Higher Rates of Delay in Starting Advanced Inflammatory Bowel Disease Therapies Linked to Insurance Delays, Intravenous Infusions, and Lack of Pharmacy Support

Abstract

INTRODUCTION:

Because biologic and small molecule therapy is expensive, payors have mandated preauthorizations for these medications, often resulting in a lengthy approval process. The aims of this study were to assess the frequency of and risk factors for delays in starting advanced therapies assessing insurance, care team, and patient-related factors.

METHODS:

Retrospective, multicenter study of patients with adult inflammatory bowel disease with prescriptions for an advanced therapy in 2 geographically distinct academic gastroenterology practices: one with and the other without a dedicated pharmacist. A priori, we defined a delay in starting therapy as >14 days between prescription and the first dose. Logistic regression analysis was performed to assess for risk factors for delay.

RESULTS:

A total of 388 patients were prescribed advanced therapies with 46.6% receiving their first dose within 14 days. Patients who were on time vs delayed were similar in baseline demographics, disease characteristics, and disease activity. After adjusting for confounders, 3 independent risk factors remained significant as predictors for delay: study site (OR = 5.2, 95% CI 2.894, 9.333), intravenous drug delivery as opposed to subcutaneous or oral (OR = 3.07, 95% CI 1.845, 5.099), and insurance denial (OR = 2.72, 95% CI 1.082, 6.825).

DISCUSSION:

In a multicenter study, we found that a delay between prescription and administration of the first dose of an advanced therapy is common, with > 50% of patients having the first dose delayed by > 2 weeks. Delays in starting therapy were significantly more likely if denied by insurance, given by intravenously induction, or at a study site without a dedicated pharmacist.

INTRODUCTION

Inflammatory bowel disease (IBD) treatment has progressed over the last 20 years since the first tumor necrosis factor alpha inhibitor (anti-TNF) was approved in 1998, and there are now many advanced therapies available to patients living with Crohn's disease (CD) and ulcerative colitis (UC). These biologic and small molecule medications help patients achieve clinical and endoscopic remission and have transformed the way that moderate-to-severe IBD is treated (1,2).

However, these treatments are also expensive (3). Insurance companies frequently attempt to contain costs by requiring prior authorization before agreeing to pay for the medication. The prior authorization process takes substantial time and effort from the prescribing clinician, office staff, and patient. The complexity of the prior authorization process has resulted in legislation of this process in certain states, publication of guidance to assist IBD clinics in navigating the process, and influenced IBD clinicians treatment options due to the perceived burden from the process (4).

The onerous process of prior authorization results in delays in insurance approval, and thus delays in initiation, for these medications in patients with IBD. In the pediatric IBD population, these delays can lead to increased utilization of health care resources and increased steroid dependence but have been less well described in the adult IBD population (5). In addition to insurance delays, there is the potential that other factors could contribute to delays with starting therapy, including patient and care team factors. To this end, the aim of our retrospective, multicenter cohort study was to assess the frequency of these delays and to identify the most likely causes or risk factors for delays in starting advanced therapies in adult patients with IBD.

METHODS

Study design

This study is a multicenter, retrospective cohort analysis of adult patients with IBD receiving outpatient prescriptions for a biologic or small molecule therapy between 2012 and 2022.

Study population

This study was conducted at 2 geographically distinct academic gastroenterology practices in the United States: Yale New Haven Health System (Site 1) and Digestive Health at the University of Texas Health Austin (Site 2). Patients included were aged 18 years or older, had a diagnosis of IBD (either CD, UC, or IBD undetermined), and had been prescribed either a biologic or a small molecule medication as an outpatient. Prior authorizations, if required by a patient's insurance company, could be for a new medication start, a medication renewal, or a medication dose escalation, whichever was the earliest in the clinic records. Patients were included if they had at least 6 months of follow-up after the initial medication order and had available records documenting medication order date and administration of the first dose. Patients were excluded if they did not receive the prescribed advanced therapy, if they did not remain on prescribed therapy for at least 6 months after prescription, if they had insufficient follow-up, or if they died within one year of medication prescription.

