In total, 48.6% (220/453) of respondents to the Associatin des Patients Immunodéficients du Québec (APIQ) survey indicated SCIg as their current IgRT infusion method and were included in our analyzes (Fig. 2). Overall, 66 (30.0%) respondents indicated vials and 120 (54.6%) indicated PFS as their current packaging method. A further 34 (15.5%) respondents reported using a mixture of vials and PFS (referred to as ‘mixed packaging’ for brevity). Of the 220 respondents, 99 (45.0%) indicated pump as their chosen administration method, with 38 (38.4%) of these using vials, 38 (38.4%) using PFS, and 23 (23.2%) using mixed packaging methods. The remaining 121 (55.0%) respondents indicated manual push as their chosen administration method, with 28 (23.1%) of these using vials, 82 (67.8%) using PFS, and 11 (9.1%) using mixed packaging methods. No significant differences were found in respondent characteristics between the vial, PFS, and mixed packaging cohorts overall (Table 1), or in the pump and manual push administration method subgroups (Additional file 1 and Additional file 2).
Table 1
Summary of respondent characteristics of the vial, PFS, and mixed packaging cohorts
Respondent characteristics (n = 220) | Vial cohort (A) | PFS cohort (B) | Mixed packaging cohort (C) | p values |
Summary | n | Summary | n | Summary | n | A vs. B | A vs. C | B vs. C |
Age (years), median [IQR] | 57 [48, 66] | 66 | 59 [49, 66] | 120 | 59 [42, 68] | 34 | 0.34 | 0.92 | 0.63 |
Age at diagnosis (years), median [IQR] | 46 [25, 59] | 65 | 42 [25, 42] | 118 | 45 [29, 60] | 34 | 0.99 | 0.84 | 0.84 |
Gender, n (%) | Female Male | 39 (60.0%) 26 (40.0%) | 65 | 76 (63.9%) 43 (36.1%) | 119 | 18 (52.9%) 16 (47.1%) | 34 | 0.61 | 0.50 | 0.25 |
Bodyweight (kg), mean (± SD) | 75.3 (± 17.2) | 58 | 76.0 (± 17.5) | 109 | 78.3 (± 15.4) | 28 | 0.79 | 0.45 | 0.53 |
Underlying condition, n (%) | CVID IgG Sub DGS SID Other* | 25 (47.2%) 13 (24.5%) 2 (3.8%) 9 (17.0%) 4 (7.6%) | 53 | 37 (40.2%) 26 (28.3%) 4 (4.4%) 10 (10.9%) 15 (16.3%) | 92 | 14 (50.0%) 6 (21.4%) 1 (3.6%) 2 (7.1%) 5 (17.9%) | 28 | 0.48 | 0.54 | 0.91 |
Years since diagnosis, n (%) | < 2 years 2–9 years ≥ 10 years | 4 (6.2%) 34 (52.3%) 27 (41.5%) | 65 | 10 (8.5%) 46 (39.0%) 62 (52.5%) | 118 | 1 (2.9%) 22 (64.7%) 11 (32.4%) | 34 | 0.22 | 0.46 | 0.03 |
Time on IgG, n (%) | < 1 year 1–2 years 2–3 years 4–6 years ≥ 6 years | 3 (4.6%) 67 (10.6%) 4 (6.1%) 21 (31.8%) 31 (47.0%) | 66 | 5 (4.2%) 21 (17.5%) 17 (14.2%) 18 (15.0%) 59 (49.2%) | 120 | 2 (5.9%) 4 (11.8%) 5 (14.7%) 9 (26.5%) 14 (41.2%) | 34 | 0.56 | 0.41 | 0.76 |
Current treatment experience, n (%) | < 2 years 2–9 years ≥ 10 years | 8 (12.1%) 44 (66.7%) 14 (21.2%) | 66 | 21 (17.8%) 64 (54.2%) 33 (28.0%) | 118 | 3 (9.1%) 24 (72.7%) 6 (18.2%) | 33 | 0.26 | 0.82 | 0.16 |
Antibiotics before IgG, n (%) | No Yes | 44 (77.2%) 13 (22.8%) | 57 | 79 (74.5%) 27 (25.5%) | 106 | 22 (71.0%) 9 (29.0%) | 31 | 0.71 | 0.52 | 0.69 |
Antibiotics since starting IgG, n (%) | No Yes | 24 (36.9%) 41 (63.1%) | 65 | 50 (43.9%) 64 (56.1%) | 114 | 11 (35.5%) 20 (64.5%) | 31 | 0.37 | 0.89 | 0.40 |
