Comparing rates of primary medication nonadherence and turnaround time among patients at a health system specialty pharmacy compared with external specialty pharmacies

Abstract

BACKGROUND:

Access to specialty medications prescribed for complex and debilitating conditions is often delayed or unsuccessful. Specialty pharmacies can streamline medication access and timely delivery.

OBJECTIVE:

To compare primary medication nonadherence (PMN), the rate at which a prescription is written but not obtained by the patient, and turnaround time (TAT) between patients filling specialty medication with an integrated health system specialty pharmacy (HSSP) and those using external specialty pharmacies (non-HSSP).

METHODS:

A retrospective single-center cohort study was performed between June 2021 and May 2022. Patients prescribed specialty medications from the Vanderbilt Health System Oncology, Inflammatory, or Multiple Sclerosis clinics were included if they had a new prescription within the study period and follow-up encounter after the prescription. The primary outcome was PMN, defined as the prescribed medication or therapeutic alternative confirmed to not be filled. TAT was defined as time from a specialty medication prescription to the first filled claim. A logistic regression model was used to test for associations with PMN and the pharmacy the medication was sent to (HSSP vs non-HSSP), controlling for age, sex, clinic, race, and insurance. An ordinal logistic regression model was used to test for associations with TAT and the filling pharmacy (HSSP vs non-HSSP), controlling for age, sex, clinic, race, and insurance.

RESULTS:

There were 1,466 patients with 1 eligible prescription (67% HSSP and 33% non-HSSP) evaluated for PMN. Of these patients, the median age was 56 (interquartile range [IQR] 41-68) years, 55% of patients were female, and 56% had commercial insurance. PMN was 6.5% for HSSP patients, and 9.9% for non-HSSP patients. Patients who had a prescription sent to non-HSSP pharmacies had 60% higher odds of having PMN (P = 0.027). For the TAT outcome, 1,188 eligible patients were included (66% HSSP and 34% non-HSSP). TAT for patients filling at an HSSP was a median 3 (1-6) days and a median 4 (1-10) days for patients filling at a non-HSSP. Patients who filled at non-HSSP pharmacies had 40% higher odds of having a longer TAT (OR = 1.4, 95% CI = 1.1-1.7, P = 0.004). Patients using commercial insurance (P = 0.018) and Black patients (P = 0.018) had higher odds of a longer TAT.

CONCLUSIONS:

Prescriptions sent to non-HSSP specialty pharmacies are more likely to not be filled or to take longer to fill than those sent to an HSSP.

Plain language summary

Patients whose prescriptions are sent to non–health system specialty pharmacies are less likely to start treatment, and when they do, they tend to start later than those using a single health system specialty pharmacy (HSSP). HSSP pharmacists efficiently help patients get through medication access barriers, such as prior authorizations and financial assistance, while communicating with the patient and prescribing provider.

Implications for managed care pharmacy

Timely access to appropriate therapy for complex conditions, such as cancer, inflammatory bowel diseases, and multiple sclerosis, is essential to improving outcomes and reducing overall health care costs. This study identified that patients able to use a HSSP are more likely to access therapy and in a shorter amount of time. Access to timely treatment should be considered when designing pharmacy access networks.

Access to high-cost specialty medications, used to treat complex conditions such as cancer, moderate to severe inflammatory conditions, and neurologic disorders, can be challenging because of payer restrictions and medication affordability. ¹ Though specialty medications are often life-saving or life-altering for those with debilitating health care conditions, they make up more than half of total pharmacy spending. ² Many patients prescribed specialty medications must undergo a complex insurance approval process ^{3 – 7} and obtain additional financial assistance to afford treatment. ^{8 , 9} High costs may also dissuade patients from attempting to access or start therapy.

Primary medication nonadherence (PMN) occurs when newly prescribed medication is not filled and obtained by the patient within a reasonable time frame. ¹⁰ The use of electronic prescriptions over recent years has enabled research on the rate and reasons for PMN in several health conditions. Many factors may cause patients not to fill an initial prescription and can be related to patients, medications, health care providers, or health care systems, ^{11 – 15} and reported PMN rates vary by condition. Oral oncolytics have PMN rates ranging from 4% to 20.5%. ^{16 – 19} Inflammatory conditions, including inflammatory bowel disease (IBD) and rheumatoid arthritis (RA), were found to have PMN rates ranging from 2.1% to 54.5%. ^{18 , 20 – 23} Studies evaluating PMN in patients with multiple sclerosis (MS) found a varied range between 7% and 46.9%. ^{18 , 24} There is a gap in the literature on PMN for dermatologic conditions, but one study found a rate of 31.6%. ²⁵ Reducing PMN and ensuring patients have access to specialty medications is a priority to realize the benefits of these therapies.

