ANNALS OF MEDICINE 2022, VOL. 54, NO. 1, 1255–1264 https://doi.org/10.1080/07853890.2022.2069851 CLINICAL STUDY Heterogeneity in pharmacological treatment and outcomes in Crohn’s disease patients in Catalonia: a population-based observational study Eduard Bruneta,b,c , Emili Velad,e, Luigi Melcarnea, Laura-Patricia Lloveta, Anna Puya, Montserrat Cl riaa,b,c, Gilaad G. Kaplanh and eriesd,e, Caridad Pontesf,g , Pilar Garcıa-Iglesiasa, Albert Villo a,b,c Xavier Calvet a Servei Aparell Digestiu, Hospital Universitari Parc Taulı, Sabadell, Spain; bDepartament de Medicina, Universitat Autonoma de Barcelona, Bellaterra, Spain; cCIBERehd, Instituto de Salud Carlos III, Madrid, Spain; dUnitat d’Informacio i Coneixement. Servei Catala de la Salut, Generalitat de Catalunya, Barcelona, Spain; eDigitalization for the Sustainability of the Healthcare System (DS3), Sistema de Salut de Catalunya, Barcelona, Spain; fGerencia del Medicament. Servei Catala de la Salut, Barcelona, Spain; gDepartament de Farmacologia, de Terapeutica i de Toxicologia. Universitat Autonoma de Barcelona, Bellaterra, Spain; hDepartments of Medicine and Community Health Sciences, University of Calgary, Calgary, Alberta, Canada ABSTRACT ARTICLE HISTORY Background: Heterogeneity in the treatment of a disease is a marker of suboptimal quality of Received 13 December 2021 care. The aim of this study is to evaluate the heterogeneity in the treatment used and the out- Revised 6 April 2022 comes for Crohn’s disease (CD) in Catalonia. Accepted 19 April 2022 Methods: All patients with CD included in the Catalan Health Surveillance System (data on KEYWORDS more than seven million individuals from 2011 to 2017) were identified. The different Catalonian Epidemiology; inflammatory health areas were grouped into 19 district groups (DG). Treatments used rates (systemic cortico- bowel diseases; Crohn’s steroids, non-biological and biological immunosuppressant) and outcomes rates (hospitalization disease; biologics; surgery and surgery) were calculated. Results: The use of systemic corticosteroids presented a decreasing trend over the study period, with an average rate of use in the different territories between 11% and 17%. The use of non- biological immunosuppressant treatment has remained stable, with an average rate of use rang- ing from 22% to 40% per year depending on the DG. The use of biological immunosuppressant treatment increased with an average rate of use in the different territories ranging from 10 to 23%. Results: Hospitalizations for any reason showed an increasing trend between 2011 and 2017 with an average rate of between 23% and 32% per year depending on the area. Hospitalizations for CD presented a decreasing trend, with an average rate of between 5% and 11% per year. Surgical treatment remained stable over time, rates per year were between 0.5% and 2%. Conclusion: A remarkable geographical heterogeneity in the use of different treatments and in outcomes of CD was observed between different geographical areas of Catalonia. KEY MESSAGE  There is a notable geographical heterogeneity in the administration of biological and immunosuppressive treatments to Crohn’s disease patients in Catalonia.  There is also a geographical heterogeneity in their rates of hospitalization and surgical intervention. 1. Introduction America where prevalence is forecasted to approxi- Inflammatory bowel disease (IBD), which includes mate 1% of the population by 2030 [2–4]. ulcerative colitis (UC) and Crohn’s Disease (CD) [1], First approved for CD in 1998, biological treatments affects more than five million people worldwide [2]. have probably changed its natural history [5]. Their The incidence and prevalence of CD are increasing use has expanded quickly and multiple effective drugs progressively, especially in Western Europe and North are available today: anti-tumor necrosis factor CONTACT Xavier Calvet xcalvet@tauli.cat Servei de Malalties Digestives, Hospital de Sabadell, Departament de Medicina, Universitat Aut onoma de Barcelona, CIBEREHD – Instituto de Salud Carlos III, Parc Taulı, 1, Sabadell, 08208, Barcelona, Spain Co-authors. ß 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 1256 E. BRUNET ET AL. (infliximab, adalimumab), anti-integrin (vedolizumab) In 2011, the CHSS was created to integrate most of and anti-interleukin 12/23 (ustekinumab) [6–10]. This those activity registers, placing the patient (instead of variety of treatments may favour heterogeneity in CD the provider) in the centre of this information system, management [11–14]. providing a more holistic and transversal view of Many studies have shown a notable geographical health problems. heterogeneity in the management of IBD [15–17]. High CHSS includes all the