Categorical variables were analyzed using chi-squared tests, and continuous variables were analyzed using Analysis of Variance (ANOVA) post-hoc tests if found to be normally distributed, or Mann Whitney tests if otherwise. Significant p-values are highlighted in bold. *Other indications are: X-linked agammaglobulinemia (vial, n = 3; PFS, n = 3; mixed packaging, n = 0), severe combined immunodeficiency (vial, n = 0; PFS, n = 4; mixed packaging, n = 2), chronic granulomatous disease (vial, n = 0; PFS, n = 0; mixed packaging, n = 1), specific antibody deficiency (vial, n = 1; PFS, n = 3; mixed packaging, n = 0), idiopathic autoimmune hemolytic Anemia (vial, n = 1; PFS, n = 0; mixed packaging, n = 1), autoimmune disease (vial, n = 1; PFS, n = 1; mixed packaging, n = 0), hypogammaglobulinemia (vial, n = 1; PFS, n = 1; mixed packaging, n = 0), chronic lymphocytic leukemia (vial, n = 1; PFS, n = 0; mixed packaging, n = 0), Waldenstrom macroglobulinemia (vial, n = 1; PFS, n = 0; mixed packaging, n = 0). CVID, common variable immune deficiency; DGS, DiGeorge syndrome; GHP, general health perception; GMH-2, global mental health 2; GPH-2, global physical health 2; IgG, immunoglobulin; IgG Sub, immunoglobulin subclass deficiency; IQR, interquartile range; kg, kilogram; PFS, pre-filled syringes; SCIg, subcutaneous immunoglobulin; SD, standard deviation; SID, secondary immunodeficiencies. |
Choosing a SCIg packaging method
To understand the factors taken into consideration by patients when choosing a SCIg packaging method, information on the most important reasons a method was chosen was collected for the three packaging cohorts (Fig. 3A). A physician’s recommendation was the most common factor for respondents in all three of the vial, PFS, and mixed packaging cohorts (66.1% [n = 41], 79.3% [84], and 93.6% [n = 29], respectively; Fig. 3A). However, respondents in the vial cohort were significantly less likely to state a physician’s recommendation as an important factor compared with respondents in the mixed packaging cohort (p = 0.004). Vial and PFS users were significantly more likely to state safety as an important factor compared with mixed packaging users (14.5% [n = 9] and 17.0% [n = 18] vs. 0.0% [n = 0], p = 0.03 and p = 0.01, respectively; Fig. 3A). Finally, respondents in the vial cohort were significantly more likely to state insurance as an important factor compared with respondents in the PFS cohort (25.8% [n = 16] vs. 12.3% [n = 13], p = 0.03, Fig. 3A).
No significant differences were found between the three packaging cohorts in the pump subgroup with respect to what they considered the most important factors (p > 0.08, Fig. 3B). Respondents using mixed packaging with manual push were significantly more likely to state a physician’s recommendation as an important factor compared with respondents using vials with manual push (100.0% [n = 10] vs. 61.5% [n = 16], respectively; Fig. 3C). No respondents in the manual push subgroup stated adverse events as an important factor when choosing SCIg packaging (Fig. 3C).
When asked how their QoL would change if they could receive their medication at a local pharmacy, the majority of respondents in the vial, PFS, and mixed packaging cohorts stated their perception of QoL would stay the same (36.4% [n = 4], 33.3% [n = 6], and 40.0% [n = 2], respectively). However, when asked how their QoL would change if they could have their medication delivered to their home, the vast majority of respondents in the vial, PFS, and mixed packaging cohorts stated that their QoL would improve or substantially improve (72.8% [n = 8], 83.3% [n = 15], and 100.0% [n = 5], respectively). Similar observations were seen in the three packaging cohorts in the pump subgroup, with no statistical difference between the three cohorts (p > 0.05 for all comparisons). Insufficient data was available for the manual push subgroup to perform similar analyzes.