Turnaround time (TAT) is a pharmacy metric commonly defined as the time from a pharmacy receiving a prescription from a health care provider to the patient obtaining the medication. ²⁶ Similar to PMN, several factors may impact the time to medication access. Insurers frequently require prior authorization (PA) completion for specialty medications, which may be denied and require further appeal, thus increasing time to medication access. ^{27 – 30} TAT has been less studied in specialty medications, with notable gaps in literature evaluating TAT in agents that treat inflammatory and dermatologic conditions. However, health system specialty pharmacists have reported improved TAT rates in some specialty areas. In oncology, a previous study at a health system specialty pharmacy (HSSP) reported a TAT of 3 days for non-limited distribution network oncolytic agents and 6 days for limited distribution oncolytics. ¹⁷ A study evaluating time to dalfampridine access found a median TAT of 22 days, which was reduced to a median of 1 day after the HSSP gained access to the manufacturer limited distribution network. ²⁴ Among patients whose specialty medications were filled with an internal HSSP, time to treatment initiation was 6 days shorter compared with patients filling at external pharmacies. ³¹

Both PMN and TAT can be important indicators of patients’ ability to access and benefit from therapy and contribute to patient experience and treatment outcomes. Evidence suggests when patients with RA wait to start therapy, they have greater joint damage. ³² Prevention of further damage and disability has also been shown with early initiation of disease-modifying therapy in MS. ³³ Patients with IBD who had delayed insurance approval for dose escalations of their biologic medication had significantly higher C-reactive protein levels than those with shorter duration to insurance approval. ³⁴ Additionally, patients with IBD who had delayed initiation of more than 15 days had higher risk of adverse events, such as infections, steroid use, emergency department visits, hospital admission, surgery, or death. ³⁵

Specialty pharmacy is a growing field with different models, including retail chain, payer-associated, independent, and integrated health systems. Integrated HSSPs are models that can improve patient care by assisting providers in the selection of appropriate medications, aiding patients in overcoming financial barriers to medication access, and closely monitoring patients during the duration of therapy. ^{36 , 37} HSSPs have the added benefit of access to the electronic health record (EHR), allowing for a streamlined process from prescribing to completing insurance requirements, ongoing clinical monitoring, and treatment optimization. ^{36 , 38 – 40} Several studies have shown the HSSP model produces low PMN rates, increases medication adherence, reduces TAT, and improves health outcomes in various disease states. ^{17 , 19 , 23 , 24 , 33 , 41 – 44} To our knowledge, there are limited comparative studies evaluating the potential impact of the HSSP model on patients’ ability to initiate new treatments in a timely manner compared with other specialty pharmacy models. ^{31 , 45} Therefore, the purpose of this study was to add to the emerging body of evidence comparing PMN and TAT between patients filling their specialty medication with an HSSP and those using external specialty pharmacies (non-HSSP).

Methods

SETTING AND STUDY DESIGN

This was a single-center retrospective cohort study of patients receiving a prescription from a large academic medical center with an integrated HSSP. Patients were from oncology, inflammatory (RA, IBD, and dermatology), or MS clinics prescribed a specialty medication between June 1, 2021, and May 31, 2022. Patients were excluded if they had any specialty medication prescribed or filled within 180 days of the index prescription (lookback window), if they used alternative filling methods (ie, provider administered, inpatient, clinical trial, manufacturer sample/patient assistance programs), if the prescription was a medication renewal, or if the prescription was sent in error. Duplicate prescriptions were removed when 2 sequential prescriptions for the same medication or therapeutic equivalent were prescribed within 30 days (duplicate window). Patients with a recorded fill date (regardless of the duration between prescription and fill date) were included in TAT analyses, excluding those requiring an intravenous (IV) loading dose. This study was approved by the Vanderbilt University Medical Center Institutional Review Board.

DATA SOURCES AND OUTCOME MEASURES

Prescription data were obtained from the EHR, and fill data were obtained from the HSSP prescription-filling software or Surescripts (non-HSSP fills). Data were stored in REDCap. ^{46 , 47} Patients without a claim for a fill within 60 days or a confirmed fill after 60 days were reviewed to confirm the ultimate outcome of the prescription (confirmed filled, confirmed not filled, unknown). This entailed a detailed chart review of the EHR, reviewing patient, pharmacist, and provider communications.