diagnoses reported by the dif- variations have been shown in the use of anti-TNF ferent providers, regardless of whether they were drugs in IBD patients in the US [18]. In a prospective recorded as the primary or secondary diagnosis. This study in North America and Canada in children, information system collects all information from the Krishnakumar et al. observed that the proportion of entire public health system, including all hospital children treated with thiopurines ranged from 6% to admissions and healthcare visits. Its automated data 89%, and reported that rates of children receiving corti- validation system checks the consistency of the data costeroids ranged from 28.6% to 96.9% [19]. Recently, and identifies potential errors. Information from pri- Zaltman et al. published a cross-sectional multicentre vate health centres was not available [29]. study observing also marked regional variations in IBD All patients with CD included in the CHSS within management in Brazil [20]. the period 2011 to 2017 were identified, and classified In CD, differences in management have been according to the ICD-9-CM codes for CD. The ICD-9 described not only between countries but also codes used are shown in Supplementary Table 1. between hospitals [19,21–23]. They can be evaluated For this study, the 43 health districts in Catalonia [30] on the basis of the rates of patients receiving specific were grouped into 19 Health Areas (HA) according to treatments. These differences may influence patients’ geographical proximity, setting (rural or urban) and refer- outcomes: for example, whether they undergo surgery, ence hospital (Figure 1 and Supplementary Table 2). The or whether they are hospitalized [24]. Regarding dis- assignment of patients to HAs was based on their home ease outcomes, epidemiological studies have also address and assignation to primary care. To allow com- reported notable differences in surgery rates between parisons, the HAs were designed so that each one would European countries [25–27], which were also observed contain a similar number of individuals. between Brazilian regions in a recent study [20]. The exposure to different CD treatments was As heterogeneity is a well-accepted marker of sub- retrieved from the electronic prescription records for the optimal care [28], detecting and analysing variability same period. CD treatments included biological drugs may help to improve care quality. The aim of this (infliximab, adalimumab, golimumab – compassionate study was to evaluate the heterogeneity in the use of use for CD – vedolizumab and ustekinumab), non-bio- the most frequent medical drugs for CD (corticoste- logic immunosuppressive drugs (azathioprine, 6-mercap- roids, non-biological immunosuppressive and bio- topurine and methotrexate) and corticosteroids. The list logical treatments) and its relationship with CD of active principles included in each pharmacological outcomes in the health districts of Catalonia, using group is appended as Supplementary Table 3. Patients administrative data from the Catalan Health were considered to have been treated with a particular Surveillance System (CHSS). biological drug in a given year if they received at least one dose of the drug during this period. As a sensitivity 2. Material and methods analysis, we also measured exposure by dividing the total number of drug vials administered throughout the 2.1. Data source and study design study period by the number of patient/years at risk. The regional health department of Catalonia, named Rates of hospitaliZation, CD-related hospitalization, CatSalut, provides universal healthcare coverage to all surgeries and ostomies per 1000 patients/year were residents and collects detailed information on health- also obtained from the CHSS according to previously care usage, including information from the minimum described criteria [31,32]. ICD-9-CM codes for the dif- basic data set registered by healthcare units (e.g. hos- ferent surgical procedures are detailed in pitals, primary care centres, nursing facilities and men- Supplementary Table 4. tal health centres). CatSalut also collects information on drug prescription and billing for services (e.g. out- 2.2. Statistical methods patient visits to specialists, emergency department vis- its, non-urgent medical transportation, outpatient All data were calculated for each year, so as to esti- rehabilitation and dialysis). mate the evolution over the study period, and for the ANNALS OF MEDICINE 1257 Figure 1. Geographical areas considered in the analysis. total study period so as to allow comparisons between risk. Final rates were expressed as number of pre- HAs. scriptions of each drug for 100 patients/year. To calculate the rates of surgical procedures and drug exposure, first we calculated the time at risk for To estimate the patient/year surgical rates, we div- each patient. The time at risk began on 1 January ided the total number of ostomies and bowel resec- 2011 (or at the date of diagnosis for incident patients) tions during the study period by their time at risk. and ended on 31 December 2017 (or the Final rates were expressed as ostomies or surgeries for patient’s death). 