) in order to improve the reliability for a diagnosis of SSc.
Population hospitalization rates and indications for hospitalization
Population hospitalization rates for SSc patients were determined for each year by gender, race and age group. Race was classified as white, black or other. Eleven states did not provide data on race and were considered missing. Age was divided into three groups: 15–44, 45–64 and 
65 yrs. Population hospitalization rates were calculated for each subgroup by dividing the number of hospitalization estimates (after applying sampling weights) by respective gender, race and age-specific population estimates obtained from the US Census (http://www.cdc.gov/nchs/). Stata version 8.0 software (Stata, College Station, TX, USA) was used to determine 95% confidence intervals (CI) for hospitalization rates.
After limiting the data set to SSc hospitalizations, we used the HCUP Clinical Classifications Software (CCS) for ICD9-CM to cluster patient diagnoses into related categories (http://www.hcup-us.ahrq.gov/toolssoftware/ccs/ccs/jsp). We then identified the 20 most common discharge diagnoses for all SSc hospitalizations and for those SSc patients who died in the hospital.
In-hospital mortality and length of stay
The overall in-hospital mortality rate was calculated as the proportion of SSc hospitalizations that resulted in death. In-hospital mortality rates were calculated by gender, race and age. Chi-square tests were used to determine if frequencies differed among the categories in each subgroup with P < 0.05 considered statistically significant. The mean LOS was compared between each category using Student's t-test or analysis of variance where appropriate. SAS version 9.1 statistical software was used to perform all analyses.
Regression models
Regression models were developed to identify factors associated with in-hospital mortality and LOS. We were unable to develop survey (SVY) models as there were too few primary sampling units (PSU). Univariate analyses were performed with age, gender and race as the independent variables. For in-hospital mortality, logistic regression analyses were performed using death as the dependent variable. Because LOS was highly skewed, we used a logarithmic transformation of LOS as the dependent variable in the linear regression models.
For multivariate models, independent variables included race, age, gender, health insurance status (private vs public), median household income by zip code (<$45 000 vs $45 000), admission type (emergency vs non-emergency), admission source (transfer from another hospital or not) and the number of diagnoses and procedures performed. The following clinically relevant SSc-related conditions were added to the model: renal failure, pulmonary fibrosis, PAH, pericarditis, oesophageal dysfunction, cachexia and myositis.
The models were adjusted for comorbid conditions using the HCUP Comorbidity Software, Version 3.1 (http://www.hcup-us.ahrq.gov/toolssoftware/comorbidity/comorbidity/jsp). This software assigns dichotomous variables that identify comorbidities in hospital discharge records using ICD9-CM codes. The following diagnoses were included: congestive heart failure (CHF), valvular disease, peripheral vascular disease, hypertension, stroke and other neurological diseases, chronic pulmonary disease, diabetes, hypothyroidism, liver disease, peptic ulcer disease, acquired immunodeficiency syndrome (AIDS), malignancy, haematological disorders, obesity, weight loss, fluid and electrolyte disorders, substance abuse and psychiatric illnesses. ICD9-CM codes that were used to define comorbid conditions are available from the authors. For each model, backward stepwise selection was performed to exclude those comorbid conditions that were not moderately associated with the dependent variable (P > 0.10). SSc-related conditions, CHF and hypertension were retained in the models as pertinent conditions in SSc patients. Sensitivity analyses were performed to account for multiple hospitalizations of the same individuals. Hospitalization records with the identical age, gender, race, income, health insurance, hospital identification code and year of admission were identified as duplicates. Regressions were repeated including only non-duplicated discharges.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Population hospitalization rates
We estimated the total number of discharges for SSc patients as 21 760 (95% CI 19 752–23 759) in 2002 and 23 885 (95% CI 21 312–26 379) in 2003. Population hospitalization rates per million population are shown in Table 1. Hospitalization rates were similar by year, with overall rates of 75.8 per million (95% CI 68.8–82.7) in 2002 and 82.4 per million (95% CI 73.6–91.1) in 2003. Rates were approximately 4.5 times higher in females than in males, with a similar pattern in each racial category. Population hospitalization rates were consistently higher for blacks compared with whites in each age group except for those aged
65 yrs.
