Digestive Diseases and Sciences

, Volume 55, Issue 6, pp 1689–1695

Incidence and Risk of Intestinal and Extra-intestinal Complications in Medicaid Patients with Inflammatory Bowel Disease: A 5-Year Population-Based Study

Authors

  • Gaurav Arora
    • Division of Gastroenterology, Hepatology and NutritionUniversity of Texas Medical School
    • Division of Gastroenterology, Hepatology and NutritionMD Anderson Cancer Center
  • Gurkirpal Singh
    • Division of Gastroenterology and HepatologyStanford University School of Medicine
    • Institute of Clinical Outcomes Research and Education
  • Shweta Vadhavkar
    • Institute of Clinical Outcomes Research and Education
  • Shamita B. Shah
    • Division of Gastroenterology and HepatologyStanford University School of Medicine
  • Ajitha Mannalithara
    • Institute of Clinical Outcomes Research and Education
  • Alka Mithal
    • Institute of Clinical Outcomes Research and Education
    • Division of Gastroenterology and HepatologyStanford University School of Medicine
Original Article

DOI: 10.1007/s10620-010-1236-z

Cite this article as:
Arora, G., Singh, G., Vadhavkar, S. et al. Dig Dis Sci (2010) 55: 1689. doi:10.1007/s10620-010-1236-z
  • 128 Views

Abstract

Background and Aim

Intestinal and extra-intestinal complications are associated with inflammatory bowel disease (IBD) but their exact incidence is not well known. In order to improve our understanding of their incidence and impact, we assessed the complications associated with ulcerative colitis (UC) and Crohn’s disease (CD) in a population-based study in Medicaid patients.

Methods

We utilized a retrospective cohort design and identified cases of UC and CD using Medi-Cal, the Medicaid program for the State of California. The disease cohort was age- and gender-matched to four controls each and the intestinal and extra-intestinal complications of CD and UC (analyzed separately) were studied over a period of 5 years following the initial diagnosis.

Results

For UC, the total number of intestinal complications, per 100 cases, was 92 observed compared to 21 expected; the total number of extra-intestinal complications was 42 observed compared to 30 expected. For CD, the number of intestinal complications was 81 observed compared to 20 expected and for extra-intestinal complications, 37 observed compared to 26 expected (all p < 0.001). For both UC and CD, bleeding was the most frequently seen intestinal complication, while the most common extra-intestinal complication was osteoporosis.

Conclusions

IBD is associated with several intestinal and extra-intestinal complications of variable incidence and risk. Success of therapeutic regimens should be measured by decreases in incidence, risks, and costs of these complications, in addition to the usual impact on disease activity.

Keywords

Inflammatory bowel diseaseCrohn’s diseaseUlcerative colitisComplicationsCo-morbidities

Abbreviations

CD

Crohn’s disease

EN

Erythema nodosum

IBD

Inflammatory bowel disease

OR

Odds ratio

UC

Ulcerative colitis

Introduction

Inflammatory bowel disease (IBD) generally encompasses two separate clinical entities: ulcerative colitis (UC) and Crohn’s disease (CD). About 10% of patients will not fall into either diagnostic category and are termed “indeterminate colitis”; such patients may subsequently develop features of UC or CD [1]. IBD, a chronic disease with significant associated morbidity, accounts for 700,000 physician visits and 100,000 hospitalizations annually in the United States, causing disability in 119,000 patients [2]. In addition to significant indirect costs in terms of loss of productivity and disability, CD and UC were responsible for $706.8 million and $387.8 million, respectively, in direct costs in the year 1998 [3].

A major aspect of care in IBD is the management of the associated intestinal complications such as fistulas, abscesses, obstruction, malabsorption, hemorrhage, etc., as well as a spectrum of complications outside the gastrointestinal tract involving organs such as the skin, eyes, joints, and biliary tract. The prevalence of extra-intestinal complications is estimated to be between 21 and 36% [48]. Furthermore, a significant percentage of patients with IBD will require surgery over their lifetime (75% of CD and 25% of UC patients) [9]. However, most studies evaluating complications of IBD have been limited by the small number of patients studied and thus may not represent true estimates or relative risk. The importance of this information cannot be overemphasized if we are to truly understand IBD and try to reduce the associated clinical, social, and economic burdens. We have, therefore, conducted this large population-based study to determine the incidence of complications in adults with IBD in a Medicaid population, in the first 5 years after their first recorded diagnosis.

