Risk of kidney failure in patients with inflammatory bowel disease undergoing colectomy: a nationwide cohort study

Abstract

Background:

Inflammatory bowel disease (IBD) is frequently accompanied by kidney complications. Potential triggers or subpopulations at high-risk of kidney problems are not well elucidated. We hypothesized that surgical interventions, specifically colectomy, might in part explain this risk.

Methods:

Nationwide Swedish cohort study comprising 82,051 individuals with biopsy-proven IBD diagnosed during 1965–2017, with follow-up until 2019. We investigated the association between incident colectomy (time-varying exposure) and future risk of acute kidney injury (AKI) and kidney failure (diagnosis of end-stage kidney disease or death due to chronic kidney disease) using Cox proportional hazard models. We also examined the impact of partial vs. total colectomy and the presence/duration of a stoma. Covariates included demographics, education level, and selected comorbidities.

Results:

Over a median follow-up of 14 years, 16,479 individuals underwent colectomy, and 2,556 AKI and 1,146 kidney failure events occurred. Colectomy was associated with an increased relative risk of both AKI (adjusted hazard ratio [aHR] 2.37; 95%CI 2.17–2.58) and kidney failure (1.54; 1.34–1.76). Compared to pre-colectomy periods, undergoing total colectomy and colectomy with prolonged stoma showed higher risks of both kidney outcomes versus partial colectomy or colectomy with a temporary stoma, respectively. Subgroup analyses suggested higher risks in patients with ulcerative colitis.

Conclusions:

In people with IBD, rates of AKI and kidney failure are higher among those undergoing colectomy, particularly among those following total colectomy, or colectomy with a prolonged stoma. This study identifies a high-risk population that may benefit from established protocols for kidney function monitoring/surveillance and referral to nephrologist care.

INTRODUCTION

Inflammatory bowel disease (IBD), comprising Crohn’s disease (CD) and ulcerative colitis (UC), is primarily a gastrointestinal disorder that also often presents with extraintestinal manifestations.¹ Among these, kidney complications such as tubulointerstitial nephritis and chronic kidney disease, are common, affecting approximately 10% of patients.²

Surgical treatment of IBD may be indicated when conventional pharmacological approaches are insufficient in managing symptoms or addressing disease complications.³ Approximately 50% of patients with CD and around 20% of those with UC undergo IBD-related surgery during their lifetime.^4–6 Colectomy is a common surgical procedure in patients with IBD that involves removal of the affected part of the colon and subsequent restoration of intestinal continuity.⁷ Although these procedures can successfully treat ongoing intestinal inflammation, post-colectomy hydro-electrolytic disorders may result in extracellular fluid volume depletion and stone crystal formation, which have the potential to lead to impairment in kidney function.⁸

To the best of our knowledge, no study has investigated the association of IBD-related colectomy and kidney complications in population-based cohorts of IBD patients (Supplementary Search Strategy). We hypothesized that the aforementioned association is plausible, given the hydro-electrolytic disorders and nephrolithiasis related to colectomy. In addition, patients requiring colectomy are a subset of sicker patients that may be more susceptible to kidney problems and/or may require intensified of IBD treatment which in turn can induce nephrotoxicity.⁹

In this nationwide cohort study, we aimed to explore the association between colectomy and subsequent acute kidney injury (AKI) and kidney failure in patients with IBD. Early identification of populations at risk of kidney complications may allow the initiation of appropriate pharmacological and lifestyle interventions to prevent or delay their onset.