Data collection and outcomes

Charts were reviewed by multiple reviewers at each site to ensure accuracy of data collection. A priori, we defined a delay in receiving the first dose of medication as more than 14 days between date of a prescription of an advanced therapy and the date of the first dose administration. A period of >14 days was chosen as the definition of delay because this represents a practical and reasonable period during which patients and providers would expect a medication to be administered.

Biologics and small molecules used during the time period of this study included infliximab (only given intravenously [IV]), adalimumab (given subcutaneously [SQ]), certolizumab (SQ), golimumab (SQ), vedolizumab (only given IV), ustekinumab (IV loading and SQ maintenance), and tofacitinib (oral).

Data regarding insurance, care team, and patient factors were collected. Insurance data included insurance type, insurance denials, requirement for peer-to-peer reviews, and/or appeals. Care team factors included delivery type of the medication prescribed (i.e. IV, SQ, or oral) whether it was a new start, renewal, or escalation, and whether a pharmacy team was embedded in the clinic. Patient-related factors collected included patient demographics, disease type and complications, and disease activity and time from prescription to getting laboratory work for prebiologic evaluations.

Disease activity was assessed by the following measures: (i) the physician global assessment at the enrollment clinic visit where the patient was prescribed the advanced therapy; (ii) fecal calprotectin within 3 months of study enrollment, where <150 mg/dL was considered remission; (iii) active inflammation on endoscopic evaluation within 3 months of study enrollment; and/or (iv) active disease on cross-sectional imaging within 3 months of study enrollment. A composite end point of disease activity was included where patients were considered to have active disease if they had elevated fecal calprotectin and active disease on endoscopy or cross-sectional imaging.

Statistical analysis

Logistic regression models were built to assess the strength of associated risk factors for delay in treatment. Adjusted odds ratios along with their corresponding confidence intervals are reported. Potential imbalances between the groups were explored in the baseline tables using univariate testing (t tests or Mann-Whitney U for continuous variables and the Fisher exact test for categorical variables). P values < 0.10 in addition to clinically plausible factors were considered as potential covariates to include in the adjusted models. Ultimately, the final model contained the variables of insurance type, drug delivery type (IV, SQ, or oral), study site, IBD behavior type, insurance denial, and type of med start (new, renewal, or escalation). Analyses were performed with Stata Statistical Software, Release 18 (College Station, TX).

Ethical considerations

This study was approved by the Institutional Review Boards of Dell Medical School at the University of Texas at Austin and the Yale University School of Medicine and was also approved by the Ascension Seton Research Enterprise.

RESULTS

A total of 388 patients (292 CD, 87 UC, and 9 IBD undetermined) met inclusion criteria and were examined in this study.

Time from advanced therapy prescription to first dose administered

Among all 388 patients, the mean time to the first dose was 25 days (median 17 days) and only 181 (46.6%) received medications within 14 days (Table 1). There was a substantial difference in the frequency of delayed starts between study sites: Site 1 achieved a start time without delay in 157 (63.3%) patients, whereas only 24 (17.2%) patients at Site 2 received their medication without delay. Site 1 had an embedded pharmacy team that processed all prescriptions including initiation and follow-up of all prior authorization prerequisites whereas site 2 did not have such a team and used standalone specialty pharmacies or infusion centers that processed the prescriptions themselves. When examining start times for all patients within 30 days, 279 (71.9%) received medication within that window.

Table 1.

Time from prescription to administration of first dose, by study sites

	Site 1 (n = 248)	Site 2 (n = 140)	Combined (n = 388)
Mean days to start	18.7	37.0	25.3
Median days to start	9	29	17
Patients within 14 d or less to first dose	157 (63.3%)	24 (17.1%)	181 (46.6%)
Patients within 30 d or less to first dose	203 (81.6%)	76 (54.3%)	279 (71.9%)

Demographic and disease-specific characteristics of study cohorts with and without delays

Patients who experienced or did not experience a delay in their first dose of therapy, were not statistically different regarding age, sex, race, or ethnicity (Table 2). In addition, there was no difference in type of IBD, age at diagnosis, disease duration, UC disease extent, CD disease location (including upper gastrointestinal), tract involvement and perianal disease), and prior IBD-related surgery or burden of comorbid disease when comparing those who experienced or did not experience a delay. Furthermore, rates of active disease (assessed by a composite of objective measures of disease activity) were not different between patients with and without a delay in the first dose of medication (Table 3). Conversely, patients with penetrating CD behavior were less likely to have delays than those with inflammatory or stricturing disease on univariate analysis.