100 patients/year. To estimate the hospitaliZation To estimate drug exposure, we used two differ- rates, we divided the total number of hospitalizations ent methods: during the study period by total time at risk. Final rates were expressed as hospitalizations for 100 1. We divided the total number of patient/years with patients/year. a given treatment during the study period by Qualitative data are given as percentages and 95% total patient/years at risk. Final rates were confidence intervals and quantitative data as expressed as the number of patients treated with means ± S.D. Annual and geographical variation of each drug for 100 patients/year. treatments, hospitalizations and surgeries, were esti- 2. As a sensitivity analysis, we also divided the total mated with Poisson regression, adjusted by age, sex dispensations for each drug during the study and socioeconomic status. The socioeconomic status period by the total number of patient/years at was stratified into four categories of pharmaceutical 1258 E. BRUNET ET AL. Figure 2. Relative risk of treatments and outcomes between 2011 and 2017. co-payment: very low (recipient of social rescue aids), complied with the ethical guidelines of the low (annual income < e18,000), moderate (annual Declaration of Helsinki. The study was reviewed and income e18,000 to e100,000) and high (annual income approved by the local ethics committee of the > e100,000). Data on drug use and outcomes were Hospital Universitari Parc Taulı in Sabadell (CEIC 2020/ displayed graphically as a colour gradient on a map of 749, 24 November 2020). Given the retrospective Catalonia. The study was performed and reported in nature of the study, the unavailability of personal accordance with the STROBE and RECORD Statement data, and the impossibility of obtaining informed con- guidelines [33,34]. sent for the whole study population, the ethics com- mittee waived the need for informed consent. 2.3. Ethical issues 3. Results The research used retrospective anonymized data from the CHSS. No personal data were used, and all As shown in our previous study [31,32], relative risk of the patient data were encrypted, so that no personal overall hospitalization increased when compared with identification was retrievable or traceable to the ori- 2011, CD-related hospitalization rates slightly ginal source from the working database. The study decreased (being statically significant in 2016 and ANNALS OF MEDICINE 1259 Figure 3. Variability in the use of medical treatments between HAs over the study period (2011–2017). (a) Mean annual rate of corticosteroid use per 100 patients. (b) Corticosteroid dispensation. Rate per 100 patients. (c) Mean annual rate of immunosup- pressive treatment per 100 patients. (d) Immunosuppressive treatment dispensation. Rate per 100 patients. (e) Mean annual rate of biological treatment per 100 patients. (f) Biological treatment dispensation. Rate per 100 patients. 2017) and surgery and ostomies tend to decrease but year and the total dispensations doses for 100 patients not significantly. Regarding drugs, biological drugs (Figure 3(a,b) and Supplementary Table 7). Thus, some increased (being statically significant over the period), areas had treated a small amount of patients with a immunosuppressant’s remained stable and salicylates low number of total doses (HA 10 presented an aver- and systemic corticosteroids significantly decreased age rate of use of 13.1% and HA 16 of 11.9%), thus (being statically significant between 2013 and 2017) suggesting overall reduction in the use of steroids. By (Figure 2, Supplementary Table 5). contrast, other areas showed a medium-range number of patients treated but with a large number of pre- scriptions (HA 1, with a 15.9% of use). The fact that 3.1. Treatment heterogeneity many prescriptions were used in a relatively small 3.1.1. Systemic corticosteroids number of patients reflects an extensive use of ste- The use of systemic corticosteroids tended to fall over roids, strongly suggesting that many patients had the study period, with a mean annual rate of use per received steroids in long courses or even as mainten- 100 patients in the different HAs ranging from 11% to ance treatment. Finally, one specific HA (HA 17 with a 17% (Figure 3(a)). The prescription rate