METHODS

Study population and cohort design

Observational cohort study from the ESPRESSO (Epidemiology Strengthened by histoPathology Reports in Sweden) cohort, which contains all biopsy-confirmed IBD diagnoses in Sweden since 1965 (Supplementary Methods).¹⁰ We included all individuals aged ≥6 years with newly diagnosed IBD during 1965–2017, and without any history of colectomy or kidney failure. IBD was defined by a histopathology code for IBD and at least one relevant International Classification of Diseases (ICD) code in either outpatient specialist or inpatient care (Supplementary Table 1).^{10, 11} The date of the first relevant histopathology or ICD code constituted the index date and study baseline. We also considered the subtypes of IBD, namely CD, UC and IBD unclassified (IBD-U). Patients with the same IBD subtype in both diagnostic listings were considered as having either CD or UC, whereas patients with discordant IBD diagnostic listings or at least 1 listing of IBD-U were defined as having IBD-U (Supplementary Table 1).^{12, 13}

Study exposure

The study exposure was colectomy, which was defined by the presence of a relevant surgical procedure code (6^th version of Swedish surgical procedure codes until 1996, and thereafter the Nordic Medico-Statistical Committee [NOMESCO] procedure codes) (Supplementary Table 2)¹⁴ during follow-up (time-varying exposure). Thus, a patient who underwent colectomy contributed person-time to the non-colectomy exposed group during the time pre-colectomy, and to the colectomy-exposed group thereafter. For simplicity, we term these two periods as pre- and post-colectomy. For our main analysis, we treated colectomy as a binary variable (pre-colectomy vs. post-colectomy). We hypothesized that the amount of colon resection would be directly associated with the risk of kidney outcomes. To that end, we repeated our analyses treating colectomy as a ternary time-varying variable, comparing pre-colectomy vs. post-partial-colectomy vs. post-total-colectomy.

Study covariates

Covariates were defined at baseline and updated at time of colectomy. Covariates included age at IBD diagnosis, sex, highest attained education, calendar year of IBD diagnosis, and selected comorbidities (listed in Table 1 and detailed definitions in Supplementary Table 3). Another covariate in the model was having a stoma formation surgical procedure (defined as a relevant surgical procedure code in Supplementary Table 2). We updated the history of stoma at the time of its first occurrence, independent of colectomy (time-varying covariate).

Table 1.

Characteristics of study participants at baseline and at the time of colectomy.

	Overall (n=82 051)	At colectomy (n=16 479)	At partial colectomy (n=7 672)	At total colectomy (n=8 807)
Age, mean±SD	40±19	43±18	43±19	43±17
<40 years	45 291 (55%)	7 957 (48%)	3 712 (48%)	4 245 (48%)
40–59 years	22 016 (27%)	4 955 (30%)	2 180 (28%)	2 775 (32%)
60–69 years	7 986 (9.7%)	1 912 (12%)	906 (12%)	1 006 (11%)
70+ years	6 758 (8.2%)	1 655 (10%)	874 (11%)	781 (8.9%)
Women, n (%)	39 578 (48%)	7 744 (47%)	4 057 (53%)	3 687 (42%)
Education, n (%)
Compulsory school (≤9 years)	17 523 (21%)	3 877 (24%)	1 901 (25%)	1 976 (22%)
Secondary school (10–12 years)	36 096 (44%)	7 622 (46%)	3 556 (46%)	4 066 (46%)
University (≥13 years)	26 878 (33%)	4 767 (29%)	2 117 (28%)	2 650 (30%)
IBD subtype, n (%)
Crohn’s disease	22 598 (28%)	6 911 (42%)	5 786 (75%)	1 125 (13%)
Ulcerative colitis	47 222 (58%)	7 600 (46%)	1 223 (16%)	6 377 (72%)
IBD unclassified	12 231 (15%)	1 968 (12%)	663 (8.6%)	1 305 (15%)
Comorbidities, n (%)
COPD	1 380 (1.7%)	504 (3.1%)	291 (3.8%)	213 (2.4%)
Alcoholism *	1 825 (2.2%)	449 (2.7%)	203 (2.6%)	246 (2.8%)
Diagnosed obesity	741 (0.9%)	232 (1.4%)	120 (1.6%)	112 (1.3%)
Hypertension	4 981 (6.1%)	1 650 (10%)	836 (11%)	814 (9.2%)
Diabetes mellitus	2 282 (2.8%)	818 (5.0%)	303 (3.9%)	515 (5.8%)
Dyslipidemia	1 185 (1.4%)	283 (1.7%)	124 (1.6%)	159 (1.8%)
Hyperuricemia & gout	211 (0.3%)	84 (0.5%)	45 (0.6%)	39 (0.4%)
Cardiovascular diseases †	3 956 (4.8%)	1 200 (7.3%)	558 (7.3%)	642 (7.3%)
Congestive heart failure	1 070 (1.3%)	343 (2.1%)	178 (2.3%)	165 (1.9%)
Myocardial infarction	1 595 (1.9%)	422 (2.6%)	199 (2.6%)	223 (2.5%)
Atrial fibrillation	1 219 (1.5%)	444 (2.7%)	213 (2.8%)	231 (2.6%)
Stroke	1 339 (1.6%)	369 (2.2%)	175 (2.3%)	194 (2.2%)
Cancer (recent 3 years) ‡	1 701 (2.1%)	1 520 (9.2%)	799 (10%)	721 (8.2%)
Colorectal cancer (recent 3 years)	107 (0.1%)	1 000 (6.1%)	495 (6.5%)	505 (5.7%)
Colorectal cancer (ever)	129 (0.2%)	1 017 (6.2%)	504 (6.6%)	513 (5.8%)
Diverticular disease	2 305 (2.8%)	968 (5.9%)	667 (8.7%)	301 (3.4%)
Volvulus	25 (<0.1%)	59 (0.4%)	55 (0.7%)	4 (<0.1%)
Chronic kidney disease §	559 (0.7%)	251 (1.5%)	118 (1.5%)	133 (1.5%)