Table 2.

Demographics

	No delay (≤14 d) n = 181	Delay (>14 d) n = 207	P value
Age (±SD)	39.8 ± 13.4	38.7 ± 14.6	0.22
Sex (% men)	87 (48.1%)	94 (45.4%)	0.61
Race
Black	14 (7.7%)	20 (9.7%)	0.59
White	148 (81.8%)	166 (80.2%)	0.70
Asian	8 (4.4%)	6 (29.0%)	0.59
Other (American Indian, Pacific Islander, Mixed)	2 (1.1%)	4 (1.9%)	0.69
Unknown	9 (5%)	11 (5.3%)	1.0
Ethnicity
Latino	15 (8.3%)	23 (11.1%)	0.39
Non-Latino	164 (90.6%)	180 (87.0%)	0.27
Unknown	2 (1.1%)	4 (1.9%)	0.69
Type of IBD
Crohn's disease	136 (75.1%)	156 (75.4%)	1.0
Ulcerative colitis	40 (22.1%)	47 (22.7%)	0.90
IBD undetermined	5 (2.8%)	4 (1.9%)	0.74
Age at diagnosis	26.9 ± 13.3	25.2 ± 13.7	0.21
Disease duration	13.0 ± 9.4	13.5 ± 9.7	0.64
UC disease extent
Proctitis	2 (5%)	3 (6.4%)	1.0
Left-sided	7 (17.5%)	10 (21.3%)	0.79
Extensive	28 (70%)	32 (68.1%)	1.0
Unknown	3 (7.5%)	2 (4.3%)	0.66
Crohn's disease location
Ileal	22 (16.2%)	36 (23.1%)	0.15
Colonic	29 (21.3%)	39 (25.0%)	0.49
Ileocolonic	84 (61.8%)	80 (51.3%)	0.08
Unknown	1 (0.7%)	1 (0.6%)	1.0
Crohn's disease behavior
Stricturing disease	29 (21.3%)	49 (31.4%)	0.06
Penetrating	58 (42.7%)	37 (23.7%)	0.0007
Inflammatory	46 (33.8%)	69 (44.2%)	0.07
Unknown	3 (2.2%)	1 (0.6%)	0.34
Upper gastrointestinal Crohn's	15 (11.0%)	16 (10.3%)	0.85
Perianal Crohn's	55 (40.0%)	53 (34.0%)	0.28
Prior IBD-related surgery	70 (38.7%)	73 (35.3%)	0.48
Charlson Comorbidity Index	0.81	0.81	0.71

Bold indicates statistical significant (P < 0.05).

IBD, inflammatory bowel disease; UC, ulcerative colitis.

Table 3.

Disease activity at time of prescription of advanced therapy

	No delay (≤14 d) n = 181	Delay (>14 d) n = 207	P value
PGA—active disease	98 (54.1%)	113 (54.6%)	1.0
PGA — mild disease	56 (30.9%)	64 (30.9%)	1.0
PGA— moderate-to-severe disease	42 (23.2%)	49 (23.7%)	1.0
Fecal calprotectin >150 µg/dLa	26/46 (56.5%)	40/72 (55.6%)	1.0
Active disease on scopea	71/94 (75.5%)	83/112 (74.1%)	0.87
Active disease on imaginga	39/57 (68.4%)	40/60 (66.7%)	0.85
Composite end point: Calprotectin >150 OR active disease on scope OR active disease on imaging	88/179 (49.2%)	91/179 (50.8%)	0.83

PGA, physician global assessment.

^aWithin 3 mo of prescription.

