Inflammatory bowel disease (IBD), which includes ulcerative colitis (UC) and Crohn disease (CD), is a chronic gastrointestinal condition with increasing pharmaceutical options and healthcare-associated costs (1). Antitumor necrosis factor (TNF) agents, such as infliximab, are considered as mainstay treatments for moderate-to-severe IBD. Although effective, biologic agents act as major cost drivers and account for a significant proportion of health plans' IBD-related total paid costs (1,2). Home infusions of infliximab are now increasingly payer-mandated and infusion-related costs are substantially less when administered at home compared with standard infusions (3,4). Other studies have also shown reduced infusion costs associated with home infliximab infusions and high patient satisfaction ratings and no serious adverse events (5,6).
However, in-home service agencies are often not affiliated with the treating physician's health system, and there is an absence of regulations regarding infusion practices (3). Recent position statements from national societies have highlighted these concerns that home infusions may contribute to fragmentation of care because home biologic infusions require higher level of care coordination and oversight to ensure optimal patient safety, efficacy, and quality (3). Furthermore, when treating pediatric patients, it is important that in-home service agencies ensure the availability of skilled pediatric nurses because delayed treatment from difficulty with intravenous access or poor management of pediatric emergencies may lead to negative clinical outcomes (3). Currently, there are limited data regarding differences in outcomes dependent on the location of biological administration. Based on a single center experience, Fenster et al. suggested that patients who started on home biologic infusions were more likely to experience adverse outcomes including discontinuation of therapy, IBD-related emergency department (ED) visits, or IBD-related hospitalizations, compared with patients receiving hospital-based biologic infusions (43.5% vs 21.7%, P = .006) (7). Although home infusions have reduced infusion-related costs, to our knowledge, there are no data demonstrating overall healthcare cost savings when considering the total cost of care for IBD patients, including the potential need for more follow-up visits, ED visits, hospitalizations, and surgeries.
Our study aimed to evaluate differences in biologic medication use, health outcomes, and the total cost of care in patients with IBD who received home-, office-, or hospital-based infliximab infusions. We hypothesized that home infusions would be associated with suboptimal outcomes, which may offset any costs savings from lower infusion-related costs.
We performed a longitudinal retrospective cohort analysis of patients with IBD in the Optum Clinformatics DataMart from 2003 to 2016. Optum contains inpatient, outpatient, and pharmaceutical claims data from 60 million deidentified commercially insured individuals in the United States. The database includes demographic and socioeconomic data (e.g., age, sex, geographical region, race, household net worth, and education level), diagnosis, and encounter data (e.g., hospital admissions, outpatient visits, emergency visits, and procedures), pharmaceutical claims data, and cost data (e.g., total cost, copayment, coinsurance, and deductibles). The Stanford University Institutional Review Board determined that this study does not meet the definition of human subject research as defined in federal regulations 45 CFR 46.102 or 21 CFR 50.3 and was exempted from further review.
Patients were included in the cohort if they met the following criteria: (1) 2 distinct inpatient or outpatient IBD-related visits (defined under the International Classification of Diseases, Ninth Revision (ICD-9) and ICD-10 codes within 1 year (8), (2) had an acceptable provider type (see Appendix I, Supplementary Digital Content 1, http://links.lww.com/AJG/B573), (3) were continuously enrolled for at least 1 year after the IBD diagnosis date, and (4) underwent at least 3 infliximab infusions within the year of the IBD diagnosis date at a home, office, or hospital. Patients were followed for 1 year after their IBD diagnosis date. Patients were classified as CD or UC if at least 80% of the ICD codes were 555.xx or K50.xx for CD and 556.xx or K51.xx for UC, respectively. Patients were classified as indeterminate colitis if neither the CD nor UC conditions were met. IBD diagnosis date was defined as the first visit with an associated IBD ICD-9 or ICD-10 diagnosis code. The pediatric cohort was defined as IBD diagnosis at younger than18 years of age.
