Temporal hospitalization trends for inflammatory bowel disease in Taiwan

Article information

Intest Res. 2026;.ir.2025.00199

Publication date (electronic) : 2026 February 24

doi : https://doi.org/10.5217/ir.2025.00199

Hao-Yu Wu^1,2, Meng-Tzu Weng^3,4, Chien-Chih Tung⁵, Yuan-Ting Chang⁶, Shu-Chen Wei^,3

¹Graduate School of Advanced Technology, Precision Health and Intelligent Medicine, National Taiwan University, Taipei, Taiwan

²Department of Internal Medicine, National Taiwan University Cancer Center, Taipei, Taiwan

³Division of Gastroenterology and Hepatology, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan

⁴Department of Medical Research, National Taiwan University Hospital, Hsin-Chu Branch, Hsinchu, Taiwan

⁵Department of Integrated Diagnostics and Therapeutics and Medical Research, National Taiwan University Hospital, Taipei, Taiwan

⁶Health Data Research Center, National Taiwan University, Taipei, Taiwan

Correspondence to Shu-Chen Wei, Division of Gastroenterology and Hepatology, Department of Internal Medicine, National Taiwan University Hospital, No.7, Chung Shan S. Rd. (Zhongshan S. Rd.), Zhongzheng Dist., Taipei City 100225, Taiwan (R.O.C.). E-mail: scwei05@gmail.com

Received 2025 August 31; Revised 2025 November 16; Accepted 2025 November 19.

Abstract

Graphical abstract

INTRODUCTION

Inflammatory bowel disease (IBD), encompassing Crohn’s disease (CD) and ulcerative colitis (UC), is a chronic immunological disorder causing persistent inflammation in the gut, resulting in symptoms, repeated hospital visits, surgeries, and potential mortality [1].

IBD prevalence rates are categorized into 4 epidemiological stages: the compounding prevalence and prevalence equilibrium stages in developed countries, the emergence stage in developing countries, and the acceleration in incidence stage for newly industrialized countries [2]. Evolving IBD incidences vary globally: higher rates in developing Asian, South American, and Middle Eastern countries, contrasting with stable or decreasing incidences in developed North American and Western European nations [3,4]. Heightened IBD prevalence poses significant economic challenges for healthcare systems [5]. Notably, developed Western nations show declining hospitalization rates for IBD patients, attributed to reduced admissions and shorter length of hospital stay [6-9]. However, limited studies explore hospitalization trends in developing nations [10]. The surge in IBD incidences in newly industrialized countries, linked to Western lifestyle adoption and urbanization, necessitates strategic healthcare management to curb rising hospitalization rates and associated costs [11-13].

Biologics revolutionized IBD care, offering improved patient outcomes not only by sustaining remission and enhancing life quality [14], but also reducing IBD-related surgical rates [11]. However, in newly developing countries with increasing IBD incidence, studies suggest the need to shift the clinical management from hospitals to community settings to alleviate healthcare burden [11,15]. Epidemiological data on IBD hospitalization in newly industrialized countries like Taiwan is limited [15]. Therefore, we performed a retrospective population-based study to explore the trends of IBD hospitalization rates and lengths of hospital stay in Taiwan and compared the clinical evolutions before and after introduction of biologics.

METHODS

RESULTS

1. Sample Characteristics

During the study period from 2001 to 2017, a total of 4,376 patients in Taiwan were newly diagnosed with IBD and received an NHI catastrophic illness certificate (CIC); 1,167 (26.7%) had CD and 3,209 (73.3%) had UC (Table 1). The ratio of patients with UC to those with CD was approximately 2.8. The male-to-female ratio was about 1.7 (2,784/1,592) for IBD, 2.2 (798/369) for CD, and 1.6 (1,986/1,223) for UC. The median age at diagnosis was 43 years (IQR, 31–55 years) for IBD, 34 years (IQR, 24–50 years) for CD, and 45 years (IQR, 34–56 years) for UC. Most patients (n=3,330; 76.1%) received their IBD diagnosis between the ages of 20 and 59 years, with the smallest proportion (n=58; 1.3%) diagnosed at age ≥ 80 years. The average age at diagnosis in the CD group was significantly lower than in the UC group (mean±SD: 37.7 ± 18.1 vs. 45.0 ± 15.6; P<0.001) (Table 1).

Table 1.