|
Indications for hospitalization
The most common principal diagnoses for SSc patients hospitalized in 2002–2003 are shown in Table 2. Diseases of the circulatory system [i.e. CHF, acute myocardial infarction (MI), dysrhythmias] accounted for 21.9% of all hospitalizations, followed by diseases of the gastrointestinal (12.7%), musculoskeletal (12.0%) (including SSc, osteoporosis/fracture and septic arthritis) and respiratory (11.5%) systems. Diseases of the genitourinary system, including kidney, bladder and genital conditions, accounted for 3.1% of hospitalizations. SSc was considered the primary reason for hospitalization in 9% of patients, 24.2% of whom had a secondary diagnosis of lung involvement or pulmonary fibrosis. Other common principal diagnoses included respiratory infections, CHF, acute MI and cancer. PAH and skin ulcers/gangrene were the 14th and 20th most common principal diagnoses, respectively. Other SSc-related conditions were less frequently identified as the primary reason for admission, with oesophageal disorder, pulmonary fibrosis, renal failure and carditis each accounting for <1.4% of hospitalizations. Systemic infection and skin/subcutaneous infection were the 8th and 16th most common principal diagnoses, respectively.
|
Diagnoses associated with in-hospital death
Diseases of the circulatory and respiratory systems were identified as the principal diagnosis in 25.8% and 23.9% of SSc patients who died in the hospital, respectively (Table 3). Diseases of the musculoskeletal (14.2%) and gastrointestinal (7.8%) systems were also common principal diagnoses in this subset, while the genitourinary system accounted for only 4.4% of in-hospital deaths. SSc was the most common principal diagnosis, followed by respiratory failure, respiratory infection and cancer. Renal failure, PAH and pulmonary fibrosis were the 9th, 10th and 14th most common principal diagnoses in this subset.
|
In-hospital mortality and length of stay
We estimated 1374 deaths in 2002 and 1473 deaths in 2003. This translates into an overall in-hospital mortality rate of 6.3% (95%CI 5.8–6.7) (Table 4). In-hospital mortality increased with each age group. Using univariate logistic regression, we found that for every 10-yr increase in age, in-hospital mortality increased by 15% (data not shown). Women had a 25% lower in-hospital mortality rate than men (P = 0.005). Blacks had the highest percentage of in-hospital deaths, while those from the other race categories had the lowest percentage; however, these differences were not significant.
|
The overall mean LOS was 6.6 days (median 4.0 days). Mean LOS increased significantly with age, with the greatest difference between the youngest and oldest age groups (P < 0.0001). Blacks had significantly longer mean LOS than whites (P < 0.0001). When comparing non-whites to whites, the LOS was 11.1% (95% CI 6.5–15.8) longer in the former (P < 0.0001) (data not shown).
Factors associated with in-hospital mortality and LOS
Results from the multivariate regression models for in-hospital mortality and LOS are shown in Tables 5 and 6, respectively. The presence of pulmonary fibrosis/interstitial lung disease (ILD) was associated with a 2.6 times increased odds of in-hospital death (P < 0.0001) and increased LOS by 7.3% (P = 0.03). In contrast, a discharge diagnosis of oesophageal dysfunction decreased the odds of in-hospital death by 43% (P = 0.0003) and decreased the LOS by 18.2% (P < 0.0001). CHF increased the odds of in-hospital death by 1.4-fold (P = 0.04), but did not significantly affect LOS.
|
|
Other significant predictors of in-hospital death (P < 0.05) included emergency admission, age, male gender and the number of diagnoses and procedures performed. In-hospital death significantly decreased the LOS by 13% (P = 0.003). Other significant predictors of LOS included transfer from another hospital, emergency admission, non-white race, public insurance and the number of diagnoses and procedures performed.