Infliximab infusions were identified by Healthcare Common Procedure Coding System J-codes. The Optum database associates a place of service to J-codes such as emergency room, skilled nursing facility, home, office, hospital, etc. J-codes that were administered with a place of service labeled as home, office, or hospital setting were included. National Drug Codes (NDCs) were not used to identify infliximab infusions because the Optum database does not have a place of service designation available for NDCs. Patients were then assigned into one of the following cohorts: home infliximab, office infliximab, or hospital infliximab. Patients were classified into the home, office, or hospital groups if 80% of their biologic therapy administrations were in the home, office, or hospital setting, respectively. Patients were excluded if they were not able to be classified into one of these groups or if they used both infliximab and adalimumab during their enrollment period.
Outcomes and variables
Demographic data used as covariates included age at IBD diagnosis, sex, ethnicity (white, African American, Hispanic, and Asian), and geographic region (Northeast/Atlantic, North Central, South, and West). The Charlson Comorbidity Index score was calculated using ICD-9 and ICD-10 codes from inpatient and outpatient claims from 1 year of the IBD diagnosis date (9).
Socioeconomic factors included household net worth, education level, and insurance type (commercial insurance or Medicare). Education level was not assessed for pediatric patients because this Optum variable represents education level of the primary subscriber (typically a parent) instead of the child.
The primary outcomes assessed included nonadherence to and discontinuation of biologic therapy. The recommended frequency of infliximab maintenance infusions is every 8 weeks (10). Nonadherence to infliximab was defined as a patient having ≥2 infusions over 10 weeks apart in 1 year, based on the clinical judgment of our institutions' gastroenterology providers and to provide a buffer for patients who may have needed to delay an infusion for a special circumstance (10). Infliximab discontinuation was defined as >10 weeks because the previous infusion with no further infusions.
Secondary outcomes of interest were IBD-related health outcomes (cumulative outpatient corticosteroid days, follow-up visits, ED visits, hospitalizations, and surgeries per year), patient out-of-pocket costs, and the total annual cost of care. Cumulative corticosteroid (see Appendix II, Supplementary Digital Content 1, http://links.lww.com/AJG/B573) days were determined by the days' supply of outpatient systemic corticosteroids prescribed in pharmaceutical claims per person in the first year after the IBD diagnosis date using NDC codes. Out-of-pocket costs included the sum of a patient's copayment, coinsurance, and deductible. Annual out-of-pocket costs were determined by the patient's average cost per month multiplied by 12 months. The mean total cost of care per patient per year was measured by summing the cost of outpatient, ED, inpatient, and pharmaceutical claims per patient during the first year after the IBD diagnosis date. The Optum database uses standardized costs to report overall costs of pharmaceuticals and services to prevent disclosure of proprietary pricing arrangements with various institutions. Inpatient financial data are not itemized; therefore, inpatient costs include medications, nursing, labor, etc. Out-of-pocket prices are not reported using standardized costs.
Chi-squared tests were used to compare categorical variables and analysis of variance tests were used to compare continuous variables. Bonferroni multiple-comparisons tests were used to assess which groups were significant from the analysis of variance tests. Statistical significance was defined as P value <.05. Missing data were considered a separate category within each variable. A Kaplan-Meier analysis was performed from the infliximab discontinuation data for the cohort. All statistics were performed using Stata/SE 14.2 (College Station, TX).
After applying inclusion and exclusion criteria, 27,396 patients were included in our study (Figure 1). Most patients received office-based infliximab infusions (84.4%); fewer patients received hospital-based (9.9%) or home-based infliximab (5.7%) infusions. Table 1 displays the demographic and socioeconomic patient characteristics of the 4 groups. The hospital infliximab patients' median age is nearly 20 years older, their median Charlson score is 1 point higher, and they are more likely to have Medicare compared with the other groups. Most infliximab patients in all groups had UC.
