Surgical Treatment in Childhood-onset Inflammatory Bowel Disease–A Nationwide Register-based Study of 4695 Incident Patients in Sweden 2002-2014

Surgical Treatment in Childhood-onset Inflammatory Bowel Disease–A Nationwide Register-based... Abstract Background and Aims The incidence of childhood-onset [< 18 years] inflammatory bowel disease [IBD] is increasing worldwide, and some studies suggest that it represents a more severe disease phenotype. Few nationwide, population-based studies have evaluated the surgical burden in patients with childhood-onset IBD, and whether the improved medical treatment has influenced the need for gastrointestinal surgery. The aim was to examine whether the surgical treatment at any age of patients with childhood-onset IBD has changed over time. Methods In a nationwide cohort study we identified 4695 children [< 18 years] diagnosed with incident IBD in 2002–2014 through the Swedish Patient Register [ulcerative colitis: n = 2295; Crohn’s disease: n = 2174; inflammatory bowel disease-unclassified: n = 226]. Abdominal [intestinal resections and colectomies] and perianal surgeries were identified through the Swedish Patient Register. The cumulative incidences of surgeries were calculated using the Kaplan-Meier method. Results In the cohort, 44% were females and 56% males. The median age at inflammatory bowel disease diagnosis was 15 years and the maximum age at end of follow-up was 31 years. The 3-year cumulative incidence of intestinal surgery was 5% in patients with ulcerative colitis and 7% in patients with Crohn’s disease, and lower in children aged < 6 years at inflammatory bowel disease diagnosis [3%] than in those aged 15–17 years at diagnosis [7%]. Calendar period of inflammatory bowel disease diagnosis was not associated with risk of surgery. Conclusion Over the past 13 years, the risk of surgery in childhood-onset inflammatory bowel disease has remained unchanged. Children, inflammatory bowel disease, surgery 1. Introduction The incidence of inflammatory bowel disease [IBD] in children is increasing worldwide.1–3 Data from Sweden and other Nordic countries have shown a similar trend,3–6 with a higher incidence of childhood-onset IBD than in other parts of the world.7–9 In 2003 to 2013, the incidence rate of childhood-onset IBD in Sweden was 17/100000 in girls and 20/100000 in boys.5 In children with ulcerative colitis [UC], surgery is needed in acute severe colitis not responding to medical therapy, and elective colectomies are performed in children with steroid-dependent disease, poor response to medical treatment, or delay of growth and/or puberty.10,11 In addition to these indications, children with Crohn’s disease [CD] require surgical treatment for strictures and fistulas.12,13 The childhood-onset IBD phenotype is characterised by more extensive disease location.14–16 Studies also suggest that childhood-onset IBD represents a more severe phenotype associated with increased use of immunomodulatory therapy and need for surgical treatment.17–19 More recent population-based studies have shown a cumulative probability of surgical treatment after 3–5 years ranging from 5–34% in children with CD and 7–20% in children with UC [Supplementary Table 1, available as Supplementary data at ECCO-JCC online].1,7,16–18,20,21 The introduction of immunomodulators [e.g. azathioprine] during the 1990s and inhibitors of tumour necrosis factor [anti-TNF] at the beginning of the 21st century has evoked hope that the need of surgical treatment would decrease.22 Whereas some studies have shown a decreased need of surgery after introduction of anti-TNFs,20,23 other studies have shown an increasing rate of intestinal resections in children with CD.24–26 Hence, it is still unclear whether the need for surgical treatment of patients with childhood-onset IBD has decreased in recent years. In order to shed further light on this question, a population-based nationwide register study was performed including all children diagnosed with IBD between 2002 and 2014 in Sweden. The aim of this study was to describe the need for surgical treatment in a modern cohort of patients with childhood-onset IBD during childhood and early adulthood. 2. Materials and Methods 2.1. Study design and setting In this population-based cohort, we included all Swedish children diagnosed with IBD before their 18th birthday during 2002 to 2014. In Sweden, paediatric gastroenterologists treat children with IBD during childhood until 18 years of age. When in need of surgery, children with IBD may be treated by paediatric surgeons as well as by general surgeons, depending on local tradition. Since 1947, residents of Sweden have been assigned a unique personal identifier including information on year, month, and day of birth, as well as sex of the person. This national personal identity number is used in all official registers and documents, also enabling linkage of registries.27 This nationwide cohort study was based on data from virtually complete Swedish nationwide registers of hospitalisations, causes of death, and emigration.28–30 Swedish inpatient diagnoses of IBD have been registered in the National Inpatient register since January 1, 1964 [increasing coverage over the years] and with nationwide full coverage since 1987.29 In addition, all Swedish IBD diagnoses in specialist [non-primary] outpatient care have been registered in the Outpatient register since January 1, 2001. 2.2. Study population In order to restrict the study population to incident cases of IBD in outpatient or inpatient care, we identified all individuals in Sweden with a first diagnosis of IBD in the Swedish patient register [inpatient or non-primary outpatient care] from January 1, 2002 until December 31, 2014 [ie children with an IBD diagnosis in 2001 or earlier were excluded]. We used the following International Classification of Diseases [ICD] codes: CD: 563,00 [ICD8]; 555 [ICD9]; K50 [ICD10]. UC: 563,10; 569,02 [ICD8]; 556 [ICD9]; K51 [ICD10], and indeterminate colitis [IBD-U]: K52.3 [ICD10]. We restricted our IBD population to only those children who had at least two IBD diagnoses recorded by a physician on two separate occasions, in order to minimise the risk of potential misclassification,31,32 and follow-up time was calculated from the first IBD diagnosis. To decrease the risk of introducing bias, given the variations in follow-up time, type of IBD was specified at study inclusion for all analyses of future surgery. Since two IBD events were required for inclusion, some children received a mix of codes for UC, CD, or the code for IBD-U, and they were all categorised as IBD-U. However, in the descriptive statistics of the study population at end of follow-up, we used a different definition of UC, CD, and IBD-U. Because different IBD diagnoses might be documented during a patient’s medical history, due to a colitis that is difficult to categorise as either UC or CD or simply due to an incorrect registration by mistake, we classified patients with a mix of codes during the past 5 years for UC, CD, or IBD-U during follow-up as IBD-U. Moreover, patients who only shifted between UC and CD [and vice versa] and who had been diagnosed with only UC or CD the past 5 years of follow-up, were classified as UC or CD in our end of follow-up classification. The same classification was also used in a sensitivity analysis calculating the risk of intestinal resection, colectomy, and perianal surgery. 2.3. IBD surgery Data regarding IBD surgery were extracted from inpatient care, and procedures were coded using the NOMESCO Classification of Surgical Procedures. Surgeries performed before IBD diagnosis and during follow-up were reported separately. In addition to describing risk of surgery through the entire follow-up [up to the 31st birthday], we also calculated risk of surgery during childhood [before the 18th birthday]. The following codes were included in the definition of IBD surgery: Colectomy [JFH], Segmental bowel resection [Small bowel: JFB00, JFB01; Large bowel: JFB20, JFB21, JFB30, JFB31, JFB40, JFB 41, JFB43, JFB 44, JFB46, JFB47, JFB50, JFB51, JFB53, JFB 54, JFB60, JFB61, JFB63; Unspecified bowel resection: JFB33, JFB34, JFB96, JFB97], Strictureplasty [JFA60], and Perianal surgery [JHA00, JHD20, JHD30, JHD33, JHD50, JHD60, JHD63, JHW96]. For reconstructive surgery, the following codes were used for ileorectal anastomosis depending on whether the procedure was done in conjunction or following the colectomy: JFH00, JFH01, and JFH00, JFH01, JFC40, JFC41, JFG26, JFG29. Likewise, the following codes for ileal pouch anal anastomosis were used: JFH30, JFH33, and JFH30, JFH33, JGB50, JGB60, JGB61. 2.4. Exposure to thiopurines and anti-TNF The Swedish Prescribed drug register started on July 1, 2005 and contains all prescribed drug dispensation data, but not medications dispensed in hospital or bought over the counter. Nationwide and complete information was obtained on the use of the thiopurines Azathioprine [L04AX01] and Mercaptopurine [L01BB02]. The following ATC codes were used to identify anti-TNF treatment; Adalimumab [L04AB04 [L04AA17 before 2008]], Infliximab [L04AB02 [L04AA12 before 2008]], Golimumab [L04AB06], and Vedolizumab [L04AA33]. Whereas the coverage of adalimumab is virtually complete, infliximab is administered as an infusion in the hospital setting, and not all infliximab doses can be found in the Prescribed drug register.33 Ever prescriptions of thiopurines, anti-TNF, or neither were evaluated at the end of follow-up. 2.5. Extra-intestinal manifestations The following diagnostic codes were used to identify presence of extra-intestinal manifestations in the skin [ICD9: 695C, ICD10: L52, L88, L98.2], eyes [ICD9: 364, ICD10: H20], or joints [ICD9: 713B, 720A, 720C, 720W, 720X, and ICD10: M07.4, M07.5, M07.6, M09.1, M09.2, M45, M460, M461, M468, M469, M139, M255] at end of follow-up. The corresponding codes for primary sclerosing cholangitis were 576B [ICD9] and K83.0 [ICD10]. 2.6. Statistical methods Cumulative and relative risks were computed for three separate outcomes: [i] intestinal surgery, [ii] colectomy, and [iii] perianal surgery. When the outcome of interest was intestinal surgery, the children were followed from study entry until the date of intestinal resection (colectomy, bowel resection, or strictureplasty [irrespective if perianal surgery had been performed prior to that date]), death, emigration or December 31, 2014, whichever came first. Similarly, in the analyses of colectomy, the children were followed from study entry until the date of colectomy [irrespective of whether other IBD surgery had been performed], death, emigration or December 31, 2014. Children in whom a colectomy was performed before study entry were excluded from these analyses. In the analyses of perianal surgery, the children were followed from study entry until the date of perianal surgery [irrespective of whether intestinal resection had been performed], death, emigration or December 31, 2014. The cumulative incidences of surgery were estimated using the Kaplan-Meier method. Potential differences in time to surgery by patient characteristics were tested using the log-rank test. Cox regression was used to estimate crude and adjusted hazard ratios of surgery with robust estimates for the standard errors.34 The proportionality of the hazard functions across covariate patterns was tested with Schoenfeld’s residuals, and there was no sign of non-proportionality. The multivariable model was adjusted for sex, age at first IBD diagnosis (< 6 [very early onset IBD; VEO-IBD], 6–9, 10–14, 15–17 years], year of first IBD diagnosis [2002–2004, 2005–2009, 2010–2014], and type of IBD [UC, CD, and IBD-U]. The analyses were carried out with Stata 14 [StataCorp, TX, USA]. Statistical tests were two-sided, and statistical significance was set at α = 0.05. The researchers only had access to de-identified data. The study had been approved by the Ethical review board in Stockholm [DNR of ethical approvals including amendments: 2007/785-31/5; 2015/1030–32]. 