- Email: [email protected]
The Burden of Inflammatory Bowel Disease in the United States: A Moving Target?
Editorial The Burden of Inflammatory Bowel Disease in the United States: A Moving Target?
A
lthough the 2 major subtypes of inflammatory bowel disease (IBD), ulcerative colitis (UC) and Crohn’s disease, are not particularly common compared with heart disease or cancer, and are associated with only minor increases in relative mortality,1 they can result in substantial morbidity and excess health care resource use.2– 4 It is therefore important to understand the overall burden of IBD in this country. However, the American health care delivery system is not an integrated one, thereby making estimates of the incidence and prevalence of conditions such as IBD difficult to perform. Most recent reports of the descriptive epidemiology of IBD in the United States have come from specific geographic regions, not necessarily representative of the entire country. For example, several reports of the incidence and prevalence of IBD in Olmsted County, Minnesota, have been published or presented,5– 8 whereas the incidence and prevalence of IBD among children residing in the state of Wisconsin has been estimated recently.9 Studies from both of these regions relied on medical record review of 100% of known potential cases for validation of diagnoses. Residents of these 2 areas may share a passion for ice fishing, Lena and Ole jokes, and Vikings vs Packers football, but such interests may not be reflective of those of the country as a whole. Furthermore, extrapolating incidence and prevalence data from these regions to the entire country may not provide accurate estimates, especially if there are significant socioeconomic, ethnic, or racial differences in the make-up of these areas compared with the entire country, and if we believe that there still are significant differences in the prevalence of IBD among various ethnic and racial groups. In this month’s edition of Clinical Gastroenterology & Hepatology, Kappelman et al10 analyze a geographically diverse health insurance claims database to estimate the prevalence of Crohn’s disease and UC among children and adults with commercial health insurance in the United States in 2003-2004. The claims database, PharMetrics (IMS Health, Watertown, MA), captured billing codes (International Classification of Diseases, 9th edition [ICD-9]) and pharmacy claims from multiple insurance plans in more than 30 states, and contained longitudinal claims histories of almost 9 million persons. The algorithm to identify IBD cases from ICD-9 codes and pharmacy claims was largely similar to that used by Bernstein et al11 to assemble a cohort from a provincial health claims database in Manitoba, Canada. In that study, the case-finding algorithm (at least 5 separate medical contacts for IBD for a follow-up duration of ⱖ2 years and at least 3 separate contacts for a follow-up duration of ⬍2 years) was validated by medical record review of a random sample of cases, and was shown to be reasonably accurate, with a sensitivity ranging from 74% to 89%, and a false-positive rate of 6% to 10%.11 Because Kappelman et al10 studied the PharMetrics database for a 2-year period, their algorithm consisted of 3 separate ICD-9 codes for IBD, or at least 1 ICD-9 code plus at least 1 pharmacy claim for an IBD-specific medication (5aminosalicylates, mercaptopurine, azathioprine, infliximab,
adalimumab, or oral delayed-release budesonide). Medical records of potential cases were not available for review. The authors performed a sensitivity analysis of sorts by concurrently identifying cases using the least stringent case-finding algorithm of at least one ICD-9 code for IBD. By using the more stringent definition, almost 14,000 Crohn’s disease cases and almost 16,000 UC cases were identified. Not surprisingly, prevalence of IBD increased with increasing age. Resource use was not reported in detail, but it appeared to be substantial, in that more than 75% of all patients had at least one pharmacy claim for an IBD medication over the 2-year period (about two thirds had at least 3 pharmacy claims), and about one quarter of Crohn’s disease patients and one seventh of UC patients required a hospitalization. Regional differences in prevalence were noted—for both children and adults, IBD seemed slightly less prevalent in the southeastern United States compared with other regions. Larger differences in prevalence were seen when the analysis was stratified by insurance status, especially in children. Those with commercial insurance were significantly more likely than those on medicaid to have IBD. The prevalence of Crohn’s disease was 201 cases per 100,000 persons among adults and 43 per 100,000 among children. For UC, the prevalence was 238 cases per 100,000 persons among adults and 28 per 100,000 among children. Extrapolating these