Abbreviations: SD, standard deviation; IBD, inflammatory bowel disease; COPD, chronic obstructive pulmonary disease.

^*Alcoholism is defined as alcohol abuse or alcohol-related liver disease.

^†Cardiovascular diseases include congestive heart failure, myocardial infarction, stroke, and atrial fibrillation.

^‡Cancer is defined as having a diagnosis of cancer in the past three years prior to study baseline.

^§Chronic kidney disease is defined as any related ICD codes at baseline (incident IBD diagnosis) or at the time of colectomy.

Study outcomes

The study outcomes were acute kidney injury (AKI) and kidney failure (chronic). AKI was defined as the first time receiving a diagnosis of acute nephritis or acute renal/kidney failure, enriched with procedural codes of acute dialysis therapy. Likewise, kidney failure included clinical diagnoses of end-stage kidney disease (including maintenance dialysis or kidney transplantation), relevant procedure codes (such as permanent implantation of a dialysis catheter) as well dying with a diagnosis of chronic kidney disease as a primary cause of death. Definitions are detailed in Supplementary Table 4. Participants were followed from IBD diagnosis until the end of follow-up (latest December 31, 2019), death, or emigration from Sweden, whichever occurred first. Date of death was obtained from the National Board of Health and Welfare’s Cause-of-death register.¹⁵

Data analyses

Descriptive statistics for continuous variables were reported as mean ± standard deviation (SD), and categorical variables as counts and percentages. We calculated incidence rates of both kidney outcomes with 95% confidence intervals (CI).

To estimate relative risks (hazard ratios) associated with colectomy (pre-colectomy vs. post-colectomy), while controlling for study covariates, we fitted a multivariable Cox proportional hazards model with colectomy as a time-varying exposure. (Analysis 1 in the study flowchart, Supplementary Figure 1). Variables adjusted for are all those mentioned in the section “covariates”. We next fitted a similar Cox model exploring the association between partial and total colectomy on subsequent kidney event risks.