3. Results A total of 45626 unique patients diagnosed with IBD in 2002 to 2014 were identified. Of those, 40931 were excluded due to age ≥ 18 years at IBD diagnosisn leaving 4695 children in the final cohort. Table 1 describes the basic characteristics of the cohort. Most children were diagnosed with UC [n = 2295, 49%] or CD [n = 2174, 46%], and 226 were classified as IBD-U [5%]. The distribution of IBD diagnoses was similar in girls and boys, as well as in different age categories, but IBD-U was more common in the latest time period. The mean age at first IBD diagnosis was 14 years [median 15 years] and almost half of the children [47%] were between 15 and 17 years at diagnosis [Figure 1]. The median age at first surgery was 17 years [range 0–28 years] and at end of follow-up 20 years [range 0–31 years][Figure 1]. Table 1. Incident cases of childhood-onset [< 18 years] inflammatory bowel disease [IBD] in Sweden 2002–2014. Patient characteristics at time of first IBD diagnosis.   UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Totalb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Girls  1056 [46]  914 [42]  94 [42]  2064 [44]  Boys  1239 [54]  1260 [58]  132 [58]  2631 [56]  Age at IBD diagnosis           Mean age [SD]  14 [4]  14 [3]  14 [4]  14 [4]   Median age [range]  15 [0–18]  14 [0–18]  15 [0–18]  15 [0–18]   < 6 years  119 [5]  85 [4]  17 [8]  221 [5]   6–9 years  203 [9]  205 [9]  16 [7]  424 [9]   10–14 years  836 [36]  947 [44]  83 [37]  1866 [40]   15–17 years  1137 [50]  937 [43]  110 [49]  2184 [47]  Year of IBD diagnosis           2002–2004  551 [24]  526 [24]  5 [2]  1082 [23]   2005–2009  1011 [44]  901 [41]  40 [18]  1952 [42]   2010–2014  733 [32]  747 [34]  181 [80]  1661 [35]  Before IBD diagnosis           Colectomy  2 [0.1]  2 [0]  -  4 [0]   Large bowel resectionc  -  12 [1]  -  12 [0]   Small bowel resection  2 [0]  3 [0]  -  5 [0]   Perianal surgery  8 [0]  101 [5]  6 [3]  115 [2]    UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Totalb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Girls  1056 [46]  914 [42]  94 [42]  2064 [44]  Boys  1239 [54]  1260 [58]  132 [58]  2631 [56]  Age at IBD diagnosis           Mean age [SD]  14 [4]  14 [3]  14 [4]  14 [4]   Median age [range]  15 [0–18]  14 [0–18]  15 [0–18]  15 [0–18]   < 6 years  119 [5]  85 [4]  17 [8]  221 [5]   6–9 years  203 [9]  205 [9]  16 [7]  424 [9]   10–14 years  836 [36]  947 [44]  83 [37]  1866 [40]   15–17 years  1137 [50]  937 [43]  110 [49]  2184 [47]  Year of IBD diagnosis           2002–2004  551 [24]  526 [24]  5 [2]  1082 [23]   2005–2009  1011 [44]  901 [41]  40 [18]  1952 [42]   2010–2014  733 [32]  747 [34]  181 [80]  1661 [35]  Before IBD diagnosis           Colectomy  2 [0.1]  2 [0]  -  4 [0]   Large bowel resectionc  -  12 [1]  -  12 [0]   Small bowel resection  2 [0]  3 [0]  -  5 [0]   Perianal surgery  8 [0]  101 [5]  6 [3]  115 [2]  SD, standard deviation. aType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, ie either K52.3 or a mixture of UC and CD diagnoses. bRow percent. cIncluding ileocaecal resection. View Large Figure 1. View largeDownload slide Age-distribution at [A] first IBD diagnosis, [B] first IBD surgery, [C] end of follow-up [first surgery, death, emigration or December 31, 2014] in 4695 incident cases of childhood-onset [< 18 years] inflammatory bowel disease. IBD, inflammatory bowel disease. Figure 1. View largeDownload slide Age-distribution at [A] first IBD diagnosis, [B] first IBD surgery, [C] end of follow-up [first surgery, death, emigration or December 31, 2014] in 4695 incident cases of childhood-onset [< 18 years] inflammatory bowel disease. IBD, inflammatory bowel disease. Table 2 describes the study population at end of follow-up, stratified by type of IBD diagnosis at start of follow-up. The IBD diagnosis may change over time, and to determine the proportion of patients who received a different IBD diagnosis during follow-up, registered IBD diagnoses during the past f5 years of follow-up were also analysed. Of patients initially given a UC diagnosis, UC was the only diagnosis during the past 5 years of follow-up in 86% of the patients. The corresponding proportion in CD patients was 83%, and in patients with IBD-U 36%. Extra-intestinal manifestations were diagnosed in 8% and primary sclerosing cholangitis in 6%. The majority of patients had received medical treatment with thiopurines [60%]. Table 2. Descriptive characteristics at the end of follow-up in 4695 incident cases of childhood-onset [< 18 years] IBD in Sweden 2002–2014.   UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  End of follow-up [first IBD surgery, death, emigration, or December 31, 2014]  Follow-up in years [median, range]  6 [0–13]  6 [0–13]  3 [0–12]  6 [0–13]  Censored due to death  13 [1]  7 [0]  2 [1]  22 [1]  Censored due to emigration  20 [1]  24 [1]  1 [0]  45 [1]  Censored due to surgery  219 [10]  422 [19]  17 [8]  658 [14]  Censored due to end of follow-up  2043 [89]  1721 [79]  206 [91]  3970 [85]  End of follow-up [death, emigration, or December 31, 2014]  Follow-up in years [median, range]  7 [0–13]  7 [0–13]  3 [0–13]  7 [0–13]  IBD diagnosis past 5 years of follow-upc           UC  1837 [86]  259 [12]  51 [2]  2147 [46]   CD  326 [15]  1815 [83]  44 [2]  2185 [47]   IBD-U  132 [36]  100 [28]  131 [36]  363 [8]  Ever diagnosed with           Extra-intestinal manifestations  175 [8]  1707 [8]  12 [5]  357 [8]   Primary sclerosing cholangitis  172 [8]  90 [4]  13 [6]  275 [6]  Ever treatment with           Thiopurines  1256 [55]  1455 [67]  114 [50]  2825 [60]   Anti-TNF  301 [13]  453 [21]  22 [10]  776 [17]   Neither  1039 [45]  719 [33]  112 [50]  1870 [40]    UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  End of follow-up [first IBD surgery, death, emigration, or December 31, 2014]  Follow-up in years [median, range]  6 [0–13]  6 [0–13]  3 [0–12]  6 [0–13]  Censored due to death  13 [1]  7 [0]  2 [1]  22 [1]  Censored due to emigration  20 [1]  24 [1]  1 [0]  45 [1]  Censored due to surgery  219 [10]  422 [19]  17 [8]  658 [14]  Censored due to end of follow-up  2043 [89]  1721 [79]  206 [91]  3970 [85]  End of follow-up [death, emigration, or December 31, 2014]  Follow-up in years [median, range]  7 [0–13]  7 [0–13]  3 [0–13]  7 [0–13]  IBD diagnosis past 5 years of follow-upc           UC  1837 [86]  259 [12]  51 [2]  2147 [46]   CD  326 [15]  1815 [83]  44 [2]  2185 [47]   IBD-U  132 [36]  100 [28]  131 [36]  363 [8]  Ever diagnosed with           Extra-intestinal manifestations  175 [8]  1707 [8]  12 [5]  357 [8]   Primary sclerosing cholangitis  172 [8]  90 [4]  13 [6]  275 [6]  Ever treatment with           Thiopurines  1256 [55]  1455 [67]  114 [50]  2825 [60]   Anti-TNF  301 [13]  453 [21]  22 [10]  776 [17]   Neither  1039 [45]  719 [33]  112 [50]  1870 [40]  TNF, tumour necrosis factor. aType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, i.e. either K52.3 or a mixture of UC and CD-diagnoses. bRow percent. cType of IBD past 5 years was classified as UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = IBD unclassified or a mixture of UC and CD diagnoses. View Large The median time from IBD diagnosis to first surgical event [among patients who had surgery] was 3 years [range 0–13 years]. All surgical procedures performed during the entire follow-up [ie not only the first event] are described in Table 3. Any surgery was performed in 10% of UC patients and 19% of CD patients, including intestinal resections in 8% of UC patients and 12% of CD patients. In patients with UC, colectomy was the most common surgical procedure [7%], whereas segmental bowel resection and perianal surgery were the most common procedures in patients with CD [10% each]. Following a colectomy, secondary reconstructions with ileorectal anastomosis and ileal pouch anal anastomosis were more common than primary reconstructions. Table 3. Descriptive characteristics of all surgical procedures performed in 4695 incident cases of childhood-onset [< 18 years] IBD [inflammatory bowel disease] in Sweden 2002–2014 during all available follow-ups.   UCa  CDa  IBD-Ua  All    n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Treated with any surgery  219 [10]  422 [19]  17 [8]  658 [14]  Type of surgeryc           Any intestinal resection  194 [8]  264 [12]  15 [7]  473 [10]   Colectomy  169 [7]  36 [2]  10 [4]  215 [5]  Primary reconstruction           Ileorectal anastomosis  10 [0]  5 [0]  -  15 [0]   Ileal pouch-anal anastomosis  9 [0]  -  1 [0]  10 [0]  Secondary reconstruction           Ileorectal anastomosis  47 [2]  15 [1]  2 [1]  64 [1]   Ileal pouch-anal anastomosis  27 [1]  3 [0]  2 [1]  32 [1]   Segmental large bowel resection  20 [1]  207 [10]  5 [2]  232 [5]   Small bowel resection  13 [1]  44 [2]  -  57 [1]   Unspecified bowel resection  2 [0]  12 [1]  -  14 [0]   Strictureplasty  -  11 [1]  -  11 [0]   Any perianal surgery  55 [2]  206 [10]  3 [1]  264 [6]    UCa  CDa  IBD-Ua  All    n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Treated with any surgery  219 [10]  422 [19]  17 [8]  658 [14]  Type of surgeryc           Any intestinal resection  194 [8]  264 [12]  15 [7]  473 [10]   Colectomy  169 [7]  36 [2]  10 [4]  215 [5]  Primary reconstruction           Ileorectal anastomosis  10 [0]  5 [0]  -  15 [0]   Ileal pouch-anal anastomosis  9 [0]  -  1 [0]  10 [0]  Secondary reconstruction           Ileorectal anastomosis  47 [2]  15 [1]  2 [1]  64 [1]   Ileal pouch-anal anastomosis  27 [1]  3 [0]  2 [1]  32 [1]   Segmental large bowel resection  20 [1]  207 [10]  5 [2]  232 [5]   Small bowel resection  13 [1]  44 [2]  -  57 [1]   Unspecified bowel resection  2 [0]  12 [1]  -  14 [0]   Strictureplasty  -  11 [1]  -  11 [0]   Any perianal surgery  55 [2]  206 [10]  3 [1]  264 [6]  aType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, ie either K52.3 or a mixture of UC and CD diagnoses. bRow percent. cAdds up to more than the number of patients treated with surgery since several procedures could be performed in one patient on different occasions. View Large 3.1. Intestinal surgery The cumulative incidence of intestinal surgery is illustrated in Figure 2. The 3-year cumulative incidence of intestinal surgery in UC patients was 5%. The corresponding proportion in CD patients was 7%. Intestinal surgery was more common in CD patients: hazard ratio [HR] 1.53 (95% CI [confidence interval] 1.27–1.85) than in UC patients [Table 4]. The 3-year cumulative incidence of intestinal surgery was 3% in children aged < 6 years at IBD diagnosis. The corresponding proportion in children aged 15 to 17 years at IBD diagnosis was 7%. The risk of intestinal surgery was increased in patients aged 15 to 17 years at IBD diagnosis compared with the youngest age group (HR 2.67 [95% CI 1.42–5.03]). There was a trend to a decreasing risk of ever having intestinal surgery with increasing calendar year at IBD diagnosis in the multivariable analysis adjusted for sex, type of IBD, and age at IBD diagnosis, with a significantly decreased risk (HR 0.71 [95% CI 0.53–0.96] for those diagnosed in year 2010 to 2014 vs those diagnosed in 2002–2004) [Table 4]. The trend was however not evident in the cumulative incidence of intestinal surgery [Supplementary Table 2, available as Supplementary data at ECCO-JCC online; and Figure 2]. Figure 2. View largeDownload slide Cumulative incidence of intestinal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value < 0.001]; [C] calendar year at first IBD diagnosis [p-value 0.13]; [D] gender [p-value 0.07]. Figure 2. View largeDownload slide Cumulative incidence of intestinal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value < 0.001]; [C] calendar year at first IBD diagnosis [p-value 0.13]; [D] gender [p-value 0.07]. Table 4. The hazard ratios [HRs] of intestinal surgery,a colectomy, and perianal surgery in 4695 children [aged < 18 years] diagnosed with inflammatory bowel disease in Sweden in 2002 to 2014, followed through adulthood, and calculated with Cox regression.   Intestinal surgerya  Colectomyb  Perianal surgery    nc  HR4  95% CI  nc  HR4  95% CI  Casesc  HRd  95% CI  Females  191  Ref    79  Ref    79  Ref    Males  282  1.17  0.97–1.41  136  1.45  1.10–1.91  185  1.84  1.42–2.40  UCe  194  Ref    169  Ref    55  Ref    CDe  264  1.53  1.27–1.85  36  0.22  0.15–0.32  206  4.15  3.08–5.60  IBD-Ue  15  1.53  0.89–2.63  10  1.11  0.57–2.15  3  0.87  0.27–2.81  Age at IBD diagnosis                     < 6 years  10  Ref    8  Ref    4  Ref     6–9 years  26  1.31  0.63–2.72  13  0.90  0.37–2.17  21  2.48  0.85–7.23   10–14 years  186  2.04  1.07–3.85  74  1.19  0.58–2.48  115  2.72  1.00–7. 