Background
Crohn’s disease (CD) and ulcerative colitis (UC), the two major forms of inflammatory bowel disease (IBD), constitute chronic inflammatory conditions of the gastrointestinal tract with a multifactorial etiology, characterized by a relapsing and remitting course [
1‐
3]. Initially described in more developed and industrialized countries (mostly in Europe and North America) [
4,
5], their incidence has been increasing worldwide in the last few decades [
6], including in newly industrialized countries, such as Brazil and other South American countries [
7‐
10].
Disease activity, drug availability and clinical response to medical therapy are known to influence the outcomes of these conditions [
11,
12], but the appropriate management of accumulated structural damage may require surgical interventions [
13‐
16]. It has been estimated that 10–30% of patients with UC [
1,
17] and approximately 38–70% of patients with CD will undergo surgery in the first 10 and 20 years after the diagnosis has been established, respectively [
1,
17]. For CD, the estimated chance of surgery throughout the patient’s lifetime has been reported to be 70–80% [
13,
18,
19]. This might considerably affect the patient’s quality of life and might also impose a relevant socioeconomic impact on patients and a burden to healthcare systems [
20,
21], especially considering that IBD frequently originates in young adults [
22].
Despite the increasing incidence and prevalence rates of IBD, recent information supports the idea that the dramatic advances in treatment, especially in the last two decades, with new drugs and therapeutic strategies [
23], have had a remarkable impact on the overall surgical interventions and hospitalization rates [
24‐
27].
However, few data exist concerning surgery, hospitalization and lethality related to IBD. Therefore, this study aimed to analyze trends of hospitalizations and IBD-related surgical procedures in Brazil. In particular, we evaluated the geographic distribution and demographic aspects in an attempt to identify associations and risks related to the geographic regions and urbanization patterns.
Discussion
In this study, we describe, for the first time to our knowledge, information on hospitalizations and surgical treatment for IBD in Brazil over a recent period of time. Data obtained from an official source of the Ministry of Health indicated that the overall IBD hospitalizations decreased twenty-four percent, while surgical treatment decreased thirty-five percent, and the lethality of IBD-related surgeries decreased almost fifty percent during this eleven-year period. The geographic distribution of IBD hospitalizations showed an overall higher prevalence of UC, with greater concentrations of hospitalizations in the South and the Southeast, where surgery rates decreased more, particularly in municipalities not integrated with MRs. Nevertheless, during the same period, lethality rates associated with IBD hospitalizations and IBD-related surgeries increased in the North region, more so in small- and medium-sized municipalities integrated with MRs.
Hospitalization rates per 100,000 inhabitants, adjusted to the totality of available hospital beds, were invariably higher for patients with UC compared to CD, and they decreased essentially in parallel during the period. Hospitalization rates for females were slightly higher. Some Brazilian regional studies indicate a slight predominance of females among the patients with IBD, but with no significant differences [
10,
33,
34], similar to the world literature [
4]. Further analysis revealed that hospitalizations for CD were proportionally higher among patients aged 20–39 years, whereas in patients with UC, the highest proportion of hospitalizations was among patients aged 40 years or older. Although hospitalizations remained stable across most age ranges, we observed an increase among young patients, more so in CD. These results appear to be in accordance with most international data showing that usually UC peaks first, followed by CD, with a lapse of approximately a decade [
5,
35]. In addition, CD preferentially affects younger individuals, whereas UC occurs in relatively older patients, with a characteristic bimodal distribution, with second peak after age 40 [
36].
Temporal analysis revealed generally decreasing numbers of surgical procedures for both IBD types. Nevertheless, CD presented a higher proportion of surgeries than UC, which is also in agreement with the literature, since several studies have shown that CD has a cumulative risk of surgery higher than UC [
37‐
39]. Altogether, patients aged 40–59 years had the highest proportion of surgical procedures in the period. However, from 2005 to 2015, there was an increase of almost fifty percent in the total number of surgeries performed in patients aged 20–39 years, a trend not observed in other age groups. This may have occurred due to the probable increase in the number of new cases emerging in younger individuals and the likely increase in the number of more complex and severe forms of disease as IBD progresses in the country. In fact, the increase in the number of surgeries among younger patients observed in this study also might be related to the global increase in IBD prevalence corroborated by recent studies carried out in the country (7, 33, 51).
Regarding length of hospital stay, both groups of patients registered an increase in the average length of hospital stay from 2005 to 2015. For CD, the increase was from 6.9 to 7.6 days, whereas for UC, it was from 6.9 to 7.2 days. As expected, the length of hospital stay was higher for surgical-related hospitalizations, with similar values for CD and UC. While the findings of this study are agreement with data from some European countries [
40,
41], they are in contrast with others [
24], probably indicating the heterogeneity of the populations affected and their respective health systems. Nonetheless, in the period analyzed here, the in-hospital lethality of IBD surgical patients decreased to almost half, whereas the general in-hospital lethality rate regarding SUS beds increased approximately thirty percent. These findings suggest that although the reduction in the number of available beds imposes a selection of the most severe cases, and probably longer hospital stays, the overall outcomes regarding IBD-related hospitalizations and surgical procedures apparently improved.
It is important to highlight the fact that even with the reduction of the in-hospital lethality of IBD surgical patients to 10.6% observed in this study, the numbers are higher than the ones from other studies, ranging from 2.5 to 6.0% [
42‐
44]. Such discrepancies may reflect differences in the efficiency of the respective health systems where data originate from, including availability of medication, access to specialists, and the time lapse before diagnosis. However, it is likely that such disparities also may reflect differences in study designs, which renders data difficult to compare. For example, some investigations analyze data from elective surgeries, while others include emergency procedures [
43,
45,
46]. On the other hand, some studies focus on the status of medical therapy and the use of biological agents, while others on the presence of comorbidities in the study population, or the duration of the disease and specific surgical procedures [
42‐
47]. Moreover, some studies may consider both intra- and extra-hospital post-operatory mortality, from one month up to three years after surgery [
42,
46,
47]. The databank used in the present study does not allow the analysis of all these individual variables, what may contribute to the apparently higher rates detected.