Materials and Methods

Patients and Data

Our study is based on longitudinal data derived from Medi-Cal, the Medicaid program for the State of California [10]. We did not study patients from a managed-care population or those with private insurance. The Medi-Cal program serves over 6.5 million beneficiaries of which approximately 3.1 million (48%) are in the fee-for-service system while the rest are enrolled in managed-care plans [11]. Medi-Cal provides comprehensive universal health care coverage to ethnically diverse, low-income, and disabled individuals who lack health insurance. Since there is no requirement for premiums or co-payments, participation in the program is practically 100%. The Medi-Cal research database contains computerized records of eligibility status of all beneficiaries, and detailed information on all medical services provided, including outpatient visits, hospital admissions, medical procedures, emergency room visits, laboratory and radiological testing, and outpatient drug prescriptions, including many over-the-counter drugs, such as aspirin [11]. As data on every patient are recorded, regardless of the visit setting, we believe it is highly unlikely that any significant complication associated with IBD would be missed. A recent audit of Medi-Cal claims found that 96.4% were medically necessary, billed appropriately, and were in concordance with the data in the claims files [11]. Due to the presence of a fiscal intermediary, fee-for-service data have been shown to be complete and valid, and were used for the present study [11].

Study Design, Cohort, and Setting

We conducted a retrospective matched-cohort study measuring the incidence of complications in the first 5 years after the first recorded diagnosis of IBD, and the risk (measured as odds ratio) of these as compared to Medicaid population controls. The case definition of IBD was at least 2 claims corresponding to the International Classification of Diseases, 9th Revision (ICD-9) code of 555.xx for CD or 556.xx for UC, at least 1 week apart; the latter to ensure accuracy of diagnosis. UC or CD patients who were enrolled in the Medi-Cal program during the period of January 1st, 1995 to June 30th, 2005 were identified. To ensure an incident cohort, we restricted entry in the cohort to only those patients who had continuous eligibility for 6 months before and 5 years after their first claim for UC or CD, and did not have a claim for UC or CD in the 6 months prior to their index date. Index date was defined as the date of the first recorded diagnosis for the subjects in the disease cohort and the index date of the respective case for individuals in the control cohort. Patients with CD and UC were studied separately. The control cohort was formed by randomly selecting individuals under observation in the database on the index date and matched in a 1:4 ratio (disease: control) for age (year of birth) and gender, separately for UC and CD. As with the disease cohort, the controls were also required to have continuous eligibility for 6 months before and 5 years after their index date. Furthermore, control subjects could not have a diagnosis of CD or UC at any time. The study outcome was identification of any complication listed in Table 1.
Table 1

ICD-9 codes used for identification of complications of ulcerative colitis and Crohn’s disease

Condition

ICD-9

Amyloidosis

277.3x

Anal/rectal hemorrhage

569.3x

Anal/rectal fistula

565.1x

Anal/rectal prolapse

569.1x

Anal/rectal stenosis

569.2x

Anal/rectal ulcer

569.41

Ankylosing spondylitis

720.xx

Cholangitis

576.1x

Episcleritis

379.xx

Erythema nodosum

695.2x

Gallstone disease

574.xx

Gastrointestinal hemorrhage

578.xx

Intestinal abscesses

569.5x

Intestinal fistula

569.81

Intestinal obstruction

560.8x, 560.9x

Malabsorption

579.xx

Malignancy of large intestine

153.xx

Malignancy of rectum

154.xx

Megacolon

564.7x

Osteoporosis

733.0x

Paralytic ileus

560.1x

Perianal abscesses

566.xx

Peripheral arthritis

713.1x

Pyoderma gangrenosum

686.01

Statistical Analysis

The cumulative incidence of complications of IBD is expressed as percentages. Odds ratios were calculated as an approximation of the relative risk of the diseased subjects having the complication as compared to the controls and 95% confidence intervals were calculated for the odds ratios to estimate the true population parameters. All p values are two-tailed. All analyses were performed using SAS 9.1.3 (SAS Institute, North Carolina, USA).