prevalence estimates to the US population, the authors estimated that slightly less than 1 million Americans had IBD in 2005.10 However, had the less stringent case definition been used, they would have identified almost 20,000 Crohn’s disease cases and almost 26,000 UC cases (increases over the more stringent definition of 50% and 64%, respectively), and the overall prevalence of IBD would have increased by 51% for children and by 59% for adults, such that there would have been approximately 1.5 million persons with IBD in the United States as of 2005. Another group recently used methodology quite similar to Kappelman et al10 to estimate the prevalence of IBD among insured persons in the United States. Both reports highlighted the strengths and difficulties of using administrative data. Herrinton et al12 recently published their estimate of the prevalence of IBD among 1.8 million members of 9 integrated health care organizations across the United States from 1999 to 2001 by developing case-finding algorithms and applying them to administrative health claims data. Herrinton et al12 had the additional advantage of having available to them in 2 of the health plans medical records for validation of their case-finding algorithm. They also tested the accuracy of the various algorithms if the duration of follow-up evaluation in the database was varied. Their case-finding algorithm varied from that of the PharMetrics study in that they required either one visit with an ICD-9 code for IBD or 1 dispensing of mesalamine, olsalazine, or balsalazide.12 The chart validation phase of the study suggested that their algorithm, when applied to 30 months’ worth of longitudinal data, had a sensitivity of 91% to 99%, depending on the health plan studied, with a positive predictive value of 81% to 86%. On a less positive note, however, applying the aforementioned algorithms to a health plan dataset with 78 months rather than 30 months of follow-up evaluation showed that the CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2007;5:1383–1384
1384
EDWARD V. LOFTUS JR
period prevalence of IBD could increase by as much as 50%.12 Presumably this arises not only from migration into and out of the health plan, but also from the fact that certain persons with IBD may not have contact with the health care delivery system for extended periods of time. For example, the UC patient who has undergone ileal pouch–anal anastomosis or Brooke ileostomy and is doing well clinically—theoretically the patient may not generate any billing codes specifically related to IBD for years. Thus, prevalence estimates from administrative databases, especially those of short longitudinal duration, represent minimal estimates. In the Herrinton et al12 study, the overall prevalence of IBD in 1999 to 2001 was 388 cases per 100,000 person-years,12 which, interestingly enough, was identical to the combined prevalence of UC and Crohn’s disease in Olmsted County in 2001.7 One hopes that it is reasonable to assume that billing practices in Manitoba11 or across the 9 HMOs in the Herrinton et al12 study will be similar enough to those of physicians in the PharMetrics database10 to allow for application of the same case-finding algorithm, but this question remains unanswered. Nevertheless, the estimate by Kappelman et al10 of 1 to 1.5 million Americans with IBD matches up reasonably well with other recent approximations. Herrinton et al12 extrapolated their 9-HMO prevalence to the United States and arrived at a figure of 1.1 million persons.12 Our group at Mayo Clinic had used Olmsted County IBD prevalence rates to estimate that there were approximately 1.1 million persons with IBD in the United States in 2000,7 and roughly 1.5 million persons in 2005.8 By 2005, Olmsted County had seen a 21% increase in the prevalence of UC to 273 cases per 100,000 persons and a 22% increase for Crohn’s disease to 222 per 100,000 over a relatively short period of time, which was thought to be caused in part by a recent increase in incidence rates.8 Although one may quibble with the different methods used in various studies to estimate the prevalence of IBD in the United States, the study by Kappelman et al10 adds to the growing body of literature on the national burden of IBD. Their finding of significant differences in IBD prevalence between commercially insured children and those on medical assistance raises questions about both the hygiene hypothesis and possible disparities in access to health care, which should be explored further. It seems reasonable to assume that there are a minimum of 1 million Americans, and as many as 1.5 million, with these chronic diseases. Hopefully, the numerous advances being made on the genetic, immunologic, epidemiologic, and therapeutic fronts will translate into real improvements in quality of life, and ultimately a cure, for all of our patients in the future.
