Comparative evaluation and preference of MixPro^® versus Mix2Vial^® reconstitution devices among people with haemophilia and caregivers

This study included people with haemophilia A or B (with or without inhibitors), or caregivers of PwH under the age of 18 years. The PwH or caregivers had to be responsible for the regular infusion of replacement factor to be eligible. Participants were permitted to have previously used both MixPro^® and Mix2Vial^® devices before the interviews were conducted.

This study was designed to evaluate the current unmet needs of today's reconstitution devices and which features of these devices are most important to PwH and caregivers. It also examined how well MixPro^® and Mix2Vial^® perform on each feature, the overall preference for one device versus the other, and key benefits and drawbacks for each device.

Qualitative structured interviews were conducted face-to-face in a fieldwork agency office in the USA, Europe, or Japan. Interviews were undertaken in person so that all respondents could interact with demonstration MixPro^® and Mix2Vial^® devices. The interviews were conducted between 22 June 2021 and 4 August 2021, and each interview lasted approximately 30 minutes. The interview flow was predetermined, with each of the two devices being evaluated separately using a random rotation. This resulted in half of the participants evaluating MixPro^® first, followed by Mix2Vial^®, and the other half evaluating the two devices vice versa. Topics included: unmet needs, importance of features, evaluation of each device including performance metrics and advantages/disadvantages, device best rated for each performance metric, overall preference, and impact on life (Figure 1).

Figure 1

To assess unmet needs, participants were asked ‘What, if any, would you say are your biggest unmet needs with your current mixing/reconstitution system?’, prior to being provided with either the MixPro^® or Mix2Vial^® device. Participants were allowed to provide their thoughts to the question with no prompts from the interviewer. The open-ended answers were then analysed and coded based on the frequency at which different needs were mentioned. Current and past devices used by each participant were also recorded, including whether they had experience using either the MixPro^® or Mix2Vial^® device.

In order to assess the importance of certain device features, a list of features was adapted from previous studies by the research team ^[3,16]; it was then further updated and refined following pilot interviews (Supplementary Table 1). The importance of each device feature was derived from a series of exercises with the respondent, followed by analysis of the data. First, participants were asked to perform a card-sort exercise to rank all 18 features from the most to least important. Results from this ranking task were then used in a win-loss analysis (also known as the Bradley-Terry model; B-T) to identify the frequency at which one feature is rated more important than each of the other features ^[27]. The raw B-T parameter estimates were transformed into a user-friendly form by dividing each attribute's B-T estimate by the sum of the B-T attribute estimates to arrive at a relative importance percentage ^[28].

Using the same list of features, respondents were asked to rate each of the two devices on the 18 features using a 7-point scale, where 1= Does not describe the device at all and 7= Completely describes the device. Data are presented as the percentage of respondents that provided a rating of either 6 or 7 for any given feature in the benefits list (top 2 box rating).

A word association exercise was also completed for each of the two devices in which participants were given a list of 26 words (Supplementary Table 2) and asked to select up to eight words that were most associated with the device being assessed. The selected list of words was based on previous studies and revised by the research team for the purpose of this study ^[3,16]. Data are presented as the percentage of participants who selected a given word. Any word selected by fewer than five participants is not shown.

Once each device was assessed individually, a direct comparison was performed. Using the same 18 device features assessed for importance, participants were asked to indicate which device performed better for each feature. Participants were asked to select one or the other device; ‘neither’ was not given as an option for the answer.

A final overall preference question was asked using a 100-point allocation exercise. Respondents were asked to allocate 100 points between the two devices to demonstrate their level of preference between them.

A paired-samples t-test of means was used to compare the data for each device for all feature metrics (where the same respondent rated both devices) to determine statistical significance at a 90% confidence level (commonly accepted for these types of studies).

When testing between two independent groups (such as PwH vs. caregivers or the USA vs. Italy), statistical testing was conducted at the 90% confidence level using an independent samples t-test of proportion.

In total, 105 participants (71 PwH and 34 caregivers) were interviewed across the USA (n=30), Italy (n=25), UK (n=30), and Japan (n=20). The mean (range) age of PwH represented in this research was 29 (3–69) years, with the highest mean being in Italy at 39 years. The mean age was 34 years for PwH self-administering injections and 16 years for the children of caregivers (Table 1). Almost all PwH (97%) had been prescribed their current medication as a prophylactic and the majority (86%) were diagnosed with haemophilia A (Table 1). Overall, seven participants reported they had previously used or were currently using the MixPro^® device and 15 participants had previous or current experience with Mix2Vial^®.