Clinical Features of Patients with Newly Diagnosed IBD from 2001 to 2017

2. Temporal Trend of Incidence and Prevalence Rates

Between 2001 and 2017, the annual crude incidence rates of CD and UC exhibited an escalating pattern, increasing from 0.14 and 0.48 per 100,000 population to 0.68 and 1.48 per 100,000 population, respectively. Additionally, age-standardized incidence rates (referred to as incidence) for CD and UC steadily increased from 0.13 and 0.54 per 100,000 population to 0.69 and 1.40 per 100,000 population, respectively (P<0.001) (Fig. 1A). These trends align with our previous study covering the years 2001 to 2015 [19]. Moreover, there was a notable rise in the IBD incidence (6.1%; 95% CI, 3.3% to 8.9%) over the study duration. Specifically, the increase in AAPC for CD incidence (11.1%; 95% CI, 8.4% to 13.9%) surpassed that of UC incidence (2.9%; 95% CI, 1.3% to 4.5%) (Fig. 1A). As the incidence increased, the crude prevalence rates for both CD and UC increased from 0.49 and 1.94 per 100,000 population in 2001 to 4.90 and 14.26 per 100,000 population in 2017, respectively. Across the study period, the age-standardized prevalence rate (referred to as prevalence) for IBD increased significantly from 2.22 to 15.17 per 100,000 population, with an increase in AAPC of 12.5% (95% CI, 11.7% to 13.4%). This upsurge was particularly pronounced in patients with CD (15.3%; 95% CI, 14.3% to 16.2%); in those with UC, the increase was slightly lower at 11.8% (95% CI, 11.0% to 12.5%) (Fig. 1B).

Fig. 1.

Age-standardized incidence (A) and prevalence (B) rates of IBD, CD, and UC from 2001 to 2017. IBD, inflammatory bowel disease; CD, Crohn’s disease; UC, ulcerative colitis; AAPC, average annual percentage change.

4. Temporal Change in All-Cause Hospitalization Rate

Between 2001 and 2017, we analyzed a total of 4,829 all-cause hospital admissions for patients with IBD, comprising 1,256 for CD and 3,573 for UC (Table 1). While the all-cause hospitalization rate remained stable in the general population throughout the study period (AAPC, 0%; 95% CI, –0.5% to 0.4%), a notable increase was observed in the hospitalization rate per 100,000 population for both CD from 0.2 to 1.8 (AAPC, 14.1%; 95% CI, 10.0% to 18.5%) and UC patients from 0.5 to 2.4 (AAPC, 10.1%; 95% CI, 8.3% to 11.8%) (Fig. 2A). However, contrary to the hospitalization rates computed using the general population as the denominator, and despite the growing prevalence of IBD, a consistent decline was noted in hospitalization rates per 100 patients with IBD during the study period. This decrease was identified as 54.4% to 40.9% for CD and 35.7% to 22.8% for UC (Fig. 2B). The downward trend in hospitalization rates per 100 IBD patients was evident with an overall AAPC of –2.6% (95% CI, –4.0% to –1.2%) for IBD and –1.8% (95% CI, –2.8% to –0.7%) for UC. However, the downward trend for patients with CD did not reach significance, registering an AAPC of −1.6% noted (95% CI, −3.5% to 0.4%; P=0.109) (Fig. 2B).

Fig. 2.

Annual hospitalization rates for IBD, CD, and UC from 2001 to 2017, expressed per 100,000 general population (A) and per 100 patients with IBD (B). IBD, inflammatory bowel disease; CD, Crohn’s disease; UC, ulcerative colitis; AAPC, average annual percentage change.

6. Temporal Trend in Length of Hospital Stay for IBD

Between 2001 and 2017, the mean length of hospital stay was 9.6 days (SD=10.9) for IBD (Table 1). Individuals with CD had a significantly longer mean length of hospital stay than those with UC (mean±SD, 10.3 ± 11.8 vs. 9.2 ± 10.2; P<0.001). Notably, we observed a substantial downward trend in length of hospital stay, with a decrease from 15.3 to 9.1 days for CD and from 11.8 to 7.8 days for UC (P<0.05) (Fig. 3).

Fig. 3.

Average length of hospital stay for IBD, CD, and UC from 2001 to 2017. IBD, inflammatory bowel disease; CD, Crohn’s disease; UC, ulcerative colitis.