Sensitivity analyses

We explored the consistency of our results across subgroups of age, sex, highest attained education, enrollment periods and IBD subtypes. To that end, we tested for signs of statistical heterogeneity by adding an interaction term between the stratifying factor and study outcome. We also explored whether risks changed over time by splitting the follow-up time at 6-month, 1-year, 10-year, 20-year, 30-year, and 40-year intervals, we censored the follow-up at the end of each time interval and modeled the relative risks during each time intervals using Cox proportional hazard regression as well as flexible parametric relative survival models.¹⁶

Finally, we explored the impact of additional covariate adjustment. This could be done for patients diagnosed since 2006, for whom we were able to enrich our control for confounding with information on use of medications recorded in the National Prescribed Drug Register.¹⁷ Specifically, we enriched the ascertainment of diabetes mellitus, hyperuricemia and gout with information on recent dispensation of related medications using anatomical therapeutic chemical (ATC) codes. Additionally, we added sequential adjustment for the use of proton pump inhibitors (PPIs) and non-steroidal anti-inflammatory drugs (NSAIDs), both of which may exert nephrotoxicity and be associated with development of IBD,^{18, 19} as well as IBD-related medications (i.e., aminosalicylates, corticosteroids, immunomodulators, and biologic agents). Algorithms to define ongoing medications are detailed in Supplementary Table 3.

Supporting analyses

We explored potential underlying mechanisms. First, we evaluated whether adjusting for stoma placement (time-varying covariate, presumably within the causal pathway) impacted our observed risks. In addition, we studied whether prolonged stomas would result in higher risk of kidney complications than temporary stomas. To that end, patients undergoing colectomy were followed for two years to evaluate if they had a stoma placed, and whether it was temporary or prolonged. We defined a prolonged (permanent) stoma as either stoma in combination with proctectomy or non-closure within 2 years of formation of stoma. Temporary stomas were those in which a closure surgical procedure was found within 2 years of stoma formation.²⁰ Conditional on surviving this 2-year period, we set a new index date at this point, updated the covariates and began follow-up (Analysis 2 in the study flowchart, Supplementary Figure 1).

Second, we performed a mediation analysis using the Baron & Kenny method²¹ to explore whether the association between colectomy and kidney failure risk was mediated by intermediate AKI (hypothesized causal diagram in Supplementary Figure 7).

Finally, we calculated the population attributable fraction (PAF) of colectomy with the “graphPAF” R package²² and 1000 times bootstrapping for the 95%CI. PAF is interpreted as the fraction of all kidney failure events that is attributable to experiencing colectomy,²³ and it is a metric used often in population health projections which is inherently affected by confounding in observational research. We performed all analyses using R software (R version 4.0.5).

RESULTS

Baseline characteristics

After applying inclusion and exclusion criteria (Supplementary Figure 1), we included 82,051 individuals with an incident IBD diagnosis (22,598 with CD, 47,222 with UC, and 12,231 with IBD-U) that did not have a history of colectomy. Their mean age was 40±19 years, 48% were women, 6.1% had diagnosed hypertension, 2.8% diabetes, and 0.7% a clinical diagnosis of chronic kidney disease at baseline (Table 1). There were 10,303 individuals enrolled between 1965–1986, 41,141 individuals between 1987–2005, and 30,607 individuals between 2006–2017 (Supplementary Tables 6–9).

During a median follow-up time of 14 years (interquartile range [IQR] 7.3–21.7) we identified 16,479 (20.1%) individuals who underwent colectomy, among them, 6,911 with CD, 7,600 with UC, and 1,968 with IBD-U. The median time from IBD diagnosis to colectomy was 3.4 years (IQR 0.7–9.4). Among colectomy cases, 7,672 had partial colectomy and 8,807 had total colectomy. At the time of colectomy, their mean age was 43±18 years, 47% were women, 10% had hypertension, 5% had diabetes, and 1.5% had a diagnosis of chronic kidney disease (Table 1).

