Table 3 Subgroup analysis, by cohort
Subgroup set and sample size (n) | Summary of the results |
|---|---|
US adolescents, N = 197 | |
Agea | Preferences were not statistically different between younger and older adolescents in the study |
Younger than median age (n = 46) | |
At or older than median age (n = 151) | |
P value = 0.738 | |
Gender | Gender was a significant driver of preferences. Improvement in symptom control was by far the most important attribute for respondents who did not identify as female, while respondents who identified as female generally placed similar importance on (1) symptom control; (2) time until the next flare-up; and (3) avoiding stomachaches, nausea, and vomiting. In addition, respondents who identified as female significantly disliked getting the treatment by IV infusion every month and preferred other modes and frequencies of administration; they also significantly preferred having no additional medicines |
Does not identify as female (n = 119) | |
Identifies as female (n = 78) | |
P value = 0.004 | |
Experience with methotrexate | Experience with methotrexate was also a significant driver of preferences |
Has no experience with methotrexate (n = 124) | Improvement in symptom control was by far the most important attribute for respondents with methotrexate experience. Respondents without methotrexate experience placed similar importance on improvement in symptom control and avoiding stomachaches, nausea, and vomiting. Preferences for mode and for combination therapy were different across these 2 groups |
Has experience with methotrexate (n = 73) | |
P value = 0.003 | |
Experience with biologics | Preferences were not statistically different between adolescents with and without experience with biologics, although respondents with biologics experience placed more importance on improvement in symptom control, relatively to other attributes |
Has no experience with biologics (n = 59) | |
Has experience with biologics (n = 138) | |
P value = 0.307 | |
Experience with injections | Preferences were not statistically different between adolescents with and without experience with injections, mainly due to the small sample size (and resulting wide confidence intervals). However, respondents with injection experience considered improvement in symptom control the most important attribute, which was not the case for respondents without injection experience |
Has no experience with injections (n = 43) | |
Has experience with injections (n = 154) | |
P value = 0.239 | |
Experience with headaches | Headache experience was a significant driver of preference heterogeneity. Improvement in symptom control was the most important attribute for respondents with headache experience, while to respondents without headache experience, efficacy and adverse event attributes generally had similar importance |
Has no experience with headaches (n = 34) | |
Has experience with headaches (n = 163) | |
P value = 0.032 | |
Experience with stomachaches | Experience with stomachache was a significant driver of preference heterogeneity as well. Improvement in symptom control was the most important attribute for respondents with stomachache experience, while respondents without stomachache experience generally considered (1) avoiding stomachaches, nausea, and vomiting and (2) time until next flare-up as most important. Respondents with no stomachache experience strongly disliked IV infusion compared to other modes of administration and preferred no additional medicine |
Has no experience with stomachaches (n = 22) | |
Has experience with stomachaches (n = 175) | |
P value = 0.024 | |
Experience with vomiting | Preferences were not statistically different between adolescents with and without experience vomiting, although respondents with vomiting experience considered improvement in symptom control the most important attribute, while those without vomiting experience placed similar importance on efficacy and on avoiding stomachaches, nausea, and vomiting |
Has no experience with vomiting (n = 51) | |
Has experience with vomiting (n = 146) | |
P value = 0.513 | |
US caregivers, N = 207 | |
Child’s median ageb | Preferences were not systematically statistically different between caregivers with younger and older children, although respondents with children at median age or older preferred treatments that delayed the time until next flare-up |
Child is younger than median age (n = 100) | |
Child is at or older than median age (n = 107) | |
P value = 0.289 | |
Child’s gender | Child’s gender was not a driver of preference heterogeneity overall; however, caregivers with a child who did not identify as female preferred no additional medicines, while caregivers with a child who identified as female preferred treatments that required additional medicines |
Child does not identify as female (n = 147) | |
Child identifies as female (n = 60) | |
P value = 0.870 | |
Caregiver’s level of education | Although preferences between these 2 groups were not systematically different at the 95% level of confidence, improvement in symptom control was the most important attribute for caregivers with a 4-year degree or higher, while caregivers with less than a 4-year degree placed the most importance on increasing the time until the next flare-up. Confidence intervals were very large for this group |