When asked about the biggest unmet needs with their existing reconstitution devices, 33% of participants reported that they were satisfied with the existing device they were using and had no unmet needs. However, 35% of participants described unmet needs relating to the overall ease of use, with specific mentions around the need for fewer parts (12%), faster reconstitution (11%), and being easier to use (7%). Eleven percent of participants stated that existing devices are too big and that less bulky devices would be preferable. Participants also expressed the need for less waste, sturdier components, a better transfer system, and alleviating the need for refrigeration to improve the overall experience with these devices (Supplementary Figure 2). This may be in part due to these participants using older devices and being less familiar with the more streamlined, modern devices used in this study. Thus, the unmet needs reported here are highly subjective and are likely to reflect the drawbacks of whichever device that a participant is currently using which may also reflect the ancillaries provided alongside the device itself.

Of the 18 listed device features that were assessed for each reconstitution device, low contamination risk was deemed the most important (importance score: 15.1) overall, scoring twice that of the next most important feature for a reconstitution device. The second most important feature was confidence in using the system correctly (8.3), followed by overall ease of use (7.5). Suitability for a person with limited strength was deemed the least important feature overall (2.2; Figure 2a).

Figure 2

When categorised by demographics (PwH or caregiver), PwH appeared to place more importance on device convenience, rating portability (7.6 vs. 5.7 for PwH and caregivers, respectively), and ease of use (6.2 vs. 2.9) highest. Conversely, caregivers were more focused on device function, rating verifying all factor is drawn up (5.8 vs. 9.7 for PwH and caregivers, respectively) and speed of mixing (5.9 vs. 8.0; Figure 2a) highest. Although low contamination risk was deemed the most important feature among all respondents (both PwH and caregivers) of all ages in the USA and the UK (12.4 and 15.0, respectively), it appeared to be of paramount importance in Japan, where respondents ranked this feature higher than in other countries (38.1). In Italy, however, device portability was deemed the most important feature (11.7), closely followed by low contamination risk (10.6; Figure 2b).

Overall, MixPro^® was rated higher than Mix2Vial^® for 17 of the 18 benefit features. Low contamination risk was the only feature rated equally for both MixPro^® and Mix2Vial^® (59% vs. 58% top 2 box rating, respectively). MixPro^® was rated most highly on good visibility of product (88%), ease of verifying all factor has been drawn up (87%), and overall ease of use (85%; Figure 3). For Mix2Vial^®, good visibility of product and confidence in using the system correctly were the features in which it performed best (59% for both), although still rating lower than MixPro^®.

Figure 3

PwH rated MixPro^® higher than caregivers did on almost all features, with the exception of low contamination risk (56% vs. 65% for PwH and caregivers, respectively) and ease of verifying all factor has been drawn up (86% vs. 88% for PwH and caregivers, respectively). With Mix2Vial^®, PwH also rated the device higher than caregivers on 13/18 features, with the exceptions relating to ease of use, ability to draw up product, and control over mixing steps.

When assessed by country, the USA and the UK generally rated MixPro^® higher than Mix2Vial^® compared with Japan and Italy.

When participants were asked to choose which device performed better for each of the listed features, MixPro^® was selected by ≥60% of participants for each of the 18 features (Figure 4a). Similarly, both PwH and caregivers selected MixPro^® as the best performer for all features (Figure 4b).

Figure 4

Regarding country-level evaluations, MixPro^® outperformed Mix2Vial^® across all features in the USA and Italy. In the UK, MixPro^® was selected as the best performer for all the listed features with the exception of the system is sturdy, where the two devices’ preference was split 50/50; this split was also seen in Japan. Further, the only listed feature for which Mix2Vial^® outperformed MixPro^® in Japan was easy to use if need more than one vial for injection, for which 55% of participants selected Mix2Vial^® as the best performer (Supplementary Figure 3).