The primary outcome of this study was PMN, defined as a prescription confirmed to not be filled, either the originally prescribed medication or a therapeutic equivalent. Therapeutic equivalence was defined as any specialty medication intended to be used for the same condition. The PMN rate was calculated by dividing the total number of prescriptions confirmed to be not filled by the total number of prescriptions in the time period, excluding those with an unknown fill outcome. The secondary outcome was TAT, defined as the time from a specialty medication prescription to the dispense date and calculated as the total number of days between these two dates (calculated as whole numbers; analyses performed with raw number of days). Lengthy TAT was defined as a TAT greater than 14 days. ^{48 – 50} Fourteen days was chosen as a benchmark based on previous literature reporting average TAT is often around 7 days or less and longer delays can impact patient outcomes in many specialty disease states. Reasons for PMN and lengthy TAT, an exploratory outcome, were also collected, based on EHR documentation.

STATISTICAL ANALYSIS

For the primary outcome of PMN, univariate comparisons of PMN rates between HSSP and non-HSSP, including differences in demographics, were conducted using Pearson chi-square test or Fisher exact test, both overall and by clinic type. HSSP or non-HSSP designation was based on the pharmacy where the prescription was sent, as this was the last available data point. A multiple logistic regression analysis was then used to assess the association between PMN and pharmacy type (HSSP vs non-HSSP), adjusting for age, sex, clinic, race, and insurance type. For TAT, univariate comparisons between HSSP and non-HSSP, defined based on the pharmacy where the medication was ultimately filled, were conducted using the Wilcoxon rank sum test. An ordinal logistic regression analysis controlling for the same variables as the PMN model was used for the adjusted analysis, which was performed using an R package rms. ^{51 , 52}

Results are reported as odds ratios (ORs) with 95% CIs. Sensitivity analyses were conducted for the PMN model, in which the PMN likely outcome was categorized to either all PMN or all not PMN to assess the impact of the uncertain PMN cases on the results.

Multiple imputation was used to account for missing data on race (5%) and insurance type (2%) in the regression analyses. Using the predictive mean matching method, 20 imputation datasets were generated, and regression results are based on pooled estimates across these datasets. Variables included in the imputation model were age, sex, pharmacy prescription was sent to, filling pharmacy, race, insurance type, clinic, index medication requiring a loading dose, and time to fill (for the TAT analysis). All analyses were performed using R version 4.4.0.

Results

PMN

There were 4,172 prescriptions from the oncology, inflammatory, and MS clinics during the study period (Figure 1). Prescriptions were first excluded because of filling in the lookback window (n = 2,080) and duplicate prescriptions (n = 33), resulting in 2,059 initial eligible prescriptions. Of those, 592 prescriptions were excluded based on chart review for the following reasons: medication renewal (n = 322), alternate filling method (n = 183), nonspecialty medication (n = 85), and other (eg, prescription was sent in error) (n = 3). Of the final 1,466 prescriptions evaluated for PMN, 1,251 (85%) had a pharmacy claim for a fill (confirmed fill) and 215 were identified as potential instances of PMN based on lack of a pharmacy claim for a fill (15%). From the potential instances of PMN, 40 (3%) were external fills found on chart review (not PMN), 105 (7%) were confirmed not to be filled (PMN), and 70 (5%) were likely to be an instance of PMN, though their outcome was unknown because of the prescription being sent to a non-HSSP pharmacy with no fill data in Surescripts and no confirmation the medication had been filled on EHR review (PMN likely). In total, 88% of the prescriptions had a confirmed fill (not PMN) (n = 1,291). The breakdown of HSSP compared with non-HSSP for each PMN category can be found in Table 1, and comparison by clinic can be found in the Supplementary Materials ^{(214.9KB, pdf)} (available in online article).

FIGURE 1.

PMN Identification and Results

LBW = lookback window; PMN = primary medication nonadherence.

TABLE 1.