Caregiver has less than a 4-year degree (n = 47) | |
Caregiver has a 4-year degree or higher (n = 160) | |
P value = 0.195 | |
Caregiver’s child has experience with methotrexate | Preferences were not systematically statistically different between caregivers with a child who has experience with methotrexate and those with a child with no experience with methotrexate |
Child has no experience with methotrexate (n = 120) | |
Child has experience with methotrexate (n = 87) | |
P value = 0.801 | |
Caregiver’s child has experience with biologics | Although the child’s experience with biologics was not a driver of preference heterogeneity among caregivers, caregivers with a child who has experience with biologics considered improvement in symptom control the most important attribute. Additionally, these respondents did not have systematically different preferences across treatments that increase the time until next flare-up by 3, 5, or 9 months but did prefer these levels to a treatment that only increases the time until next flare-up by 1 month |
Child has no experience with biologics (n = 76) | |
Child has experience with biologics (n = 131) | |
P value = 0.248 | |
Caregiver’s child has experience with injections | Although preferences were not systematically different between these 2 groups at the 95% level of confidence, caregivers with a child who has no experience with injections considered improvement in symptom control the most important attribute (although confidence intervals were very large), while caregivers with a child who has experience with injections considered improvement in symptom control and time until the next flare-up the most important attributes in the study |
Child has no experience with injections (n = 59) | |
Child has experience with injections (n = 148) | |
P value = 0.115 | |
Caregiver’s child has experience with headaches | Preferences were similar across this subgroup set, although caregivers with a child who has not experienced headaches considered increasing the time until the next flare-up and improvement in symptom control the most important attributes and strongly disliked injections every 2 weeks, while caregivers with a child who has experience with headaches placed the most importance on improvements in symptom control and strongly disliked IV infusion every month |
Child has no experience with headaches (n = 59) | |
Child has experience with headaches (n = 148) | |
P value = 0.095 | |
Caregiver’s child has experience with stomachaches | Preferences across these 2 groups were not systematically different; however, time until the next flare-up and improvement in symptom control were the most important attributes for caregivers with a child who has not experienced stomachaches, while improvement in symptom control was the most important attribute for respondents with a child who has experienced stomachaches. In addition, preferences for mode and frequency of administration were different across the 2 groups |
Child has no experience with stomachaches (n = 47) | |
Child has experience with stomachaches (n = 160) | |
P value = 0.262 | |
Caregiver’s child has experience with vomiting | The child’s experience with vomiting significantly drove preference heterogeneity. Caregivers with a child with no experience with vomiting placed most importance on improvement in symptom control and increasing the time until the next flare-up. These respondents preferred a treatment by oral tablet or liquid to a treatment by injection every week, but were indifferent between tablets, injection every 2 weeks, or IV infusion every month. Improvement in symptom control was the most important attribute for caregivers with a child who has experienced vomiting as a side effect. These respondents strongly disliked a treatment by IV infusion every month |
Child has no experience with vomiting (n = 97) | |
Child has experience with vomiting (n = 110) | |
P value = 0.037 | |
UK adolescents, N = 100 | |
Adolescent at median age or oldera | Those younger than median age placed significantly more importance on improvement in symptom control, whereas those at median age or older placed more importance on avoiding headaches than on the other attributes included in the study |
Younger than median age (n = 26) | |
At or older than median age (n = 74) | |
P value = 0.005 | |
Additionally, the preference weight estimates indicate that adolescents younger than median age preferred an injection every week to tablets or syrup twice a week | |
Adolescent identifies as female | Adolescents who did not identify as female placed more importance on avoiding headaches, whereas adolescents who identified as female placed more importance on improvement in symptom control than on the other attributes included in the study. Additionally, adolescents who did not identify as female preferred tablets, syrup, or injections to IV |
Does not identify as female (n = 80) | |
Identifies as female (n = 20) | |
P value = 0.080 | |
Adolescent has experience with methotrexate | Preferences were not statistically systematically different across the 2 groups, probably due to small sample size; however, respondents with methotrexate experience considered avoiding headaches the most important attribute, while respondents without methotrexate experience considered improving symptom control the most important attribute |