40   15–17 years  251  2.67  1.42–5.03  120  1.61  0.79–3.30  124  3.06  1.13–8.28  Year of IBD diagnosis                     2002–2004  188  Ref    86  Ref    94  Ref     2005–2009  213  0.86  0.70–1.06  94  0.84  0.61–1.15  118  0.94  0.69–1.21   2010–2014  72  0.71  0.53–0.96  35  0.82  0.52–1.28  52  1.00  0.70–1.47    Intestinal surgerya  Colectomyb  Perianal surgery    nc  HR4  95% CI  nc  HR4  95% CI  Casesc  HRd  95% CI  Females  191  Ref    79  Ref    79  Ref    Males  282  1.17  0.97–1.41  136  1.45  1.10–1.91  185  1.84  1.42–2.40  UCe  194  Ref    169  Ref    55  Ref    CDe  264  1.53  1.27–1.85  36  0.22  0.15–0.32  206  4.15  3.08–5.60  IBD-Ue  15  1.53  0.89–2.63  10  1.11  0.57–2.15  3  0.87  0.27–2.81  Age at IBD diagnosis                     < 6 years  10  Ref    8  Ref    4  Ref     6–9 years  26  1.31  0.63–2.72  13  0.90  0.37–2.17  21  2.48  0.85–7.23   10–14 years  186  2.04  1.07–3.85  74  1.19  0.58–2.48  115  2.72  1.00–7. 40   15–17 years  251  2.67  1.42–5.03  120  1.61  0.79–3.30  124  3.06  1.13–8.28  Year of IBD diagnosis                     2002–2004  188  Ref    86  Ref    94  Ref     2005–2009  213  0.86  0.70–1.06  94  0.84  0.61–1.15  118  0.94  0.69–1.21   2010–2014  72  0.71  0.53–0.96  35  0.82  0.52–1.28  52  1.00  0.70–1.47  CI, confidence interval; ref, reference. aIncluding colectomy, segmental large bowel resection, small bowel resection, unspecified bowel resection and strictureplasty. bPatients in whom colectomy was performed before study entry were excluded. cNumber of patients in whom surgery was performed. dMultivariable analyses adjusted for all listed variables. eType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, ie either K52.3 or a mixture of UC and CD diagnoses. View Large 3.2. Colectomy The 3 -year cumulative incidence of colectomy was higher in patients with UC [4%] compared with patients with CD [1%] [Figure 3]. In the multivariable model, male gender was associated with an increased risk of colectomy (HR 1.45 [95% CI 1.10–1.91]). Age at IBD diagnosis and calendar year of IBD diagnosis were not significantly associated with the risk of colectomy [Table 4, and Supplementary Tables 2 and 3, available as Supplementary data at ECCO-JCC online]. Figure 3. View largeDownload slide Cumulative incidence of colectomy during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.02]; [C] calendar year at first IBD diagnosis [p-value 0.74]; [D] gender [p-value 0.02]. Figure 3. View largeDownload slide Cumulative incidence of colectomy during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.02]; [C] calendar year at first IBD diagnosis [p-value 0.74]; [D] gender [p-value 0.02]. 3.3. Perianal surgery The 3-year cumulative incidence of perianal surgery was 1% in UC patients and 7% in CD patients [Figure 4]. The risk of perianal surgery was higher in males than in females (HR 1.84 [95% CI 1.42–2.40]). There was no negative trend in the cumulative incidence or risk of perianal surgery over time [Table 4, and Supplementary Table 2]. VEO-IBD was associated with a decreased risk of perianal surgery compared with the older age groups [Table 4, and Supplementary Table 3]. Figure 4. View largeDownload slide Cumulative incidence of perianal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.07]; [C] calendar year at first IBD diagnosis [p-value 0.85]; [D] gender [p-value < 0.001]. Figure 4. View largeDownload slide Cumulative incidence of perianal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.07]; [C] calendar year at first IBD diagnosis [p-value 0.85]; [D] gender [p-value < 0.001]. 3.4. Sensitivity analysis using the IBD diagnoses during the past 5 years Given the shift in IBD diagnoses during follow-up, a sensitivity analysis was performed using the IBD diagnoses given during the past 5 years of follow-up instead of the IBD diagnosis at study entry. This analysis revealed that the risks of intestinal resection, colectomy, and perianal surgery, and type of IBD, were not significantly different from the original analyses [data not shown]. 3.5. Surgery during childhood The surgical burden during childhood [before the age of 18 years] is illustrated in Supplementary Table 4 and Supplementary Figure 1, available as Supplementary data at ECCO-JCC online. In 4695 children, 245 [5%] intestinal resections were performed before the age of 18, including 100 [2%] colectomies. Calendar year of IBD diagnosis was not associated with the risk of abdominal surgery. Perianal surgery was performed in 154 [3%] children, and the risk of perianal surgery was higher in boys (HR 1.95 [95% CI 1.37–2.79]) [Supplementary Table 5, available as Supplementary data at ECCO-JCC online]. 4. Discussion In a large nationwide cohort of 4695 incident patients with childhood-onset IBD in Sweden during 13 years, the cumulative incidences of intestinal surgery during childhood and early adulthood [up to the age of 31 years] was lower than previously reported. Patients diagnosed with VEO-IBD were at the lowest risk of intestinal as well as perianal surgery. The risk of IBD surgery did not change during the studied period. In our study, the 3-year cumulative incidence of intestinal surgery in patients with childhood-onset CD was 7%, which is lower than most previously reported estimates. Population-based cohorts have presented proportions ranging from 5% to 20% after 3 years [Supplementary Table 1].1,7,18 Similarly, the 3-year cumulative incidence of colectomy in patients with childhood-onset UC in our study was lower than in previous studies [4%], where the corresponding percentages were between 7% and 15%.1,17 The present study is the most recent one, and differences in the cumulative risk of intestinal surgery compared with older studies could to some extent reflect the increased use of biological therapy in children over recent years, as it has been shown that the surgical rates are lower in responders to treatment with anti-TNF.35 Unfortunately, data on medical therapy in the present study were not complete and therefore not included in the regression analyses. Sweden has a high and increasing incidence of childhood-onset IBD,4,16 and it is possible that a childhood-onset core phenotype is diluted over time by patients characterised by a less severe disease course. This might explain the relatively low proportion of patients in need of surgical treatment in our cohort. The differences in the cumulative incidence of surgery could possibly be explained by inclusion criteria. In the present study, we used a validated method to capture all IBD cases, and required two IBD episodes in outpatient or inpatient care.31 The Canadian study used a validated algorithm including both outpatient and inpatient visits,1 whereas the Danish study was restricted to inpatient data, which may explain their high proportion of surgery despite the inclusion of patients in recent years.20 4.1. Year of diagnosis The present study showed no convincing decrease in the need for surgical treatment in patients with childhood-onset IBD. In one multivariable adjusted model, there was a significant decrease in the need for intestinal resection for those diagnosed in 2010 to 2014, compared with those diagnosed in 2002 to 2004. However, no such trend was evident in cumulative incidence of intestinal resection, either for the entire cohort or in the analyses stratified for type of IBD. In addition, calendar year was not associated with the risk of colectomy or perianal surgery. Previous studies have shown conflicting results. Canadian data from 1994–2004 showed a a decrease in the need for surgery in patients with childhood-onset CD, whereas no such trend was observed in patients with childhood-onset UC.24 A Danish study showed a decreased need for surgery in patients with childhood-onset UC in parallell to an increasing use of anti-TNF from 2005 and onward, whereas no such trend was seen in patients with childhood-onset CD.20 4.2. Age at diagnosis It has been discussed whether age at diagnosis is associated with different phenotypes of IBD. Our cohort consisted of 5% with VEO-IBD [< 6 years of age at diagnosis]. In previous studies with upper age limitis of 15 to 18 years, the proportions of children with VEO-IBD have varied between 3% and 15%.36–41 Children with VEO-IBD have been shown to have an even more extensive disease than patients afflicted in later childhood.37 Nevertheless, our data showed that the cumulative incidences and risks of intestinal and perianal surgery increased with age at diagnosis, which is similar to previous results.1,42,43 There are several possible explanations of the association of VEO-IBD and low risk of surgery in the present study. One possibility could be that in younger children disease course was milder or responded better to medical therapy than in those diagnosed at an older age, and therefore did not require surgical treatment.44 Another explanation could be a hesistant attitude among the treating physicians, who may be more willing to perform surgery in older children and adults. Older children who do not respond to medical therapy might be more urgently directed towards abdominal surgery in order to catch up on growth and puberty. 4.3. Gender Male gender was associated with a greater risk of colectomy and perianal surgery. The data regarding gender and colectomy are contradictory,17,23 and have not always been reported.7,20 Similar to our results, two studies found an increased risk of colectomy in males with paediatric-onset UC,21,45 and hypothesised that surgeons might be more hesistant to perform abdominal surgery in young women due to the risk of adhesions.21 However, more research is needed to confirm this association and explore potential mechanisms. The information regarding gender and perianal surgery is sparse, even though perianal disease has been described to be more common in boys.42,43 4.4. Strengths and limitations The greatest advantage with this study is its nationwide nature, with a complete registration of in-hospital as well as outpatient visits, and the validated method of capturing the IBD cases.31 Despite available data from earlier years, we chose to restrict the study population to 2002–2014 [1 year after the introduction of outpatient visits in the patient register] to be sure to capture only incident cases. This is the largest European population-based study of patients with childhood-onset IBD,16,18,20,38 and one of two with nationwide coverage.20 Given the current study’s national setting, we believe that the results are generalisable to countries with a similar population and health care system. However, the low cumulative risk of IBD surgery in Sweden could partly be explained by the publicly funded Swedish health care system, available to all residents, and to local traditions. It should be noted that the results may not be applicable to nations with other health care and medical insurance systems. It is possible that the threshold for performing surgery is different in countries where a health care unit’s financial results are dependent on the number of surgeries performed and/or where there is limited funding for medical therapies such as biologics. The main limitation of this study is the lack of detailed clinical information, and the inability to describe the extent of the disease in a reliable way. The clinical considerations, including disease activity and the indication for surgery, and the effects of given medications, are all unknown. In order to avoid introducing differential misclassification, the categorisation of IBD was based on the knowledge received at study entry. As a consequence, there is a mixture of diagnoses in the different groups which is reflected in the number of children with UC treated for perianal diseases. However, the misclassification was independent of IBD class, and using information on IBD diagnoses during the past 5 years of follow-up, instead of the initial IBD diagnosis, did not change the results. In conclusion, this nationwide cohort study of childhood-onset IBD showed that the cumulative risk of surgery was relatively low, but did not change over the studied time. The lowest risk of intestinal and perianal surgery was seen in those aged < 6 years at IBD diagnosis. Funding This work was supported by the Bengt Ihre research fellowship, Swedish Medical Society, Karolinska Institutet Foundations, Stockholm County Council, Jane and Dan Olsson Foundation, Kempe Carlgrenska Foundation, Mjölkdroppen Foundation, and Swedish Research Council, and by grants provided by the Stockholm County Council [ALF project]. None of the funders had any influence on this study. Conflict of Interest The authors declare no conflict of interest. Author Contributions All authors designed the study. The data were acquired by OO. CN and MB performed the statistical analysis. The manuscript was drafted by CN and OO. All authors participated in the interpretation of the data and approved the final manuscript. Supplementary Data Supplementary data to this article can be found at ECCO-JCC online. References 1. Benchimol EI, Mack DR, Nguyen GCet al.   