7. Temporal Trend of IBD Surgical Rates

In contrast to the growth of IBD incidence and prevalence, a significant reduction in IBD-related surgical rates was observed, with rates decreasing from 8.6% to 3.4% for IBD, 19.6% to 8.5% for CD, and 5.8% to 1.6% for UC from 2001 to 2017 (P<0.001) (Fig. 4A). The IBD-related surgical rates consistently declined, indicating an overall AAPC of −5.6% (95% CI, −7.5% to −3.6%) for IBD, −5.3% (95% CI, −7.3% to −3.2%) for CD, and −7.8% (95% CI, −9.7% to −5.8%) for UC. Patients diagnosed with CD were significantly more likely to undergo surgical procedures compared to those diagnosed with UC (66.2% vs. 21.1%; P<0.001) (Table 1).

Fig. 4.

Annual surgical (A) and mortality (B) rates for IBD, CD, and UC from 2001 to 2017. IBD, inflammatory bowel disease; CD, Crohn’s disease; UC, ulcerative colitis; AAPC, average annual percentage change.

9. Comparison between the Prebiologic and Postbiologic Eras

We observed a significant reduction in all-cause hospitalization rates during the postbiologic era relative to the prebiologic era: –0.46% (standard error [SE], 0.11) for IBD and −0.44% (SE, 0.14) for UC (P<0.05) (Table 2). A nonsignificant downward trend in hospitalization rates was observed: −0.22% for individuals with CD (SE, 0.19; P=0.241). Although the length of hospital stay for patients with IBD was lower in the postbiologic era than in the prebiologic era, this reduction was nonsignificant. During the postbiologic era, surgical rates related to IBD were lower in both patients with CD and those with UC (P<0.05) relative to during the prebiologic era. Furthermore, the mortality rate for those with CD was significantly lower during the postbiologic era than during the prebiologic era (P<0.05); however, the reduction in mortality rates for the patients with UC was nonsignificant (P=0.476) (Table 2).

Table 2.

Regression Analysis of IBD, CD, and UC Clinical Data Comparing Prebiologic (2001–2010) to Postbiologic (2011–2017) Eras

DISCUSSION

Our study represents the first population-based cohort investigation into the temporal trends of IBD hospitalization rates in Asia. Moreover, through updated incidence and prevalence data, we provide insight into IBD hospitalization rates using different denominators in Taiwan spanning from 2001 to 2017. Diverse patterns in all-cause hospitalization rates can emerge when employing different denominators, especially for denominators as diverse as 100,000 person-years and 100 patients with IBD [10]. In our study, although the all-cause hospitalization rates per 100,000 person-years remained consistent in the general population, these rates increased significantly in the IBD population between 2001 and 2017, using Taiwanese general population as the denominator. Conversely, all-cause hospitalization rates when utilizing the IBD population as the denominator exhibited a declining trend. The incidence of IBD has notably risen in newly industrialized countries in Asia and Latin America, potentially due to increased disease awareness and the influence of Westernized societies [20,21]. In Taiwan, along with the rising incidence, the prevalence of IBD has grown exponentially in the general population. Rising patients with IBD in Taiwan pose challenges for gastroenterologists, necessitating enhanced care due to increased complexity and comorbidities [22].

Similar patterns in hospitalization rates for IBD have been observed in various countries. In Czechia, all-cause hospitalization rates in patients with IBD increased from 48.11 to 53.83 per 100,000 person-years between 2007 and 2015 (AAPC, 2.52%; 95% CI, 1.55% to 3.48%), while decreasing from 14.31 to 12.18 per 100 individuals with IBD during the same period (AAPC, −1.11%; 95% CI, −2.15% to −0.06%) [23]. Similarly, a study in Catalonia, Spain, noted an increase in CD-related hospitalization rates per 100,000 population from 10.56 to 14.8 (AAPC, 5.34%; 95% CI, 4.07% to 6.62%), whereas a decrease from 9.27 to 7.22 per 100 patients with CD between 2011 and 2017 (AAPC, −4.38%; 95% CI, −5.50% to −3.25%) [24]. In Alberta, Canada, there was an increase in all-cause hospitalization rates from 110.65 to 142.18 per 100,000 population (AAPC, 1.56%; 95% CI, 1.32% to 1.80%), while a decrease from 36.57 to 16.72 per 100 patients with IBD was observed between 2002–2003 and 2018–2019 (AAPC, −4.18%; 95% CI, −4.69% to −3.66%) [17]. This study also highlighted reductions in both IBD-related and non-IBD-related hospitalization rates. However, not all countries follow this pattern. In Scotland, despite a rising IBD prevalence similar to Canada, there was a decrease in both primary hospitalization rates relative to the general population and in rates relative to the IBD population [10,25]. Differences in hospitalization reporting and healthcare systems may explain discrepancies. Researchers should use IBD population as a benchmark to analyze hospitalization trends, gaining clearer insights into resource use and intervention effectiveness. Factors like biologic therapy coverage, improved clinical care, and reduced IBD-related surgeries contributed to lower hospitalization rates within the IBD population [26].