Has no experience with methotrexate (n = 58) | |
Has experience with methotrexate (n = 42) | |
P value = 0.206 | |
Adolescent has experience with biologics | Although the sample size did not allow for identification of systematic differences in preferences, respondents without biologics experience valued efficacy and adverse event attributes about the same, while respondents who have experience with biologics placed the most value on avoiding headaches and improving symptom control. Respondents with biologics experience preferred tablets or syrup twice a day or an injection once a week to IV |
Has no experience with biologics (n = 17) | |
Has experience with biologics (n = 83) | |
P value = 0.109 | |
Adolescent has experience with injections | Adolescents who do not have experience with injections did not have strong preferences for any specific attribute included in the survey. In fact, there were no statistically significant differences between symptom control; time until next flare-up; stomachaches, nausea, and vomiting; and mode and frequency of administration. Additionally, adolescents who have experience with injections placed more relative importance on improvement in symptom control and avoiding headaches and less relative importance on mode and frequency of administration. Interestingly, adolescents with injection experience preferred injection every week to tablets or syrup twice a day |
Has no experience with injections (n = 25) | |
Has experience with injections (n = 75) | |
P value = 0.002 | |
Adolescent has experience with headaches | Adolescents who have no experience with headaches placed more relative importance on improvement in symptom control and avoiding headaches, whereas adolescents who have experienced headaches placed more relative importance on improvement in symptom control and mode and frequency of administration than on avoiding headaches |
Has no experience with headaches (n = 47) | |
Has experience with headaches (n = 53) | |
P value = 0.007 | |
Adolescent has experience with stomachaches | Adolescents who have no experience with stomachaches as a side effect of treatment placed more relative importance on improvement in symptom control and avoiding headaches, and adolescents who have experience with stomachaches as a side effect of their treatment did not have strong preferences for any specific attribute included in the survey |
Has no experience with stomachaches (n = 41) | |
Has experience with stomachaches (n = 59) | |
P value = 0.007 | |
Adolescent has experience with vomiting | Adolescents who have no experience with vomiting placed more relative importance on improvements in symptom control and avoiding headaches than the other attributes included in the study. This group of adolescents did not have statistically different preferences across varying modes and frequencies of administration |
Has no experience with vomiting (n = 62) | |
Has experience with vomiting (n = 38) | |
P value = 0.003 | Although it may appear that respondents with vomiting experience placed more relative value on mode and frequency of administration and symptom control than other attributes in the study, there were no statistical differences across any of the CRI estimates for any attribute. However, those with vomiting experience preferred treatment that is taken as tablets or syrup twice a day or injection every week over an IV infusion every month |
UK caregivers, N = 200 | |
Caregiver’s child’s median ageb | Those respondents with a child below the median age generally valued improvement in symptom control, avoiding stomachaches, and combination therapy |
Child is younger than median age (n = 100) | |
Child is at or older than median age (n = 100) | Those respondents with a child at or above the median age valued the improvement in symptom control the most relative to the other attributes in the study. They did not have statistically different preferences between no additional medicines and combination therapy, but this group preferred injections, tablets, or syrup to IV infusions |
P value = 0.001 | |
Caregiver’s child’s gender | Respondents with a child who did not identify as female generally placed the most importance on improvements in symptom control relative to the other attributes in the study. These respondents preferred a treatment that is taken by tablet, syrup, or injection over a treatment by IV infusion |
Child does not identify as female (n = 131) | |
Child identifies as female (n = 69) | |
P value = 0.006 | Respondents with a child who identified as female generally placed more relative importance on improvements in symptom control, avoiding stomachaches, and having an additional medicine. These respondents preferred a combination therapy regimen over no additional medicines |
Caregiver’s level of education | Caregivers with less than an undergraduate degree, on average, valued all of the attributes included in the study. These respondents preferred a regimen with additional medicines over a regimen that does not involve combination therapy, and they preferred tablets or syrup twice a day over an IV infusion every month and an injection every week (but no more than an injection every 2 weeks) |