Payor and drug characteristics of study cohorts with and without delays

On univariate analysis, the type of payor influenced the rates of delay. Having commercial insurance was associated with higher likelihood for delays (P = 0.0066) whereas having Medicaid was associated with a lower likelihood for delays (P = 0.0002). There was no difference for patients with Medicare (Table 4).

Table 4.

Insurance and drug type information

	No delay (≤14 d) n = 181	Delay (>14 d) n = 207	P value
Insurance type
Commercial (any)	118 (65.2%)	161 (77.8%)	0.0066
Medicaid	44 (24.3%)	21 (10.1%)	0.0002
Medicare	15 (8.3%)	20 (9.7%)	0.72
Other/Uninsured/self-pay	4 (2.2%)	5 (2.4%)	1.0
Drug prescribed
Infliximab	15 (8.3%)	44 (21.3%)	0.0004
Adalimumab	101 (55.8%)	59 (28.5%)	<0.0001
Certolizumab	2 (1.1%)	4 (1.9%)	0.69
Golimumab	1 (0.6%)	1 (0.5%)	1.0
Vedolizumab	6 (3.3%)	21 (10.1%)	0.009
Ustekinumab	41 (22.7%)	74 (35.8%)	0.0053
Tofacitinib	15 (8.3%)	4 (1.9%)	0.0043
Type of medication order
New	83 (45.9%)	100 (48.3%)	0.68
Renewal	82 (45.3%)	98 (47.3%)	0.76
Dose escalation	16 (8.8%)	9 (4.4%)	0.10
Drug delivery type (based on first dose type received)
Intravenous	43 (23.8%)	115 (55.7%)	<0.0001
Subcutaneous	123 (67.9%)	88 (42.5%)	<0.0001
Oral	15 (8.3%)	4 (1.8%)	0.004
Study site
Site 1	157 (86.7%)	91 (44.0%)	<0.0001
Site 2	24 (13.3%)	116 (56.0%)

Bold indicates statistical significant (P < 0.05).

Delays were also examined by type of prescription (new, renewal, or dose escalation), which drug was prescribed, and delivery method (IV, SQ, or oral). The status of a prescription as new, renewal, or dose escalation did not correlate with higher likelihood of delay (Table 4). However, the method of delivery of the drug being prescribed was strongly associated with delays, with IV dosed medications being more likely to be delayed. Delay was found for IV dosed medications regardless of prescription type (i.e. new, renewal, or escalation) on analysis, with IV induction doses of newly started medications being delayed 74.7% of the time and IV maintenance dosed medication renewals delayed 71.9% of the time. Both sites exhibited higher rates of delay in IV dosed medications. Similarly, when analyzed by specific drugs, those with IV dosing (infliximab, vedolizumab, and ustekinumab induction) were more likely to be delayed whereas those with subcutaneous or oral dosing (adalimumab, tofacitinib) were less likely to be delayed.

Identifying delays within the complicated prior authorization process

We collected data regarding what steps of the prescription-authorization-drug delivery process were more likely to lead to delays. Patients who had delays compared with those without delays did not have a longer time completing requested pretreatment laboratory results. However, those with delays were more likely to have experienced initial denials from their insurance company, require peer-to-peer reviews, and/or require appeals on univariate analysis (Table 5).

Table 5.

Process between decision to start, escalate, or renew medication and first dose received

	No delay (≤14 d) n = 181	Delay (>14 d) n = 207	P value
Time between laboratory order and laboratory completion (d)	14.6	10.7	0.009
Initial denial by insurance company	7 (3.9%)	46 (22.2%)	<0.0001
Peer-to-peer required	3 (1.7%)	15 (7.2%)	0.0134
Appeal required	4 (2.2%)	24 (11.6%)	0.0003

Bold indicates statistical significant (P < 0.05).

