Knowledge, attitude and practice of health care providers toward prescribing factor replacement at federally funded haemophilia treatment centres in the United States

Study participants were exclusively health care providers currently working at federally funded HTCs in the US. This included haematologists, nurse practitioners (NPs), physician assistants (PAs), nurse coordinators, and staff nurses.

Data was collected in 2019 using a one-time electronic, anonymous survey created through the University of Miami's Qualtrics secure web-based survey application. Potential participants received an e-mail invitation with a unique embedded link for self-administering the survey.

The survey was a non-validated tool developed by the authors and comprised ten (10) demographic questions, five (5) questions specific to current prescribing practices, and five (5) questions on attitudes toward industry. The estimated survey completion time was 15–20 minutes.

An email distribution list was created using the Center for Disease Control and Prevention's (CDC) HTC directory, which was validated by HTC regional coordinators and the Partners in Bleeding Disorders Education organisation ^[5,6]. Survey invitations were distributed to 775 potential participants. Thirty-one (31) surveys were omitted as a result of incorrect emails, or providers responding that they either no longer worked at an HTC, or solely specialised in thrombosis.

The survey response data was downloaded directly from the Qualtrics survey software into the Statistical Package for the Social Sciences (IBM SPSS, Armonk, New York), Version 25, for data computation and analysis. Analysis included simple descriptive analysis of demographic information and descriptive/frequency statistics to assess for missing data. To examine the driving forces for changing a patient's factor concentrate, mean (M) responses for each driving force were calculated, and the factors were ranked by level of importance, with one (1) being not important and five (5) being very important. The level of importance of HCPs’ perceptions of patient motivation leading to a request for a change in prescribed factor concentrate was also analysed by calculating mean responses and ranking by level of importance. To determine whether provider level of importance or patient level of importance differed based on provider characteristics and industry practices, a series of Chi-Square analyses were conducted. When assumptions of Chi-Square were violated, Fisher's Exact Test was used instead.

A summary of study participant demographic characteristics can be found in Table 1. The highest response rates were from the Great Lakes (16.9%), Southeast (15.3%) and Great Plains (14.4%) regions. The majority of respondents reported working at an HTC associated with an academic centre (59.3%), with haematologists accounting for 36.6% of responses, followed by nurse coordinators (25.4%) and NPs (16.1%). Survey participants had an average age of 50.9 years (SD = 12.0, range 28–77). When further broken down by category, the average age of haematologists was 51.4 years (SD = 10.7, range 34–72), NPs 43.9 years (SD = 10.9, range 28–65), PAs 35.0 years (n=1), nurse coordinators 52.4 years (SD = 12.7, range 28–77), and staff nurses 53.0 years (SD = 17.3, range 33–63). The mean number of years working as a HCP was 23.68 years (SD = 11.67, range 5–45) and mean length of time working in a HTC was 12.0 years (SD = 10.59, range 1–40), with 2.5% of providers having worked at a HTC for less than one year. Thirty-six percent (36.4%) of HCPs reported being responsible for paediatric patients; 16.1% cared for adults only, and 28.8% cared for both children and adults. Seventy-one per cent (71.2%) of respondents reported having a 340B Drug Pricing Program at their HTC ^[7].

Provider responses to the level of importance assigned to various driving forces for changing a patient's factor concentrate is summarised in Table 2. The most common driving forces for changing a patient's factor concentrate included: poor response to current therapy (breakthrough bleeds) (M = 4.84); non-adherence to current therapy (M = 4.39); inadequate pharmacokinetic (PK) studies; ‘Other’ (M = 4.25); and less frequent dosing option (M = 4.20).

Table 3 summarises provider perceptions of patient motivations leading to a request to change their prescribed factor concentrate. These included less frequent dosing option (M = 4.55); poor response to current therapy (M = 4.36); factor characteristics (volume to be infused, mixing device, vial size) (M = 4.14); lower out-of-pocket costs (M = 3.98); and patient/caregiver brand loyalty (M = 3.82).

Models also examined whether particular demographic factors predicted the likelihood of identifying a driving force as being ‘important’ or ‘very important’, both from the perspective of the provider and the patient. These indicated that age (above or below median age), experience (above or below median experience) and position (haematologist compared to NP/PA) were not significantly associated with a higher likelihood of endorsing any particular reason as important or very important.

When evaluating who typically initiates the conversation regarding switching a patient's factor therapy, the HCP (haematologist, NP, PA, nurse coordinator) was most common (56.0%) followed by the patient/caregiver (23.8%). An ‘Other’ option was included for this question, and several respondents reported a combination of both patient and provider initiating the conversations.

