Inflammatory Bowel Disease Population Analyzer Tool for Health Systems (IBD PATH): a case study risk stratifying patients with inflammatory bowel disease for clinical outcomes

Introduction

Inflammatory bowel disease (IBD) including Crohn’s disease (CD) and ulcerative colitis (UC) are chronic diseases with relapsing and remitting cycles. IBD affects 1.2–3.1 million individuals in the United States (U.S.) and the incidence is steadily increasing worldwide (1-3). In the U.S., the cost of IBD care is estimated at $23,000 per patient per year, underscoring the substantial financial burden (4).

According to the 2018 American College of Gastroenterology (ACG) guideline, 70–80% of patients with CD will have progressive disease based on data from long periods of observation (5). For patients with UC, approximately 13% experience cumulative 5-year risk of progression (6). The choice of treatment is influenced by the severity of the disease and risk of progressive disease. The American Gastroenterology Association’s (AGA’s) clinical pathways for CD and UC determine the appropriate treatment options based whether a patient has low or moderate/high risk disease (7,8). Thus, identifying patients who may be prone to disease progression is important to help tailor their healthcare and could potentially reduce the rising expenses linked to managing IBD.

The Institute for Healthcare Improvement’s (IHI’s) Triple Aim—improving the health of populations, improving experience of care, and reducing per capita cost of health care—has informed population health strategies included in the Patient Protection and Affordable Care Act and serves as a guide for population health management (PHM) processes (9,10). Recently, IHI has endorsed expansion of the Triple Aim to a Quintuple Aim by adding the following aims: recognizing and reducing burnout among the health care workforce and advancing health equity (11). In response to the call for a focus on quality, value, and equity, health care organizations have incorporated PHM strategies to achieve the goals of the fifth aim. These strategies encompass various activities, such as identifying at-risk populations and disparities, developing targeted, evidence-based interventions, monitoring health outcomes over time, and investing in equity measurement.

For effective PHM, data analytics is critical to identifying patterns and trends in health data (including disparities), which can then inform the development of targeted interventions and improve equity and care coordination. However, conducting these analyses can be complex and beyond the resources of many healthcare organizations. The IBD Population Analyzer Tool for Health Systems (IBD PATH) was developed to allow organizations to evaluate the risk of disease progression in their IBD population and to identify gaps in care based on retrospective medication and medical record analysis. We used IBD PATH to conduct a real-world case study to identify data standardization gaps and facilitate PHM efforts across all sites at Ochsner Health (Ochsner).

Methods

Tool development process

The development process, detailed below, consisted of the following steps: (I) select data and analysis parameters; (II) develop storyboard framework and model; and (III) program and test software.

For step I, EPI-Q evaluated and incorporated data parameters based on current guidelines for IBD management in the U.S., literature, and expert opinion such as risk assessments, medication utilization, and quality indicators. Based on the data parameters selected, step II was to construct a storyboard and Excel model to detail the design and data output. The storyboard was developed by EPI-Q and the Janssen Scientific Affairs, LLC, Real World Evidence Team. The outline of the data screens and variables included in the tool are listed below.

• IBD population overview:
  • Total number of patients with an IBD diagnosis code and proportion with a CD and/or UC diagnosis code.
    • If a patient has a diagnosis code for both CD and UC in the data, then they will be represented in both conditions.
  • Proportion of patient at moderate/high risk and low risk of poor outcomes according to AGA risk factors for CD (7).
    • Proportion of patients with documented AGA risk factors for moderate/high CD which include: “age <30 years at diagnosis, extensive anatomic involvement, perianal disease, severe rectal disease, deep ulcers, previous surgical resection, stricturing behavior, and penetrating lesion behavior” (7).
  • Proportion of patient at moderate/high risk and low risk of poor outcomes according to AGA risk factors for UC (8).
    • Proportion of patients with documented AGA risk factors for moderate/high UC which include: “age <40 years, extensive colitis, steroid-requiring disease, deep ulcers, history of hospitalization, high C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), Clostridium difficile infection, and cytomegalovirus infection” (8).
• Medication utilization for CD & UC (7,8):
  • Proportion of patients with a CD or UC treatment medication order, reported by class and drug.
    • Medication utilization represents all medication records in the follow-up period, not just current medications. Therefore, medication utilization may be overestimated, especially combination therapies and steroid use.
• Quality indicators for CD & UC:
  • Proportion of patients with IBD-related healthcare resource utilization [hospitalizations, emergency department (ED) visits, surgeries, and procedures].
  • Proportion of patients with and without evidence of steroid therapy.
  • Proportion of patients with evidence of psychosocial screening (e.g., PHQ-9).
  • Proportion of patients with evidence of narcotic analgesic use.
• Patient level actionable report:
  • Provides a listing of the data for each patient.