Regression models using the subset limited to the states that provided race data (24 states in 2002, 26 in 2003) showed similar results to the entire group (data not shown). Sensitivity analyses showed that the results remained robust after accounting for multiple hospitalizations of the same individuals.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Our results provide further evidence that hospitalization patterns of SSc patients in the USA differ by gender, age and race. Women had higher hospitalization rates than men, consistent with gender differences in prevalence of SSc. Although the peak age of onset of SSc is in the fourth and fifth decades [18], hospitalization rates continued to increase with age beyond 65 yrs, likely related to comorbid illnesses or progression of the underlying disease. As in the Nietert study [10], we showed that blacks had higher hospitalization rates than whites in those who were younger than 65 yrs old. This agrees with studies showing an earlier, more severe presentation of disease in blacks compared with whites [1, 18].
Unlike the Nietert study, we evaluated common indications for hospitalization and diagnoses associated with in-hospital mortality in SSc patients. Several mortality studies in SSc have been performed in the USA and other countries over the past several decades [5–9]. In these studies, SSc-related internal organ involvement (including renal failure, PAH, pulmonary fibrosis and cachexia/oesophageal dysfunction) as well as ischaemic heart disease, CHF, infections and malignancies were the most common causes of death. In our study, we found that SSc was the most common principal diagnosis in all SSc hospitalizations, and the most common principal diagnosis in those who died in the hospital. Patients with a principal diagnosis of SSc likely had SSc-related internal organ involvement, as the most common secondary diagnoses in this subgroup included lung involvement in SSc or pulmonary fibrosis followed by CHF, acute renal failure and PAH.
Although SSc was the most common indication for hospitalization, PAH and gangrene were the only SSc-related conditions that were identified in the 20 most common reasons for admission. It is possible that improvements in the management of specific SSc-related conditions, in contrast to the overall disease process, account for our observations. For example, renal failure was identified as the principal diagnosis in only 1.1% of admissions and in 3.7% of patients who died in the hospital. This suggests that early recognition and treatment of renal crisis with ACE inhibitors has not only improved survival in patients with SSc, but has also improved hospitalization outcomes.
Cardiopulmonary involvement was identified in half of those SSc patients who died in the hospital and respiratory failure was the second most common diagnosis in this subgroup. Pulmonary fibrosis increased the odds of in-hospital mortality by 2.6-fold and the LOS by 7.3%. These findings highlight the persistent negative impact of ILD on hospitalization outcomes in patients with SSc, despite recent trends toward aggressive immunosuppression for early lung disease. Our results are consistent with a recent study of the Pittsburgh Scleroderma Databank showing that pulmonary fibrosis was the most common cause of SSc-related deaths from 1997 to 2001 [19]. In our study, infections were also common causes of hospitalization and in-hospital death, possibly related to the increased use of immunosuppressive agents in the treatment of SSc. Oesophageal dysfunction was not identified as a common principal diagnosis for SSc patients admitted to the hospital or for those who died in the hospital. It was also associated with lower in-hospital mortality and shorter LOS in the regression models. However, aspiration pneumonitis, a complication of oesophageal dysfunction, was the eighth most common principal diagnosis in the SSc patients who died in the hospital. This suggests that the effects of oesophageal dysfunction on in-hospital mortality may be higher than estimated by the regression models.
The overall in-hospital mortality rate for SSc patients was 6.3% with a mean LOS of 6.6 days in 2002–2003. In 1995, Nietert et al. [10] estimated the overall in-hospital mortality rate as 7.1% and mean LOS as 7.5 days. The trends towards improvement that we observed in our study may reflect the advances in the care of patients with SSc over the past decade. Alternatively, changes in insurance regulations may have resulted in shorter hospitalizations and more deaths outside of the hospital, for instance, at home or in the hospice setting. Lastly, it is possible that the additional states included in the 2002–2003 NIS data sets allowed more accurate national estimates than in the previous study.