Patient Characteristics

Characteristic	HSSP (n = 986)	Non-HSSP (n = 480)	Total (N = 1,466)	P value
Age, years, median (IQR)	60 (45-70)	50 (35-61)	56 (41-68)	<0.001
Female sex, n (%)	516 (52.3)	295 (61.5)	811 (55.3)	<0.001
Race, n (%)
White	797 (80.8)	393 (81.9)	1,190 (81.2)	0.873 a
Black	94 (9.5)	48 (10.0)	142 (9.7)
Other	44 (4.5)	19 (4.0)	63 (4.3)
Missing	51 (5.2)	20 (4.2)	71 (4.8)
Insurance, n (%)
Commercial	495 (50.2)	327 (68.1)	822 (56.1)	<0.001 a
Government/Other	467 (47.4)	143 (29.8)	610 (41.6)
Missing	24 (2.4)	10 (2.1)	34 (2.3)
Clinic, n (%)
Multiple sclerosis	58 (5.9)	48 (10.0)	106 (7.2)	<0.001
Inflammatory	306 (31.0)	268 (55.8)	574 (39.2)
Oncology	622 (63.1)	164 (34.2)	786 (53.6)
PMN category, n (%)
Not PMN	908 (92.1)	383 (79.8)	1,291 (88.1)	<0.001
PMN likely	15 (1.5)	55 (11.5)	70 (4.8)
Confirmed PMN	63 (6.4)	42 (8.8)	105 (7.2)

^a Missing values were not included in the test.

PMN = primary medication nonadherence.

In the 1,466 patients evaluated, the median age was 56 (interquartile range [IQR] 41-68) years, 55% were female, and 56% had commercial insurance. There were 986 HSSP patients and 480 non-HSSP patients. Most patients were from the oncology clinic (54%, n = 786), 39% (n = 574) were from the inflammatory clinics, and 7% (n = 106) were from the MS clinics (Table 1). Patients filling at an HSSP were older (median 60 years [IQR, 45-70] vs 50 years [IQR, 35-61], P < 0.001), were more commonly male (48% vs 39%, P < .001), and more frequently had non-commercial insurance (49% vs 30%, P < 0.001) (Table 1).

Of the 1,396 prescriptions with a confirmed outcome of PMN or not PMN (uncertain cases of PMN removed), unadjusted PMN rates were lower for prescriptions sent to HSSPs compared with non-HSSPs (6.5%, n = 63 vs 9.9%, n = 42, P = 0.027) (Table 2). In the MS clinic (n = 97), 3% of HSSP prescriptions (n = 2) resulted in PMN vs 21% of non-HSSP prescriptions (n = 8) (P = 0.013). In the oncology clinic (n = 762), 7% of HSSP prescriptions (n = 42) resulted in PMN vs 6% of non-HSSP prescriptions (n = 9) (P = 0.654). In the inflammatory clinics (n = 537), 6% of HSSP prescriptions (n = 19) resulted in PMN vs 11% of non-HSSP prescriptions (n = 25) (P = 0.061). In a multiple logistic regression model, patients who had a prescription sent to non-HSSP pharmacies had 60% higher odds of having PMN (OR = 1.6, 95% CI = 1.1-2.5, P = 0.027). Patients with commercial insurance had 30% lower odds of experiencing PMN compared with patients with noncommercial insurance (OR = 0.7, 95% CI = 0.4-1.1, P = 0.110) (Figure 2). Results from the sensitivity analysis were consistent with the original model. Categorizing the uncertain PMN cases to PMN yielded an OR of 2.9 (95% CI = 2.0-4.0, P < 0.001) comparing non-HSSP patients to HSSP patients, whereas categorizing uncertain PMN to not PMN resulted in an OR of 1.5 (0.95-2.2, P = 0.086).

TABLE 2.

Unadjusted Analysis Results for PMN and TAT by Clinic

	Overall			Multiple sclerosis			Oncology			Inflammatory
	HSSP	Non-HSSP	P value	HSSP	Non-HSSP	P value	HSSP	Non-HSSP	P value	HSSP	Non-HSSP	P value
PMN (n = 1,396)	n = 971	n = 425		n = 58	n = 39		n = 609	n = 153		n = 304	n = 233
n (%)	63 (6.5)	42 (9.9)	0.027	2 (3.5)	8 (20.5)	0.013	42 (6.9)	9 (5.9)	0.654	19 (6.3)	25 (10.7)	0.061
TAT (n = 1,188)	n = 779	n = 409		n = 56	n = 31		n = 503	n = 208		n = 220	n = 170
Days, median (IQR)	3 (1-6)	4 (1-10)	<0.001	2 (1-5)	3 (1-10)	0.103	3 (1-6)	3 (1-7)	0.620	2 (1-5)	6 (1-18)	<0.001

HSSP = health system specialty pharmacy; IQR = interquartile range; PMN = primary medication nonadherence; TAT = turnaround time.