The tool also designed to allow the data to be stratified by AGA moderate/high and low risk patients (7,8) or race/ethnicity groups.

In the final development step, EPI-Q’s software developers developed and programed the software tool. Figures S1-S7 illustrate how the data is uploaded into the tool and mapped for analyses by the tool. The data structure and formats from sample administrative claims and electronic medical record (EMR) datasets informed the data construct and requirements. Additionally, EPI-Q created sample data that was used to validate the calculations and continue to reside in the tool for the purposes of demonstrating functionality. The tool was tested by members of the EPI-Q Development Team, and Janssen Scientific Affairs, LLC, Real World Evidence Team. The completed IBD PATH can now be accessed on the Janssen Science website, https://www.janssenscience.com/therapeutic-areas/immunology/population-health-tools.

Ochsner use case study methods

Ochsner was engaged for the use case study to evaluate the feasibility and usability of the tool. Ochsner EMR data included patients with outpatient IBD visits between January 2020 and December 2021. Medication name and ordered date, visit dates and associated diagnosis and procedure codes were structured EMR data that was extracted from their EMR into a .CSV file. Additionally, Ochsner used a standardized template included in IBD PATH to collect the unstructured AGA clinical risk factor data, as listed above, from IBD procedure notes in the EMR on a subset of patients. Both data files were formatted per the tool specifications and uploaded into the tool (Figures S5-S7). All results reported are based on the descriptive analyses performed and generated by IBD PATH. The utility of the tool was assessed through interviews with a supervisor project lead, clinical informatics specialist, specialty pharmacist, IBD gastroenterologist, and the Assistant Vice President of Outcomes Research at Ochsner. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Ochsner Clinic Foundation IRB (No. FWA00002050) and individual consent for this retrospective analysis was waived.

Results

Population overview

The analyses of the Ochsner population data revealed AGA risk was not documented in the EMR. Based on the structured EMR data and data collected for the risk factors, a total of 164 patients (124 CD, 71 UC) were included in the analysis (Table 1). There were 31 patients with diagnosis codes for both CD and UC. Given the tool counts patients based on diagnosis records, these patients are counted in each condition.

Risk stratification

The majority of cases (82% CD, 77% UC) were classified as moderate/high risk. The primary risk factors for those with moderate/severe CD were perianal disease (81%) followed by severe rectal disease (31%). For those with moderate/severe UC, 78% of patients were noted to have previous steroid-requiring disease and 60% had extensive colitis as primary risk factors. Nearly a fourth of patients with CD and UC were also considered moderate/severe based on age at diagnosis and current age, respectively.

Of the patients at moderate/high risk, 34% of patients with CD (n=35) and 29% of patients with UC (n=16) had a biologic medication order (Table 2). Additionally, only up to 39 (38%) of patients with moderate/high risk CD and 28 (51%) of patients with moderate/high risk UC had an IBD medication order for a biologic, immunomodulator, or 5-aminosalicyclic acid. Thus, over 50% of the patients at moderate/high risk were untreated or did not have medication orders within the follow-up period. Overall, the recorded steroid orders were low; however, majority of patients that had recorded steroid orders were moderate/high risk (Figure 1). Narcotic analgesic use was also high in those with moderate/high risk.

Figure 1 Healthcare resource and medication quality indicators among patients with CD and UC stratified by AGA risk level. The quality indictors percentages for the moderate/high and low risk columns are based on total number of patients in the row, not on the total number of moderate/high or low risk patients. CD, Crohn’s disease; mod, moderate; UC, ulcerative colitis; IBD, inflammatory bowel disease; AGA, American Gastroenterology Association.

Patients at moderate/high risk were also more likely to have had IBD-related hospitalizations, ED visits, and IBD-related procedures in the follow-up period (Figure 1). Overall, 20% of the CD patients and 11% of patients with UC had an IBD-related hospitalization. Of the patients with a hospitalization, the patients at moderate/high risk constituted 88% of the patients with CD and UC. Additionally, of the patients with an IBD-related ED visit (44% CD, 30% UC), over 80% were patients with moderate/severe disease.