Incidence, outcomes, and health services burden of very early onset inflammatory bowel disease. Gastroenterology  2014; 147: 803– 13.e7; quiz e14–5. Google Scholar CrossRef Search ADS PubMed  2. Benchimol EI, Fortinsky KJ, Gozdyra P, Van den Heuvel M, Van Limbergen J, Griffiths AM. Epidemiology of pediatric inflammatory bowel disease: a systematic review of international trends. Inflamm Bowel Dis  2011; 17: 423– 39. Google Scholar CrossRef Search ADS PubMed  3. Virta LJ, Saarinen MM, Kolho KL. Inflammatory bowel disease incidence is on the continuous rise among all paediatric patients except for the very young: a nationwide registry-based study on 28-year follow-up. J Crohns Colitis  2017; 11: 150– 6. Google Scholar CrossRef Search ADS PubMed  4. Malmborg P, Grahnquist L, Lindholm J, Montgomery S, Hildebrand H. Increasing incidence of paediatric inflammatory bowel disease in northern Stockholm County, 2002-2007. J Pediatr Gastroenterol Nutr  2013; 57: 29– 34. Google Scholar CrossRef Search ADS PubMed  5. Everhov ÅH, Olén O, Ludvigsson JF. Editorial: importance of definition of inflammatory bowel disease and an increased incidence in children. Aliment Pharmacol Ther  2017; 45: 1369– 70. Google Scholar CrossRef Search ADS PubMed  6. Vind I, Riis L, Jess Tet al.  ; DCCD study group. Increasing incidences of inflammatory bowel disease and decreasing surgery rates in Copenhagen City and County, 2003-2005: a population-based study from the Danish Crohn colitis database. Am J Gastroenterol  2006; 101: 1274– 82. Google Scholar CrossRef Search ADS PubMed  7. Adamiak T, Walkiewicz-Jedrzejczak D, Fish Det al.   Incidence, clinical characteristics, and natural history of pediatric IBD in Wisconsin: a population-based epidemiological study. Inflamm Bowel Dis  2013; 19: 1218– 23. Google Scholar CrossRef Search ADS PubMed  8. Müller KE, Lakatos PL, Arató Aet al.  ; Hungarian IBD Registry Group [HUPIR]. Incidence, Paris classification, and follow-up in a nationwide incident cohort of pediatric patients with inflammatory bowel disease. J Pediatr Gastroenterol Nutr  2013; 57: 576– 82. Google Scholar CrossRef Search ADS PubMed  9. Auvin S, Molinié F, Gower-Rousseau Cet al.   Incidence, clinical presentation and location at diagnosis of pediatric inflammatory bowel disease: a prospective population-based study in northern France [1988-1999]. J Pediatr Gastroenterol Nutr  2005; 41: 49– 55. Google Scholar CrossRef Search ADS PubMed  10. Turner D, Levine A, Escher JCet al.  ; European Crohn’s and Colitis Organisation; European Society for Paediatric Gastroenterology, Hepatology, and Nutrition. Management of paediatric ulcerative colitis: joint ECCO and ESPGHAN evidence-based consensus guidelines. J Pediatr Gastroenterol Nutr  2012; 55: 340– 61. Google Scholar CrossRef Search ADS PubMed  11. Turner D, Travis SP, Griffiths AMet al.  ; European Crohn’s and Colitis Organisation; Porto IBD Working Group, European Society of Paediatric Gastroenterology, Hepatology, and Nutrition. Consensus for managing acute severe ulcerative colitis in children: a systematic review and joint statement from ECCO, ESPGHAN, and the Porto IBD Working Group of ESPGHAN. Am J Gastroenterol  2011; 106: 574– 88. Google Scholar CrossRef Search ADS PubMed  12. Ruemmele FM, Veres G, Kolho KLet al.  ; European Crohn’s and Colitis Organisation; European Society of Paediatric Gastroenterology, Hepatology and Nutrition. Consensus guidelines of ECCO/ESPGHAN on the medical management of paediatric Crohn’s disease. J Crohns Colitis  2014; 8: 1179– 207. Google Scholar CrossRef Search ADS PubMed  13. Amil-Dias J, Kolacek S, Turner Det al.  ; IBD Working Group of ESPGHAN [IBD Porto Group]. Surgical management of Crohn disease in children: guidelines from the paediatric IBD Porto Group of ESPGHAN. J Pediatr Gastroenterol Nutr  2017; 64: 818– 35. Google Scholar CrossRef Search ADS PubMed  14. Van Limbergen J, Russell RK, Drummond HEet al.   Definition of phenotypic characteristics of childhood-onset inflammatory bowel disease. Gastroenterology  2008; 135: 1114– 22. Google Scholar CrossRef Search ADS PubMed  15. Jakobsen C, Bartek JJr, Wewer Vet al.   Differences in phenotype and disease course in adult and paediatric inflammatory bowel disease–a population-based study. Aliment Pharmacol Ther  2011; 34: 1217– 24. Google Scholar CrossRef Search ADS PubMed  16. Malmborg P, Grahnquist L, Ideström Met al.   Presentation and progression of childhood-onset inflammatory bowel disease in Northern Stockholm County. Inflamm Bowel Dis  2015; 21: 1098– 108. Google Scholar CrossRef Search ADS PubMed  17. Gower-Rousseau C, Dauchet L, Vernier-Massouille Get al.   The natural history of pediatric ulcerative colitis: a population-based cohort study. Am J Gastroenterol  2009; 104: 2080– 8. Google Scholar CrossRef Search ADS PubMed  18. Vernier-Massouille G, Balde M, Salleron Jet al.   Natural history of pediatric Crohn’s disease: a population-based cohort study. Gastroenterology  2008; 135: 1106– 13. Google Scholar CrossRef Search ADS PubMed  19. Pigneur B, Seksik P, Viola Set al.   Natural history of Crohn’s disease: comparison between childhood- and adult-onset disease. Inflamm Bowel Dis  2010; 16: 953– 61. Google Scholar CrossRef Search ADS PubMed  20. Larsen MD, Qvist N, Nielsen J, Kjeldsen J, Nielsen RG, Nørgård BM. Use of anti-TNFα agents and time to first-time surgery in paediatric patients with ulcerative colitis and Crohn’s disease. J Crohns Colitis  2016; 10: 650– 6. Google Scholar CrossRef Search ADS PubMed  21. Malaty HM, Abraham BP, Mehta S, Garnett EA, Ferry GD. The natural history of ulcerative colitis in a pediatric population: a follow-up population-based cohort study. Clin Exp Gastroenterol  2013; 6: 77– 83. Google Scholar CrossRef Search ADS PubMed  22. Svensson M, Lindberg E, Ludvigsson JF. Biologiska läkemedel etablerad terapi för barn med IBD i Sverige (Treatment with biological drugs is an established method in Swedish children with inflammatory bowel disease). Lakartidningen  2017; 114: piiD71X 23. Aloi M, D’Arcangelo G, Pofi Fet al.   Presenting features and disease course of pediatric ulcerative colitis. J Crohns Colitis  2013; 7: e509– 15. Google Scholar CrossRef Search ADS PubMed  24. Benchimol EI, Guttmann A, To T, Rabeneck L, Griffiths AM. Changes to surgical and hospitalization rates of pediatric inflammatory bowel disease in Ontario, Canada [1994-2007]. Inflamm Bowel Dis  2011; 17: 2153– 61. Google Scholar CrossRef Search ADS PubMed  25. Debruyn JC, Soon IS, Hubbard J, Wrobel I, Panaccione R, Kaplan GG. Nationwide temporal trends in incidence of hospitalization and surgical intestinal resection in pediatric inflammatory bowel diseases in the United States from 1997 to 2009. Inflamm Bowel Dis  2013; 19: 2423– 32. Google Scholar CrossRef Search ADS PubMed  26. Sandberg KC, Davis MM, Gebremariam A, Adler J. Increasing hospitalizations in inflammatory bowel disease among children in the United States, 1988-2011. Inflamm Bowel Dis  2014; 20: 1754– 60. Google Scholar CrossRef Search ADS PubMed  27. Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, Ekbom A. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol  2009; 24: 659– 67. Google Scholar CrossRef Search ADS PubMed  28. Nyrén O, McLaughlin JK, Gridley Get al.   Cancer risk after hip replacement with metal implants: a population-based cohort study in Sweden. J Natl Cancer Inst  1995; 87: 28– 33. Google Scholar CrossRef Search ADS PubMed  29. Ludvigsson JF, Andersson E, Ekbom Aet al.   External review and validation of the Swedish national inpatient register. BMC Public Health  2011; 11: 450. Google Scholar CrossRef Search ADS PubMed  30. Ludvigsson JF, Almqvist C, Bonamy AKet al.   Registers of the Swedish total population and their use in medical research. Eur J Epidemiol  2016; 31: 125– 36. Google Scholar CrossRef Search ADS PubMed  31. Jakobsson GL, Sternegård E, Olén Oet al.   Validating inflammatory bowel disease [IBD] in the Swedish National Patient Register and the Swedish Quality Register for IBD [SWIBREG]. Scand J Gastroenterol  2017; 52: 216– 21. Google Scholar CrossRef Search ADS PubMed  32. Ludvigsson JF, Büsch K, Olén Oet al.   Prevalence of paediatric inflammatory bowel disease in Sweden: a nationwide population-based register study. BMC Gastroenterol  2017; 17: 23. Google Scholar CrossRef Search ADS PubMed  33. Neovius M, Sundström A, Simard Jet al.  ; ARTIS Study Group. Small-area variations in sales of TNF inhibitors in Sweden between 2000 and 2009. Scand J Rheumatol  2011; 40: 8– 15. Google Scholar CrossRef Search ADS PubMed  34. Lin DY, Wei LJ. The robust inference for the Cox proportional hazards model. J Am Stat Assoc  1989; 84: 1074– 8. Google Scholar CrossRef Search ADS   35. Crombé V, Salleron J, Savoye Get al.   Long-term outcome of treatment with infliximab in pediatric-onset Crohn’s disease: a population-based study. Inflamm Bowel Dis  2011; 17: 2144– 52. Google Scholar CrossRef Search ADS PubMed  36. Sawczenko A, Sandhu BK. Presenting features of inflammatory bowel disease in Great Britain and Ireland. Arch Dis Child  2003; 88: 995– 1000. Google Scholar CrossRef Search ADS PubMed  37. Aloi M, Lionetti P, Barabino Aet al.  ; SIGENP IBD Group. Phenotype and disease course of early-onset pediatric inflammatory bowel disease. Inflamm Bowel Dis  2014; 20: 597– 605. Google Scholar CrossRef Search ADS PubMed  38. Gupta N, Cohen SA, Bostrom AGet al.   Risk factors for initial surgery in pediatric patients with Crohn’s disease. Gastroenterology  2006; 130: 1069– 77. Google Scholar CrossRef Search ADS PubMed  39. Paul T, Birnbaum A, Pal DKet al.   Distinct phenotype of early childhood inflammatory bowel disease. J Clin Gastroenterol  2006; 40: 583– 6. Google Scholar CrossRef Search ADS PubMed  40. Heyman MB, Kirschner BS, Gold BDet al.   Children with early-onset inflammatory bowel disease [IBD]: analysis of a pediatric IBD consortium registry. J Pediatr  2005; 146: 35– 40. Google Scholar CrossRef Search ADS PubMed  41. Bequet E, Sarter H, Fumery Met al.   Incidence and phenotype at diagnosis of very-early-onset compared with later-onset paediatric inflammatory bowel disease: a population-based study [1988–2011]. J Crohns Colitis  2017; 11: 519– 26. Google Scholar PubMed  42. Adler J, Dong S, Eder SJet al.   Perianal Crohn’s disease in a large multicenter pediatric collaborative. J Pediatr Gastroenterol Nutr  2017; 64: 117– 24. Google Scholar CrossRef Search ADS   43. Zwintscher NP, Shah PM, Argawal Aet al.   The impact of perianal disease in young patients with inflammatory bowel disease. Int J Colorectal Dis  2015; 30: 1275– 9. Google Scholar CrossRef Search ADS PubMed  44. Hyams J, Crandall W, Kugathasan Set al.  ; REACH Study Group. Induction and maintenance infliximab therapy for the treatment of moderate-to-severe Crohn’s disease in children. Gastroenterology  2007; 132: 863– 73. Google Scholar CrossRef Search ADS PubMed  45. Rinawi F, Assa A, Eliakim Ret al.   Risk of colectomy in patients with pediatric-onset ulcerative colitis. J Pediatr Gastroenterol Nutr  2017; 65: 410– 5. Google Scholar CrossRef Search ADS PubMed  Copyright © 2017 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Crohn's and Colitis Oxford University Press

Surgical Treatment in Childhood-onset Inflammatory Bowel Disease–A Nationwide Register-based Study of 4695 Incident Patients in Sweden 2002-2014

Loading next page...
 
/lp/ou_press/surgical-treatment-in-childhood-onset-inflammatory-bowel-disease-a-z713qU8wSI
Publisher
Oxford University Press
Copyright
Copyright © 2017 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
1873-9946
eISSN
1876-4479
D.O.I.