Table 2 displays primary outcome results for the full sample. Patients receiving home infliximab were more likely to be nonadherent compared with patients receiving both office-based infliximab (22.2% vs 19.8%, P = .044) and hospital-based infliximab (22.2% vs 21.2%, P < .001). The home infliximab group also had significantly more patients discontinue infliximab compared with both office infusions (44.7% vs 33.7%, P < .001) and hospital infusions (44.7% vs 33.4%, P < .001). Kaplan-Meier analysis for freedom from infliximab discontinuation is shown in Figure 2. The probabilities of remaining on infliximab by day 200 of therapy were 64.4%, 74.2%, and 79.3% for home-, hospital-, and office-based infusions, respectively (P < .001).
IBD-related health outcomes.
The home infliximab group had the highest mean cumulative corticosteroid days (238.2 ± 489.8 days) among the 3 groups (Table 2). Home infliximab patients used significantly more corticosteroids compared with office-based infliximab patients (238.2 vs 189.7 days, P = .001); however, corticosteroid use between home-based and hospital-based infliximab patients was not significantly different (238.2 vs 208.5 days, P = .627).
The home infliximab group had the fewest follow-up visits with the primary gastrointestinal (GI) provider per year (2.7 ± 5.4 visits, P < .001) among all groups and more IBD-related ED visits and hospitalizations per year compared with office infliximab (P = .009, P < .001, respectively).
Figure 3 displays the total paid claims per member per year for the full sample. Although the home infliximab group had higher total costs compared with the group receiving office-based infliximab infusions ($49,149 vs $43,466, P < .001), their cost of care was $2,021 lower than patients receiving hospital-based infusions ($49,149 vs $51,170, P < .001). The proportion of patients with >$500 a year of out-of-pocket costs were similar between the home, office, and hospital infliximab groups (2.6% vs 1.9% vs 2.3%).
Pediatric subsample analysis
The pediatric subsample included 1,839 patients younger than 18 years of age. Similar to adults, most pediatric patients received office-based infliximab (77.4%); fewer patients received home-based infliximab infusions (12.1%) or hospital-based infliximab (10.4%). Table 1 displays the characteristics of the pediatric patient cohort. The home, office, and hospital infliximab groups had a similar median age. Unlike the full sample, most pediatric patients receiving infliximab had CD.
Pediatric patients receiving home-based infliximab were significantly more nonadherent than pediatric patients receiving both office-based (27.1% vs 21.1%, P = .047) and hospital-based infliximab infusions (27.1% vs 14.4%, P < .001) (Table 3). The discontinuation rate of the home infliximab group was significantly higher than hospital infusions (46.6% vs 26.5%, P < .001) but not significantly different when compared with office infusions (46.6% vs 31.0%, P = .609).
Unlike the full sample, pediatric home infliximab patients did not differ in the amounts of corticosteroids used when compared with patients receiving either office-based or hospital-based infliximab (P = .560 and P = .940, respectively). The home infliximab group had fewer follow-up visits with the primary GI provider compared with patients receiving office-based infusions (5.3 vs 14.2 visits, P < .001) but more compared with patients receiving hospital-based infusions (5.3 vs 4.0 visits, P = .850). There were no significant differences in rates of IBD-related surgeries, ED visits, or hospitalizations by location of infliximab infusion.
Figure 4 displays the total paid claims per member per year for the pediatric sample. The total cost of care was around $10,000 higher for pediatric patients receiving home infliximab infusions compared with office-based ($48,703 vs $38,179, P < .001) or hospital-based ($48,703 vs $39,159, P < .001) infliximab infusions. Annual out-of-pocket costs of home infliximab patients were similar compared with office infliximab patients; however, fewer patients paid >$500 per year in the hospital infliximab group.
From an analysis of over 27,000 patients with IBD, including over 1,800 pediatric patients with IBD, we showed that although less than 6% of the patients are using home infliximab infusions, these patients were more likely to be nonadherent to appropriate therapy plans compared with patients using either office- and hospital-based infusions. We report nonadherence rates of approximately 20%. This is lower than previously described; however, direct comparison between studies is difficult because of the lack of consensus on the definition of adherence to biologics (11). Relatedly, patients receiving home infusions were more likely to discontinue infliximab therapy. In our full sample, discontinuation rates were approximately 30%–40% within the first year, which is in range with reported rates from the recent analyses (12–14). Although the durability of infliximab is shorter when patients receive therapy at home, patients in the home infliximab group had statistically higher cumulative corticosteroid exposure and fewest follow-up visits with their treating gastroenterologist on a per-year basis. Despite this, patients receiving home infusions had higher overall annual care costs compared with patients who had infusions in office-based infusion settings. The subgroup analysis of pediatric patients showed similar results as the overall population.