Results

Ulcerative Colitis

A description of baseline demographic variables for UC is given in Table 2. The disease cohort included 1,731 patients and the control cohort contained 6,746 individuals. The mean age for both of these groups was about 48 years, and 66% were women. Patients less than 18 years of age accounted for 7.8% of the disease cohort and a similar percentage in the control cohort. In the <18 years group, approximately 52% were females. Overall, Caucasians accounted for the majority of patients.
Table 2

Baseline demographics of the disease and control cohorts of ulcerative colitis

 

Disease (n = 1,731)

Control (n = 6,746)

Age

Mean (SD)

48.3 (19)

48.5 (19)

Age < 18 (% of total)

135 (7.8) [52% female]

512 (7.6) [52% female]

Gender (%)

Women

1141 (66)

4439 (66)

Men

590 (34)

2307 (34)

Race/ethnicity (%)

Caucasian

871 (50)

2757 (41)

Black

156 (9)

883 (13)

Hispanic

327 (19)

1298 (19)

Asian

150 (9)

904 (13)

Other

57 (3)

257 (4)

Missing

170 (10)

647 (10)

Overall, the total number of intestinal complications, per 100 cases, was 92 observed compared to 21 expected (p < 0.001). The cumulative incidence and risk of complications of UC is shown in Table 3, sorted in descending order of incidence in the disease cohort. Intestinal complications that occurred with the greatest incidence in the first 5 years after UC diagnosis include hemorrhage followed by intestinal obstruction and paralytic ileus. The highest risk as compared to general population was seen for ano-rectal ulcers (20-fold) and stenosis (10-fold). Risk of megacolon was six times higher. The risk of developing intestinal and perianal abscesses was 15- and 8-fold higher, respectively. A clearly higher incidence and risk of colon (2.72%, OR 2.87) and rectal (1.21%, OR 2.58) malignancy were noted. The total number of extra-intestinal complications was 42 observed compared to 30 expected (p < 0.001); the highest incidence was seen for osteoporosis (14.56%), gallstone disease (9.94%), and episcleritis (9.82%), whereas the highest risk was noted for peripheral arthritis (10-fold). When stratified by gender and age (<65 vs. ≥65 years), females less than 65 were significantly more likely to have osteoporosis compared to controls (OR 1.91 [95% CI 1.55–2.36]); no significant increased risk was found in the other strata: females ≥65 years, OR 1.02 (0.74–1.41), males <65 years, OR 1.30 (0.84–2.00), and males ≥65 years, OR 1.96 (0.98–3.93).
Table 3

Cumulative incidence (at 5 years) and risk of complications of ulcerative colitis

 

Disease count (%)

Control count (%)

Odds ratio (95% CI)

p value

Intestinal complications

Anal/rectal hemorrhage

491 (28.37)

320 (4.74)

7.95 (6.82–9.27)

<0.001

Gastrointestinal hemorrhage

490 (28.31)

550 (8.15)

4.45 (3.88–5.10)

<0.001

Intestinal obstruction

198 (11.44)

215 (3.19)

3.92 (3.21–4.80)

<0.001

Paralytic ileus

106 (6.12)

120 (1.78)

3.60 (2.76–4.70)

<0.001

Malabsorption

51 (2.95)

45 (0.67)

4.52 (3.02–6.77)

<0.001

Perianal abscesses

50 (2.89)

23 (0.34)

8.69 (5.29–14.29)

<0.001

Malignancy of large intestine

47 (2.72)

65 (0.96)

2.87 (1.96–4.19)

<0.001

Anal/rectal ulcer

41 (2.37)

8 (0.12)

20.43 (9.56–43.67)

<0.001

Anal/rectal fistula

23 (1.33)

16 (0.24)

5.66 (2.99–10.74)

<0.001

Malignancy of rectum

21 (1.21)

32 (0.47)

2.58 (1.48–4.48)

0.002

Anal/rectal stenosis

18 (1.04)

7 (0.10)

10.12 (4.22–24.26)

<0.001

Anal/rectal prolapse

17 (0.98)

13 (0.19)

5.14 (2.49–10.60)

<0.001

Megacolon

17 (0.98)

11 (0.16)

6.07 (2.84–12.99)

<0.001

Intestinal fistula

15 (0.87)

14 (0.21)

4.20 (2.03–8.72)

<0.001

Intestinal abscesses

8 (0.46)

2 (0.03)

15.66 (3.32–73.79)

<0.001

Extra-intestinal complications

    

Osteoporosis

252 (14.56)

679 (10.07)

1.52 (1.30–1.78)

<0.001

Gallstone disease

172 (9.94)

586 (8.69)

1.16 (0.97–1.39)