EDWARD V. LOFTUS JR Miles & Shirley Fiterman Center for Digestive Diseases Mayo Clinic Rochester, Minnesota
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 5, No. 12
References 1. Loftus EV. Crohn’s disease: why the disparity in mortality? Gut 2006;55:447– 449. 2. Longobardi T, Jacobs P, Bernstein CN. Utilization of health care resources by individuals with inflammatory bowel disease in the United States: a profile of time since diagnosis. Am J Gastroenterol 2004;99:650 – 655. 3. Longobardi T, Bernstein CN. Health care resource utilization in inflammatory bowel disease. Clin Gastroenterol Hepatol 2006;4: 731–743. 4. Bewtra M, Su CY, Lewis JD. Trends in hospitalization rates for inflammatory bowel disease in the United States. Clin Gastroenterol Hepatol 2007;5:597– 601. 5. Loftus EV Jr, Silverstein MD, Sandborn WJ, et al. Crohn’s disease in Olmsted County, Minnesota, 1940 –1993: incidence, prevalence, and survival. Gastroenterology 1998;114:1161–1168. 6. Loftus EV Jr, Silverstein MD, Sandborn WJ, et al. Ulcerative colitis in Olmsted County, Minnesota, 1940 –1993: incidence, prevalence, and survival. Gut 2000;46:336 –343. 7. Loftus CG, Loftus EV Jr, Harmsen WS, et al. Update on the incidence and prevalence of Crohn’s disease and ulcerative colitis in Olmsted County, Minnesota, 1940 –2000. Inflamm Bowel Dis 2007;13:254 –261. 8. Ingle SB, Loftus EV, Tremaine WJ, et al. Increasing incidence and prevalence of inflammatory bowel disease in Olmsted County, Minnesota, during 2001–2004 (abstr). Gastroenterology 2007; 132:A19 –A20. 9. Kugathasan S, Judd RH, Hoffmann RG, et al, Wisconsin Pediatric Inflammatory Bowel Disease A. Epidemiologic and clinical characteristics of children with newly diagnosed inflammatory bowel disease in Wisconsin: a statewide population-based study. J Pediatr 2003;143:525–531. 10. Kappelman MD, Rifas-Shiman SL, Kleinman K, et al. The prevalence and geographical distribution of Crohn’s disease and ulcerative colitis in the United States. Clin Gastroenterol Hepatol 2007;1424 –1429. 11. Bernstein CN, Blanchard JF, Rawsthorne P, et al. Epidemiology of Crohn’s disease and ulcerative colitis in a central Canadian province: a population-based study. Am J Epidemiol 1999;149: 916 –924. 12. Herrinton LJ, Liu L, Lafata JE, et al. Estimation of the period prevalence of inflammatory bowel disease among nine health plans using computerized diagnoses and outpatient pharmacy dispensings. Inflamm Bowel Dis 2007;13:451– 461.