Race stratification

IBD PATH allows users to stratify the results by race. Of the 164 patients with race documented, 105 (64%) were White, 54 (33%) Black, and the remaining patients were noted as American Indian/Alaskan Native, other, or unknown. Due to the small numbers of these other racial groups, they were not included in the analysis.

For CD, 86% of Black patients and 80% of White patients were noted to have moderate/severe disease (Table 3). Perianal disease was the most prevalent risk factor among the groups; however, Black patients had greater percentages of extensive anatomic involvement (11% Black, 3% White), severe rectal disease (36% Black, 29% White), and penetrating behavior (31% Black, 22% White). Deep ulcers were documented in 11% of White patients with CD and only 3% of Black patients. Majority of the Black and White patients with UC analyzed also had moderate/severe disease (88% and 70%, respectively). The two most common risk factors reported were steroid-requiring disease (95% Black, 68% White) and extensive colitis (55% Black, 61% White).

In the follow-up period, a higher percent of Black patients with CD and UC received a prescription order for biologic medications (38% Black CD and 30% White CD; 28% Black UC and 24% White UC; Table 4). Evidence of narcotic analgesic orders was also higher in the Black patients for both CD and UC, while steroid orders was similar.

Regarding healthcare resource utilization, 64% of Black patients with CD and 48% of Black patients with UC utilized the ED for IBD-related issues whereas 34% of White patients with CD and 21% of White patients with UC had an IBD-related ED visit in the follow-up period (Table 4). Only 20% of patients with CD and 11% of patients with UC had an IBD-related hospitalization. The percent of Black patients with an IBD-related hospitalization was 14% for CD and 16% for UC while 24% of the White patients with CD and 10% of the White patients with UC had an IBD-related hospitalization. IBD-related procedures among the groups with UC were 40% in the Black patients with UC versus 31% of White patients with UC.

Discussion

In this case study, we found that nearly 80% of IBD cases were in the moderate/high risk category according to the AGA clinical pathways. The AGA clinical pathways were created to help clinicians identify a patient’s risk level and determine the appropriate treatment to control their IBD and prevent progression (7,8). However, the AGA risk stratification can only be determined based on reviewing and interpreting IBD procedure notes and other clinical and medication data versus a yes/no structured EMR field for each factor or the risk category. The Disease Severity Index (DSI) for IBD is another measure developed and recently validated to predict poor (12). Compared to the AGA, the DSI incorporates additional clinical factors, biologic use, steroid use for CD, and patient-centered factors such as the impact of disease on daily activity and symptoms. While the inclusion of patient-centered attributes is important, they present additional data assessment and documentation challenges. This lack of standardization assessing and documenting a patient’s risk could in part explain why only a third of patients in the moderate/high risk category had a record for use of a biologic medication and the higher frequency of IBD-related healthcare resource utilization (i.e., ED use and hospitalization). In IHI’s 7-year assessment of Triple Aim, they identified several critical elements for successful PHM which include identifying the relevant population and population segments, use of population-level measures, and conducting iterative testing (10). IHI noted that several organizations could not meet the per capita cost dimension of Triple Aim and was likely due to not identifying the appropriate population to focus on. A key component of the expanded “Quintuple Aim” is identifying disparities and designing and implementing evidence-based interventions to reduce them (11). There is a growing body of evidence suggesting that there are disparities in care for Black patients with IBD including delays in diagnosis, lower rates of appropriate medication management, and reduced access to specialty care (13-17). Other studies have found that Black patients were more likely to have been evaluated in the ED and had significantly higher frequency of ED visits compared with White patients (16,18). Race and ethnicity stratifications in PHM can help further pinpoint care gaps, thus identifying disparities and engaging health care providers in PHM efforts that may drive improved achievement of the Quintuple Aim. The profound health inequities in healthcare today highlight the need for healthcare organizations to actively work to eliminate healthcare disparities in order to achieve the original goals of the Triple Aim, particularly in individuals and communities who need them the most (11). The IBD PATH, as demonstrated in this use case study, could aid the process of identifying actionable areas for improvement in care, including areas to advance health equity.