Our study confirms reduced IBD-related and non-IBD-related hospitalization rates among the IBD population, aligning with prior research [17,27]. However, the more rapid decline in non-IBD-related hospitalization rates was noted. Our findings reveal that hospitalizations related to IBD constitute a significant portion of the overall annual hospitalization rates: 81% for IBD in general, 90% for CD, and 76% for UC. As a result, while the overall trend in hospitalizations for all causes has been declining annually, this reduction may be slower when it comes to primary IBD-related hospitalizations. The decline in IBD-related admissions contributed substantially to the decrease in all-cause hospitalization rates, potentially easing healthcare strain and patient costs [28]. Clinical factors including early diagnosis, proactive monitoring, treatment advancements, and management strategies likely contribute [29,30]. Moreover, the classification of hospitalizations as IBD-related or non-IBD-related was based on discharge codes that included ICD codes associated with symptoms or diseases of IBD (Supplementary Table 1). This method introduces the possibility of both underestimating and overestimating hospitalization reasons due to potential inaccuracies in discharge coding. Further research is needed for comprehensive insight.

Our study showed a mean all-cause length of hospital stay of 9.6 days for IBD patients in Taiwan, longer compared to 4 days in Scotland and 4.5 days in Dallas [25,31]. Several factors are associated with prolonged length of hospital stay including Clostridioides difficile infection, malnutrition, and immunosuppression [32,33]. Specifically, NHI program provides comprehensive, universal coverage with very low co-payments for inpatient services. This system minimizes the financial barrier to admission and may reduce the economic pressure for early discharge that exists in other healthcare systems. This, combined with local clinical practice norms, may contribute to the longer mean length of hospital stay for IBD patients observed in our study compared to those reported in Scotland and Dallas. However, advancements in treatments have decreased overall length of hospital stay for patients with IBD [31]. It is noteworthy that while the overall 17-year trend (Fig. 3) showed a significant decline (P<0.05), the era-to-era comparison (Table 2) did not show a statistically significant break (P=0.779). This suggests the reduction in length of stay was a gradual, consistent improvement over the entire study period rather than a sudden drop specifically after 2011.

We observed significant drops in surgical and mortality rates among individuals with IBD, consistent with trends in England and the United States [34,35]. It should be noted, however, that the lack of age-specific mortality rate limits a direct comparison of our mortality trends to other population-based studies. In the early years of cohort (2001–2002), there was a tendency to include transitional epidemiologic dynamics or data instability due to the restricted look-back period within the NHI database cohort. Consequently, surgical and mortality rates during these initial years were likely exaggerated, causing a sharp decline in trends. However, the sensitivity analysis findings demonstrate that the overall trend estimates remained stable even when the initial 2 years were excluded from the analysis. The lack of significant difference supports the temporal consistency of observed trends and suggests that earlier data do not meaningfully bias long-term conclusions for these outcome measures. Improved management and treatment may reduce the need for surgeries and mortality in individuals with IBD. Despite declining mortality, monitoring risk factors like comorbidities, infectious diseases, and perioperative care remains crucial [36]. In Taiwan, a decline in hospitalization rates, surgical rates, and mortality rates was observed among patients with IBD between the prebiologic era (2001–2010) and the postbiologic era (2011–2017). Current studies have highlighted an association between increased use of biologic therapy and the decline of surgical rates in patients with IBD [24,29,37]. One retrospective study reported a decrease in emergency department visits in individuals with IBD after the introduction of biologics [38].