Self-infusion training experience
To understand the factors associated with self-infusion training experience and satisfaction, information regarding training characteristics of the three packaging method cohorts were collected (summarized in Table 2 for all SCIg respondents, in Additional file 3 for the pump subgroup, and in Additional file 4 for the manual push subgroup). The statistical comparisons suggested no significant differences between the three packaging method cohorts for any of the measures examined overall (p > 0.05, Table 2). However, when analyzing training characteristics of the pump subgroup, the responders who use pump with vials required significantly fewer training sessions on average compared with the responders who use pump with PFS (p = 0.007, Additional file 3). When analyzing training characteristics of the manual push subgroup, the distribution of the location of training differed between the vial and PFS cohorts (p = 0.02, Additional file 4) - a lower proportion of respondents using manual push with vials were trained at a hospital compared with respondents using manual push with PFS (61.5% [n = 16] vs. 85.9% [n = 61], respectively; Additional file 4).
Table 2
Self-infusion training experiences of the vial, PFS, and mixed packaging cohorts
Training characteristics (n = 220) | Vial cohort (A) | PFS cohort (B) | Mixed packaging cohort (C) | p values |
Summary | n | Summary | n | Summary | n | A vs. B | A vs. C | B vs. C |
Number of training sessions, n (%) | 1 2 3 4 5 > 5 | 32 (51.6%) 17 (27.4%) 9 (14.5%) 1 (1.6%) 2 (3.2%) 1 (1.6%) | 62 | 53 (49.5%) 32 (29.9%) 11 (10.3%) 6 (5.6%) 0 (0.0%) 5 (4.7%) | 107 | 16 (50.0%) 8 (25.0%) 6 (18.8%) 1 (3.1%) 0 (0.0%) 1 (3.1%) | 32 | 0.79 | 0.80 | 0.95 |
Location of training, n (%) | Physician’s office Home Hospital Infusion center Other* | 2 (3.2%) 24 (38.7%) 32 (51.6%) 3 (4.8%) 1 (1.6%) | 62 | 3 (2.8%) 27 (25.2%) 71 (66.4%) 2 (1.9%) 4 (3.7%) | 107 | 0 (0.0%) 14 (43.8%) 18 (56.3%) 0 (0.0%) 0 (0.0%) | 32 | 0.21 | 0.51 | 0.28 |
Length of training session (hours), median [IQR] | 1.3 [1.0, 2.0] | 60 | 2.0 [1.0, 2.0] | 102 | 1.0 [1.0, 2.0] | 29 | 0.08 | 0.66 | 0.50 |
Ease of learning to infuse, n (%) | Very difficult Difficult Neither Easy Very easy | 2 (3.2%) 3 (4.8%) 11 (17.7%) 21 (33.9%) 25 (40.3%) | 62 | 3 (2.8%) 12 (11.2%) 14 (13.1%) 39 (36.5%) 39 (36.5%) | 107 | 0 (0.0%) 1 (3.1%) 4 (12.5%) 16 (50.0%) 11 (34.4%) | 32 | 0.28 | 0.85 | 0.51 |
Satisfaction with training, n (%) | Very dissatisfied Dissatisfied Neither Satisfied Very satisfied | 0 (0.0%) 1 (1.6%) 0 (0.0%) 12 (19.7%) 48 (78.7%) | 61 | 0 (0.0%) 1 (0.9%) 4 (3.7%) 20 (18.7%) 82 (76.6%) | 107 | 0 (0.0%) 1 (3.1%) 0 (0.0%) 3 (9.4%) 28 (87.5%) | 32 | 0.70 | 0.33 | 0.20 |
Concerns during training, n (%) | Drawing drug Inserting needle Using pump Prime tube Other† No concerns | 9 (15.0%) 24 (40.0%) 3 (5.0%) 3 (5.0%) 0 (0.0%) 21 (35.0%) | 62 | 13 (12.2%) 47 (43.9%) 0 (0.0%) 5 (4.7%) 1 (0.9%) 41 (38.3%) | 107 | 5 (15.6%) 14 (43.8%) 1 (3.1%) 3 (9.4%) 1 (3.1%) 8 (25.0%) | 32 | 0.29 | 0.68 | 0.20 |
Data were compared using Mann-Whitney test (ease of SCIg training, number of SCIg training sessions) or Fisher’s exact test (SCIg training location, type of SCIg trainer). Significant p values are in bold. *Other training locations: Institut de recherches cliniques de Montréal (vials, n = 1; PFS, n = 3; mixed packaging, n = 0) or a local community service center (vials, n = 0; PFS, n = 1; mixed packaging, n = 0); †Other concerns: applying needles to tubing (vials, n = 0; PFS, n = 1; mixed packaging, n = 0), tipto-tip transfer (vials, n = 0; PFS, n = 0; mixed packaging, n = 1). PFS, pre-filled syringes; SCIg, subcutaneous immunoglobulin. |