Haematologists (55.3%) were primarily responsible for informing patients about factor therapy options, followed by NPs/PAs (23.5%) and nurse coordinators (15.3%). The majority of HTC providers reported patients/caregivers having a strong influence (65.5%) or some influence (33.3%) over which factor therapy is prescribed. When asked how much influence providers have over which factor therapy is prescribed, 58.8% reported having a strong influence while 36.5% had some influence.

When examining whether provider level of importance differed based on provider characteristics and industry practices, significant differences were found with regard to how much influence a patient or caregiver was perceived to have over which factor therapy is prescribed (p = .047) and which patients requested a switch in factor products most often (p = .050). Specifically, in practices where patients or caregivers were perceived to have a strong influence over which factor therapy is prescribed, providers were more likely to rate less frequent dosing options as important or very important. Additionally, as patient or caregiver influence increased, providers rated patient/caregiver requests as more important. Differences approaching significance were noted in how much influence providers have over which factor therapy is prescribed (p = .064), where providers with stronger influences were more likely to cite inadequate PK studies as an important or very important driving force for changing a patient's factor therapy regimen.

When reviewing whether the perceived patient level of importance differed based on provider characteristics and industry practices, significant differences were found in how much perceived influence a patient or caregiver has over which factor therapy is prescribed (p = .013). Specifically, providers who perceived that patients had a strong influence over which factor therapy is prescribed were more likely to rate less frequent dosing options as being very important.

With regard to which population of patients switched factor products most, the adolescent age range (12–18 years old) was the most common (34.1%), followed by 25–45 year olds (30.5%), 18–25 year olds (19.5%), and patients <12 years (15.9%). As the age of the patient decreased, less frequent dosing options were more likely to be rated as a very important driving force for switching factor therapy: 61.5% ranked less frequent dosing options as very important for those aged <12 years, 46.4% for 12–18 years old, 20.0% for 18–25 years old, and 36.0% for 25–45 years old.

The majority of HTCs did not have standard protocols in place for previously untreated patients (PUPs), prophylaxis treatment, and immune tolerance induction (ITI): 43% had protocols for PUPs, 31% had protocols for choice of factor concentrate for prophylaxis treatment, and 45.9% had protocols for ITI therapy.

With regard to the scope of practice for non-physician prescribers (NP/PA), of those who responded, only 21.7% were able to initiate and/or change prescriptions independently. Forty-seven percent (47.0%) were able to submit prescriptions; however, any changes had to be discussed with the haematologist before submission. Thirty-one per cent (31.3%) mentioned the haematologist being the primary prescriber. Haematologists, NPs, PAs, nurse coordinators and staff nurses had different patterns of responses to this question (p = .001). Of the haematologists who responded, 32.4% reported the haematologist being the primary prescriber, 18.9% allowed NPs and PAs to prescribe independently, and 48.6% required changes to be discussed prior to submission. Of the NP/PA respondents, 52.9% reported being able to prescribe independently, and 47.1% needed to discuss changes prior to submission. The majority of nurse coordinators (48.1%) and staff nurses (50.0%) reported haematologists being the primary prescriber.

Eighty-six percent (86.2%) of respondents indicated that their institution allows pharmaceutical representatives to visit the office for in-person meetings. The frequency of these visits is most commonly yearly (49.3%) or every six months (29.6%). About half (52.4%) of the HTCs sampled allow industry to interact with patients. Most respondents (77.4%) do not participate in any manufacturer's speakers’ bureaus. When asked about the primary source of information regarding new or existing therapies, respondents indicated the most common primary source of information being in-person symposium/presentations at medical conferences (39.3%), medical journals (16.7%), meeting with pharmaceutical sales representatives (11.9%), meeting with pharmaceutical medical science liaisons (10.7%), and their own research (9.5%). When further evaluating the percentage of HCPs receiving information at medical conferences, 79.1% of haematologists, 94.4% of NPs, 100% of PAs, and 83.3% of nurse coordinators reported this as being their primary source.

This study was a self-directed survey with self-reported response biases. The study response rate of 15.9% resulted in a small sample size with the limitations of self-selection ^[18]. The number of valid responses also varied significantly by geographic region resulting in an uneven distribution that did not allow for regional comparisons.

The response rate was also impacted after discovering the CDC HTC directory is not routinely updated and many e-mail addresses belonged to HCPs who no longer worked at a particular HTC or were invalid. The CDC and regional coordinators need to maintain a reliable directory to better understand the staff delivering care at HTCs, and more importantly, provide patients with a dependable resource for contacting HTC providers, which is common when patients are moving home or traveling. Due to the lower response rate, we were unable to draw statistically significant conclusions.

Despite the low response rate, this is the first study to examine prescribing practices among HTC providers in the US and provides interesting insight.