In 2018, Kochar et al. (4) performed a retrospective cohort study of 13,745 adult patients with IBD and reported that rates of home infliximab range from 4.7% to 7.2% per year, with the highest prevalence of 20.6% in the western region of the United States. Despite this prevalence, literature evaluating home infliximab infusions are limited. A small prospective study in 13 adult patients with CD receiving infliximab demonstrated that median patient satisfaction was 8 of 10 for both home and hospital treatment settings and that home infusions were associated with cost savings (6). The data are even more limited in pediatrics, with the only identified study being a small retrospective chart review by Condino et al. (5). They evaluated 10 pediatric patients with IBD enrolled in their center's home-based infliximab infusion program and report an average patient satisfaction rating of 9 of 10 with home infusions, as well as decreased school absenteeism that they attribute to increased flexibility of scheduling infusion on any day of the week (5). However, evening and weekend home infusion availability is not consistent across home infusion service agencies, and Kuin et al. (6) reported that 3 of 16 patients who discontinued home infliximab did so because of the lack of scheduling flexibility. This is a potential contributor to the higher rates of nonadherence and discontinuation of infliximab found in our study. Lack of care coordination and treatment oversight may also be potential contributors, as discussed in the recent North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition position statement (3). A recent publication further highlights the necessary steps to overcome these barriers, such as a tracking system to assure regular follow-up visits, increased provider support for care coordination and insurance authorization, and standardized procedure for medication ordering and laboratory testing to safely implement a home infusion program for pediatric patients with IBD (15). Overall, the reasons for patient discontinuation of infliximab and the overall safety are not discussed in our analysis. It is thus unknown what the overall severe adverse event rate is in these different populations (i.e., home-, office-, and hospital-based infusions). Further studies are needed to evaluate these areas of importance.
Although review of the existing literature on home infliximab corroborates our findings, our study sheds further light on linking differences in health outcomes with site of infusion services. A recent single-center retrospective cohort study by Fenster et al. (7) evaluated adverse outcomes with home vs hospital biologic infusions in patients with IBD. Forty-six home infliximab patients and 23 vedolizumab patients were compared with matched control subjects. The authors observed that patients with IBD receiving home infusions were significantly more likely compared with control patients to experience a composite of adverse outcomes including stopping biologic therapy, IBD-related ED visits, and IBD-related hospitalizations (43.5% vs 21.7%, P = .006) (7). When evaluating these endpoints individually, although, the study likely did not have adequate statistical power and results for discontinuation rates, and hospitalization rates were not statistically significant. However, our study did find statistical significance that home infusions were associated with higher rates of each of these outcomes. The study by Fenster et al. (7) corroborates our study showing that patients persisting on home infusions had significantly fewer IBD clinic visits compared with control subjects in the year after home infusion initiation. Of note, our study demonstrating an association of home infliximab infusions with increased cumulative corticosteroid exposure is a previously undescribed finding. Previous studies have also described high corticosteroid use in patients with IBD, especially in patients who have fewer gastroenterology visits (16–19).
Previous literature support reduced infusion costs when infliximab infusions are given at home. Condino et al. (5) reported an average cost savings of $1,335 per 100 mg of infliximab at their institution's home infusion program in Colorado. Kochar et al.'s (4) retrospective cohort analysis of patients with IBD in the United States reported median charges for infliximab home infusions to be $5,700 compared with $6,200 for standard infusions. In Europe, the study performed by Kuin et al. (6) report €55 (∼$65) cost savings per infliximab home infusion, excluding drug costs. However, these reports do not account for other noninfusion-related costs. Our study, to our knowledge, is the first to evaluate total cost of care for patients with IBD receiving their infusions at various sites of care. Cost savings with home infliximab infusions is not apparent in our analysis when taking into account noninfusion-related costs, including outpatient, inpatient, ED, diagnostic, and pharmacy costs. Interestingly, home infusion patients have higher outpatient infusion costs despite having fewer visits with their primary GI provider. The reasons for this are unknown, and further studies are warranted to investigate these specific contributing factors. It is important to note, however, that the Optum database reports standardized costs to avoid revealing proprietary and contract-based cost information. Therefore, we cannot derive conclusions regarding the absolute value of the total per member per year cost.