Risk factors for delay in starting advanced therapy

We then performed a multivariate logistic regression analysis to better understand which factors were most important for leading to delays in receiving medications (Table 6). After adjustment for potential confounders between groups, the factors that remained significant as risk factors for delays were study site (OR = 5.20, 95% CI 2.894, 9.333), receiving IV therapy as opposed to SQ or oral (OR = 3.07, 95% CI 1.845, 5.099), and receiving an insurance denial (OR = 2.72, 95% CI 1.082, 6.825). Finally, to explore whether these findings could have been driven by the patients with extreme delays within the cohort, we performed a sensitivity analysis for patients with ≤ 30 days to the first dose of their drug. In this analysis, study site (OR = 5.69, 95% CI 2.812, 11.504) and receiving IV therapy as opposed to SQ or oral therapy (OR = 2.11, 95% CI 1.146, 3.887) remained significant while insurance denial was no longer significant (OR = 1.43, 95% CI 0.475, 4.338).

Table 6.

Multivariate logistic regression analysis adjusting for baseline differences between groups with and without delay to evaluate factors that affect delays

Outcome	Adjusted OR	P value	95% CI
IV delivery	3.07	<0.0001	1.845, 5.099
Study site	5.20	<0.0001	2.894, 9.333
Insurance denial	2.72	0.033	1.082, 6.825

DISCUSSION

In this first multicenter study to evaluate delays in receiving advanced therapy in IBD, we found that delays in starting outpatient treatment were common, with fewer than 50% of patients receiving their first dose of an advanced therapy within 14 days from the prescription. The mean time to the first dose was 25 days with a median of 17 days. On average, this is comparable with the time reported in other single-center studies (5–7). We defined a delay as longer than 14 days because 2 weeks is a long time for symptomatic patients to continue to feel poorly before starting the first dose of a therapy and is a reasonable time within which for an insurance company to make a determination for approval. Many patients are initiated or continued on corticosteroids during this time, which carry significant side effects. Thus, the current time required to start therapy seen in our multicenter study and replicated in other single-center studies all indicate that the process can be quite lengthy and needs to be better understood so that we can shorten the duration of symptomatic disease by shortening the time between submitting a prescription and the first dose of these medications.

Prior studies have focused primarily on insurance approval rates and other insurance-related factors when looking at delays in starting appropriate treatment for patients with IBD (5–7). Our study is unique in that we aimed to evaluate a variety of factors that may contribute to delays including insurance-related factors in addition to care -team-related factors and patient-related factors. On univariate analysis, we identified several significant potential contributors including insurance type, study site, drug delivery method, insurance denials, peer-to-peer reviews, and appeals. On multivariate analysis after adjusting for potential confounders, we found 3 significant risk factors for delays in treatment starts remained: the study site, intravenous vs SQ or oral administration, and insurance denial during the prior authorization process.

Although both study sites were at academic medical centers, they were in geographically disparate locations of the country, one in the south (Texas) and the other in the northeast (Connecticut), with different state laws and insurance coverage policies. Although significant differences were seen overall in delays based on insurance type and denials from insurance companies, on subgroup analysis these differences, while still numerically higher in the delayed group, were no longer significant. One notable difference between the programs is that site 1 has a well-established embedded pharmacy program that oversees all advanced therapy prescriptions from ordering to approval whereas site 2 does not have such a pharmacy program and relies on a variety of unaffiliated specialty pharmacies and standalone infusion centers to fill and administer advanced therapy prescriptions. We speculate that this difference contributed significantly to the delays in time to the first dose in site 2.

Second, independent of site, we found that advanced therapies given through intravenous infusion had a 3-fold increased odds of delay compared with subcutaneous or oral medications. Moreover, this same finding was seen in the individual medications given IV vs SQ, regardless of whether the medication was new or maintenance therapy. We speculate that this is because IV dosed medication requires additional coordination with an infusion center and therefore adds an extra layer of necessary cooperation and communication to an already convoluted process involving the prescribing practice, the insurer, the specialty pharmacy, and often patient assistance programs. This becomes further complicated for medications that are given as IV for induction and then SQ for maintenance, which requires separate processes for approval through both pharmacy and medical benefits.