When analyzing who provided training, a hospital nurse was found to be the most common provider for each of the vial, PFS, and mixed packaging cohorts (56.5% [n = 35], 61.0% [n = 64], and 56.3% [n = 18], respectively), with no significant difference overall between the three cohorts (p > 0.05). In all three packaging cohorts, the trainer was graded as ‘knowledgeable’ or ‘very knowledgeable’ by over 95.0% of respondents. The majority of respondents (vials: 93.5% [n = 57], PFS: 80.4% [n = 86], and mixed packaging: 93.8% [n = 30]) felt they were ‘confident’ or ‘very confident’ in selfadministering SCIg following training. Similar observations were made in the manual push subgroup (vials: 96.0% [n = 24], PFS: 87.4% [n = 62], and mixed packaging: 100.0% [n = 10]). Overall, over 95.0% of respondents in all packaging cohorts were satisfied with their training, the majority of which were very satisfied (Table 2). When asked what their greatest training concern was, the most commonly reported answer was inserting the needle (vials: 40.0% [n = 24], PFS: 43.9% [n = 47], and mixed packaging: 43.8% [n = 14]; Table 2).
Self-infusion characteristics
SCIg dosing
PFS users were associated with a significantly lower median (interquartile range [IQR]) SCIg dose compared with vial users (10 [8, 12] vs. 12 [9, 16] g/week, respectively; p = 0.02; Table 3). Further, the PFS cohort was also associated with a lower median (IQR) SCIg dose adjusted by patient bodyweight compared with the vial cohort (0.13 [0.10, 0.17] vs. 0.15 [0.11, 0.19] g/week/kg, respectively; Table 3), albeit with borderline significance (p = 0.06, Table 3). The mean (± SD) bodyweight of respondents in the three packaging cohorts were similar (all p > 0.05, Table 1).
Table 3
SCIg dosing of the vial, PFS, and mixed packaging cohorts, stratified by administration method (pump and manual push).
Dosing | Vial cohort (A) | PFS cohort (B) | Mixed packaging cohort (C) | p values |
Summary | n | Summary | n | Summary | n | A vs. B | A vs. C | B vs. C |
All respondents (n = 220) |
SCIg dose (g/week) (median [IQR]) | 12 [9, 16] | 63 | 10 [8, 12] | 110 | 12 [8, 17] | 31 | 0.02 | 0.79 | 0.12 |
SCIg dose per bodyweight (g/week/kg) (median [IQR]) | 0.15 [0.11, 0.19] | 56 | 0.13 [0.10, 0.17] | 100 | 0.16 [0.11, 0.25] | 25 | 0.06 | 0.68 | 0.11 |
Pump subgroup (n = 99) |
SCIg dose (g/week) (median [IQR]) | 12 [10, 14] | 37 | 10 [8, 16] | 37 | 13 [10, 24] | 22 | 0.30 | 0.30 | 0.17 |
SCIg dose per bodyweight (g/week/kg) (median [IQR]) | 0.15 [0.12, 0.19] | 32 | 0.15 [0.11, 0.20] | 34 | 0.19 [0.13, 0.54] | 18 | 0.66 | 0.36 | 0.25 |
Manual push subgroup (n = 121) |
SCIg dose (g/week) (median [IQR]) | 10 [8, 16] | 26 | 10 [8, 12] | 73 | 8 [6, 12] | 9 | 0.18 | 0.15 | 0.47 |
SCIg dose per bodyweight (g/week/kg) (median [IQR]) | 0.14 [0.11, 0.30] | 24 | 0.12 [0.10, 0.16] | 66 | 0.12 [0.08, 0.19] | 7 | 0.14 | 0.54 | 0.95 |
All data were compared using the MannWhitney (non-parametric) test due to the non-normality (skewness) of the distributions. Significant p-values are highlighted in bold. g, grams; IQR, interquartile range; kg, kilogram; PFS, pre-filled syringes; SCIg, subcutaneous immunoglobulin. |
No significant differences were found between the self-infusion characteristics of the vial, PFS, and mixed packaging cohorts in the pump and manual push administration method subgroups (all p > 0.05, Table 3).