We acknowledge that the retrospective nature of our analysis limits our ability to control for confounding factors. Unfortunately, disease severity and phenotype matching were not possible in our analysis. This is because markers for disease severity such as Mayo Clinic Scores, Simple Endoscopic Scores for Crohn Disease, Crohn Disease Activity Index scores, and calprotectin are incomplete or unavailable in the Optum database. However, given the largeness of our dataset representing the commercially insured IBD population in the United States, we postulate that whether there were bias for more disease-active or higher-risk patients with IBD, they would be in the hospital-based infusion group. As previously mentioned, although this group had lower corticosteroid exposure, drug discontinuation, and nonadherence rates. In addition, a root-cause analysis was not possible in our analysis, given markers for infliximab efficacy (infliximab levels) and immunogenicity (anti-drug antibodies) are also unavailable or incomplete in the database. Finally, the use of the Optum database does not capture patients with government-subsidized insurance. This is of significance especially for the pediatric population because 44% of children with special care needs are covered by public insurance (20).
In conclusion, our study represents the most comprehensive investigation of home-based infliximab infusions to date, although extending previous findings in patients with IBD concerning suboptimal patient outcomes, nonadherence, and drug durability when receiving maintenance infliximab at home. This is the first study evaluating health outcomes in pediatric patients with IBD and site of infusion services. Furthermore, our analysis suggests that home infusions may not be associated with cost savings when considering the total cost of care of the patients with IBD.
CONFLICTS OF INTEREST
Guarantor of the article: K.T. Park, MD, MS.
Specific author contributions: Study concept and design: N.G.K., M.W., M.D., L.A.S., and K.T.P.; Data acquisition: M.D.; Data analysis and interpretation: N.G.K., M.W., M.D., L.A.S., and K.T.P.; Drafting the work or revising it critically for important intellectual content: N.G.K., M.W., M.D., L.A.S., and K.T.P.; Obtaining funding: K.T.P.
Financial support: Support for this study comes from an institutional grant from Stanford University School of Medicine (Department of Pediatrics), the Bacher-English Family Fund, and NIDDK094868 (K.T.P.). Data for this project were accessed using the Stanford Center for Population Health Sciences Data Core, supported by a National Institutes of Health National Center for Advancing. Translational Science Clinical and Translational Science Award (UL1 TR001085).
Potential competing interests: N.G.K., M.W., and M.D.: None declared. L.A.S. is supported by a National Institutes of Health National Center for Advancing Translational Science, Clinical and Translational Science Award (KL2TR001083 and UL1TR001085). K.T.P. had no conflicts relevant to this manuscript; at time of publication, is an employee at Genentech, Inc. and shareholder in the Roche Group.
WHAT IS KNOWN
- ✓ Home biologic infusions for treatment of inflammatory bowel disease are increasingly payer-mandated due to non-drug related costs.
- ✓ Home biologic infusions require higher level of care coordination and oversight to ensure optimal safety, efficacy, and quality.
- ✓ It is unknown whether health outcome differences exist when patients receive biologic medications at home-, office-, or hospital-based infusion settings.
WHAT IS NEW HERE
- ✓ Home biologic infusions led to higher non-adherence to scheduled therapy plans, leading to increased discontinuation rates and decreased durability of infliximab over time.
- ✓ Home biologic infusions did not result in lower cost of care, but patients receiving home infusions had higher overall annual health care costs compared to patients receiving infusions in office-based settings.