When asked to select which of the two devices they preferred overall, 74% of participants chose MixPro^® over Mix2Vial^® (Figure 5a), stating reasons such as MixPro^®'s small, compact design; speed and ease of use; and good visibility when drawing up the mixed factor into the syringe. MixPro^® was the preferred device for caregivers (73%) and PwH of all ages (74%), across all countries (Figure 5b), with the highest degree of preference being in Italy (78%) and the USA (77%).

Figure 5

Although MixPro^® was the preferred device by PwH and caregivers in all countries, in the USA and the UK, PwH had a higher degree of preference for MixPro^® compared with caregivers (78% vs. 74% in the USA, and 72% vs. 66% in the UK, for PwH and caregivers, respectively); the opposite was true in Italy and Japan (78% vs. 80% in Italy, and 68% vs. 75% in Japan, for PwH and caregivers, respectively).

When participants were asked to select words that they most closely associated with each device, words such as quick (54%), user-friendly (47%), easy (46%), convenient (40%), simple (40%), compact (38%), and portable (34%) were most commonly associated with MixPro^® (Figure 6a). For Mix2Vial^®, words such as easy (33%), safe (32%), awkward (30%), simple (30%), fiddly (29%), and user-friendly (29%) were most commonly used (Figure 6b). Further, word association for MixPro^® was generally positive, with few participants associating this device with ‘negative’ words such as fragile or complicated; word associations for Mix2Vial^® were more balanced, with participants using both ‘positive' and ‘negative’ phrases more equally (range 10–33% for Mix2Vial^® vs. 5–54% for MixPro^®).

All words, with the exception of user-friendly, were associated with MixPro^® at a similar frequency for PwH and caregivers. However, caregivers associated the term user-friendly with MixPro^® more than PwH (37% vs. 68% for PwH and caregivers, respectively; Figure 6a). For Mix2Vial^®, confusing was the only word used more by caregivers than PwH (11% vs. 26% for PwH and caregivers, respectively; Figure 6b).

Figure 6

When participants who preferred MixPro^® were asked what impact the device would have on their life, unaided responses included saving time, ease of portability, and general confidence in using the system. Participant responses regarding MixPro^® included such remarks as “That slight chance of mix-up with product Mix2Vial^® is concerning, no chances with MixPro^®, I'll use whatever product reduces that chances of mix error” (Caregiver, US) and “Nice, easy, quick, better for holidays, less packing, less to explain to customs, assume small package, more inclined to take to work so better for emergency situations” (PwH, UK). Only 7% of participants indicated that MixPro^® would have no impact on their life.

The overall commentary was limited for Mix2Vial^® due to its lower preference overall, with only nine responses. Of those nine, five suggested the device would not have any impact on their life, primarily as it is similar to existing reconstitution devices they are using. Participant remarks on Mix2Vial^® included “It is similar to the device I am using, and feel no difficulty” (PwH, Japan) and “Similar to what I am using now but not any better so not much difference” (Caregiver, UK).

Due to the nature of this comparative, semi-qualitative study, there are some limitations. First, as this study did not exclude participants with prior experience of using either MixPro^® or Mix2Vial^®, or stratify participants based on level of experience with these devices, potential variations in between groups may have skewed the data. Although participants were permitted to have previously used the devices, they may have been more familiar with one or other, or have never used either device, which may have affected their view. This could also be reflected in their reported unmet needs, as participant responses are likely to vary significantly depending on which type of device they currently use. As these data were not considered in this analysis, we cannot fully interpret the responses. The same may also be true of participant age, given that adolescents tend to assess devices differently to adults, possibly skewing the results.

Second, the study design included both qualitative and quantitative data collection. Qualitative data are innately subjective and open to interpretation, and it is possible that errors or bias were introduced when analysing these data (specifically on unmet needs and impact on life). Additionally, as data were gathered via face to-face interviews, it is possible that some degree of interview bias was introduced. However, the order of features and answers in most tasks were rotated, and interviews were conducted at multiple sites by different interviewers to help eliminate the possibility of bias.

Another limitation is that the paired comparison for device features involved a forced choice, meaning that participants had to choose only one device and could not indicate if they believed the devices performed equally well. This type of questioning could skew results in a particular direction and may not reflect the actual preference.

Finally, although recruitment took place over a number of weeks, the total number of participants in this analysis was small. This limited the type of statistical analyses that could be performed at the segment level and, therefore, precludes any conclusive statements for specific segment analyses (e.g. a caregiver in Italy).