10.1093/ecco-jcc/jjx132
Publisher site
See Article on Publisher Site

Abstract

Abstract Background and Aims The incidence of childhood-onset [< 18 years] inflammatory bowel disease [IBD] is increasing worldwide, and some studies suggest that it represents a more severe disease phenotype. Few nationwide, population-based studies have evaluated the surgical burden in patients with childhood-onset IBD, and whether the improved medical treatment has influenced the need for gastrointestinal surgery. The aim was to examine whether the surgical treatment at any age of patients with childhood-onset IBD has changed over time. Methods In a nationwide cohort study we identified 4695 children [< 18 years] diagnosed with incident IBD in 2002–2014 through the Swedish Patient Register [ulcerative colitis: n = 2295; Crohn’s disease: n = 2174; inflammatory bowel disease-unclassified: n = 226]. Abdominal [intestinal resections and colectomies] and perianal surgeries were identified through the Swedish Patient Register. The cumulative incidences of surgeries were calculated using the Kaplan-Meier method. Results In the cohort, 44% were females and 56% males. The median age at inflammatory bowel disease diagnosis was 15 years and the maximum age at end of follow-up was 31 years. The 3-year cumulative incidence of intestinal surgery was 5% in patients with ulcerative colitis and 7% in patients with Crohn’s disease, and lower in children aged < 6 years at inflammatory bowel disease diagnosis [3%] than in those aged 15–17 years at diagnosis [7%]. Calendar period of inflammatory bowel disease diagnosis was not associated with risk of surgery. Conclusion Over the past 13 years, the risk of surgery in childhood-onset inflammatory bowel disease has remained unchanged. Children, inflammatory bowel disease, surgery 1. Introduction The incidence of inflammatory bowel disease [IBD] in children is increasing worldwide.1–3 Data from Sweden and other Nordic countries have shown a similar trend,3–6 with a higher incidence of childhood-onset IBD than in other parts of the world.7–9 In 2003 to 2013, the incidence rate of childhood-onset IBD in Sweden was 17/100000 in girls and 20/100000 in boys.5 In children with ulcerative colitis [UC], surgery is needed in acute severe colitis not responding to medical therapy, and elective colectomies are performed in children with steroid-dependent disease, poor response to medical treatment, or delay of growth and/or puberty.10,11 In addition to these indications, children with Crohn’s disease [CD] require surgical treatment for strictures and fistulas.12,13 The childhood-onset IBD phenotype is characterised by more extensive disease location.14–16 Studies also suggest that childhood-onset IBD represents a more severe phenotype associated with increased use of immunomodulatory therapy and need for surgical treatment.17–19 More recent population-based studies have shown a cumulative probability of surgical treatment after 3–5 years ranging from 5–34% in children with CD and 7–20% in children with UC [Supplementary Table 1, available as Supplementary data at ECCO-JCC online].1,7,16–18,20,21 The introduction of immunomodulators [e.g. azathioprine] during the 1990s and inhibitors of tumour necrosis factor [anti-TNF] at the beginning of the 21st century has evoked hope that the need of surgical treatment would decrease.22 Whereas some studies have shown a decreased need of surgery after introduction of anti-TNFs,20,23 other studies have shown an increasing rate of intestinal resections in children with CD.24–26 Hence, it is still unclear whether the need for surgical treatment of patients with childhood-onset IBD has decreased in recent years. In order to shed further light on this question, a population-based nationwide register study was performed including all children diagnosed with IBD between 2002 and 2014 in Sweden. The aim of this study was to describe the need for surgical treatment in a modern cohort of patients with childhood-onset IBD during childhood and early adulthood. 2. Materials and Methods 2.1. Study design and setting In this population-based cohort, we included all Swedish children diagnosed with IBD before their 18th birthday during 2002 to 2014. In Sweden, paediatric gastroenterologists treat children with IBD during childhood until 18 years of age. When in need of surgery, children with IBD may be treated by paediatric surgeons as well as by general surgeons, depending on local tradition. Since 1947, residents of Sweden have been assigned a unique personal identifier including information on year, month, and day of birth, as well as sex of the person. This national personal identity number is used in all official registers and documents, also enabling linkage of registries.27 This nationwide cohort study was based on data from virtually complete Swedish nationwide registers of hospitalisations, causes of death, and emigration.28–30 Swedish inpatient diagnoses of IBD have been registered in the National Inpatient register since January 1, 1964 [increasing coverage over the years] and with nationwide full coverage since 1987.29 In addition, all Swedish IBD diagnoses in specialist [non-primary] outpatient care have been registered in the Outpatient register since January 1, 2001. 2.2. Study population In order to restrict the study population to incident cases of IBD in outpatient or inpatient care, we identified all individuals in Sweden with a first diagnosis of IBD in the Swedish patient register [inpatient or non-primary outpatient care] from January 1, 2002 until December 31, 2014 [ie children with an IBD diagnosis in 2001 or earlier were excluded]. We used the following International Classification of Diseases [ICD] codes: CD: 563,00 [ICD8]; 555 [ICD9]; K50 [ICD10]. UC: 563,10; 569,02 [ICD8]; 556 [ICD9]; K51 [ICD10], and indeterminate colitis [IBD-U]: K52.3 [ICD10]. We restricted our IBD population to only those children who had at least two IBD diagnoses recorded by a physician on two separate occasions, in order to minimise the risk of potential misclassification,31,32 and follow-up time was calculated from the first IBD diagnosis. To decrease the risk of introducing bias, given the variations in follow-up time, type of IBD was specified at study inclusion for all analyses of future surgery. Since two IBD events were required for inclusion, some children received a mix of codes for UC, CD, or the code for IBD-U, and they were all categorised as IBD-U. However, in the descriptive statistics of the study population at end of follow-up, we used a different definition of UC, CD, and IBD-U. Because different IBD diagnoses might be documented during a patient’s medical history, due to a colitis that is difficult to categorise as either UC or CD or simply due to an incorrect registration by mistake, we classified patients with a mix of codes during the past 5 years for UC, CD, or IBD-U during follow-up as IBD-U. Moreover, patients who only shifted between UC and CD [and vice versa] and who had been diagnosed with only UC or CD the past 5 years of follow-up, were classified as UC or CD in our end of follow-up classification. The same classification was also used in a sensitivity analysis calculating the risk of intestinal resection, colectomy, and perianal surgery. 2.3. IBD surgery Data regarding IBD surgery were extracted from inpatient care, and procedures were coded using the NOMESCO Classification of Surgical Procedures. Surgeries performed before IBD diagnosis and during follow-up were reported separately. In addition to describing risk of surgery through the entire follow-up [up to the 31st birthday], we also calculated risk of surgery during childhood [before the 18th birthday]. The following codes were included in the definition of IBD surgery: Colectomy [JFH], Segmental bowel resection [Small bowel: JFB00, JFB01; Large bowel: JFB20, JFB21, JFB30, JFB31, JFB40, JFB 41, JFB43, JFB 44, JFB46, JFB47, JFB50, JFB51, JFB53, JFB 54, JFB60, JFB61, JFB63; Unspecified bowel resection: JFB33, JFB34, JFB96, JFB97], Strictureplasty [JFA60], and Perianal surgery [JHA00, JHD20, JHD30, JHD33, JHD50, JHD60, JHD63, JHW96]. For reconstructive surgery, the following codes were used for ileorectal anastomosis depending on whether the procedure was done in conjunction or following the colectomy: JFH00, JFH01, and JFH00, JFH01, JFC40, JFC41, JFG26, JFG29. Likewise, the following codes for ileal pouch anal anastomosis were used: JFH30, JFH33, and JFH30, JFH33, JGB50, JGB60, JGB61. 2.4. Exposure to thiopurines and anti-TNF The Swedish Prescribed drug register started on July 1, 2005 and contains all prescribed drug dispensation data, but not medications dispensed in hospital or bought over the counter. Nationwide and complete information was obtained on the use of the thiopurines Azathioprine [L04AX01] and Mercaptopurine [L01BB02]. The following ATC codes were used to identify anti-TNF treatment; Adalimumab [L04AB04 [L04AA17 before 2008]], Infliximab [L04AB02 [L04AA12 before 2008]], Golimumab [L04AB06], and Vedolizumab [L04AA33]. Whereas the coverage of adalimumab is virtually complete, infliximab is administered as an infusion in the hospital setting, and not all infliximab doses can be found in the Prescribed drug register.33 Ever prescriptions of thiopurines, anti-TNF, or neither were evaluated at the end of follow-up. 2.5. Extra-intestinal manifestations The following diagnostic codes were used to identify presence of extra-intestinal manifestations in the skin [ICD9: 695C, ICD10: L52, L88, L98.2], eyes [ICD9: 364, ICD10: H20], or joints [ICD9: 713B, 720A, 720C, 720W, 720X, and ICD10: M07.4, M07.5, M07.6, M09.1, M09.2, M45, M460, M461, M468, M469, M139, M255] at end of follow-up. The corresponding codes for primary sclerosing cholangitis were 576B [ICD9] and K83.0 [ICD10]. 2.6. Statistical methods Cumulative and relative risks were computed for three separate outcomes: [i] intestinal surgery, [ii] colectomy, and [iii] perianal surgery. When the outcome of interest was intestinal surgery, the children were followed from study entry until the date of intestinal resection (colectomy, bowel resection, or strictureplasty [irrespective if perianal surgery had been performed prior to that date]), death, emigration or December 31, 2014, whichever came first. Similarly, in the analyses of colectomy, the children were followed from study entry until the date of colectomy [irrespective of whether other IBD surgery had been performed], death, emigration or December 31, 2014. Children in whom a colectomy was performed before study entry were excluded from these analyses. In the analyses of perianal surgery, the children were followed from study entry until the date of perianal surgery [irrespective of whether intestinal resection had been performed], death, emigration or December 31, 2014. The cumulative incidences of surgery were estimated using the Kaplan-Meier method. Potential differences in time to surgery by patient characteristics were tested using the log-rank test. Cox regression was used to estimate crude and adjusted hazard ratios of surgery with robust estimates for the standard errors.34 The proportionality of the hazard functions across covariate patterns was tested with Schoenfeld’s residuals, and there was no sign of non-proportionality. The multivariable model was adjusted for sex, age at first IBD diagnosis (< 6 [very early onset IBD; VEO-IBD], 6–9, 10–14, 15–17 years], year of first IBD diagnosis [2002–2004, 2005–2009, 2010–2014], and type of IBD [UC, CD, and IBD-U]. The analyses were carried out with Stata 14 [StataCorp, TX, USA]. Statistical tests were two-sided, and statistical significance was set at α = 0.05. The researchers only had access to de-identified data. The study had been approved by the Ethical review board in Stockholm [DNR of ethical approvals including amendments: 2007/785-31/5; 2015/1030–32]. 