FIGURE 2.

Multiple Regression Analysis Results for PMN (Top) and TAT (Bottom)

Solid dots represent the OR, and horizontal lines represent the 95% CI. The numerical values of the ORs and corresponding 95% CIs are also shown in parentheses. For categorical variables, the reference group used in the analysis is specified. For age, which is a continuous variable, the OR reflects the comparison of a patient aged 70 years to a patient aged 40 years.

HSSP = health system specialty pharmacy; OR = odds ratio; PMN = primary medication nonadherence; ref = reference; TAT = turnaround time.

Among patients with a prescription sent to an HSSP, the most common reason for PMN was patient decision (30%, n = 19). PMN also occurred because of patients being unable to afford the specialty medication because the copay was too high (patients did not qualify for additional assistance) (19%, n = 12), and the medication was never approved by insurance (14%, n = 9). For non-HSSP patients, the most common reasons for PMN were medication not approved by insurance (31%, n = 13), the reason for PMN was undocumented (patient never confirmed receipt through the EHR) (24%, n = 10), and patient decision (19%, n = 8). For HSSP and non-HSSP patients, factors influencing patient decision included the patient not feeling ready to start the medication (n = 10), patient choosing a different medication (n = 8), concerns about side effects (n = 3), addressing other health issues first (n = 2), deferring treatment as this time (n = 2), did not want additional treatment (polypharmacy) (n = 1), and no longer needed (n = 1) (Figure 3).

FIGURE 3.

Reasons for PMN and for Lengthy TAT (>14 days)

HSSP = health system specialty pharmacy; PMN = primary medication nonadherence; TAT = turnaround time.

TAT

For the TAT outcome, the population studied included 1,188 eligible patients after excluding patients with an IV loading dose. There were 779 HSSP patients and 409 non-HSSP patients. Most patients were from the oncology clinic (60%, n = 711), 33% (n = 390) were from the inflammatory clinics, and 7% (n = 87) were from the MS clinics. Median TAT for HSSP patients was 3 (IQR 1-6) days and for non-HSSP patients was 4 (1-10) days (P < 0.001). Only 8% of HSSP patients experienced lengthy TAT, compared with 18% of non-HSSP patients (P < 0.001). The reasons for lengthy TAT among HSSP and non-HSSP patients are illustrated in Figure 3.

When stratified by clinic, there was a significant difference in TAT for inflammatory patients (HSSP 2 (1-5) days vs non-HSSP 6 (1-18) days, P < 0.001). Turnaround time was similar between HSSP and non-HSSP for oncology or MS patients.

In an ordinal logistic regression model for having a longer TAT, holding all other variables in the model constant, patients who filled at non-HSSP pharmacies had 40% higher odds of having a longer TAT (OR = 1.4, 95% CI = 1.1-1.7, P = 0.004). Patients using commercial insurance had 1.3 times higher odds of a longer TAT than those with non-commercial insurance (OR = 1.3, 95% CI = 1.1-1.7, P = 0.018). Additionally, the odds of having a longer TAT were observed by race, with Black patients having the highest odds of a longer TAT, followed by White patients and then patients grouped together as Other (2 degrees of freedom P = 0.018).

Discussion

Patients with a prescription generated at an HSSP and sent to an external (non-HSSP) pharmacy were significantly less likely to start treatment. When treatment was received, patients filling at a non-HSSP were more likely to have lengthy TAT compared with patients able to use the HSSP. These data are among the first to provide comparative results between the HSSP and non-HSSP model of care across multiple specialty areas and demonstrate the benefit of using an HSSP to have patients initiate life-altering therapy more often and faster.

IMPROVING ACCESS THROUGH THE HSSP MODEL

Low rates of PMN and shorter TAT among prescriptions filled at HSSPs are consistent with previous literature highlighting benefits of the HSSP model. ^{37 , 38 , 39} HSSP pharmacists are embedded in the clinic with specialty providers and other services and have access to the EHR in which patient clinical information is documented. This access enables HSSP staff to overcome barriers to medication access by thoroughly completing PAs and appeals with relevant clinical data. ³⁶ In the 2022 ASHP HSSP Clinical Services survey, 99% of HSSP respondents reported using the EHR to obtain pretreatment screening information and 51% coordinated with the clinic and patient to complete the necessary workup. ³⁶ As a member of the patient’s health care team, pharmacists are wholly invested in helping patients access appropriate and needed therapy. Access to the EHR and pharmacists’ integration in the clinic are key differentiators of HSSPs that enable timely access to medication for patients.