SCIg infusion efficiency
To assess the burden of treatment infusion, information on infusion duration and the number of infusion sites was collected for the three packaging cohorts (Table 4). The mean duration of infusion preparation time was significantly shorter in the PFS cohort compared with the vial cohort (16.7 minutes vs. 19.2 minutes, p = 0.02, Table 4), although the median time was the same in both cohorts (15 minutes, Table 4). On average, the PFS cohort was associated with a trend for shorter median (IQR) actual infusion times compared with the vial cohort (60 [35, 90] vs. 70 [48, 90] minutes, respectively; p = 0.07; Table 4) and had significantly shorter median actual infusion times compared with the mixed packaging cohort (60 [35, 90] vs. 83 [49, 120] minutes, respectively; p = 0.02; Table 4). In addition, the PFS cohort used a lower number of infusion sites on average than both the vial and mixed packaging cohorts, although this was only statistically different between the PFS and mixed packaging cohorts (p = 0.02, Table 4). Actual infusion times were numerically lower in all three packaging cohorts in the manual push subgroup compared with the pump subgroup (vials: 60 minutes, PFS: 50 minutes, and mixed packaging: 46 minutes vs. vials: 75 minutes, PFS: 90 minutes, and mixed packaging: 90 minutes; respectively; Table 4).
Table 4
SCIg infusion efficiency of the vial, PFS, and mixed packaging cohorts, stratified by administration method (pump and manual push).
Infusion characteristics | Vial cohort (A) | PFS cohort (B) | Mixed packaging cohort (C) | p values |
Summary | n | Summary | n | Summary | n | A vs. B | A vs. C | B vs. C |
All respondents (n = 220) |
Infusion preparation time (mins) (median [IQR]) | 15 [10, 30] | 64 | 15 [10, 20] | 110 | 15 [10, 30] | 32 | 0.02* | 0.91 | 0.05 |
Actual infusion time (mins) (median [IQR]) | 70 [48, 90] | 64 | 60 [35, 90] | 110 | 83 [49, 120] | 32 | 0.07 | 0.31 | 0.02 |
Post-infusion clean up time(mins) (median [IQR]) | 5 [5, 10] | 60 | 5 [5, 10] | 108 | 5 [5, 10] | 32 | 0.89 | 0.37 | 0.46 |
Number of infusion sites, n (%) | 1 2 or 3 ≥ 4 | 2 (3.3%) 41 (67.2%) 18 (29.5%) | 61 | 6 (5.7%) 83 (78.3%) 17 (16.0%) | 106 | 1 (3.1%) 18 (56.3%) 13 (40.6%) | 32 | 0.12 | 0.66 | 0.02 |
Pump subgroup (n = 99) |
Infusion preparation time (mins) (median [IQR]) | 20 [15, 30] | 37 | 15 [10, 20] | 37 | 15 [10, 30] | 22 | 0.20 | 0.49 | 0.79 |
Actual infusion time (mins) (median [IQR]) | 75 [60, 90] | 37 | 90 [60, 105] | 37 | 90 [60, 120] | 22 | 0.33 | 0.14 | 0.49 |
Post-infusion clean up time(mins) (median [IQR]) | 6 [5, 10] | 34 | 9 [5, 15] | 36 | 5 [5, 10] | 22 | 0.54 | 0.73 | 0.36 |
Number of infusion sites, n (%) | 1 2 or 3 ≥ 4 | 0 (0.0%) 21 (58.3%) 15 (41.7%) | 36 | 2 (5.6%) 25 (69.4%) 9 (25.0%) | 36 | 0 (0.0%) 12 (54.5%) 10 (45.5%) | 22 | 0.15 | 0.79 | 0.24 |
Manual push subgroup (n = 121) |
Infusion preparation time (mins) (median [IQR]) | 15 [10, 20] | 27 | 10 [8, 15] | 73 | 18 [15, 20] | 10 | 0.23 | 0.30 | 0.09 |
Actual infusion time (mins) (median [IQR]) | 60 [40, 90] | 27 | 50 [30, 75] | 73 | 46 [38, 120] | 10 | 0.06 | 0.59 | 0.50 |
Post-infusion clean up time(mins) (median [IQR]) | 5 [2, 5] | 26 | 5 [3, 10] | 72 | 6 [5, 10] | 10 | 0.46 | 0.16 | 0.33 |
Number of infusion sites, n (%) | 1 2 or 3 ≥ 4 | 2 (8.0%) 20 (80.0%) 3 (12.0%) | 25 | 4 (5.7%) 58 (82.9%) 8 (11.4%) | 70 | 1 (10.0%) 6 (60.0%) 3 (30.0%) | 10 | 0.92 | 0.46 | 0.14 |
All data were compared using the MannWhitney (non-parametric) test due to the non-normality (skewness) of the distributions. Significant p-values are highlighted in bold. *Although the median infusion preparation times are numerically similar for the vial and PFS cohorts, the difference between the two distributions is significant and is demonstrated by the mean (± standard deviation) values of 19.2 (± 12.4) mins and 16.7 (± 15.6) mins, respectively. IQR, interquartile range; mins, minutes; PFS, pre-filled syringes; SCIg, subcutaneous immunoglobulin. |
When asked how important respondents perceived the duration of infusions, nearly all respondents from all packaging cohorts in all infusion method subgroups reported it as ‘important’ or ‘very important’ in their response, except for one (11.1%) PFS respondent in the manual push subgroup, who reported their perception as neither ‘important’ or ‘unimportant’.