However, our subgroup analyses revealed key discrepancies. First, the nonsignificant drop in CD hospitalization rates (P=0.241), compared to the significant drop for UC (P=0.002), may be explained by several factors. CD’s transmural nature means biologics cannot reverse pre-existing structural damage (e.g., fibrotic strictures) that still requires hospitalization, whereas they can effectively heal UC’s mucosal inflammation [39]. Furthermore, “confounding by indication” likely masked the true effect, as the most severe CD patients were prioritized for treatment. Finally, CD biologics were approved in 2007, “contaminating” the pre-2011 baseline data, while UC reimbursement lagged, providing a cleaner “before-and-after” comparison that resulted in a statistically significant finding. Second, while those with CD experienced a notable decrease in mortality rates (P=0.012), patients with UC did not show a statistically significant reduction (P=0.476) (Table 2). This finding for UC likely stems from: (1) delayed NHI reimbursement for biologics, postponing their full clinical impact; (2) our use of all-cause mortality in an older-onset UC population (median; 45 years vs. 34 years) more influenced by non-IBD comorbidities; and (3) while our study found reduced overall UC surgical rates, this benefit may not extend to the most severe cases. A recent Taiwanese ASUC study using data beyond our 2017 observation period reported no significant difference in 3-year colectomy rates for severe cases [40]. This persistent surgical burden in high-risk subgroups, combined with the higher comorbidity load in an older-onset UC population, likely explains the nonsignificant change in all-cause mortality.

Further analysis is needed to explore the correlation between different biologic therapies, including TNF-α blockers, integrin blockers, and interleukin 12/23 blockers. However, due to the limited availability of these medications and the relatively smaller number of patients, we were unable to further analyze these conditions in this study.

The increasing IBD population in Taiwan presents a clinical challenge to gastroenterologists. Owing to the chronic nature and low mortality rate of the disease, both the prevalence and the average age of patients with IBD are increasing. Clinicians can thus expect increased clinical complexity and comorbidities in such patients. The optimization of practice support and a multidisciplinary approach in IBD units have contributed substantially to the improvement of clinical outcomes and the quality of IBD care [22].

This population-based cohort study has several limitations. First, the hospitalization numbers may be underestimated because an IBD diagnosis is required for a patient to apply for the CIC. Thus, patients without this certificate, potentially including those with early-stage IBD or with mild disease severity, were not represented. Second, we were unable to investigate the clinical status of the disease (e.g., laboratory data), activity index scores, smoking history, comorbidities, and medical therapies. Third, we defined 2011 as the commencement of the postbiologic era, as this was when anti-TNF therapies were broadly covered under the NHI program for IBD. However, we acknowledge that the specific reimbursement approval for UC lagged behind that for CD due to the drugs’ availabilities. This discrepancy in timing may have confounded the comparison of outcomes for UC, particularly length of stay in hospital and mortality rates, in the early postbiologic period. A further limitation is that our mortality analysis relies on all-cause mortality. Due to the nature of the administrative claims database, we were unable to verify the specific, direct cause of death (i.e., IBD-specific mortality vs. other causes). In conclusion, aside from specific endpoints like mortality outcomes, the observed long-term trends in IBD burden, hospitalization, and surgical rates are statistically consistent and stable across analytic intervals, supporting the validity of the primary findings.

One of the key strengths of this study is the use of a nationwide cohort comprising individuals with IBD listed on databases that cover nearly the entire population of Taiwan. This extensive population-based data helped to minimize selection bias.

In conclusion, a persistent upward trend can be observed in the incidence and prevalence of IBD in Taiwan that is similar to observations in other industrialized areas. The all-cause hospitalization rate in individuals with IBD relative to the general population increased because the IBD population grew more quickly than the general population did. Despite this rise in IBD prevalence, the all-cause hospitalization rates relative to the IBD population decreased. We also noted a significant decrease in the length of hospital stay in the patients with IBD over our study period. The reduction in hospitalization rates per 100 patients with IBD and length of hospital stay could be ascribed to advancements in clinical therapies. Our findings—demonstrating that the hospitalization burden per patient can be controlled and even reduced through advances in clinical management despite a rapidly rising IBD prevalence—offer important policy insights for other Asian regions that are undergoing similar epidemiological transitions. This highlights the value of health policy planning and ensuring access to advanced therapies to manage the evolving burden of IBD.

Supplementary Material

Supplementary materials are available at the Intestinal Research website (https://www.irjournal.org).

Article information Continued