3. Results A total of 45626 unique patients diagnosed with IBD in 2002 to 2014 were identified. Of those, 40931 were excluded due to age ≥ 18 years at IBD diagnosisn leaving 4695 children in the final cohort. Table 1 describes the basic characteristics of the cohort. Most children were diagnosed with UC [n = 2295, 49%] or CD [n = 2174, 46%], and 226 were classified as IBD-U [5%]. The distribution of IBD diagnoses was similar in girls and boys, as well as in different age categories, but IBD-U was more common in the latest time period. The mean age at first IBD diagnosis was 14 years [median 15 years] and almost half of the children [47%] were between 15 and 17 years at diagnosis [Figure 1]. The median age at first surgery was 17 years [range 0–28 years] and at end of follow-up 20 years [range 0–31 years][Figure 1]. Table 1. Incident cases of childhood-onset [< 18 years] inflammatory bowel disease [IBD] in Sweden 2002–2014. Patient characteristics at time of first IBD diagnosis.   UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Totalb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Girls  1056 [46]  914 [42]  94 [42]  2064 [44]  Boys  1239 [54]  1260 [58]  132 [58]  2631 [56]  Age at IBD diagnosis           Mean age [SD]  14 [4]  14 [3]  14 [4]  14 [4]   Median age [range]  15 [0–18]  14 [0–18]  15 [0–18]  15 [0–18]   < 6 years  119 [5]  85 [4]  17 [8]  221 [5]   6–9 years  203 [9]  205 [9]  16 [7]  424 [9]   10–14 years  836 [36]  947 [44]  83 [37]  1866 [40]   15–17 years  1137 [50]  937 [43]  110 [49]  2184 [47]  Year of IBD diagnosis           2002–2004  551 [24]  526 [24]  5 [2]  1082 [23]   2005–2009  1011 [44]  901 [41]  40 [18]  1952 [42]   2010–2014  733 [32]  747 [34]  181 [80]  1661 [35]  Before IBD diagnosis           Colectomy  2 [0.1]  2 [0]  -  4 [0]   Large bowel resectionc  -  12 [1]  -  12 [0]   Small bowel resection  2 [0]  3 [0]  -  5 [0]   Perianal surgery  8 [0]  101 [5]  6 [3]  115 [2]    UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Totalb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Girls  1056 [46]  914 [42]  94 [42]  2064 [44]  Boys  1239 [54]  1260 [58]  132 [58]  2631 [56]  Age at IBD diagnosis           Mean age [SD]  14 [4]  14 [3]  14 [4]  14 [4]   Median age [range]  15 [0–18]  14 [0–18]  15 [0–18]  15 [0–18]   < 6 years  119 [5]  85 [4]  17 [8]  221 [5]   6–9 years  203 [9]  205 [9]  16 [7]  424 [9]   10–14 years  836 [36]  947 [44]  83 [37]  1866 [40]   15–17 years  1137 [50]  937 [43]  110 [49]  2184 [47]  Year of IBD diagnosis           2002–2004  551 [24]  526 [24]  5 [2]  1082 [23]   2005–2009  1011 [44]  901 [41]  40 [18]  1952 [42]   2010–2014  733 [32]  747 [34]  181 [80]  1661 [35]  Before IBD diagnosis           Colectomy  2 [0.1]  2 [0]  -  4 [0]   Large bowel resectionc  -  12 [1]  -  12 [0]   Small bowel resection  2 [0]  3 [0]  -  5 [0]   Perianal surgery  8 [0]  101 [5]  6 [3]  115 [2]  SD, standard deviation. aType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, ie either K52.3 or a mixture of UC and CD diagnoses. bRow percent. cIncluding ileocaecal resection. View Large Figure 1. View largeDownload slide Age-distribution at [A] first IBD diagnosis, [B] first IBD surgery, [C] end of follow-up [first surgery, death, emigration or December 31, 2014] in 4695 incident cases of childhood-onset [< 18 years] inflammatory bowel disease. IBD, inflammatory bowel disease. Figure 1. View largeDownload slide Age-distribution at [A] first IBD diagnosis, [B] first IBD surgery, [C] end of follow-up [first surgery, death, emigration or December 31, 2014] in 4695 incident cases of childhood-onset [< 18 years] inflammatory bowel disease. IBD, inflammatory bowel disease. Table 2 describes the study population at end of follow-up, stratified by type of IBD diagnosis at start of follow-up. The IBD diagnosis may change over time, and to determine the proportion of patients who received a different IBD diagnosis during follow-up, registered IBD diagnoses during the past f5 years of follow-up were also analysed. Of patients initially given a UC diagnosis, UC was the only diagnosis during the past 5 years of follow-up in 86% of the patients. The corresponding proportion in CD patients was 83%, and in patients with IBD-U 36%. Extra-intestinal manifestations were diagnosed in 8% and primary sclerosing cholangitis in 6%. The majority of patients had received medical treatment with thiopurines [60%]. Table 2. Descriptive characteristics at the end of follow-up in 4695 incident cases of childhood-onset [< 18 years] IBD in Sweden 2002–2014.   UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  End of follow-up [first IBD surgery, death, emigration, or December 31, 2014]  Follow-up in years [median, range]  6 [0–13]  6 [0–13]  3 [0–12]  6 [0–13]  Censored due to death  13 [1]  7 [0]  2 [1]  22 [1]  Censored due to emigration  20 [1]  24 [1]  1 [0]  45 [1]  Censored due to surgery  219 [10]  422 [19]  17 [8]  658 [14]  Censored due to end of follow-up  2043 [89]  1721 [79]  206 [91]  3970 [85]  End of follow-up [death, emigration, or December 31, 2014]  Follow-up in years [median, range]  7 [0–13]  7 [0–13]  3 [0–13]  7 [0–13]  IBD diagnosis past 5 years of follow-upc           UC  1837 [86]  259 [12]  51 [2]  2147 [46]   CD  326 [15]  1815 [83]  44 [2]  2185 [47]   IBD-U  132 [36]  100 [28]  131 [36]  363 [8]  Ever diagnosed with           Extra-intestinal manifestations  175 [8]  1707 [8]  12 [5]  357 [8]   Primary sclerosing cholangitis  172 [8]  90 [4]  13 [6]  275 [6]  Ever treatment with           Thiopurines  1256 [55]  1455 [67]  114 [50]  2825 [60]   Anti-TNF  301 [13]  453 [21]  22 [10]  776 [17]   Neither  1039 [45]  719 [33]  112 [50]  1870 [40]    UCa  CDa  IBD-Ua  All  n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  End of follow-up [first IBD surgery, death, emigration, or December 31, 2014]  Follow-up in years [median, range]  6 [0–13]  6 [0–13]  3 [0–12]  6 [0–13]  Censored due to death  13 [1]  7 [0]  2 [1]  22 [1]  Censored due to emigration  20 [1]  24 [1]  1 [0]  45 [1]  Censored due to surgery  219 [10]  422 [19]  17 [8]  658 [14]  Censored due to end of follow-up  2043 [89]  1721 [79]  206 [91]  3970 [85]  End of follow-up [death, emigration, or December 31, 2014]  Follow-up in years [median, range]  7 [0–13]  7 [0–13]  3 [0–13]  7 [0–13]  IBD diagnosis past 5 years of follow-upc           UC  1837 [86]  259 [12]  51 [2]  2147 [46]   CD  326 [15]  1815 [83]  44 [2]  2185 [47]   IBD-U  132 [36]  100 [28]  131 [36]  363 [8]  Ever diagnosed with           Extra-intestinal manifestations  175 [8]  1707 [8]  12 [5]  357 [8]   Primary sclerosing cholangitis  172 [8]  90 [4]  13 [6]  275 [6]  Ever treatment with           Thiopurines  1256 [55]  1455 [67]  114 [50]  2825 [60]   Anti-TNF  301 [13]  453 [21]  22 [10]  776 [17]   Neither  1039 [45]  719 [33]  112 [50]  1870 [40]  TNF, tumour necrosis factor. aType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, i.e. either K52.3 or a mixture of UC and CD-diagnoses. bRow percent. cType of IBD past 5 years was classified as UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = IBD unclassified or a mixture of UC and CD diagnoses. View Large The median time from IBD diagnosis to first surgical event [among patients who had surgery] was 3 years [range 0–13 years]. All surgical procedures performed during the entire follow-up [ie not only the first event] are described in Table 3. Any surgery was performed in 10% of UC patients and 19% of CD patients, including intestinal resections in 8% of UC patients and 12% of CD patients. In patients with UC, colectomy was the most common surgical procedure [7%], whereas segmental bowel resection and perianal surgery were the most common procedures in patients with CD [10% each]. Following a colectomy, secondary reconstructions with ileorectal anastomosis and ileal pouch anal anastomosis were more common than primary reconstructions. Table 3. Descriptive characteristics of all surgical procedures performed in 4695 incident cases of childhood-onset [< 18 years] IBD [inflammatory bowel disease] in Sweden 2002–2014 during all available follow-ups.   UCa  CDa  IBD-Ua  All    n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Treated with any surgery  219 [10]  422 [19]  17 [8]  658 [14]  Type of surgeryc           Any intestinal resection  194 [8]  264 [12]  15 [7]  473 [10]   Colectomy  169 [7]  36 [2]  10 [4]  215 [5]  Primary reconstruction           Ileorectal anastomosis  10 [0]  5 [0]  -  15 [0]   Ileal pouch-anal anastomosis  9 [0]  -  1 [0]  10 [0]  Secondary reconstruction           Ileorectal anastomosis  47 [2]  15 [1]  2 [1]  64 [1]   Ileal pouch-anal anastomosis  27 [1]  3 [0]  2 [1]  32 [1]   Segmental large bowel resection  20 [1]  207 [10]  5 [2]  232 [5]   Small bowel resection  13 [1]  44 [2]  -  57 [1]   Unspecified bowel resection  2 [0]  12 [1]  -  14 [0]   Strictureplasty  -  11 [1]  -  11 [0]   Any perianal surgery  55 [2]  206 [10]  3 [1]  264 [6]    UCa  CDa  IBD-Ua  All    n [%]  n [%]  n [%]  n [%]  Total cohortb  2295 [49]  2174 [46]  226 [5]  4695 [100]  Treated with any surgery  219 [10]  422 [19]  17 [8]  658 [14]  Type of surgeryc           Any intestinal resection  194 [8]  264 [12]  15 [7]  473 [10]   Colectomy  169 [7]  36 [2]  10 [4]  215 [5]  Primary reconstruction           Ileorectal anastomosis  10 [0]  5 [0]  -  15 [0]   Ileal pouch-anal anastomosis  9 [0]  -  1 [0]  10 [0]  Secondary reconstruction           Ileorectal anastomosis  47 [2]  15 [1]  2 [1]  64 [1]   Ileal pouch-anal anastomosis  27 [1]  3 [0]  2 [1]  32 [1]   Segmental large bowel resection  20 [1]  207 [10]  5 [2]  232 [5]   Small bowel resection  13 [1]  44 [2]  -  57 [1]   Unspecified bowel resection  2 [0]  12 [1]  -  14 [0]   Strictureplasty  -  11 [1]  -  11 [0]   Any perianal surgery  55 [2]  206 [10]  3 [1]  264 [6]  aType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, ie either K52.3 or a mixture of UC and CD diagnoses. bRow percent. cAdds up to more than the number of patients treated with surgery since several procedures could be performed in one patient on different occasions. View Large 3.1. Intestinal surgery The cumulative incidence of intestinal surgery is illustrated in Figure 2. The 3-year cumulative incidence of intestinal surgery in UC patients was 5%. The corresponding proportion in CD patients was 7%. Intestinal surgery was more common in CD patients: hazard ratio [HR] 1.53 (95% CI [confidence interval] 1.27–1.85) than in UC patients [Table 4]. The 3-year cumulative incidence of intestinal surgery was 3% in children aged < 6 years at IBD diagnosis. The corresponding proportion in children aged 15 to 17 years at IBD diagnosis was 7%. The risk of intestinal surgery was increased in patients aged 15 to 17 years at IBD diagnosis compared with the youngest age group (HR 2.67 [95% CI 1.42–5.03]). There was a trend to a decreasing risk of ever having intestinal surgery with increasing calendar year at IBD diagnosis in the multivariable analysis adjusted for sex, type of IBD, and age at IBD diagnosis, with a significantly decreased risk (HR 0.71 [95% CI 0.53–0.96] for those diagnosed in year 2010 to 2014 vs those diagnosed in 2002–2004) [Table 4]. The trend was however not evident in the cumulative incidence of intestinal surgery [Supplementary Table 2, available as Supplementary data at ECCO-JCC online; and Figure 2]. Figure 2. View largeDownload slide Cumulative incidence of intestinal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value < 0.001]; [C] calendar year at first IBD diagnosis [p-value 0.13]; [D] gender [p-value 0.07]. Figure 2. View largeDownload slide Cumulative incidence of intestinal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value < 0.001]; [C] calendar year at first IBD diagnosis [p-value 0.13]; [D] gender [p-value 0.07]. Table 4. The hazard ratios [HRs] of intestinal surgery,a colectomy, and perianal surgery in 4695 children [aged < 18 years] diagnosed with inflammatory bowel disease in Sweden in 2002 to 2014, followed through adulthood, and calculated with Cox regression.   Intestinal surgerya  Colectomyb  Perianal surgery    nc  HR4  95% CI  nc  HR4  95% CI  Casesc  HRd  95% CI  Females  191  Ref    79  Ref    79  Ref    Males  282  1.17  0.97–1.41  136  1.45  1.10–1.91  185  1.84  1.42–2.40  UCe  194  Ref    169  Ref    55  Ref    CDe  264  1.53  1.27–1.85  36  0.22  0.15–0.32  206  4.15  3.08–5.60  IBD-Ue  15  1.53  0.89–2.63  10  1.11  0.57–2.15  3  0.87  0.27–2.81  Age at IBD diagnosis                     < 6 years  10  Ref    8  Ref    4  Ref     6–9 years  26  1.31  0.63–2.72  13  0.90  0.37–2.17  21  2.48  0.85–7.23   10–14 years  186  2.04  1.07–3.85  74  1.19  0.58–2.48  115  2.72  1.00–7. 