Abbreviations used in this paper: IBD, inflammatory bowel disease; ICD-9, International Classification of Diseases, 9th edition; UC, ulcerative colitis. Address requests for reprints to: Edward V. Loftus Jr, MD, Mayo Clinic, 200 First Street, SW, Rochester, Minnesota 55905. e-mail: [email protected]; fax: (507) 266-0335. Dr Loftus has received research support from Procter & Gamble, Schering-Plough, Abbott Labs, UCB Pharma, and PDL Biopharma. Dr. Loftus has consulted for Abbott Labs, UCB Pharma, Elan Pharmaceuticals, PDL Biopharma, Procter & Gamble, Shire, and Salix. doi:10.1016/j.cgh.2007.10.016
A
lthough the 2 major subtypes of inflammatory bowel disease (IBD), ulcerative colitis (UC) and Crohn’s disease, are not particularly common compared with heart disease or cancer, and are associated with only minor increases in relative mortality,1 they can result in substantial morbidity and excess health care resource use.2– 4 It is therefore important to understand the overall burden of IBD in this country. However, the American health care delivery system is not an integrated one, thereby making estimates of the incidence and prevalence of conditions such as IBD difficult to perform. Most recent reports of the descriptive epidemiology of IBD in the United States have come from specific geographic regions, not necessarily representative of the entire country. For example, several reports of the incidence and prevalence of IBD in Olmsted County, Minnesota, have been published or presented,5– 8 whereas the incidence and prevalence of IBD among children residing in the state of Wisconsin has been estimated recently.9 Studies from both of these regions relied on medical record review of 100% of known potential cases for validation of diagnoses. Residents of these 2 areas may share a passion for ice fishing, Lena and Ole jokes, and Vikings vs Packers football, but such interests may not be reflective of those of the country as a whole. Furthermore, extrapolating incidence and prevalence data from these regions to the entire country may not provide accurate estimates, especially if there are significant socioeconomic, ethnic, or racial differences in the make-up of these areas compared with the entire country, and if we believe that there still are significant differences in the prevalence of IBD among various ethnic and racial groups. In this month’s edition of Clinical Gastroenterology & Hepatology, Kappelman et al10 analyze a geographically diverse health insurance claims database to estimate the prevalence of Crohn’s disease and UC among children and adults with commercial health insurance in the United States in 2003-2004. The claims database, PharMetrics (IMS Health, Watertown, MA), captured billing codes (International Classification of Diseases, 9th edition [ICD-9]) and pharmacy claims from multiple insurance plans in more than 30 states, and contained longitudinal claims histories of almost 9 million persons. The algorithm to identify IBD cases from ICD-9 codes and pharmacy claims was largely similar to that used by Bernstein et al11 to assemble a cohort from a provincial health claims database in Manitoba, Canada. In that study, the case-finding algorithm (at least 5 separate medical contacts for IBD for a follow-up duration of ⱖ2 years and at least 3 separate contacts for a follow-up duration of ⬍2 years) was validated by medical record review of a random sample of cases, and was shown to be reasonably accurate, with a sensitivity ranging from 74% to 89%, and a false-positive rate of 6% to 10%.11 Because Kappelman et al10 studied the PharMetrics database for a 2-year period, their algorithm consisted of 3 separate ICD-9 codes for IBD, or at least 1 ICD-9 code plus at least 1 pharmacy claim for an IBD-specific medication (5aminosalicylates, mercaptopurine, azathioprine, infliximab,
adalimumab, or oral delayed-release budesonide). Medical records of potential cases were not available for review. The authors performed a sensitivity analysis of sorts by concurrently identifying cases using the least stringent case-finding algorithm of at least one ICD-9 code for IBD. By using the more stringent definition, almost 14,000 Crohn’s disease cases and almost 16,000 UC cases were identified. Not surprisingly, prevalence of IBD increased with increasing age. Resource use was not reported in detail, but it appeared to be substantial, in that more than 75% of all patients had at least one pharmacy claim for an IBD medication over the 2-year period (about two thirds had at least 3 pharmacy claims), and about one quarter of Crohn’s disease patients and one seventh of UC patients required a hospitalization. Regional differences in prevalence were noted—for both children and adults, IBD seemed slightly less prevalent in the southeastern United States compared with other regions. Larger differences in prevalence were seen when the analysis was stratified by insurance status, especially in children. Those with commercial insurance were significantly more likely than those on medicaid to have IBD. The prevalence of Crohn’s disease was 201 