- ✓ Compared to patients who received biologic medications at office- or hospital-based settings, patients receiving home infusions had higher corticosteroid use and fewer follow-up visits with their primary gastrointestinal provider.
1. Park KT, Colletti RB, Rubin DT, et al. Health insurance paid costs and drivers of costs for patients with Crohn's disease in the United States:. Am J Gastroenterol 2016;111(1):15–23.
2. Yu H, MacIsaac D, Wong JJ, et al. Market share and costs of biologic therapies for inflammatory bowel disease in the USA. Aliment Pharmacol Ther 2018;47:364–70.
3. Barfield E, Sockolow R, Hoffenberg E, et al. Assuring quality for non-hospital based biologic infusions in pediatric inflammatory bowel disease: A clinical report from the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. J Pediatr Gastroenterol Nutr 2018;66(4):680–6.
4. Kochar B, Jiang Y, Martin CF, et al. P140 current status of home infusions for inflammatory bowel disease in the United States. Gastroenterology 2018;154(1):S72–3.
5. Condino A, Fidanza S, Hoffenberg E. A home infliximab infusion program. J Pediatr Gastroenterol Nutr 2005;40(1):67–9.
6. Kuin S, Stolte S, van den Brink G, et al. Short article: Remicade infusions at home an alternative setting of infliximab therapy for patients with Crohn's disease. Eur J Gastroenterol Hepatol 2016;28(2):222–5.
7. Fenster M, Ungaro RC, Hirten R, et al. Home vs hospital infusion of biologic agents for patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol 2020;18(1):257–8.
8. Rezaie A, Quan H, Fedorak RN, et al. Development and validation of an administrative case definition for inflammatory bowel diseases. Can J Gastroenterol 2012;26(10):711–7.
9. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson Comorbidity Index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol 2011;173(6):676–82.
10. Remicade (infliximab) [prescribing information]. Janssen Biotech, Inc: Horsham, PA, 2018.
11. Wentworth BJ, Buerlein RCD, Tuskey AG, et al. Nonadherence to biologic therapies in inflammatory bowel disease. Inflamm Bowel Dis 2018;24(9):2053–61.
12. Chen C, Hartzema AG, Xiao H, et al. Real-world pattern of biologic use in patients with inflammatory bowel disease: Treatment persistence, switching, and importance of concurrent immunosuppressive therapy. Inflamm Bowel Dis 2019;25(8):1417–27.
13. Mevius A, Brandes A, Hardtstock F, et al. Persistence with biologic treatment in patients with inflammatory bowel disease: A German claims data analysis. Digestion 2019:1–11.
14. Sprakes MB, Ford AC, Warren L, et al. Efficacy, tolerability, and predictors of response to infliximab therapy for Crohn's disease: A large single centre experience. J Crohns Colitis 2012;6(2):143–53.
15. Gupta SR, Crandall WV, Donegan A, et al. A quality improvement approach to external infliximab infusions in pediatric inflammatory bowel disease. J Pediatr Gastroenterol Nutr 2019;69(5):544–50.
16. Waljee AK, Wiitala WL, Govani S, et al. Corticosteroid use and complications in a US inflammatory bowel disease cohort. PLoS One 2016;11(6):e0158017.
17. Dorrington AM, Selinger CP, Parkes GC, et al. The historical role and contemporary use of corticosteroids in inflammatory bowel disease. J Crohns Colitis 2020.
18. Armuzzi A, daCosta DiBonaventura M, Tarallo M, et al. Treatment patterns among patients with moderate-to-severe ulcerative colitis in the United States and Europe. PLoS One 2020;15(1):e0227914.
19. Selinger CP, Parkes GC, Bassi A, et al. Assessment of steroid use as a key performance indicator in inflammatory bowel disease—Analysis of data from 2385 UK patients. Aliment Pharmacol Ther 2019;50(9):1009–18.
20. American Academy of Pediatrics. Medicaid Facts [Facts Sheet]. AAP: Itasca, IL, 2017. (https://www.aap.org/en-us/Documents/federaladvocacy_medicaidfactsheet_all_states.pdf
). Accessed April 3, 2019.