Finally, insurance denials clearly delay the time to receive the first dose of medication and in our study led independently to an almost 3-fold increased risk of the medication being delayed. This confirms the findings in other studies that have also shown delays in receiving the first dose of medication after prescription related to the prior authorization process (5,7). Interestingly, when we performed a sensitivity analysis for patients who received their medications within 30 days (i.e. removed those with more extreme delays), the multivariate finding of insurance denials as a predictor is no longer significant. This suggests that denials are likely playing a bigger part for patients with longer delays. This process is even more frustrating for prescribing practitioners because insurance company practices align with IBD treatment guidelines only 5.8%–58.5% of the time (8). One study even reported that despite these denials and complicated authorization processes (appeals, peer-to-peer reviews, and requirements for step therapy) that 97% of their prescriptions were eventually approved, highlighting whether the complexity of the process is necessary (7). Moreover, a study in a pediatric population demonstrated that these delays in approvals have been associated with increased health care utilization (5) whereas a study in adult patients with IBD showed that delays were associated with more disease activity (9). Our data combined with these studies highlight the combination of delay in appropriate care, the burden of time and manpower needed to get these approvals, and the cost to patients due to these delays.

There are a few limitations to this study. Owing to the retrospective nature of the study, it is possible that some of the details of the insurance authorization process are less well characterized or misclassified. This was minimized by using multiple reviewers, including the review of medical and pharmacy records, to confirm the data for each chart and adjudicating any questionable data. Although our multicenter design with and without sites using a dedicated pharmacy team is a strength, it is possible that these data may not be generalizable to practices outside of an academic center where delays may be even longer or appeals for denials sought less often. Moreover, it is possible that site-specific differences may exist beyond pharmacy services that could have contributed to differences in delay times between sites. Finally, while we performed a multivariate analysis with adjustment for confounders, it is possible there were still unmeasured biases in our results given the retrospective design.

In conclusion, in a multicenter study, we found that a delay between prescription and administration of the first dose of an advanced therapy is common, with more than 50% of patients having the first dose delayed more than 2 weeks. We analyzed insurance-related, care-team-related factors, and patient-related factors and observed that a delay in starting therapy was significantly more likely if the therapy is given by IV infusion, is prescribed at a study site without a dedicated pharmacy team, or is initially denied by the insurance company. Although we understand the need to manage the costs of care, requiring additional processes, and frequently additional staffing dedicated to managing medication approvals, is not in the best interest of our patients with IBD because these requirements do not align with IBD treatment guidelines and appear unnecessary as most requests are approved eventually. Our research highlights areas of friction in this process that could be the areas for targeted intervention. For example, while certain aspects of care such as insurance approval and denial may require policy changes on a broader level, establishing relationships between IV infusion centers and pharmacy on an institutional or regional level where possible can allow for more effective patient-centered care. The importance of this topic was recently highlighted in an American Gastroenterological Association white paper because of the impact of the delays resulting from the prior authorization on IBD care (10).

CONFLICTS OF INTEREST

Guarantor of the article: Linda A. Feagins, MD, FACG.

Specific author contributions: S.G.: data collection; analysis of the data; drafting of the article; K.X.: data collection; analysis of the data; H.N., E.H., E.S., P.J., F.J., J.D., K.T.: data collection; Z.T.: statistical analysis; J.K.J.G.: conception and design; interpretation of the data; revision of the article for important intellectual content; final approval of the article. L.A.F.: conception and design; analysis and interpretation of the data; drafting of the article; final approval of the article. All authors have approved the final manuscript.

Financial support: This work was supported in part with institutional funds by Ascension Seton under a partnership with Dell Medical School at the University of Texas at Austin.

Potential competing interests: None relevant to this manuscript for any author. L.A.F. has research funding for clinical trial participation from Takeda and CorEvitas. J.K.J.G. has research support for clinical trial participation from AbbVie, Janssen, Merck, Takeda and CorEvitas. There was no outside writing assistance used in the preparation of this manuscript.

Study Highlights.

WHAT IS KNOWN

• ✓ Starting treatment with an advanced inflammatory bowel disease therapy is often associated with delays because of the insurance authorization process.

WHAT IS NEW HERE

• ✓ Assessing patient-related, care-team-related, and insurance-related factors, we identified 3 significant factors that delay therapy: insurance denials, intravenous therapy, and lack of a dedicated pharmacy team.

• ✓ We highlight these areas of friction in the prescription-authorization process that could be targeted for intervention to reduce delays.