remained fairly 16.3% of use) presented a marked increase in both unchanging for each HA over time (Supplementary the number of patients treated and the number of Table 6 and Supplementary Figure 1). prescriptions, thus indicating a generalized overuse Sensitivity analysis showed big differences between (Figure 3(a,b)). In most areas, high steroid use was the rate of patients treated with corticosteroids each associated with a trend towards poorer outcomes 1260 E. BRUNET ET AL. Figure 4. Variability in the outcomes between over the study period (2011–2017). (a) Mean annual rate of overall hospitalization per 100 patients. (b) Mean annual rate of IBD-related hospitalization per 100 patients. (c) Mean annual rate of ostomies per 100 patients. (d) Mean annual rate of surgical resections per 100 patients. (Figure 4) and a fall in the use of immunosuppressive HAs ranged from 10% to 23% (Figure 3(e,f)). Some and biological treatments (Figure 3c–f). HA’s presented a significant and sustained increase in the use of this treatment (HA’s 10, 16 and 17 with a 3.1.2. Non-biological immunosuppressive treatment 21.5%, 20.5% and 22.1% of use respectively), but The use of non-biological immunosuppressive treat- others HA’s presented a significant and sustained ment remained steady between 2011 and 2017, with a reduction (HA’s 5, 6 and 11 with a 12.2%, 11.9% and mean rate ranging from 22% to 40% per year depend- 12.2% of use respectively). Detailed data for HA and ing on the HA (Figure 3(c, d)). Some HA’s presented a year are shown in Supplementary Table 10 and significant and elevated use of this treatment (HA’s 9 Supplementary Figure 1. and 10 with a 36.1% and 39.3% of use respectively), As a chronic treatment, the sensitivity analysis did whereas others had significantly lower rates of use not show marked differences (Supplementary Table 11). compared with the whole population (HA’s 5, 7 and 11 with a 23.3%, 21.3% and 21.4% of use respectively). 3.2. Hospitalization rates Detailed data for HA and year are shown in Supplementary Table 8 and Supplementary Figure 1. The mean total hospitalization rate rose from 23% in As a chronic treatment, the sensitivity analysis did 2011 to 28% in 2017 (Figure 3). The mean rate ranged not show marked differences (Supplementary Table 9). from 22% to 32%, depending on the HA: the lowest rate of total hospitalization was found in HA 18. In 3.1.3. Biological treatment contrast, CD-related hospitalizations remained con- The use of biological treatment tended to increase stant over time, ranging between 8% in 2011 and 7% over the study period. The mean rate of use in the in 2017 (Figure 4(a,b)). The rate of CD-related ANNALS OF MEDICINE 1261 hospitalization ranged from 5% to 11% between the differences may depend on the availability of special- different HA’s, but no sustained trends were observed. ized care or IBD units. Thus, in the US, Detailed data for HA and year are shown in Ananthakrishnan et al. observed much lower differen- Supplementary Tables 12 and 13 and Supplementary ces in the use of biologicals when comparing seven Figure 2. high volume centres [36]. The variability, however, seemed to persist, even between very specialized centres. This and other studies [16,17,20] suggest that 3.3. Surgery rates there is a significant variation in the management of Surgical treatment (both resections and ostomies) CD disease even between experts. remained unchanging over time. Rates per year were Rates of variability may depend on the equilibrium low, between 0.5% and 2% (Figure 4(c,d)). There were between factors that favour the standardization of no significant differences between HAs. Detailed data management and those that may induce heterogen- for HA and year are shown in Supplementary Tables eity. In our opinion, there were two factors that 14 and 15 and Supplementary Figure 3 reduce the variability to the moderate (though signifi- cant) rates we observed. First of all, the presence of a widespread and relatively homogeneous public health- 3.4. Correlations between treatment use care structure may help to reduce heterogeneity. and outcomes Thus, our study found similar rates of heterogeneity to A significant positive correlation was observed those observed in other countries such as Korea and between both total and CD-related hospitalizations Canada. In the three main cities in Korea (Seoul, and corticosteroid use (0.53; p ¼ 0.01 and 0.45 Deagu and Busan), Han et al. found rates of biological p ¼ 0.05, respectively). The correlation was non-signifi- treatment use of 20.7%, 22.9% and 14.6% respectively cant for non-biological immunosuppressive treatments [37]. Second, GETECCU, the Spanish Crohn’s disease (0.03; p ¼ 0.88 and 0.16; p ¼ 0.49 respectively) and