Patient-reported treatment satisfaction (TSQM9)
TSQM9 effectiveness
There were no significant differences between the vial cohort and the PFS or mixed packaging cohorts in terms of the Treatment Satisfaction for Medication Questionnaire (TSQM-9) effectiveness domain scores (p = 0.54 and p = 0.34, respectively; Fig. 4A). Similarly, there was no significant difference between the three packaging cohorts in the scores for any item on the TSQM9 effectiveness domain (all p > 0.05, Fig. 4B).
When comparing TSQM9 effectiveness domain scores between the three packaging cohorts in the pump subgroup, the vial cohort scored significantly higher compared with the mixed packaging cohort (p = 0.04, Fig. 4A), but not compared with the PFS cohort (p = 0.11, Fig. 4A). When examining item specific evidence on the TSQM9 effectiveness domain in the pump subgroup, the vial cohort was associated with a significantly greater satisfaction score for the way the medication relieves symptoms compared with the mixed packaging cohort (p = 0.03, Fig. 4B) but not compared with the PFS cohort (p = 0.12, Fig. 4B).
Finally, when comparing TSQM9 effectiveness domain scores between the three packaging cohorts in the manual push subgroup, there were no significant differences between the vial cohort and the PFS or mixed packaging cohorts (p = 0.19 and p = 0.22, respectively; Fig. 4A). Similarly, there was no significant difference between the three cohorts in the scores for any item specific evidence on the TSQM9 effectiveness domain in the manual push subgroup (all p > 0.05, Fig. 4B).
TSQM9 convenience
With respect to the TSQM9 convenience domain, the vial and PFS cohorts scored numerically higher compared with the mixed packaging cohort, but the difference was only significant between the vial and mixed packaging cohorts (p = 0.008, Fig. 5A) and not between the vial and PFS cohorts (p = 0.20, Fig. 5A). When examining item specific evidence on the TSQM9 convenience domain, the vial cohort was associated with significantly greater satisfaction scores for the plan of medication use and the ease of use of the medication compared with the mixed packaging cohort (p = 0.01 and p = 0.02, respectively; Fig. 5B) but not compared with the PFS cohort (p > 0.05, Fig. 5B).
When comparing TSQM‑9 convenience domain scores between the three packaging cohorts in the pump subgroup, the vial cohort was associated with a trend for a higher score compared with the PFS cohort (p = 0.07, Fig. 5A) and scored significantly higher compared with the mixed packaging cohorts (p = 0.04, Fig. 5A). When examining item specific evidence on the TSQM9 convenience domain, vial users were associated with a significantly greater satisfaction score for the ability to take medication as instructed compared with PFS users (p = 0.03, Fig. 5B) but not compared with mixed packaging users (p = 0.09, Fig. 5B).
Finally, in the manual push subgroup, both the vial and PFS cohorts scored numerically higher for the TSQM9 convenience domain compared with the mixed packaging cohorts, but the differences were not significant in either comparison (p = 0.72. and p = 0.11, respectively; Fig. 5A). Similarly, there was no significant difference between the three cohorts in the scores for any item specific evidence on the TSQM9 convenience domain in the manual push subgroup (all p > 0.05, Fig. 5B).