0.108

Episcleritis

170 (9.82)

553 (8.20)

1.22 (1.02–1.46)

0.034

Ankylosing spondylitis

55 (3.18)

58 (0.86)

3.78 (2.61–5.49)

<0.001

Uveitis

35 (2.02)

96 (1.42)

1.43 (0.97–2.11)

0.080

Cholangitis

21 (1.21)

18 (0.27)

4.59 (2.44–8.63)

<0.001

Peripheral arthritis

8 (0.46)

3 (0.04)

10.44 (2.77–39.38)

<0.001

Erythema nodosum

4 (0.23)

7 (0.10)

2.23 (0.65–7.63)

0.251

Pyoderma gangrenosum

2 (0.12)

0 (0)

0.042

Amyloidosis

2 (0.12)

6 (0.09)

1.30 (0.26–6.44)

0.669

Crohn’s Disease

A description of demographic variables for CD is shown in Table 4. The disease cohort included 1,411 patients while the corresponding number in the control cohort was 5,512. The mean age for both was approximately 43 years (significantly less than that for UC) and 67% were women. Patients less than 18 years of age accounted for 16% in both cohorts, significantly more than that seen for UC; 46% of these were females. As with UC, Caucasians accounted for the majority of patients.
Table 4

Baseline demographics of the disease and control cohorts of Crohn’s disease

 

Disease (n = 1,411)

Control (n = 5,512)

Age

Mean (SD)

42.9 (22)

43.1 (22)

Age <18 (% of total)

234 (16.6)

[46% female]

908 (16.5)

[46% female]

Gender (%)

Women

938 (67)

3,658 (66)

Men

473 (33)

1,854 (34)

Race/ethnicity (%)

Caucasian

738 (52)

2,073 (38)

Black

145 (10)

722 (13)

Hispanic

268 (19)

1,284 (23)

Asian

94 (7)

707 (13)

Other

40 (3)

220 (4)

Missing

126 (9)

506 (9)

Overall, the number of intestinal complications was 81 observed compared to 20 expected (p < 0.001) and for extra-intestinal complications, 37 observed compared to 26 expected (all p < 0.001). The cumulative incidence and risk of complications is shown in Table 5. As with UC, the highest incidence was seen for hemorrhage, followed by intestinal obstruction. The risk of ano-rectal fistulas and ulcers (22- and 16-fold, respectively), intestinal fistulas (15-fold), as well as intestinal abscesses (14-fold) was significantly increased in CD patients, compared to controls. While the cumulative 5-year incidence of colonic and rectal malignancy was 1.63 and 1%, the risk compared to controls was 2.5- and 2.6-fold, respectively. Among extra-intestinal complications, osteoporosis was the most common, followed closely by gallstone disease and episcleritis. Each of these extra-intestinal complications occurred with significantly increased risk compared to controls. For osteoporosis, when stratified by gender and age (<65 vs. ≥65 years), females less than 65 were significantly more likely to have osteoporosis compared to controls (OR 1.69 [1.33–2.15]); no significant increased risk was found in the other strata: females ≥65 years, OR 1.05 (0.69–1.59), males <65 years, OR 0.98 (0.56–1.71) and males ≥65 years, OR 1.51 (0.57–4.03). Interestingly, amyloidosis risk was increased fourfold. Risk of peripheral arthritis was significantly increased, though odds ratio could not be calculated secondary to lack of this diagnosis in controls.
Table 5

Cumulative incidence (at 5 years) and risk of complications of Crohn’s disease

Intestinal complications

Disease count (%)

Control count (%)

Odds ratio (95% CI)

p value

Gastrointestinal hemorrhage

300 (21.26)

421 (7.64)

3.27 (2.78–3.84)

<0.001

Intestinal obstruction

255 (18.07)

182 (3.30)

6.46 (5.29–7.89)

<0.001

Anal/rectal hemorrhage

231 (16.37)

258 (4.68)

3.99 (3.30–4.81)

<0.001

Paralytic ileus

100 (7.09)

91 (1.65)

4.54 (3.40–6.07)

<0.001

Anal/rectal fistula

55 (3.90)

10 (0.18)

22.32 (11.35–43.89)

<0.001

Malabsorption

54 (3.83)

39 (0.71)

5.58 (3.68–8.47)

<0.001

Perianal abscesses

40 (2.83)

21 (0.38)