for and ulcerating colitis working group has been very the use of biological treatment (0.19; p ¼ 0.43 and active in training IBD specialists [38] in developing 0.16; p ¼ 0.51 respectively). standards of care [39], and in accrediting IBD units The correlation between the use of corticosteroids according to these standards [40]. Although these and the two surgical procedures, ostomies and resec- homogeneizing influences have probably reduced vari- tions, had a positive estimate that did not reach signi- ability, geographical differences remain an issue. In a fication (0.33; p ¼ 0.16 and 0.25; p ¼ 0.28 respectively). Canadian population study, Kuenzig et al. observed Biological agents showed a non-significant negative that patients from rural zones were less likely to be correlation estimate with ostomies and resections treated by a specialist gastroenterologist and pre- (0.18; p ¼ 0.45 and 0.12; p ¼ 0.62 respectively), as sented a higher rate of comorbidities and infectious did the use of non-biological immunosuppressive diseases than patients from urban areas [41]. Even in a drugs with ostomies (0.22; p ¼ 0.35). (Detailed data smaller territory like Catalonia, where distances to on correlations are shown in Supplementary Table 16 acute hospital care is in average less than 20 Km [42], and Supplementary Figure 4). the access to IBD specialists and IBD units differs in rural and urban areas. Indeed, our study described non-significant lower rates of use of biological drugs 4. Discussion and higher ostomy and surgery rates in patients living Data on the prevalence, incidence and mortality on in rural areas located far from a major hospital. CD and demographical variables have been described Most correlations between treatment and outcomes in detail in previous articles [31,32]. in our study are non-significant. This finding may have This study shows that despite there is heterogen- many different (and possibly complementary) explana- eity in the pharmacological treatment and the out- tions. First, as the variability in medical treatment is comes of CD patients in Catalonia (the differences moderate, using treatments between the observed may be as high as 200% in the use of biological drugs ranges may have had a limited effect on outcomes. or in the rates of surgery), these differences, are not Second, other factors such as the availability of sur- as high as those reported in other settings. For gery, patients’ or medical team’s preferences or other example, the differences in the use of biological drugs structural factors (for example, access to a day care in the US were found to range from less than 5% to unit for early treatment of flares) may also play a role more than 40% depending on postal code [35]. The in deciding the choice of hospitalization or surgery, 1262 E. BRUNET ET AL. which may be even more important than pharmaco- included in the analysis due to their potential higher logical treatment. The only significant correlation rate of misclassification using ICD-9-CM codes. observed was between corticosteroid use and both A plausible explanation of the lack of huge varia- total and CD-related hospitalizations. There is a global tions across the territory is that the organization of trend to decrease the use of corticoids that have been care allows derivation of cases, so that escalation of progressively substituted by safest therapies [32]. The severe cases for care to tertiary sites with wider current study, however, shows that there is still a high experience in both surgery and advanced medical care rate of steroid overuse in some areas. The increased is foreseen and implemented. In Catalonia 98% of rate of hospitalization associated to steroids may have the population has a distance lower than 20 km to the different possible explanations: Steroids may be closest acute care hospital, and roughly half of the related, per se, to a poorer control of the disease or population lives in the metropolitan area, where six more adverse events leading to hospitalization or, high technology public hospitals are placed [43]. The alternatively, a high rate of corticosteroid use may be organization of healthcare ensures that access to a marker for a more general situation of suboptimal highly specialized care is granted to patients regard- care – for example, unavailability of timely on demand less of where they live in the territory. attention during flares – leading to an increase hospi- The data provided by this study and by previous research may offer an example of a rather successful talization rates. This study has certain limitations. First, the CHSS care of this disease. However, some measures may help to further improve treatment. For example, non-special- database