Main analyses

During follow-up, 2,556 events of AKI and 1,146 events of kidney failure occurred. Compared with pre-colectomy periods, the incidence rates (IR) of both AKI and kidney failure were higher post-colectomy, resulting in a higher adjusted hazard ratio (aHR) of 2.37 (95%CI 2.17–2.58) and 1.54 (95%CI 1.34–1.76), respectively (Figure 1). In subgroup analyses, risks were higher in UC patients compared with CD patients (for example, risks of AKI were 2.88 [95%CI 2.55–3.24] and 1.76 [95%CI 1.50–2.07] respectively, P for interaction 0.011), but remained elevated in both IBD subtypes (Supplementary Figure 3).

Figure 1. Relative risk of acute kidney injury and kidney failure (chronic) undergoing colectomy.

Abbreviations: IR, incidence rate; py, person-years; aHR, adjusted hazard ratio; CI, confident interval.

* aHR was calculated from Cox proportional hazard model and adjusted for age at IBD diagnosis, sex, highest attained education, calendar year, and COPD, diagnosed obesity, chronic kidney disease, hypertension, diabetes, dyslipidemia, hyperuricemia and gout, cardiovascular diseases (including congestive heart failure, myocardial infarction, stroke, and atrial fibrillation), cancer in the past three years. Participants who were prevalent cases were excluded for each study outcome.

We observed a “dose-response” relationship when evaluating the amount of colon resection: For both AKI and kidney failure, compared to pre-colectomy periods, the relative risks were higher post-partial-colectomy (AKI: aHR 1.64, [95%CI 1.44–1.86]; kidney failure: aHR 1.19 [95%CI 0.98–1.44]), and highest post-total-colectomy (AKI: aHR 2.98 [95%CI 2.70–3.28]; kidney failure: aHR 1.89 [95%CI 1.60–2.22]; both P for trend <0.001) (Figure 1). We observed a similar “dose-response” across IBD subtypes (Supplementary Figure 4).

Sensitivity analyses

Results were consistent across subgroups of sex, age at IBD diagnosis, and education. There was suggestion of heterogeneity (P for interaction <0.05) between IBD subtypes, with AKI risk being higher for UC patients than CD patients. However, risk magnitudes were elevated across all IBD subtypes (Supplementary Figure 3).

Absolute and relative risks of AKI and kidney failure were the highest within the first 6–12 months post-colectomy, and AKI risks remained elevated (compared with pre-colectomy periods) throughout the entire follow-up (Supplementary Table 13 and Supplementary Table 14), and kidney risks were higher after undergoing total colectomy compared to partial colectomy (Figure 3). We found similar time trends with flexible parametric models (Supplementary Figure 6).

Figure 3.

Cumulative incidence plots of acute kidney injury (Panel A) and kidney failure (Panel B) in patients with IBD undergoing partial or total colectomy.

Adjusting for ongoing use of medications in the more recent subset of patients where this information was available did not modify our conclusions: The associations between colectomy and subsequent AKI were strong and statistically significant. The associations with the rarer event of kidney failure (n=28) remained of elevated magnitude (on the range of 45% higher risk, like previous time-periods) but affected by broad confidence intervals (Supplementary Table 11).

Supporting analyses

Adjusting within the causal pathway for the presence of a stoma reduced the magnitude of AKI risks and abrogated the association with kidney failure risks (Supplementary Figure 2). Among 16,391 IBD patients that underwent colectomy and had a follow-up ≥2 years, 10,889 did not have a stoma, 4,438 had a stoma with closure within 2 years (temporary stoma), and 1,064 had a permanent stoma (Supplementary Table 10). Compared to those who did not have a stoma, the relative risks of both kidney outcomes were higher among those with a temporary stoma (AKI: aHR 1.35 [95%CI 1.16–1.58]; kidney failure: aHR 1.30 [95%CI 1.00–1.69]), and highest among those with a permanent stoma (AKI: aHR 2.26 [95%CI 1.82–2.81]; kidney failure: aHR 2.36 [95%CI 1.66–3.36]; both P for trend <0.001) (Figure 2). We observed similar “dose-response” relationships in AKI risks across IBD subtypes (Supplementary Figure 5).