The most frequently identified profile among the surgical patients was CD (71% in 2015), with a slight predominance of males and a predominant age range of 20–59 years (40–59 years in 2005, and 20–39 years in 2015). The predominance of young male patients (not including the pediatric population, which has the lowest surgery rates) has also been described before [
37]. The most common type of surgical procedure was major surgery, with abdominal laparotomy, which may be related to the longer length of hospital stay for surgical patients. Surgeries involving the small bowel, especially resections and treatment of complications of CD (e.g., fistulas) were the most commonly performed procedures, in agreement with the higher rates of surgeries observed in CD patients (small bowel involvement).
In assessing the country according to its macro regions, the highest hospitalization rates and the largest proportions of surgeries were concentrated in the South and Southeast, which could be justified by the fact that these are the most developed regions of the country, with more equipped hospitals, that are associated with large MRs. During the study period, the in-hospital lethality of IBD patients decreased, whereas the lethality of surgical patients had an even more marked reduction. Together, these findings may reflect improvements in the care of patients with IBD, including pre-and postoperative management, and the technical and quality care in the centers where most of these procedures took place.
Since Brazil is a country with continental proportions and great heterogeneity regarding demographic and socioeconomic characteristics, in this study, we also analyzed data from the smallest administrative unit, the municipalities. The analysis of municipalities showed an overall reduction in hospitalization rates, but the reduction was greater for those not integrated with MRs. This finding may reflect recent improvements in health care in the latter, tending to technically approach the care offered in large centers in 2015. Following this line of reasoning, we observed a reduction in lethality of hospitalized patients in municipalities not integrated with MRs, while practically no changes were observed in the integrated municipalities. However, when assessing surgical treatment, we note that large municipalities associated or not with MRs present the highest proportions of surgeries, and large ones associated with MRs maintained the highest rate of all in 2015. A possible explanation is the concentration of IBD referral centers in these areas. In contrast to the overall reduction in IBD-related surgeries, small municipalities associated with MRs showed an increase, suggesting a tendency towards the higher concentration of these procedures in more developed regions, such as those associated with MRs. This may also indicate a possible migratory movement of patients from underresourced areas to referral hospitals. However, there was also an increase in proportions of surgeries in small municipalities, where large, specialized centers are not normally allocated. This may reflect a trend towards decentralization of surgeries, although still in greater concentration in more developed regions of the country. Corroborating this fact, we observed that the lethality of operated patients was higher in municipalities not associated with MR, especially in small ones, where there are probably fewer resources and qualified professionals for more complex procedures.
The last two decades witnessed the progressive development of new drugs and diagnostic tools for IBD, an increased early referral to specialists, and changes related to diagnostic and treatment strategies, such as combined drug therapy with immunosuppressants and biologics [
48,
49]. In this scenario, accumulated information has led to the interpretation of a probable relationship between these changes in patient management and the improvement in major outcomes, notably with fewer surgeries [
50,
51] and, in some populations, fewer hospitalizations [
39,
52]. However, there is a lack of studies with appropriate follow-up data to confirm these direct relations, particularly in the long term [
25,
53]. Furthermore, it has been suggested that the decrease in surgical rates occurs mostly due to emergency procedures, with no significant changes in elective procedures [
54]. The results from this study are in agreement with international trends, showing progressive reductions in surgeries and hospitalizations, including surgical hospitalizations, and decreased surgery-related lethality among the patients with IBD throughout the country [
23,
55,
56]. The widespread use of immunosuppressants, followed by the introduction of biologic therapies in the public health system in the last decade, might have had an important impact on the overall outcomes of patients with IBD in the country, as pointed out previously [
57]. Nevertheless, temporal association with our findings may not constitute a sufficient explanation. Even before biologic therapy availability, hospitalizations and surgeries were already reducing, which led us to think of other possible factors, such as improvements in diagnosis and patient follow-up. In a previous study from our group, using a databank from the social security benefits, we demonstrated that IBD frequently leads disability for prolonged periods and contributes to early retirement, but with clear reduction trends between 2010 and 2014 in the country [
20]. Although the databank used in the current study does not contain information on the socioeconomic or work status of the patients, the general trend towards reductions in hospitalizations and IBD-related surgeries follow a similar pattern and appear to be in accordance with reductions in work disability, previously observed. Taken together, these data appear to support the notion of probable ongoing improvements in the management of IBD in the country.
Although the findings presented here offer insights into time trends and the geographic distribution of IBD hospitalizations and IBD-related surgeries and lethality for the first time in Brazil, several limitations of this study need to be addressed. The inherent characteristics of this type of study and the possibility of an ecological fallacy [
58] have been greatly diminished due to the application of an objective and consistent methodology, allowing a straightforward analysis of the electronic data entered in a single national database. However, the large amount of data from the whole country in this database does not have specific information on disease details, medical follow-up or comorbidities. Moreover, hospital-based documentation does not cover outpatient procedures and medical therapy. Another potential caveat of this study involves the concept of the municipality. Typically diverse and heterogeneous, Brazil has cities with broad differences in terms of population densities. For example, some municipalities may display populations ranging from one thousand to twenty million inhabitants and would still be classified on the same level in the database. However, an analysis based on municipalities may compensate for system defects, such as underreporting or equivocal registry, because the single and official database covers the whole country in its minimal administrative units.