includes only patients who use the public ist physicians in rural areas would benefit from having health system. This means that, conceivably, some access to IBD specialists by teleconference or e-mail. patients with mild disease might not have been The results of the study may also increase the pressure recorded. However, the public system is used by over to develop value-based medicine tools as, for instance, 80% of the population of Catalonia; the rate of use is, Care Improvement communities, which have shown therefore, much higher, since the 20% of non-users their ability to quickly improve patients’ outcomes. For also includes healthy individuals who also have access example, within a year of launch of the paediatric IBD to the system in case of need. Furthermore, as CD is a ImproveCareNow project [44], the rate of children in chronic disease and pharmacological (and especially remission increased from 77% to 83% [45]. Future stud- biological) treatments are relatively expensive, most ies should address the effect of heterogeneity in patients request financial support from the public sys- patient outcomes. It is also necessary to initiate meas- tem. In very expensive treatments such as biologicals, ures like local guidelines, surveys, courses or conferen- public prescription approaches 100%. It is also import- ces to unify the management of CD in our territory. ant to consider that the tool to collect data on bio- In conclusion, in this register-based, population-wide logical treatments by the CHSS database was not study, we observed a remarkable geographical hetero- mandatory for invoicing until 2014. Until then, some geneity in the use of different treatments and in the biological treatments were invoiced as inhospital med- rate of hospitalizations and surgeries for CD between ications in certain sites. Thus, heterogeneity in bio- different geographical areas of Catalonia. Reducing this logical treatments used before 2014 may be partly variability represents a clinical and organizational chal- influenced by different speed in recording data lenge, but is likely to increase equity and quality of amongst sites that were late uptakers of the voluntary care. In the same sense, it is essential to improve clin- registry. This fact, however, has probably no effect in ical records and standardize results measurement in the results. Although the use of biological treatments order to carry out good-quality research. increased steadily between 2011 and 2017, the increase was greater between 2015 and 2017 when Acknowledgments recording data on biological treatments were manda- tory. A final limitation is that it is also known that The authors thank Michael Maudsley for his help with patients with aggressive CD (i.e. penetrating or struc- the English. turing patterns) need more aggressive treatment and have higher rates of biological treatments, hospitaliza- Author contributions tion and surgery. Unfortunately, information on the Eduard Brunette, Emili Vela and Xavier Calvet designed the baseline severity of the disease was not available for study, analysed data and wrote the manuscript; designed analysis. In addition, perianal-CD and UC was not the study, analysed data and wrote the manuscript. ANNALS OF MEDICINE 1263 Xavier Calvet, Eduard Brunette, Emili Vela, Luigi Melcarne, [5] Guo Y, Lu N, Bai A. Clinical use and mechanisms of Laura Llovet, Pilar Garcia-Iglesias, Montserrat Cleries, Caridad infliximab treatment on inflammatory bowel disease: Pontes, Albert Villoria, and Gilaad G. Kaplan critically a recent update. Biomed Res Int. 2013;2013:581631. reviewed the text and provided important intellec- [6] Hanauer SB, Feagan BG, Lichtenstein GR, et al. tual content. Manteniance infliximab in Crohn’s disease: the All authors definitively approved the submitted version. ACCENT I randomised trial. Lancet. 2002;359(9317): 1541-1549. [7] Colombel JF. The CHARM trial of adalimumab in Disclosure statement crohn’s disease. Gastroenterol Hepatol (N Y). 2006; 2(7):486–488. Albert Villoria has served as a speaker and consultant from [8] Sandborn WJ, Feagan BG, Marano C, et al. for MSD and Abbvie. Eduard Brunette, Emili Vela, Luigi Subcutaneous golimumab maintains clinical response Melcarne, Laura-Patricia Llovet, Anna Puy, Montserrat Cleries, in patients with moderate-to-severe ulcerative colitis. Caridad Pontes and Pilar Garcıa-Iglesias have not conflicts of Gastroenterology. 