Figure 2. Relative risks of acute kidney injury and kidney failure (chronic) in IBD patients experiencing colectomy, stratified by the presence of stomas.

Abbreviations: IR, incidence rate; py, person-years; aHR, adjusted hazard ratio; CI, confident interval.

* aHR was calculated from Cox proportional hazard model and adjusted for age at IBD diagnosis, sex, highest attained education, calendar year, and COPD, diagnosed obesity, chronic kidney disease, hypertension, diabetes, dyslipidemia, hyperuricemia and gout, cardiovascular diseases (including congestive heart failure, myocardial infarction, stroke, and atrial fibrillation), cancer in the past three years. Participants who were prevalent cases were excluded for each study outcome.

Mediation analyses showed a total effect of 1.57 (95%CI 1.36–1.80) and a direct effect of 1.13 (95%CI 1.05–1.22), indicating that 27.9% (95%CI 1.6–143.9%) of the effect of colectomy on kidney failure events was mediated through experiencing AKI (Supplementary Table 12). Finally, PAF analyses suggested that 9.2% (95%CI 4.7–16.2%) of kidney failure events could be attributed to undergoing colectomy.

DISCUSSION

In this large nationwide cohort study, we provide robust observational evidence that individuals with IBD who undergo colectomy have higher risks of AKI and kidney failure. The risk of kidney complications (both acute and chronic) was higher among individuals undergoing total colectomy, and among those with a prolonged stoma. Results were consistent across subgroups, and robust to a range of sensitivity analyses. Collectively, our study identifies a high-risk population who may benefit from established protocols for kidney function monitoring/surveillance and referral to nephrological care.

To the best of our knowledge, this is the sole study to date suggesting an association between colectomy and subsequent risk of AKI and kidney failure among individuals with IBD. Precise underlying mechanisms are unknown and are likely multifactorial. Colectomy – especially in case of stoma with high ileostomy output – may induce metabolic disorders which may further increase kidney impairment. A major function of the colon is the reabsorption of water and electrolytes. After colectomy, decreased urine pH, together with low urine volume, contribute to urine uric acid supersaturation and stone formation.²⁴ In addition, beyond nephrolithiasis risks, chronic loss of water and sodium lead to chronic extracellular fluid volume depletion, which in turn, induces stimulation of renin-angiotensin system and may add to the risk of kidney impairment.²⁵ Lastly, serving as a crucial digestive organ, the colon hosts a substantial microbial population. Alterations in the composition of this microbiota result in the overproduction of uremic toxins, along with a reduced presence of reno-protective metabolites. These imbalances have been implicated in various processes, including oxidative stress, uremia, inflammation, the decline of kidney function, and the progression of chronic kidney disease.²⁶ Our analyses add to this knowledge in several ways: Stomas may be an important player in this risk, as we observed the association between colectomy and kidney failure being sensitive to the adjustment for stoma. We also showed that patients with prolonged stomas were at highest risk of kidney complications. Second, people experiencing AKI may progress to kidney failure,²⁷ and we could identify AKI as an important mediator in our analysis, but it did not fully explain the direct effect of colectomy on kidney failure risks. While AKI risks were highest in the months following colectomy, they remained elevated over time. It is thus possible that colon dissection has long-term sequelae that may predispose to adverse kidney outcomes (both acute and chronic) in the long-term.