7.63 (4.48–12.98)

<0.001

Intestinal fistula

35 (2.48)

9 (0.16)

15.5 (7.46–32.43)

<0.001

Malignancy of large intestine

23 (1.63)

36 (0.65)

2.52 (1.49–4.27)

<0.001

Anal/rectal ulcer

16 (1.13)

4 (0.07)

15.79 (5.27–47.32)

<0.001

Malignancy of rectum

14 (0.99)

21 (0.38)

2.62 (1.33–5.17)

0.009

Anal/rectal prolapse

9 (0.64)

7 (0.13)

5.05 (1.88–13.58)

0.002

Anal/rectal stenosis

8 (0.57)

6 (0.11)

5.23 (1.81–15.11)

0.003

Intestinal abscesses

7 (0.50)

2 (0.04)

13.74 (2.85–66.19)

<0.001

Extra-intestinal complications

    

Osteoporosis

165 (11.69)

477 (8.65)

1.40 (1.16–1.69)

<0.001

Gallstone disease

138 (9.78)

394 (7.15)

1.41 (1.15–1.73)

0.001

Episcleritis

133 (9.43)

395 (7.17)

1.35 (1.10–1.66)

0.006

Ankylosing spondylitis

28 (1.98)

54 (0.98)

2.05 (1.29–3.24)

0.003

Uveitis

23 (1.63)

68 (1.23)

1.33 (0.82–2.14)

0.240

Peripheral arthritis

13 (0.92)

0 (0)

<0.001

Amyloidosis

8 (0.57)

8 (0.15)

3.92 (1.47–10.47)

0.008

Cholangitis

5 (0.35)

11 (0.20)

1.78 (0.62–5.13)

0.346

Erythema nodosum

3 (0.21)

8 (0.15)

1.47 (0.39–5.53)

0.477

Pyoderma gangrenosum

2 (0.14)

1 (0.02)

7.82 (0.71–86.33)

0.108

Discussion

IBD remains an important clinical challenge for the practicing gastroenterologist because of the difficulties encountered in controlling disease activity and maintaining remission. While intestinal inflammation is the main focus of management, the complications resulting from IBD deserve equal attention since they can severely affect health-related quality-of-life and impact morbidity and mortality. To the best of our knowledge, this population-based study is the largest to date, reporting a cumulative incidence and risk of complications of IBD in a Medicaid population at 5 years after initial diagnosis.

Mild gastrointestinal bleeding is a common manifestation of IBD but severe hemorrhage is rare, with an incidence of 2–3% in various prior series [1214]. Despite such reported low incidence, hemorrhage accounts for nearly 10% of all urgent colectomies in UC [14], and thus may represent an important but under-recognized complication. Our data indicates that gastrointestinal hemorrhage in IBD is more common than previously thought. Unfortunately, the database does not permit us to classify its severity. The higher incidence of hemorrhage in UC versus CD may be explained by the fact that the former frequently leads to larger areas of ulcerated mucosa compared to more localized bleeding sources in the latter. Besides, inflammation may manifest as stricturing or fistulizing disease in CD, rather than mucosal ulceration or inflammation, and the disease may just be limited to the small bowel as well.

Osteoporosis is a prominent extra-intestinal feature in many IBD patients. Disruption of bone metabolism and bone mineral content may result from disease activity, genetic factors, and adverse effects of therapy, chiefly glucocorticoid use [15]. We found the incidence of osteoporosis at 15 and 12% and the increase in odds of 50 and 40%, respectively, in UC and CD. This incidence is within the range of 2–16% shown previously in larger studies focusing on osteoporosis in IBD [1618]. Importantly, our results show that younger women with IBD are significantly more likely to have osteoporosis than the corresponding age-matched controls. If confirmed in other studies, this will imply that these women should be screened sooner for osteoporosis than their non-IBD-afflicted counterparts.

In smaller studies, the prevalence of cholelithiasis in CD was 11–24%, and in UC was 4.6–7.5% [1921]. The higher prevalence in CD is likely related to ileal resection, bypass, or active ileitis; all of which may disrupt normal enterohepatic circulation of bile salts and lead to cholesterol crystallization in the gallbladder and biliary tree. Our study highlights the 5-year incidence of gallstone disease to be about 10% in both CD and UC, with a statistically increased risk only in CD (OR 1.41, p < 0.001). Based on these findings, it is clear that gallstone disease as a complication of Crohn’s disease occurs soon after initial diagnosis.