The use of the IBD PATH tool to stratify by race and ethnicity did find differences in biologic prescription orders between Black and White patients. For example, the percent of patients with CD who received a prescription for biologic medications was 38% among Black patients and 30% among White patients. The corresponding percentages for patients with UC were 28% for Black patients and 23% for White patients. These disparities could be explained by a higher proportion of moderate/severe disease in Black patients or various factors such as: differences in healthcare access and insurance coverage, cultural belief and attitudes towards biologics, variation in physician prescribing practices and treatment recommendations, as well as potential bias in healthcare delivery. However, this was in a population that was actively receiving care and the analyses were not stratified by disease severity at baseline and are thus susceptive to confounding by indication. Despite these complexities, our findings are consistent with existing literature indicating that Black patients are more likely to have moderate to severe disease and utilized the emergency room more than White patients.

Many organizations face challenges in the implementation of these PHM strategies such as cross-sector collaboration, workforce, and access to and analysis of population data (19,20). Considering IBD PHM specifically, previous studies indicate that improvements in IBD management, adherence, and quality measure compliance reporting are still needed (21). IBD PATH was designed to assist health systems in identifying high risk patients by race and ethnicity for interventions or targeted programs, promoting patient and provider-based disease and treatment management initiatives. This case study demonstrated that there are gaps in IBD care between the IBD guidelines and actual care received and disparities in care between Black and White patients with IBD which can be evaluated using IBD PATH.

IBD PATH is evidence-based, and the content is aligned with current guidelines and care pathways, as possible. Quality control measures were taken during the development process through clinical input, software programming review, and comparative analytics (analyzing the same dataset by a statistician and with the tool). Additionally, the software structure and download feature ensures that no data can be shared with or accessible to Janssen Scientific Affairs, LLC through the use of the tool. Finally, the software includes a Symantec Code Signing Certificate which informs users that they can trust the software download by verifying code integrity and company legitimacy.

Limitations

There are important limitations when reviewing data from IBD PATH. First, common limitations of EMR data, such as the possibility of errors in diagnosis and coding, have the potential to affect the inaccuracies of the results. For instance, a situation where a patient is assigned both UC and CD diagnoses, which could result in overlapping counts. The tool aims to capture the intricate nature of real-world data such as highlighting the challenges faced by clinicians differentiating between CD and UC, especially during the early stages of disease or when disease characteristics overlap. However, it is important to note the accuracy of the results is reliant on the quality of input data. Furthermore, EMR data reflect patient care within a specific healthcare setting, potentially excluding data from patients who received care outside the system. Moreover, IBD PATH relies on the data entry form to populate and stratify the patients into moderate/high and low risk categories.

All limitations are clearly stated in the disclaimer and/or user guide to ensure users make appropriate conclusions when reviewing the data.

Conclusions

IBD PATH is a user-friendly tool that provides descriptive and actionable measures at the patient, provider, and population level. We demonstrate the importance of risk stratification among patients with IBD to ensure they are receiving appropriate therapy thus underscoring the need for structured field in the EMR data. Additionally, the use case highlights the racial disparities continue to exist in IBD care. Tools such as IBD PATH can inform PHM of patients diagnosed with IBD, facilitating the identification of potential population level gaps and disparities in care for further assessment and intervention.

Acknowledgments

Funding: This study was sponsored by the Janssen Scientific Affairs, LLC (No. 70-1036-052).