2014;146(1):96–109.e1. interest to declare. [9] Feagan BG, Sandborn WJ, Gasink C, et al. Gilaad G. Kaplan has received honoraria for speaking or Ustekinumab as induction and maintenance therapy consultancy from Abbvie, Janssen, Pfizer, Amgen, and for Crohn’s disease. N Engl J Med. 2016;375(20): Takeda. He has received research support from Ferring, 1946–1960. Janssen, Abbvie, GlaxoSmith Kline, Merck, and Shire. He has [10] Mao EJ, Hazlewood GS, Kaplan GG, et al. Systematic been a consultant for Gilead. He shares ownership of a review with Meta-analysis: comparative efficacy of patent: TREATMENT OF INFLAMMATORY DISORDERS, immunosuppressants and biologics for reducing hos- AUTOIMMUNE DISEASE, AND PBC. UTI Limited Partnership, pitalisation and surgery in Crohn’s disease and ulcera- assignee. Patent WO2019046959A1. PCT/CA2018/051098. 7 tive colitis. Aliment Pharmacol Ther. 2017;45(1):3–13. Sept. 2018. [11] Magro F, Gionchetti P, Eliakim R, European Crohn’s Xavier Calvet has received grants for research from and Colitis Organisation [ECCO], et al. Third european Abbott, MSD, and Vifor, and fees for advisory boards services evidence-based consensus on diagnosis and manage- form Abbott, MSD, Takeda and Vifor. He and has also given ment of ulcerative colitis. Part 1: Definitions, diagno- lectures for Abbott, MSD, Takeda, Shire and Allergan. sis, extra-intestinal manifestations, pregnancy, cancer Antonio Obrador grant support has been received surveillance, surgery, and ileo-anal pouch disorders. J from GETECCU. Crohns Colitis. 2017;11(6):649–670. [12] Harbord M, Eliakim R, Bettenworth D, et al. ORCID Corrigendum: Third European evidence-based consen- sus on diagnosis and management of ulcerative col- Eduard Brunet http://orcid.org/0000-0001-8661-0608 itis. Part 2: current management. J Crohns Colitis. Caridad Pontes http://orcid.org/0000-0002-3274-6048 2017;11(12):1512. [13] Gomollo n F, Dignass A, Annese V, et al. 3rd European evidence-based consensus on the diagnosis and man- Data availability statements agement of Crohn’s disease 2016: part 1: diagnosis Data cannot be shared for ethical/privacy reasons. The data and medical management. J Crohns Colitis. 2017; underlying this article cannot be shared publicly due to is 11(1):3–25. individual date from a population registry from Catalan [14] Gionchetti P, Dignass A, Danese S, ECCO, et al. 3rd Health Surveillance System and only can be accessed from european evidence-based consensus on the diagnosis Catsalut. The data will be shared on reasonable request to and management of Crohn’s disease 2016: part 2: sur- the corresponding author. gical management and special situations. J Crohns Colitis. 2017;11(2):135–149. [15] Spiegel BMR, Ho W, Esrailian E, et al. Controversies in References ulcerative colitis: a survey comparing decision making of experts versus community gastroenterologists. [1] Abraham C, Cho JH. Inflammatory bowel disease. N Clin Gastroenterol Hepatol. 2009;7(2):168–174. Engl J Med. 2009;361(21):2066–2078. [16] Esrailian E, Spiegel BMR, Targownik LE, et al. [2] Kaplan GG, Windsor JW. The four epidemiological Differences in the management of Crohn’s disease stages in the global evolution of inflammatory bowel among experts and community providers, based on a disease. Nat Rev Gastroenterol Hepatol. 2021;18(1): national survey of sample case vignettes. Aliment 56–66. Pharmacol Ther. 2007;26(7):1005–1018. [3] Coward S, Clement F, Benchimol EI, et al. Past and [17] Benchimol EI, Cook SF, Erichsen R, et al. International future burden of inflammatory bowel diseases based variation in medication prescription rates among eld- on modeling of population-based data. erly patients with inflammatory bowel disease. J Gastroenterology. 2019;156(5):1345–1353.e4. Crohn’s Colitis. 2013;7(11):878–889. [4] Jones GR, Lyons M, Plevris N, et al. IBD prevalence in [18] David G, et al. Geographic variation in care of Lothian, Scotland, derived by capture-recapture meth- patients with IBD suggests unequal quality of care in odology. Gut. 2019;68(11):1953–1960. the United States. Present DDW 2013. Orlando, FL. 1264 E. BRUNET ET AL. [19] Krishnakumar C, Ballengee CR, Liu C, et al. Variation [32] Brunet E, Vela E, Melcarne L, et al. Time trends of in care in the management of children with crohn’s Crohn’s disease in Catalonia from 2011 to 2017. disease: data from a multicenter inception cohort Increasing use of biologics correlates with a reduced study. Inflamm Bowel Dis. 2019;25(7):1208–1217. need for surgery. JCM. 2020;9(9):2896. [20] Zaltman C, Parra RS, Sassaki LY, et al. Real-world dis- [33] ThePLOS MedicineEditors. Observational studies: get- ease activity and sociodemographic, clinical and treat- ting clear about transparency. PLoS Med. 2014;11(8): ment characteristics of moderate-to-severe e1001711. inflammatory bowel disease in Brazil. World J [34] Benchimol EI, Smeeth L, Guttmann A, et al. The Gastroenterol. 