All observational studies are inherently affected by confounding. It is possible that IBD patients undergoing colectomy represent a sicker population more prone to immune-mediated kidney manifestations of IBD per se, or in need of more frequent use or higher dosages of IBD-specific medications that may exert nephrotoxicity. Our observation that both AKI and kidney failure risks were the highest during the first 6–12 months post-colectomy may reflect this, but we also demonstrate that risks were maintained elevated during the subsequent 40 years of follow-up. Furthermore, adjustment for ongoing use of IBD-specific nephrotoxic medications did not abrogate the associations. Interestingly, we observed risk of kidney events post-colectomy that were higher in magnitude for UC patients compared with CD patients, consistent with previous studies.² We attribute this to the fact that patients with UC were more likely to undergo total colectomy than those with CD.

While this observation is novel in the field of IBD, we note that some studies in the setting of colon cancer have shown that the presence of a high-output stoma is correlated with an elevated risk of AKI and chronic kidney disease.^28–30 Of these, a recent Canadian study investigated relationships between ileostomy and subsequent kidney diseases in 19,889 patients. They found that the presence of ileostomy formation was associated with community onset AKI (adjusted odds ratio [aOR] of 4.08), and new-onset chronic kidney disease (aOR of 4.99 and 2.45 for those with and without previous community-onset AKI episodes).²⁸

Strengths of our study include a nationwide study population of individuals with biopsy-confirmed IBD and >40 years of follow-up that provide power to quantify the long-term and rare event of kidney failure. We also used validated definitions of IBD and surgical procedure codes, with a positive predictive value of 95%.^{11, 14} Our study also has some limitations: we used data spanning from 1965 to 2017, a long period during which both ICD codes and procedure codes underwent evolution, and this may have introduced challenges to the accurate identification of our exposures, outcomes, and comorbidities; We lacked information on prescribed medications before 2005; The algorithms to define AKI and kidney failure align with previous work, but have not been validated in Swedish settings; We modelled AKI as a single outcome, and did not account for the possibility that some patients may experience multiple AKI events over time; The study population was mainly western, limiting the generalizability to ethnically diverse cohorts; Finally, we lacked information on body mass index, dietary habits, lipid profiles or IBD severity.

In conclusion, this observational study credibly illustrates that people with IBD experiencing colectomy are at increased risk of both AKI and kidney failure, particularly those with total colectomy and those needing (prolonged) stomas. Our PAF analysis suggested that up to 9% of kidney failure events could be attributed to colectomy in this nationwide cohort. As a clinical implication, patients undergoing colectomy represent a particular high-risk group that may benefit from more close monitoring of kidney biomarkers. This may help to early identify signs of kidney damage that can prompt minimizing the exposure to nephrotoxic medications, initiating antiproteinuric therapies, or timely referring to nephrological care.³¹

What you need to know:

Background

Kidney complications are common in inflammatory bowel disease. We explored the hypothesis that colectomy may, in part, explain this risk.

Findings

In a nationwide sample of 82,051 people with newly diagnosed inflammatory bowel disease, we report that those undergoing colectomy were at a higher risk of both acute kidney injury and kidney failure, particularly following total colectomy or colectomy with a prolonged stoma.

Implications for patient care

Our study identifies a population at high risk of kidney complications that may benefit from established protocols for kidney function surveillance and referral to nephrologist care.

Abbreviations:

aHR

Adjusted hazard ratio

AKI

Acute kidney injury

aOR

adjusted odds ratio

ATC

Anatomical therapeutic chemical

CD

Crohn’s disease

CI

Confidence interval

COPD

Chronic obstructive pulmonary disease

ESKD

End stage kidney disease

ESPRESSO

Epidemiology Strengthened by HistoPathology Reports in Sweden

IBD

Inflammatory bowel disease

IBD-U

IBD unclassified

ICD

International Classification of Diseases

IQR

Interquartile range

IR

Incidence rate

LISA

Longitudinal integrated database for health insurance and labor market studies

NOMESCO

Nordic Medico-Statistical Committee

NSAIDs

Non-steroidal anti-inflammatory drugs

PPIs

Proton pump inhibitors

SD

Standard deviation

SNOMED

Systematized Nomenclature of Medicine

UC

Ulcerative colitis