cases per 100,000 persons among adults and 43 per 100,000 among children. For UC, the prevalence was 238 cases per 100,000 persons among adults and 28 per 100,000 among children. Extrapolating these prevalence estimates to the US population, the authors estimated that slightly less than 1 million Americans had IBD in 2005.10 However, had the less stringent case definition been used, they would have identified almost 20,000 Crohn’s disease cases and almost 26,000 UC cases (increases over the more stringent definition of 50% and 64%, respectively), and the overall prevalence of IBD would have increased by 51% for children and by 59% for adults, such that there would have been approximately 1.5 million persons with IBD in the United States as of 2005. Another group recently used methodology quite similar to Kappelman et al10 to estimate the prevalence of IBD among insured persons in the United States. Both reports highlighted the strengths and difficulties of using administrative data. Herrinton et al12 recently published their estimate of the prevalence of IBD among 1.8 million members of 9 integrated health care organizations across the United States from 1999 to 2001 by developing case-finding algorithms and applying them to administrative health claims data. Herrinton et al12 had the additional advantage of having available to them in 2 of the health plans medical records for validation of their case-finding algorithm. They also tested the accuracy of the various algorithms if the duration of follow-up evaluation in the database was varied. Their case-finding algorithm varied from that of the PharMetrics study in that they required either one visit with an ICD-9 code for IBD or 1 dispensing of mesalamine, olsalazine, or balsalazide.12 The chart validation phase of the study suggested that their algorithm, when applied to 30 months’ worth of longitudinal data, had a sensitivity of 91% to 99%, depending on the health plan studied, with a positive predictive value of 81% to 86%. On a less positive note, however, applying the aforementioned algorithms to a health plan dataset with 78 months rather than 30 months of follow-up evaluation showed that the CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2007;5:1383–1384
1384
EDWARD V. LOFTUS JR
period prevalence of IBD could increase by as much as 50%.12 Presumably this arises not only from migration into and out of the health plan, but also from the fact that certain persons with IBD may not have contact with the health care delivery system for extended periods of time. For example, the UC patient who has undergone ileal pouch–anal anastomosis or Brooke ileostomy and is doing well clinically—theoretically the patient may not generate any billing codes specifically related to IBD for years. Thus, prevalence estimates from administrative databases, especially those of short longitudinal duration, represent minimal estimates. In the Herrinton et al12 study, the overall prevalence of IBD in 1999 to 2001 was 388 cases per 100,000 person-years,12 which, interestingly enough, was identical to the combined prevalence of UC and Crohn’s disease in Olmsted County in 2001.7 One hopes that it is reasonable to assume that billing practices in Manitoba11 or across the 9 HMOs in the Herrinton et al12 study will be similar enough to those of physicians in the PharMetrics database10 to allow for application of the same case-finding algorithm, but this question remains unanswered. Nevertheless, the estimate by Kappelman et al10 of 1 to 1.5 million Americans with IBD matches up reasonably well with other recent approximations. Herrinton et al12 extrapolated their 9-HMO prevalence to the United States and arrived at a figure of 1.1 million persons.12 Our group at Mayo Clinic had used Olmsted County IBD prevalence rates to estimate that there were approximately 1.1 million persons with IBD in the United States in 2000,7 and roughly 1.5 million persons in 2005.8 By 2005, Olmsted County had seen a 21% increase in the prevalence of UC to 273 cases per 100,000 persons and a 22% increase for Crohn’s disease to 222 per 100,000 over a relatively short period of time, which was thought to be caused in part by a recent increase in incidence rates.8 Although one may quibble with the different methods used in various studies to estimate the prevalence of IBD in the United States, the study by Kappelman et al10 adds to the growing body of literature on the national burden of IBD. Their finding of significant differences in IBD prevalence between commercially insured children and those on medical assistance raises questions about both the hygiene hypothesis and possible disparities in access to health care, which should be explored further. It seems reasonable to assume that there are a minimum of 1 million Americans, and as many as 1.5 million, with these chronic diseases. Hopefully, the numerous advances being made on the genetic, immunologic, epidemiologic, and therapeutic fronts will translate into real improvements in quality of life, and ultimately a cure, for all of our patients in the future.