The most common reason for PMN among HSSP patients was patient decision. These results align with previous studies that have found nonadherence to be in part due to concerns about side effects, perceived redundancy, and questionable effectiveness. ^{4 , 53 – 55} Over half of HSSPs (57%) report sometimes, frequently, or always providing pretreatment patient counseling prior to a medication being selected. ³⁶ To support patients in making informed decision that may reduce PMN, HSSP pharmacists should continue to provide thorough education and counseling, specifically on the potential benefits and risks associated with the medication and addressing the initial concerns that arise. HSSPs should also use tailored monitoring strategies based on patient preferences and desired level of support. Patients may be more likely to start therapy if they are confident that they will receive support while receiving treatment. ⁵⁶ Conversely, inability to access or afford the prescribed specialty medication was the most common reason for PMN among non-HSSP patients. Specialty medications are complex and require a time-consuming process to obtain payer approval. ^{5 , 30} In the 2022 ASHP HSSP Clinical Services survey, 84% of HSSPs reported assisting all patients with insurance PAs, 80% assisted with PA denials, and 77% helped patients enroll in financial assistance when needed for all patients referred to the HSSP regardless of their ability to fill the medication. ³⁶

However, these practices differ by site, and many HSSPs do not complete access requirements for medications they are unable to fill to prevent redundancy with external specialty pharmacies that will be filling the medication. At the health system in the current study, embedded HSSP pharmacists provide medication access services if the pharmacy has access to dispense the medication, regardless of whether the patient fills with the HSSP. It is likely that the difference in PMN and TAT between the HSSP and non-HSSPs would be greater if the HSSP pharmacists did not provide this service to all patients. This study demonstrated that PMN is lower for patients prescribed a specialty medication who filled at an HSSP compared with patients who filled at a non-HSSP.

TAT

Treatment delays are often experienced with specialty medications because of an often complex medication approval process. ^{4 , 6 , 57} TAT for HSSP patients was significantly less than that for non-HSSP patients, and significantly fewer HSSP patients had lengthy TAT. This study builds on previous HSSP research demonstrating shorter TAT compared with that of non-HSSPs. Russell et al found patients filling with an internal HSSP had a shorter time to initiation compared with those filling externally (12 days, HSSP patients vs 18 days, non-HSSP patients, P < 0.0001). ²¹ Another study found delays in starting IBD therapies were significantly more likely to occur when the site did not have a dedicated pharmacist. ⁵⁸ HSSPs are uniquely positioned to reduce TAT by assisting patients with initiating treatment, coordinating PAs, navigating financial assistance challenges, and coordinating timely medication delivery. ^{27 , 59}

The most common reasons for lengthy TAT were due to insurance delays and pharmacy logistics barriers. This was true for both HSSP and non-HSSP patients. The results are similar to a recent study, which found delays in starting therapy (>14 days from prescription to first dose) were significantly more likely if the prescription was denied by insurance. ⁵⁸ Notably, 18% of reasons for lengthy TAT in non-HSSP patients were unable to be identified because of no communication from the external pharmacy to the provider, highlighting the limited visibility into the prescription journey when sent externally. Drivers for increasing TAT in the current study align with a previous qualitative study assessing factors that could lengthen specialty pharmacy TAT, including PA delays, barriers with health benefit formulary management, differences between managed care organizations, and miscommunication with physicians. ¹⁷

In addition to the HSSP model impact, patients using commercial insurance had significantly higher odds of a longer TAT and lower odds of PMN than those with non-commercial insurance. Patients with commercial insurance often have the opportunity to use copayment assistance to help afford their high-cost specialty medications. ^{60 , 61} However, this can lead to delays in treatment initiation. ⁶² A recent study of oncology patients filling oral anticancer drugs found that patients who used a copay assistance program had longer time to medication receipt compared with patient who did not use copay assistance. Additionally, patients filling oral anticancer drugs with commercial insurance or Medicare had longer time to medication receipt compared with patients with Medicaid. ⁶⁵ Differences in access rates and TAT may vary by clinic or medication depending on payer formularies and requirements.