40   15–17 years  251  2.67  1.42–5.03  120  1.61  0.79–3.30  124  3.06  1.13–8.28  Year of IBD diagnosis                     2002–2004  188  Ref    86  Ref    94  Ref     2005–2009  213  0.86  0.70–1.06  94  0.84  0.61–1.15  118  0.94  0.69–1.21   2010–2014  72  0.71  0.53–0.96  35  0.82  0.52–1.28  52  1.00  0.70–1.47    Intestinal surgerya  Colectomyb  Perianal surgery    nc  HR4  95% CI  nc  HR4  95% CI  Casesc  HRd  95% CI  Females  191  Ref    79  Ref    79  Ref    Males  282  1.17  0.97–1.41  136  1.45  1.10–1.91  185  1.84  1.42–2.40  UCe  194  Ref    169  Ref    55  Ref    CDe  264  1.53  1.27–1.85  36  0.22  0.15–0.32  206  4.15  3.08–5.60  IBD-Ue  15  1.53  0.89–2.63  10  1.11  0.57–2.15  3  0.87  0.27–2.81  Age at IBD diagnosis                     < 6 years  10  Ref    8  Ref    4  Ref     6–9 years  26  1.31  0.63–2.72  13  0.90  0.37–2.17  21  2.48  0.85–7.23   10–14 years  186  2.04  1.07–3.85  74  1.19  0.58–2.48  115  2.72  1.00–7. 40   15–17 years  251  2.67  1.42–5.03  120  1.61  0.79–3.30  124  3.06  1.13–8.28  Year of IBD diagnosis                     2002–2004  188  Ref    86  Ref    94  Ref     2005–2009  213  0.86  0.70–1.06  94  0.84  0.61–1.15  118  0.94  0.69–1.21   2010–2014  72  0.71  0.53–0.96  35  0.82  0.52–1.28  52  1.00  0.70–1.47  CI, confidence interval; ref, reference. aIncluding colectomy, segmental large bowel resection, small bowel resection, unspecified bowel resection and strictureplasty. bPatients in whom colectomy was performed before study entry were excluded. cNumber of patients in whom surgery was performed. dMultivariable analyses adjusted for all listed variables. eType of IBD at diagnosis was defined by the first two diagnoses: UC = ulcerative colitis only; CD = Crohn’s disease only; IBD-U = unclassified inflammatory bowel disease, ie either K52.3 or a mixture of UC and CD diagnoses. View Large 3.2. Colectomy The 3 -year cumulative incidence of colectomy was higher in patients with UC [4%] compared with patients with CD [1%] [Figure 3]. In the multivariable model, male gender was associated with an increased risk of colectomy (HR 1.45 [95% CI 1.10–1.91]). Age at IBD diagnosis and calendar year of IBD diagnosis were not significantly associated with the risk of colectomy [Table 4, and Supplementary Tables 2 and 3, available as Supplementary data at ECCO-JCC online]. Figure 3. View largeDownload slide Cumulative incidence of colectomy during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.02]; [C] calendar year at first IBD diagnosis [p-value 0.74]; [D] gender [p-value 0.02]. Figure 3. View largeDownload slide Cumulative incidence of colectomy during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.02]; [C] calendar year at first IBD diagnosis [p-value 0.74]; [D] gender [p-value 0.02]. 3.3. Perianal surgery The 3-year cumulative incidence of perianal surgery was 1% in UC patients and 7% in CD patients [Figure 4]. The risk of perianal surgery was higher in males than in females (HR 1.84 [95% CI 1.42–2.40]). There was no negative trend in the cumulative incidence or risk of perianal surgery over time [Table 4, and Supplementary Table 2]. VEO-IBD was associated with a decreased risk of perianal surgery compared with the older age groups [Table 4, and Supplementary Table 3]. Figure 4. View largeDownload slide Cumulative incidence of perianal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.07]; [C] calendar year at first IBD diagnosis [p-value 0.85]; [D] gender [p-value < 0.001]. Figure 4. View largeDownload slide Cumulative incidence of perianal surgery during all available follow-ups in 4695 patients with childhood-onset [< 18 years] inflammatory bowel disease [IBD], stratified by [A] type of IBD [UC: ulcerative colitis, CD: Crohn’s disease, p-value < 0.001]; [B] age at first IBD diagnosis [in years, p-value 0.07]; [C] calendar year at first IBD diagnosis [p-value 0.85]; [D] gender [p-value < 0.001]. 3.4. Sensitivity analysis using the IBD diagnoses during the past 5 years Given the shift in IBD diagnoses during follow-up, a sensitivity analysis was performed using the IBD diagnoses given during the past 5 years of follow-up instead of the IBD diagnosis at study entry. This analysis revealed that the risks of intestinal resection, colectomy, and perianal surgery, and type of IBD, were not significantly different from the original analyses [data not shown]. 3.5. Surgery during childhood The surgical burden during childhood [before the age of 18 years] is illustrated in Supplementary Table 4 and Supplementary Figure 1, available as Supplementary data at ECCO-JCC online. In 4695 children, 245 [5%] intestinal resections were performed before the age of 18, including 100 [2%] colectomies. Calendar year of IBD diagnosis was not associated with the risk of abdominal surgery. Perianal surgery was performed in 154 [3%] children, and the risk of perianal surgery was higher in boys (HR 1.95 [95% CI 1.37–2.79]) [Supplementary Table 5, available as Supplementary data at ECCO-JCC online]. 4. Discussion In a large nationwide cohort of 4695 incident patients with childhood-onset IBD in Sweden during 13 years, the cumulative incidences of intestinal surgery during childhood and early adulthood [up to the age of 31 years] was lower than previously reported. Patients diagnosed with VEO-IBD were at the lowest risk of intestinal as well as perianal surgery. The risk of IBD surgery did not change during the studied period. In our study, the 3-year cumulative incidence of intestinal surgery in patients with childhood-onset CD was 7%, which is lower than most previously reported estimates. Population-based cohorts have presented proportions ranging from 5% to 20% after 3 years [Supplementary Table 1].1,7,18 Similarly, the 3-year cumulative incidence of colectomy in patients with childhood-onset UC in our study was lower than in previous studies [4%], where the corresponding percentages were between 7% and 15%.1,17 The present study is the most recent one, and differences in the cumulative risk of intestinal surgery compared with older studies could to some extent reflect the increased use of biological therapy in children over recent years, as it has been shown that the surgical rates are lower in responders to treatment with anti-TNF.35 Unfortunately, data on medical therapy in the present study were not complete and therefore not included in the regression analyses. Sweden has a high and increasing incidence of childhood-onset IBD,4,16 and it is possible that a childhood-onset core phenotype is diluted over time by patients characterised by a less severe disease course. This might explain the relatively low proportion of patients in need of surgical treatment in our cohort. The differences in the cumulative incidence of surgery could possibly be explained by inclusion criteria. In the present study, we used a validated method to capture all IBD cases, and required two IBD episodes in outpatient or inpatient care.31 The Canadian study used a validated algorithm including both outpatient and inpatient visits,1 whereas the Danish study was restricted to inpatient data, which may explain their high proportion of surgery despite the inclusion of patients in recent years.20 4.1. Year of diagnosis The present study showed no convincing decrease in the need for surgical treatment in patients with childhood-onset IBD. In one multivariable adjusted model, there was a significant decrease in the need for intestinal resection for those diagnosed in 2010 to 2014, compared with those diagnosed in 2002 to 2004. However, no such trend was evident in cumulative incidence of intestinal resection, either for the entire cohort or in the analyses stratified for type of IBD. In addition, calendar year was not associated with the risk of colectomy or perianal surgery. Previous studies have shown conflicting results. Canadian data from 1994–2004 showed a a decrease in the need for surgery in patients with childhood-onset CD, whereas no such trend was observed in patients with childhood-onset UC.24 A Danish study showed a decreased need for surgery in patients with childhood-onset UC in parallell to an increasing use of anti-TNF from 2005 and onward, whereas no such trend was seen in patients with childhood-onset CD.20 4.2. Age at diagnosis It has been discussed whether age at diagnosis is associated with different phenotypes of IBD. Our cohort consisted of 5% with VEO-IBD [< 6 years of age at diagnosis]. In previous studies with upper age limitis of 15 to 18 years, the proportions of children with VEO-IBD have varied between 3% and 15%.36–41 Children with VEO-IBD have been shown to have an even more extensive disease than patients afflicted in later childhood.37 Nevertheless, our data showed that the cumulative incidences and risks of intestinal and perianal surgery increased with age at diagnosis, which is similar to previous results.1,42,43 There are several possible explanations of the association of VEO-IBD and low risk of surgery in the present study. One possibility could be that in younger children disease course was milder or responded better to medical therapy than in those diagnosed at an older age, and therefore did not require surgical treatment.44 Another explanation could be a hesistant attitude among the treating physicians, who may be more willing to perform surgery in older children and adults. Older children who do not respond to medical therapy might be more urgently directed towards abdominal surgery in order to catch up on growth and puberty. 4.3. Gender Male gender was associated with a greater risk of colectomy and perianal surgery. The data regarding gender and colectomy are contradictory,17,23 and have not always been reported.7,20 Similar to our results, two studies found an increased risk of colectomy in males with paediatric-onset UC,21,45 and hypothesised that surgeons might be more hesistant to perform abdominal surgery in young women due to the risk of adhesions.21 However, more research is needed to confirm this association and explore potential mechanisms. The information regarding gender and perianal surgery is sparse, even though perianal disease has been described to be more common in boys.42,43 4.4. Strengths and limitations The greatest advantage with this study is its nationwide nature, with a complete registration of in-hospital as well as outpatient visits, and the validated method of capturing the IBD cases.31 Despite available data from earlier years, we chose to restrict the study population to 2002–2014 [1 year after the introduction of outpatient visits in the patient register] to be sure to capture only incident cases. This is the largest European population-based study of patients with childhood-onset IBD,16,18,20,38 and one of two with nationwide coverage.20 Given the current study’s national setting, we believe that the results are generalisable to countries with a similar population and health care system. However, the low cumulative risk of IBD surgery in Sweden could partly be explained by the publicly funded Swedish health care system, available to all residents, and to local traditions. It should be noted that the results may not be applicable to nations with other health care and medical insurance systems. It is possible that the threshold for performing surgery is different in countries where a health care unit’s financial results are dependent on the number of surgeries performed and/or where there is limited funding for medical therapies such as biologics. The main limitation of this study is the lack of detailed clinical information, and the inability to describe the extent of the disease in a reliable way. The clinical considerations, including disease activity and the indication for surgery, and the effects of given medications, are all unknown. In order to avoid introducing differential misclassification, the categorisation of IBD was based on the knowledge received at study entry. As a consequence, there is a mixture of diagnoses in the different groups which is reflected in the number of children with UC treated for perianal diseases. However, the misclassification was independent of IBD class, and using information on IBD diagnoses during the past 5 years of follow-up, instead of the initial IBD diagnosis, did not change the results. In conclusion, this nationwide cohort study of childhood-onset IBD showed that the cumulative risk of surgery was relatively low, but did not change over the studied time. The lowest risk of intestinal and perianal surgery was seen in those aged < 6 years at IBD diagnosis. Funding This work was supported by the Bengt Ihre research fellowship, Swedish Medical Society, Karolinska Institutet Foundations, Stockholm County Council, Jane and Dan Olsson Foundation, Kempe Carlgrenska Foundation, Mjölkdroppen Foundation, and Swedish Research Council, and by grants provided by the Stockholm County Council [ALF project]. None of the funders had any influence on this study. Conflict of Interest The authors declare no conflict of interest. Author Contributions All authors designed the study. The data were acquired by OO. CN and MB performed the statistical analysis. The manuscript was drafted by CN and OO. All authors participated in the interpretation of the data and approved the final manuscript. Supplementary Data Supplementary data to this article can be found at ECCO-JCC online. References 1. Benchimol EI, Mack DR, Nguyen GCet al.   