We found that men had significantly higher in-hospital death rates compared with women in both univariate and multivariate analyses. Previous mortality studies have shown conflicting results with regards to gender differences. Two studies have shown that men with SSc have higher mortality rates than women, but with similar standardized mortality ratios (SMR) [6, 7]. These findings could be explained by the fact that men are more likely to die of causes other than SSc. However, Ferri et al. [20] showed that men with SSc had a much lower 10-yr survival rate than women with SSc, despite similar survival rates by gender in the general population. In contrast, Krishnan and Furst [3] showed that women with SSc had a 3.5-fold higher age-adjusted mortality rate compared with men.
We did not find a significant effect of race on in-hospital mortality. Blacks did have a higher in-hospital death rate than whites, but the other race category had the lowest in-hospital mortality rate. Our data conflicts with a study using the South Carolina Office of Research and Statistics database [21]. This study showed that blacks and other races had 1.7-fold and 2.1-fold higher in-hospital death rates than whites, respectively [21]. Since this study only evaluated hospitalizations in South Carolina, the results are not generalizable to the entire USA. The ability to detect racial differences in in-hospital mortality in our study and the Nietert NIS study may have been limited by the missing race data. Alternatively, our findings could be explained by the fact that inpatient care of SSc patients, as opposed to ambulatory care, does not differ by race.
Non-white race was a significant predictor of LOS. This finding may be related to a combination of more severe disease and fewer outpatient resources. The impact of public insurance on increasing LOS supports the hypothesis that lower socioeconomic status is associated with prolonged hospitalizations for patients with SSc.
Our study had several limitations related to the database we used. First, we were unable to validate discharge diagnoses that were provided by treating physicians, and therefore misclassification of SSc may have occurred. Clinical verification of the ICD9-CM code 710.1 for a diagnosis of SSc has recently been reported to be as low as 45% [22]. We excluded subjects with a concomitant diagnosis of SLE or RA to improve the specificity of the diagnosis of SSc, but it is possible that patients with localized scleroderma, overlap syndromes or other fibrosing conditions were included in our analyses. The second limitation is that all hospitalizations were considered independent events as no unique patient identifiers are provided in the NIS database. Our hospitalization rates are likely inflated, while our in-hospital death rates likely underestimate the true rates, as we were unable to account for patients with readmissions. Third, the database provides no information on important confounders such as medications, autoantibody status and smoking history. Lastly, since limited and diffuse SSc have the same ICD9-CM code, we were not able to stratify our analyses according to SSc subtype.
Despite these limitations, we estimated hospitalization rates, in-hospital mortality and LOS from the largest number of SSc hospitalizations studied to date. Our study is the first to investigate reasons for hospitalization and in-hospital death in patients with SSc. Our results suggest that improvements in the care of SSc patients in the past several years have had a positive impact on hospitalization patterns; however, pulmonary fibrosis continues to be a major predictor of poor outcomes.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The authors thank Gary Friedman, MD, MS and Thomas A. Medsger Jr, MD for their critical review of the manuscript. The authors also thank Bharathi Lingala, PhD for her assistance in the statistical analyses.
Disclosure statement: L.C. has served as a paid consultant and on the speakers’ bureau for Encysive Pharmaceuticals, Inc. All other authors have declared no conflicts of interest.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
- Laing TJ, Gillespie BW, Toth MB, et al. Racial differences in scleroderma among women in Michigan. Arthritis Rheum (1997) 40:734–42.[Web of Science][Medline]
- Mayes MD, Lacey JV, Beebe-Dimmer J, et al. Prevalence, incidence, survival, and disease characteristics of systemic sclerosis in a large US population. Arthritis Rheum (2003) 48:2246–55.[CrossRef][Web of Science][Medline]
- Krishnan E, Furst DE. Systemic sclerosis mortality in the United States: 1979–1998. Eur J Epidemiol (2005) 20:855–61.[CrossRef][Web of Science][Medline]
- Ioannidis JPA, Vlachoyiannopoulos PG, Haidich A-B, et al. Mortality in systemic sclerosis: an international meta-analysis of individual patient data. Am J Med (2005) 118:2–10.[CrossRef][Web of Science][Medline]
- Abu-Shakra M, Lee P. Mortality in systemic sclerosis: a comparison with the general population. J Rheumatol (1995) 22:2100–2.[Web of Science][Medline]
- Bryan C, Howard Y, Brennan P, Black C, Silman A. Survival following the onset of scleroderma: results from a retrospective inception cohort study of the UK patient population. Br J Rheumatol (1996) 35:1122–6.