2021;27(2):208–223. REporting of studies conducted using observational [21] Yu H, MacIsaac D, Wong JJ, et al. Market share and routinely-collected health data (RECORD) statement. costs of biologic therapies for inflammatory bowel PLoS Med. 2015;12(10):e1001885–22. disease in the USA. Aliment Pharmacol Ther. 2018; [35] Siegel CA, Yang F, Cai J. P1304 – significant variation 47(3):364–370. in biologic use across the United States for Crohn’s [22] Kurti Z, Vegh Z, Golovics PA, et al. Nationwide preva- disease and ulcerative colitis. World Congr lence and drug treatment practices of inflammatory Gastroenterol. 2017;112. bowel diseases in Hungary: a population-based study [36] Ananthakrishnan AN, Kwon J, Raffals L, et al. Variation based on the national health insurance fund data- in treatment of patients with inflammatory bowel dis- base. Dig Liver Dis. 2016;48(11):1302–1307. eases at major referral centers in the United States. [23] Lirhus SS, Høivik ML, Moum B, et al. Regional differen- Clin Gastroenterol Hepatol. 2015;13(6):1197–1200. ces in anti-TNF-a therapy and surgery in the treat- [37] Han M, Jung YS, Cheon JH, et al. Regional variations ment of inflammatory bowel disease patients: a in the use of biologics and immunomodulators norwegian nationwide cohort study. Scand J among korean patients with inflammatory bowel dis- Gastroenterol. 2018;53(8):952–957. eases. J Gastroenterol Hepatol. 2019;34(7):1166–1174. [24] King JA, Underwood FE, Panaccione N, et al. Trends [38] Cursos y reuniones GETECCU [Internet]. Grupo Espan ~ol de Trabajo en Enfermedad de Crohn y Colitis in hospitalisation rates for inflammatory bowel dis- Ulcerosa. 2021. [citado 26 abril 2021]. Disponible en: ease in Western versus newly industrialised countries: https://geteccu.org/cursos-y-reuniones-organizadas- a population-based study of countries in the organ- por-geteccu. isation for economic Co-operation and development. [39] Van Den Brink G, Van Gaalen MAC, De Ridder L, et al. Lancet Gastroenterol Hepatol. 2019;4(4):287–295. Health care transition outcomes in inflammatory [25] King JA, Underwood FE, Panaccione N, et al. Trends bowel disease: a multinational delphi study. J Crohns in hospitalisation rates for inflammatory bowel dis- Colitis. 2019;13(9):1163–1172. ease in Western versus newly industrialised countries: [40] Acosta MB, Gutierrez A, Zabana Y, et al. I nflammatory a population-based study of countries in the organ- bowel disease integral care units: Evaluation of a isation for economic Co-operation and development. nationwide quality certification program. The Lancet Gastroenterol Hepatol. 2019;4(4):287–295. GETECCU experience. United Eur Gastroenterol J. [26] Frolkis AD, Dykeman J, Negro n ME, et al. Risk of sur- 2021;9(7):766–772. gery for inflammatory bowel diseases has decreased [41] Kuenzig ME, Nguyen GC, Benchimol EI. Rural and over time: a systematic review and meta-analysis of urban differences in the risk of inflammatory bowel population-based studies. Gastroenterology. 2013; disease and subsequent health services utilization in 145(5):996–1006. Ontario. Healthc Q. 2019;22(2):6–9. [27] Tsai L, Ma C, Dulai PS, et al. Contemporary risk of sur- [42] Prat E, Sanchez J, Pesquer L, et al. Estudio sobre la gery in patients with ulcerative colitis and Crohn’s accesibilidad de los centros sanitarios p ublicos de disease: a meta-analysis of population-based cohorts. Catalun ~a. Cent Recer Ecolo gica i Apl For (CREAF) Fac Clin Gastroenterol Hepatol. 2021;19(10):2031–2045. Ciencies Univ Auto noma Barcelona [Internet]. 2008. [28] Strohl M, Gonczi L, Kurt Z, et al. Quality of care in Available from: http://www.creaf.uab.es/MiraMon/pub- inflammatory bowel diseases: What is the best way to licat/papers/XIIICNTIG/ComunicacionAccesibilidad.pdf. better outcomes? World J Gastroenterol. 2018;24(22): [43] Departament de Salut Generalitat de Catalunya. Pla 2363–2372. salut de Catalunya 2011- 2015. General Catalunya [29] Vela E, Cleries M, Vella VA, et al. Analisis poblacional [Internet]. 2012;22–3. Available from: http://www.coill. del gasto en servicios sanitarios en catalun ~a (espan~a): org/Ficheros/Novedades/plasalut2011-15.pdf. >que y quien consume mas recursos? Gac Sanit. [44] Crandall W, Kappelman MD, Colletti RB, et al. 201733(1):24–31. ImproveCareNow: the development of a pediatric [30] Catsalut. Servei Catala de la Salut. Cataleg d’arees de inflammatory bowel disease improvement network. gestio assistencial (AGA). 2018. Inflamm Bowel Dis. 2011;17(1):450–457. [31] Brunet E, Roig-Ramos C, Vela E, et al. Prevalence, inci- [45] Savarino JR, Kaplan JL, Winter HS, et al. Improving dence and mortality of inflammatory bowel disease in clinical remission rates in pediatric inflammatory Catalonia. A population-based analysis. Ann Med. bowel disease with previsit planning. BMJ Qual 2018;50(7):613–619. Improv Report. 2016;5(1):u211063.w4361.