Incidence, outcomes, and health services burden of very early onset inflammatory bowel disease. Gastroenterology  2014; 147: 803– 13.e7; quiz e14–5. Google Scholar CrossRef Search ADS PubMed  2. Benchimol EI, Fortinsky KJ, Gozdyra P, Van den Heuvel M, Van Limbergen J, Griffiths AM. Epidemiology of pediatric inflammatory bowel disease: a systematic review of international trends. Inflamm Bowel Dis  2011; 17: 423– 39. Google Scholar CrossRef Search ADS PubMed  3. Virta LJ, Saarinen MM, Kolho KL. Inflammatory bowel disease incidence is on the continuous rise among all paediatric patients except for the very young: a nationwide registry-based study on 28-year follow-up. J Crohns Colitis  2017; 11: 150– 6. Google Scholar CrossRef Search ADS PubMed  4. Malmborg P, Grahnquist L, Lindholm J, Montgomery S, Hildebrand H. Increasing incidence of paediatric inflammatory bowel disease in northern Stockholm County, 2002-2007. J Pediatr Gastroenterol Nutr  2013; 57: 29– 34. Google Scholar CrossRef Search ADS PubMed  5. Everhov ÅH, Olén O, Ludvigsson JF. Editorial: importance of definition of inflammatory bowel disease and an increased incidence in children. Aliment Pharmacol Ther  2017; 45: 1369– 70. Google Scholar CrossRef Search ADS PubMed  6. Vind I, Riis L, Jess Tet al.  ; DCCD study group. Increasing incidences of inflammatory bowel disease and decreasing surgery rates in Copenhagen City and County, 2003-2005: a population-based study from the Danish Crohn colitis database. Am J Gastroenterol  2006; 101: 1274– 82. Google Scholar CrossRef Search ADS PubMed  7. Adamiak T, Walkiewicz-Jedrzejczak D, Fish Det al.   Incidence, clinical characteristics, and natural history of pediatric IBD in Wisconsin: a population-based epidemiological study. Inflamm Bowel Dis  2013; 19: 1218– 23. Google Scholar CrossRef Search ADS PubMed  8. Müller KE, Lakatos PL, Arató Aet al.  ; Hungarian IBD Registry Group [HUPIR]. Incidence, Paris classification, and follow-up in a nationwide incident cohort of pediatric patients with inflammatory bowel disease. J Pediatr Gastroenterol Nutr  2013; 57: 576– 82. Google Scholar CrossRef Search ADS PubMed  9. Auvin S, Molinié F, Gower-Rousseau Cet al.   Incidence, clinical presentation and location at diagnosis of pediatric inflammatory bowel disease: a prospective population-based study in northern France [1988-1999]. J Pediatr Gastroenterol Nutr  2005; 41: 49– 55. Google Scholar CrossRef Search ADS PubMed  10. Turner D, Levine A, Escher JCet al.  ; European Crohn’s and Colitis Organisation; European Society for Paediatric Gastroenterology, Hepatology, and Nutrition. Management of paediatric ulcerative colitis: joint ECCO and ESPGHAN evidence-based consensus guidelines. J Pediatr Gastroenterol Nutr  2012; 55: 340– 61. Google Scholar CrossRef Search ADS PubMed  11. Turner D, Travis SP, Griffiths AMet al.  ; European Crohn’s and Colitis Organisation; Porto IBD Working Group, European Society of Paediatric Gastroenterology, Hepatology, and Nutrition. Consensus for managing acute severe ulcerative colitis in children: a systematic review and joint statement from ECCO, ESPGHAN, and the Porto IBD Working Group of ESPGHAN. Am J Gastroenterol  2011; 106: 574– 88. Google Scholar CrossRef Search ADS PubMed  12. Ruemmele FM, Veres G, Kolho KLet al.  ; European Crohn’s and Colitis Organisation; European Society of Paediatric Gastroenterology, Hepatology and Nutrition. Consensus guidelines of ECCO/ESPGHAN on the medical management of paediatric Crohn’s disease. J Crohns Colitis  2014; 8: 1179– 207. Google Scholar CrossRef Search ADS PubMed  13. Amil-Dias J, Kolacek S, Turner Det al.  ; IBD Working Group of ESPGHAN [IBD Porto Group]. Surgical management of Crohn disease in children: guidelines from the paediatric IBD Porto Group of ESPGHAN. J Pediatr Gastroenterol Nutr  2017; 64: 818– 35. Google Scholar CrossRef Search ADS PubMed  14. Van Limbergen J, Russell RK, Drummond HEet al.   Definition of phenotypic characteristics of childhood-onset inflammatory bowel disease. Gastroenterology  2008; 135: 1114– 22. Google Scholar CrossRef Search ADS PubMed  15. Jakobsen C, Bartek JJr, Wewer Vet al.   Differences in phenotype and disease course in adult and paediatric inflammatory bowel disease–a population-based study. Aliment Pharmacol Ther  2011; 34: 1217– 24. Google Scholar CrossRef Search ADS PubMed  16. Malmborg P, Grahnquist L, Ideström Met al.   Presentation and progression of childhood-onset inflammatory bowel disease in Northern Stockholm County. Inflamm Bowel Dis  2015; 21: 1098– 108. Google Scholar CrossRef Search ADS PubMed  17. Gower-Rousseau C, Dauchet L, Vernier-Massouille Get al.   The natural history of pediatric ulcerative colitis: a population-based cohort study. Am J Gastroenterol  2009; 104: 2080– 8. Google Scholar CrossRef Search ADS PubMed  18. Vernier-Massouille G, Balde M, Salleron Jet al.   Natural history of pediatric Crohn’s disease: a population-based cohort study. Gastroenterology  2008; 135: 1106– 13. Google Scholar CrossRef Search ADS PubMed  19. Pigneur B, Seksik P, Viola Set al.   Natural history of Crohn’s disease: comparison between childhood- and adult-onset disease. Inflamm Bowel Dis  2010; 16: 953– 61. Google Scholar CrossRef Search ADS PubMed  20. Larsen MD, Qvist N, Nielsen J, Kjeldsen J, Nielsen RG, Nørgård BM. Use of anti-TNFα agents and time to first-time surgery in paediatric patients with ulcerative colitis and Crohn’s disease. J Crohns Colitis  2016; 10: 650– 6. Google Scholar CrossRef Search ADS PubMed  21. Malaty HM, Abraham BP, Mehta S, Garnett EA, Ferry GD. The natural history of ulcerative colitis in a pediatric population: a follow-up population-based cohort study. Clin Exp Gastroenterol  2013; 6: 77– 83. Google Scholar CrossRef Search ADS PubMed  22. Svensson M, Lindberg E, Ludvigsson JF. Biologiska läkemedel etablerad terapi för barn med IBD i Sverige (Treatment with biological drugs is an established method in Swedish children with inflammatory bowel disease). Lakartidningen  2017; 114: piiD71X 23. Aloi M, D’Arcangelo G, Pofi Fet al.   Presenting features and disease course of pediatric ulcerative colitis. J Crohns Colitis  2013; 7: e509– 15. Google Scholar CrossRef Search ADS PubMed  24. Benchimol EI, Guttmann A, To T, Rabeneck L, Griffiths AM. Changes to surgical and hospitalization rates of pediatric inflammatory bowel disease in Ontario, Canada [1994-2007]. Inflamm Bowel Dis  2011; 17: 2153– 61. Google Scholar CrossRef Search ADS PubMed  25. Debruyn JC, Soon IS, Hubbard J, Wrobel I, Panaccione R, Kaplan GG. Nationwide temporal trends in incidence of hospitalization and surgical intestinal resection in pediatric inflammatory bowel diseases in the United States from 1997 to 2009. Inflamm Bowel Dis  2013; 19: 2423– 32. Google Scholar CrossRef Search ADS PubMed  26. Sandberg KC, Davis MM, Gebremariam A, Adler J. Increasing hospitalizations in inflammatory bowel disease among children in the United States, 1988-2011. Inflamm Bowel Dis  2014; 20: 1754– 60. Google Scholar CrossRef Search ADS PubMed  27. Ludvigsson JF, Otterblad-Olausson P, Pettersson BU, Ekbom A. The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research. Eur J Epidemiol  2009; 24: 659– 67. Google Scholar CrossRef Search ADS PubMed  28. Nyrén O, McLaughlin JK, Gridley Get al.   Cancer risk after hip replacement with metal implants: a population-based cohort study in Sweden. J Natl Cancer Inst  1995; 87: 28– 33. Google Scholar CrossRef Search ADS PubMed  29. Ludvigsson JF, Andersson E, Ekbom Aet al.   External review and validation of the Swedish national inpatient register. BMC Public Health  2011; 11: 450. Google Scholar CrossRef Search ADS PubMed  30. Ludvigsson JF, Almqvist C, Bonamy AKet al.   Registers of the Swedish total population and their use in medical research. Eur J Epidemiol  2016; 31: 125– 36. Google Scholar CrossRef Search ADS PubMed  31. Jakobsson GL, Sternegård E, Olén Oet al.   Validating inflammatory bowel disease [IBD] in the Swedish National Patient Register and the Swedish Quality Register for IBD [SWIBREG]. Scand J Gastroenterol  2017; 52: 216– 21. Google Scholar CrossRef Search ADS PubMed  32. Ludvigsson JF, Büsch K, Olén Oet al.   Prevalence of paediatric inflammatory bowel disease in Sweden: a nationwide population-based register study. BMC Gastroenterol  2017; 17: 23. Google Scholar CrossRef Search ADS PubMed  33. Neovius M, Sundström A, Simard Jet al.  ; ARTIS Study Group. Small-area variations in sales of TNF inhibitors in Sweden between 2000 and 2009. Scand J Rheumatol  2011; 40: 8– 15. Google Scholar CrossRef Search ADS PubMed  34. Lin DY, Wei LJ. The robust inference for the Cox proportional hazards model. J Am Stat Assoc  1989; 84: 1074– 8. Google Scholar CrossRef Search ADS   35. Crombé V, Salleron J, Savoye Get al.   Long-term outcome of treatment with infliximab in pediatric-onset Crohn’s disease: a population-based study. Inflamm Bowel Dis  2011; 17: 2144– 52. Google Scholar CrossRef Search ADS PubMed  36. Sawczenko A, Sandhu BK. Presenting features of inflammatory bowel disease in Great Britain and Ireland. Arch Dis Child  2003; 88: 995– 1000. Google Scholar CrossRef Search ADS PubMed  37. Aloi M, Lionetti P, Barabino Aet al.  ; SIGENP IBD Group. Phenotype and disease course of early-onset pediatric inflammatory bowel disease. Inflamm Bowel Dis  2014; 20: 597– 605. Google Scholar CrossRef Search ADS PubMed  38. Gupta N, Cohen SA, Bostrom AGet al.   Risk factors for initial surgery in pediatric patients with Crohn’s disease. Gastroenterology  2006; 130: 1069– 77. Google Scholar CrossRef Search ADS PubMed  39. Paul T, Birnbaum A, Pal DKet al.   Distinct phenotype of early childhood inflammatory bowel disease. J Clin Gastroenterol  2006; 40: 583– 6. Google Scholar CrossRef Search ADS PubMed  40. Heyman MB, Kirschner BS, Gold BDet al.   Children with early-onset inflammatory bowel disease [IBD]: analysis of a pediatric IBD consortium registry. J Pediatr  2005; 146: 35– 40. Google Scholar CrossRef Search ADS PubMed  41. Bequet E, Sarter H, Fumery Met al.   Incidence and phenotype at diagnosis of very-early-onset compared with later-onset paediatric inflammatory bowel disease: a population-based study [1988–2011]. J Crohns Colitis  2017; 11: 519– 26. Google Scholar PubMed  42. Adler J, Dong S, Eder SJet al.   Perianal Crohn’s disease in a large multicenter pediatric collaborative. J Pediatr Gastroenterol Nutr  2017; 64: 117– 24. Google Scholar CrossRef Search ADS   43. Zwintscher NP, Shah PM, Argawal Aet al.   The impact of perianal disease in young patients with inflammatory bowel disease. Int J Colorectal Dis  2015; 30: 1275– 9. Google Scholar CrossRef Search ADS PubMed  44. Hyams J, Crandall W, Kugathasan Set al.  ; REACH Study Group. Induction and maintenance infliximab therapy for the treatment of moderate-to-severe Crohn’s disease in children. Gastroenterology  2007; 132: 863– 73. Google Scholar CrossRef Search ADS PubMed  45. Rinawi F, Assa A, Eliakim Ret al.   Risk of colectomy in patients with pediatric-onset ulcerative colitis. J Pediatr Gastroenterol Nutr  2017; 65: 410– 5. Google Scholar CrossRef Search ADS PubMed  Copyright © 2017 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com

Journal

Journal of Crohn's and ColitisOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off