[Abstract/Free Full Text] - Jacobsen S, Halberg P, Ullman S. Mortality and causes of death of 344 Danish patients with systemic sclerosis (scleroderma). Br J Rheumatol (1998) 37:750–5.
[Abstract/Free Full Text] - Hesselstrand R, Scheja A, Akesson A. Mortality and causes of death in a Swedish series of systemic sclerosis patients. Ann Rheum Dis (1998) 57:682–6.
[Abstract/Free Full Text] - Simeón CP, Armadans L, Fonollosa V, et al. Mortality and prognostic factors in Spanish patients with systemic sclerosis. Rheumatology (2003) 42:71–5.
[Abstract/Free Full Text] - Nietert PJ, Silverstein MD, Silver RM. Hospital admissions, length of stay, charges, and in-hospital death among patients with systemic sclerosis. J Rheumatol (2001) 28:2031–7.
[Abstract/Free Full Text] - Steen VD, Costantino JP, Shapiro AP, Medsger TA. Outcome of renal crisis in systemic sclerosis: relation to availability of angiotensin converting enzyme (ACE) inhibitors. Ann Intern Med (1990) 113:352–7.
[Abstract/Free Full Text] - DeMarco PJ, Weisman MH, Seibold JR, et al. Predictors and outcomes of scleroderma renal crisis: the high-dose versus low-dose D-penicillamine in early diffuse systemic sclerosis trial. Arthritis Rheum (2002) 46:2983–9.[CrossRef][Web of Science][Medline]
- Sahhar J, Littlejohn G, Conron M. Fibrosing alveolitis in systemic sclerosis: the need for early screening and treatment. Intern Med J (2004) 34:626–38.[CrossRef][Web of Science][Medline]
- Tashkin DP, Elashoff R, Clements PJ, et al. Cyclophosphamide versus placebo in scleroderma lung disease. N Engl J Med (2006) 354:2655–66.
[Abstract/Free Full Text] - Hoyles RK, Ellis RW, Wellsbury J, et al. A multicenter, prospective, randomized, double-blind, placebo-controlled trial of corticosteroids and intravenous cyclophosphamide followed by oral azathioprine for the treatment of pulmonary fibrosis in scleroderma. Arthritis Rheum (2006) 54:3962–70.[CrossRef][Web of Science][Medline]
- Steen V. Targeted therapy for systemic sclerosis. Autoimmun Rev (2006) 5:122–4.[CrossRef][Web of Science][Medline]
- Williams MH, Das C, Handler CE, et al. Systemic sclerosis associated pulmonary hypertension: improved survival in the current era. Heart (2006) 92:926–32.
[Abstract/Free Full Text] - Mayes MD. Scleroderma epidemiology. Rheum Dis Clin North Am (2003) 29:239–54.[CrossRef][Web of Science][Medline]
- Steen VD, Medsger TA. Changes in causes of death in systemic sclerosis, 1972–2002. Ann Rheum Dis. (2007) 66:940–4.
[Abstract/Free Full Text] - Ferri C, Valentini G, Cozzi F, et al. Systemic sclerosis: demographic, clinical, and serologic features and survival in 1,012 Italian patients. Medicine (2002) 81:139–53.[CrossRef][Medline]
- Nietert PJ, Silver RM, Mitchell HC, Shaftman SR, Tilley BC. Demographic and clinical factors associated with in-hospital death among patients with systemic sclerosis. J Rheumatol (2005) 32:1888–92.
[Abstract/Free Full Text] - Khurma V, Furst DE, Krishnan E, Khanna D. Verification of ICD-CM-9 coding for the diagnosis of systemic sclerosis [abstract]. Arthritis Rheum (2006) 54(Suppl 9):S344.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||