CLINIC DIFFERENCES

When broken down by clinic, patients with MS were more likely to experience PMN when filling with non-HSSPs. Although previous studies have shown high rates of secondary adherence for patient with MS, ^{63 – 65} there is limited literature on PMN to compare these results with previous studies. PMN results in the current study were lower compared with a study reporting noninitiation for Medicare beneficiary patients, in which MS was included in the immune-condition patient group. ¹² A large amount of research in MS specifically has shown that HSSP pharmacists practicing in MS clinics improve patient care and coordination of care and provide clinically meaningful intervention to support access, adherence, and persistence to disease-modifying therapy. ^{29 , 66}

There was no significant difference in PMN for patients in the inflammatory clinics filling their prescription at an HSSP or non-HSSP. Similar to MS, there are limited data on PMN for patients with IBD. Although the population is different, one study reported a PMN rate of 46.9% for Medicare patients prescribed immune systems medications (ie, MS, RA, and IBD). ¹² Additionally, limited literature is available on PMN rates for dermatologic conditions. Beukelman et al (2023) found for patients prescribed disease-modifying antirheumatic drug medications from 2007 to 2020 for RA, PMN at 3 months was 33% for Medicare patients and 38% for patients with commercial insurance. However, PMN was calculated for both specialty and nonspecialty medications. One possible reason for PMN (not mutually exclusive) was due to insurance denial. ²² A recent study found patients prescribed dermatologic medications had higher rates of noninitiation if a PA was denied. ⁶⁷ Similarly, PA denial was a top reason for PMN in the current study. Additional research is needed to further evaluate rates and reasons for PMN in these conditions.

Overall, PMN for oncolytic medication was 7% for HSSP prescriptions and 6% for non-HSSP prescriptions, which is similar to prior studies. ^{6 – 12} We hypothesize that these similarities are due to the clinical pharmacist role within those clinics. Pharmacists help patients access and afford specialty medication treatment regardless of whether patients can fill at the HSSP or not, which likely brings down the overall PMN rate for all patients seen at the HSSP. There may be a sense of urgency to fill oncology medications since earlier treatment can impact clinical outcomes, ^{5 , 68} whereas MS and inflammatory medications may require a more lengthy approval process to demonstrate medical necessity or step therapy before the prescription is filled. The most common reason for PMN for oncology patients was patient decision, highlighting the importance of shared decision-making conversations in this population.

LIMITATIONS

This was a single-center study, and the data may not be generalizable to the population or other specialty pharmacies. The final outcome of some prescriptions, and the reason for PMN and lengthy TAT, were not clearly documented after prescriptions were sent to external pharmacies. Therefore, we were not able to fully capture the reasons for PMN among patients filling at external pharmacies. The pharmacy the prescription was sent to was used to categorize as HSSP or non-HSSP for the PMN calculation. Typically, prescriptions are triaged to the HSSP for review of appropriateness prior to being sent to an external pharmacy if the HSSP is not filling the prescription. Though uncommon, there could have been instances in which a prescription was sent to the external pharmacy and later returned to the HSSP to be filled. There could be unmeasured confounders and differences in patient characteristics (eg, socioeconomic status) that were not accounted for in the analysis that led to patients filling with the HSSP or external pharmacy. Since these were not captured, their potential impact on PMN and TAT cannot be ruled out. Another limitation is the use of Surescripts data for external pharmacy fills, which captures approximately 80% of all external fill data through pharmacy and insurer claims and reports these claims back to the EHR. Therefore, not all external fills were captured. However, chart review was performed for all patients without a fill. Patients categorized as PMN were confirmed to not have filled a prescription, whereas those for which a final determination could not be made were categorized as PMN likely. Another limitation to this study was the smaller sample size of patients in the MS clinic and generalizability to Medicare patients due to changes made in 2025. The changes made to cost shares for Medicare beneficiaries could impact PMN because of lower cost shares (reduce PMN) or potentially increase PA requirements (increase PMN). Finally, HSSP clinical pharmacists provide comprehensive care and help patients fill specialty medications to the best of their ability, despite the filling pharmacy. This could have led to reduced PMN and TAT for external fills.

Conclusions

This study demonstrated that patients with a prescription sent to a non-HSSP specialty pharmacy are more likely to experience PMN and lengthy TAT. Low rates of PMN at HSSPs may be attributed to integrated specialty pharmacists who assist with potential medication access barriers, such as PA and financial assistance, while maintaining transparent communication with the patient and prescribing provider.