EDWARD V. LOFTUS JR Miles & Shirley Fiterman Center for Digestive Diseases Mayo Clinic Rochester, Minnesota
CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 5, No. 12
References 1. Loftus EV. Crohn’s disease: why the disparity in mortality? Gut 2006;55:447– 449. 2. Longobardi T, Jacobs P, Bernstein CN. Utilization of health care resources by individuals with inflammatory bowel disease in the United States: a profile of time since diagnosis. Am J Gastroenterol 2004;99:650 – 655. 3. Longobardi T, Bernstein CN. Health care resource utilization in inflammatory bowel disease. Clin Gastroenterol Hepatol 2006;4: 731–743. 4. Bewtra M, Su CY, Lewis JD. Trends in hospitalization rates for inflammatory bowel disease in the United States. Clin Gastroenterol Hepatol 2007;5:597– 601. 5. Loftus EV Jr, Silverstein MD, Sandborn WJ, et al. Crohn’s disease in Olmsted County, Minnesota, 1940 –1993: incidence, prevalence, and survival. Gastroenterology 1998;114:1161–1168. 6. Loftus EV Jr, Silverstein MD, Sandborn WJ, et al. Ulcerative colitis in Olmsted County, Minnesota, 1940 –1993: incidence, prevalence, and survival. Gut 2000;46:336 –343. 7. Loftus CG, Loftus EV Jr, Harmsen WS, et al. Update on the incidence and prevalence of Crohn’s disease and ulcerative colitis in Olmsted County, Minnesota, 1940 –2000. Inflamm Bowel Dis 2007;13:254 –261. 8. Ingle SB, Loftus EV, Tremaine WJ, et al. Increasing incidence and prevalence of inflammatory bowel disease in Olmsted County, Minnesota, during 2001–2004 (abstr). Gastroenterology 2007; 132:A19 –A20. 9. Kugathasan S, Judd RH, Hoffmann RG, et al, Wisconsin Pediatric Inflammatory Bowel Disease A. Epidemiologic and clinical characteristics of children with newly diagnosed inflammatory bowel disease in Wisconsin: a statewide population-based study. J Pediatr 2003;143:525–531. 10. Kappelman MD, Rifas-Shiman SL, Kleinman K, et al. The prevalence and geographical distribution of Crohn’s disease and ulcerative colitis in the United States. Clin Gastroenterol Hepatol 2007;1424 –1429. 11. Bernstein CN, Blanchard JF, Rawsthorne P, et al. Epidemiology of Crohn’s disease and ulcerative colitis in a central Canadian province: a population-based study. Am J Epidemiol 1999;149: 916 –924. 12. Herrinton LJ, Liu L, Lafata JE, et al. Estimation of the period prevalence of inflammatory bowel disease among nine health plans using computerized diagnoses and outpatient pharmacy dispensings. Inflamm Bowel Dis 2007;13:451– 461.
Abbreviations used in this paper: IBD, inflammatory bowel disease; ICD-9, International Classification of Diseases, 9th edition; UC, ulcerative colitis. Address requests for reprints to: Edward V. Loftus Jr, MD, Mayo Clinic, 200 First Street, SW, Rochester, Minnesota 55905. e-mail: [email protected]; fax: (507) 266-0335. Dr Loftus has received research support from Procter & Gamble, Schering-Plough, Abbott Labs, UCB Pharma, and PDL Biopharma. Dr. Loftus has consulted for Abbott Labs, UCB Pharma, Elan Pharmaceuticals, PDL Biopharma, Procter & Gamble, Shire, and Salix. doi:10.1016/j.cgh.2007.10.016