Caregiver has less than an undergraduate degree (n = 51) | |
Caregiver has an undergraduate or higher degree (n = 149) | |
Those caregivers with an undergraduate degree or higher placed the most relative importance on improvements in symptom control. These respondents also preferred an additional medicine to no combination therapy, but they had no statistically greater preferences for tablets or syrup twice a day to another mode and frequency of administration | |
P value = 0.011 | |
Caregiver’s child has experience with methotrexate | Caregivers with a child who had no experience with methotrexate placed most relative importance on improvements in symptom control; avoiding stomachaches, nausea, and vomiting; and having a combination therapy |
Child has no experience with methotrexate (n = 108) | |
Caregivers with a child who does have experience with methotrexate generally placed the most relative importance on improvements in symptom control and avoiding headaches. Unlike those without methotrexate experience, they did not differentiate between no additional medicines or combination therapy | |
Child has experience with methotrexate (n = 92) | |
P value = < 0.001 | |
Caregiver’s child has experience with biologics | Caregivers with a child with no experience with biologics generally cared about improvements in symptom control and mode and frequency of administration. These respondents strongly preferred a treatment by tablet, syrup, or injection over a treatment by IV infusion, but they did not have statistically different preferences between a treatment with or without combination therapy |
Child has no experience with biologics (n = 48) | |
Child has experience with biologics (n = 152) | |
P value = < 0.001 | |
Respondents with a child with biologics experience generally valued improvements in symptom control the most. These respondents did not prefer one mode of administration over another, but they did prefer a treatment with combination therapy | |
Caregiver’s child has experience with injections | Those respondents with a child with no experience with injections strongly valued improvements in symptom control over the other attributes included in the study. These respondents preferred tablets or syrup twice a day or an injection every 2 weeks over an IV infusion every month, but they did not prefer an injection every week over an IV infusion |
Child has no experience with injections (n = 29) | |
Child has experience with injections (n = 171) | |
P value = 0.001 | |
Respondents with a child who does have experience with injections did not have preferences for any one attribute that dominated over preferences for other attributes. These respondents had a preference for an injection every week over an IV infusion every month and also preferred a treatment with combination therapy | |
Caregiver’s child has experience with headaches | Caregivers with a child without experience with headaches generally place more relative importance on improvements in symptom control and avoiding headaches than other attributes in the study |
Child has no experience with headaches (n = 44) | |
Child has experience with headaches (n = 156) | Caregivers with a child who does have experience with headaches also placed a great deal of relative importance on improvements in symptom control and avoiding headaches. Also, they preferred a treatment regimen that included additional medicines over one without additional medicines |
P value = < 0.001 | |
Caregiver’s child has experience with stomachaches | Respondents with a child who has not experienced stomachaches as a side effect placed greater relative importance on improvements in symptom control, time until next flare-up, and avoiding headaches. These respondents preferred a treatment by syrup or tablet over a treatment by infusion and preferred a treatment with no additional medicines to one that needs additional medicines |
Child has no experience with stomachaches (n = 45) | |
Child has experience with stomachaches (n = 155) | |
Respondents with a child who has experienced stomachaches as a side effect similarly valued improvements in symptom control, but they also valued avoiding stomachaches, nausea, and vomiting. These respondents also preferred a treatment by syrup, tablet, or injection over a treatment by infusion, but they preferred a treatment with additional medicines to one without additional medicines | |
P value = < 0.001 | |
Caregiver’s child has experience with vomiting | Respondents with a child who has not experienced vomiting as a side effect strongly preferred improvements in symptom control relative to the other variables included in the study |
Child has no experience with vomiting (n = 69) | |
Child has experience with vomiting (n = 131) | |
Respondents with a child who had experience with vomiting did not, on average, have any preferences that dominated other attributes in terms of conditional relative importance. Unlike those with no vomiting experience, these respondents preferred a treatment regimen with additional medicines over one that does not include additional medicines | |
P value = < 0.001 | |