Prior larger studies have focused on the prevalence of extra-intestinal complications in IBD and report overall prevalence rates of 5–20% for peripheral arthritis [7, 22, 23] and 4–15% for erythema nodosum (4–10% in UC, 15% in CD) [5, 24]. However, the incidence of these extra-intestinal complications in the initial years after diagnosis of IBD has not been well defined. We found the cumulative 5-year incidence of peripheral arthritis to be between 0.5 and 1%. Thus, only a small proportion of patients develop peripheral arthritis within a few years of diagnosis. In our study, the incidence of erythema nodosum was very low (0.2%) in the initial few years after diagnosis and was similar to pyoderma gangrenosum.

A recent meta-analysis (five cohort studies) [25] of CD patients reported that the pooled standardized incidence ratio for colon cancer was 2.5 (95% CI 1.7–3.5) and for rectal cancer was 1.4 (95% CI 0.8–2.6), the latter not being statistically significant. In our study, the increase in risk was similar for colon cancer in CD. Additionally, we found a statistically significant increased risk for rectal cancer. In a separate meta-analysis, the prevalence of colorectal carcinoma in UC was estimated at 3.7% (95% CI 3.2–4.2%) [26]. We report that at 5 years after diagnosis, the incidence of colon cancer was 2.7% and that of rectal cancer was 1.2%; both were associated with significantly increased odds. Future studies to evaluate the optimal interval for screening colonoscopy in this population should take into account this relatively high incidence even in the early years after diagnosis found in our study.

Overall, our study reports a significant burden of the complications related to IBD in a Medicaid population in the initial 5 years after diagnosis. This underscores the importance of being vigilant for these complications as timely management may be the key for many of these. We are cognizant of the possibility that some complications of IBD may not manifest until later in the course of the disease and consequently may dilute our estimates by not being included in the 5-year period that we chose for studying each subject. However, the focus of our study was to detect the earlier-onset complications and not to capture the lifetime incidence or risk of all complications. It is also possible that some subjects may have manifested the complications related to IBD even before the actual diagnosis of IBD and thus our estimates may not truly reflect the earlier complications of IBD. Yet another possibility is that physicians, particularly gastroenterologists, are more likely to bill for the main diagnosis such as Crohn’s disease and less likely to bill for complications, such as peripheral arthritis. As our detection of the complications is based only on the claims data, this may lead to under-detection of the complications.

There are other limitations to our study. As our sample is a Medicaid population, it represents data from people who are typically sicker and less affluent; therefore, we may not be able to generalize our findings to the non-Medicaid population. Although it has recently been shown that IBD may be more common in people with private insurance as compared to those without [27], we are not aware of any data suggesting that complications of IBD may differ according to socioeconomic status. A possibility of selection bias by indication exists in our study as patients with severe manifestations of IBD may become disabled, leading to enrollment in Medicaid. Detection bias may be potentially present as well, as IBD patients are more likely to be screened for asymptomatic conditions, such as osteoporosis. This may increase the strength of the association between IBD and some complications. As with the use of any administrative database, the diagnosis of IBD and identification of the various complications are dependent on the accuracy of the coding procedures. This is unlikely to be a significant issue in our study as an audit of Medi-Cal claims found that 96.4% were medically necessary, billed appropriately, and were in concordance with the data in the claims files [11]. The main strength of our study is the large sample size involved at the population-level, thereby giving us good estimates for the true incidence of the complications of IBD. Additionally, we have also shown the odds of having the various complications in IBD as compared to a general population cohort.

In summary, this large population-based, matched cohort study determines the intestinal and extra-intestinal complications in patients with IBD and provides a framework for the further understanding of the disease burden to the US population suffering from the disease and undoubtedly causing a substantial economic impact on the Medi-Cal system and to society in general. Since IBD is associated with several complications of variable prevalence and impact and in turn affecting disease activity indices, success of therapeutic regimens should be measured by decreases in incidence, risks, and costs of these complications, in addition to the usual impact on disease activity.

Competing interests

The authors have no conflicts of interest to declare. This study was sponsored by the Institute of Clinical Outcomes Research and Education (ICORE). In the past 3 years, ICORE has received grant support from Altana, Novartis, Centocor, Pfizer, and Astra-Zeneca.

Copyright information

© Springer Science+Business Media, LLC 2010