TSQM9 global satisfaction
The vial cohort scored statistically higher in the TSQM9 global satisfaction domain compared with both the PFS and mixed packaging cohorts (p = 0.02 and p = 0.01, respectively; Fig. 6A). When examining item specific evidence on the TSQM9 global satisfaction domain, the vial cohort was associated with a significantly greater confidence score when asked if taking medication is a good thing compared with both the PFS and mixed packaging cohorts (p = 0.03, and p = 0.01, respectively; Fig. 6B).
There were no significant differences in the score for the TSQM9 global satisfaction domain between the three packaging cohorts in the pump subgroup (all p > 0.05, Fig. 6A). However, when examining item specific evidence on the TSQM9 global satisfaction domain, the vial cohort was associated with a significantly greater confidence score when asked if taking medication is a good thing compared with the mixed packaging cohort (p = 0.02, Fig. 6B) but not compared with the PFS cohort (p = 0.23, Fig. 6B).
Finally, in the manual push subgroup, the vial cohort scored numerically higher for the TSQM9 global satisfaction domain compared with both the PFS and mixed packaging cohorts, but the difference was significant only between the vial and PFS cohorts (p = 0.02, Fig. 6A) and not between the vial and mixed packaging cohorts (p = 0.27, Fig. 6A). There was no significant difference between the three cohorts in the scores for any item specific evidence on the TSQM9 convenience domain in the manual push subgroup (all p > 0.05, Fig. 6B).
Patient-reported outcomes
There were no significant differences in the measures of PROs, including Patient Acceptable Symptom State (PASS), General Health Perception (GHP), and Patient Reported Outcomes Measurement Information System (PROMIS) Global Physical Health (GPH-2) and Global Mental Health (GMH-2) in PROs between the vial, PFS, and mixed packaging cohorts (all p > 0.05, Fig. 7A, Fig. 7B and Fig. 7C). However, when examining the two administration method subgroups (pump and manual push), respondents using vials with manual push scored significantly higher for PROMIS GPH-2 compared with respondents using PFS with manual push (51.1 vs. 46.5, respectively, p = 0.007; Fig. 7C). There were no significant differences in any other measures of PROs between the three cohorts in the two administration method cohorts (all p > 0.05, Fig. 7A, Fig. 7B and Fig. 7C).
Switching between SCIg packaging methods
Of the 220 respondents currently receiving SCIg, 57 had switched to their current SCIg packaging method from a previously used, different packaging method. Of these, 16 (28.1%) respondents switched to vial packaging, 35 (61.4%) switched to PFS packaging, and eight (14.0%) switched to mixed packaging. Easier administration was the most common reason given by respondents in the vial cohort (63.6% [n = 7]), whereas a physician’s recommendation was the most common reason given by respondents who switched to PFS (44.4% [n = 8]) and respondents who switched to mixed packaging (66.7% [n = 4]). The majority of respondents in all three cohorts reported improved or substantially improved QoL following their switch (vial: 75.0% [n = 9], PFS: 66.6% [n = 16], and mixed packaging: 87.5% [n = 7]). Similarly, most of the respondents in all three cohorts reported improved or substantially improved treatment satisfaction following their switch (75.0% [n = 9] of the vial cohort, 83.4% [n = 20] of the PFS cohort, and 75.0% [n = 6] of the mixed packaging cohort). The positive impact of switching packaging methods on mental health was significantly greater in the PFS cohort compared with the vial cohort (p = 0.03), with 31.8% (n = 7) of responders who switched to PFS reporting improved mental health following switch, compared with 0.0% (n = 0) of the vial cohort. Positive impacts on productivity, treatment compliance, and physical health were also observed for each cohort after switching packaging methods, but to a lesser extent than those reported above, with 58.3% (n = 7) of the vial cohort, 54.2% (n = 13) of the PFS cohort, and 50% (n = 4) of the mixed packaging cohort reporting an improved or substantially improved impact on productivity. In addition, 33.3% (n = 4) of the vial cohort, 45.8% (n = 11) of the PFS cohort, and 62.5% (n = 5) of the mixed packing cohort reported an improved or substantially improved impact on treatment compliance. Meanwhile, 41.6% (n = 5) of the vial cohort, 45.9% (n = 11) of the PFS cohort, and 37.5% (n = 3) of the mixed packaging cohort reported an improved or substantially improved impact on physical health. No differences were observed between current and previous training experiences in the three packaging cohorts in any characteristics measured.