Footnote

Peer Review File: Available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-49/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jhmhp.amegroups.com/article/view/10.21037/jhmhp-24-49/coif). A.A.P. is a current employee of Janssen Scientific Affairs, LLC. K.L.D. is an employee of EPI-Q Inc., which received payment from Janssen Scientific Affairs, LLC associated with the development and execution of this study. S.B.S., E.K.K., J.C., and S.B. were employees of Ochsner Health, which received payment from Janssen Scientific Affairs, LLC associated with the participation as a site for this study. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the Ochsner Health system IRB (No. FWA00002050) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Ye Y, Manne S, Treem WR, et al. Prevalence of Inflammatory Bowel Disease in Pediatric and Adult Populations: Recent Estimates From Large National Databases in the United States, 2007-2016. Inflamm Bowel Dis 2020;26:619-25. [PubMed]
2. Xu F, Dahlhamer JM, Zammitti EP, et al. Health-Risk Behaviors and Chronic Conditions Among Adults with Inflammatory Bowel Disease - United States, 2015 and 2016. MMWR Morb Mortal Wkly Rep 2018;67:190-5. [Crossref] [PubMed]
3. Kaplan GG. The global burden of IBD: from 2015 to 2025. Nat Rev Gastroenterol Hepatol 2015;12:720-7. [Crossref] [PubMed]
4. Park KT, Ehrlich OG, Allen JI, et al. The Cost of Inflammatory Bowel Disease: An Initiative From the Crohn's & Colitis Foundation. Inflamm Bowel Dis 2020;26:1-10. Erratum in: Inflamm Bowel Dis 2020;26:1118. [Crossref] [PubMed]
5. Lichtenstein GR, Loftus EV, Isaacs KL, et al. ACG Clinical Guideline: Management of Crohn's Disease in Adults. Am J Gastroenterol 2018;113:481-517. Erratum in: Am J Gastroenterol 2018;113:1101. [Crossref] [PubMed]
6. Fumery M, Singh S, Dulai PS, et al. Natural History of Adult Ulcerative Colitis in Population-based Cohorts: A Systematic Review. Clin Gastroenterol Hepatol 2018;16:343-356.e3. [Crossref] [PubMed]
7. Sandborn WJ. Crohn's disease evaluation and treatment: clinical decision tool. Gastroenterology 2014;147:702-5. [Crossref] [PubMed]
8. Dassopoulos T, Cohen RD, Scherl EJ, et al. Ulcerative Colitis Care Pathway. Gastroenterology 2015;149:238-45. [Crossref] [PubMed]
9. Institute for Healthcare Improvement. Triple Aim and Population Health. 2023 [cited 2024 May 14]. Available online: https://www.ihi.org/Engage/Initiatives/TripleAim/Pages/default.aspx
10. Whittington JW, Nolan K, Lewis N, et al. Pursuing the Triple Aim: The First 7 Years. Milbank Q 2015;93:263-300. [Crossref] [PubMed]
11. Nundy S, Cooper LA, Mate KS. The Quintuple Aim for Health Care Improvement: A New Imperative to Advance Health Equity. JAMA 2022;327:521-2. [Crossref] [PubMed]
12. Swaminathan A, Fulforth JM, Frampton CM, et al. The Disease Severity Index for Inflammatory Bowel Disease Is a Valid Instrument that Predicts Complicated Disease. Inflamm Bowel Dis 2023;izad294. [Crossref] [PubMed]
13. Walker C, Allamneni C, Orr J, et al. Socioeconomic Status and Race are both Independently associated with Increased Hospitalization Rate among Crohn's Disease Patients. Sci Rep 2018;8:4028. [Crossref] [PubMed]
14. Straus WL, Eisen GM, Sandler RS, et al. Crohn's disease: does race matter? The Mid-Atlantic Crohn's Disease Study Group. Am J Gastroenterol 2000;95:479-83. [Crossref] [PubMed]
15. Alchirazi KA, Mohammed A, Eltelbany A, et al. S1045 Racial Disparities in Utilization of Medications in Inflammatory Bowel Disease Patients. Am J Gastroenterol 2022;117:e758-9. [Crossref]
16. Nguyen GC, LaVeist TA, Harris ML, et al. Racial disparities in utilization of specialist care and medications in inflammatory bowel disease. Am J Gastroenterol 2010;105:2202-8. [Crossref] [PubMed]
17. Flasar MH, Johnson T, Roghmann MC, et al. Disparities in the use of immunomodulators and biologics for the treatment of inflammatory bowel disease: a retrospective cohort study. Inflamm Bowel Dis 2008;14:13-9. [Crossref] [PubMed]
18. Li D, Collins B, Velayos FS, et al. Racial and ethnic differences in health care utilization and outcomes among ulcerative colitis patients in an integrated health-care organization. Dig Dis Sci 2014;59:287-94. [Crossref] [PubMed]
19. Berwick DM, Nolan TW, Whittington J. The triple aim: care, health, and cost. Health Aff (Millwood) 2008;27:759-69. [Crossref] [PubMed]
20. Conway PH, Mostashari F, Clancy C. The future of quality measurement for improvement and accountability. JAMA 2013;309:2215-6. [Crossref] [PubMed]
21. Berry SK, Melmed GY. Quality indicators in inflammatory bowel disease. Intest Res 2018;16:43-7. [Crossref] [PubMed]