Difelikefalin for treating pruritus in people having haemodialysis

Company: Merged severe and very severe populations in mapping study, resulting in identical utility scores and costs for both groups due to small numbers in each group

ERG: Not explicitly stated

Committee: Questioned appropriateness but ultimately accepted that model should include all 5 health states of CKD-aP severity

ICER impact: uncertain_direction

Company: Used 5-D itch scores for modelling treatment efficacy in base case because WI-NRS was only used for first 12 weeks, whereas 5-D itch scale was used for up to 64 weeks

ERG: Felt that model structure adequately reflected clinical issues

Committee: Accepted model structure but noted uncertainty in using 5-D itch scores given WI-NRS was primary outcome in trials

ICER impact: uncertain_direction

Company: Used simulated approach based on mean change from baseline in itch scores by CKD-aP severity. Simulated approach only allowed for improvement, not deterioration

ERG: Preferred to estimate transition probabilities from directly observed patient-level data

Committee: Agreed that observed data better reflected clinical practice. Preferred EAG's approach of estimating transitions using observed patient-level data as this better reflected transitions in practice

ICER impact: uncertain_direction

Company: Pooled all randomised participants from KALM-1 and KALM-2, resulting in odds ratio of 1.93 (95% CI 1.44, 2.57) for at least 3-point WI-NRS reduction at week 12

ERG: Conducted meta-analysis of individual studies rather than simple data pooling, resulting in odds ratio of 2.07 (95% CI 1.24, 3.45). Suggested that company's method of adding individual participant data together might lead to over-precise results and bias

Committee: Noted uncertainty in pooled data; concluded evidence suggested difelikefalin reduces itch severity but with uncertainty

ICER impact: uncertain_direction

Company: Used 20 multiple imputations with specific covariates (baseline WI-NRS, anti-itch medication use, medical conditions, region for KALM-2 only) in logistic regression

ERG: Unclear how transition matrices identified; uncertain whether transition probabilities were based on averages or individual datasets; company did not directly test between-arm dataset variability; should include more variables as standard practice in imputation

Committee: Multiple imputation approach more appropriate than direct observations given 279 missing observations, but insufficient information provided about imputation methodology

ICER impact: uncertain_direction

Company: KALM trials recruited 20 UK participants across 5 UK treatment centres and data represents UK population well; effect of difelikefalin not statistically significantly different in Black participants vs other backgrounds

ERG: KALM trials had larger proportion of Black participants (29.2%) than UK target population (12.8%); family background may be effect modifier leading to overestimation of efficacy in UK population

Committee: Trials broadly generalisable to UK population; large health inequalities mean high proportion of Black and Asian patients in UK clinical practice; trial population representative of UK clinical population

ICER impact: uncertain_direction

Company: Used 5-D itch scale in base-case analysis because it was collected for up to 64 weeks in open-label phases, whereas WI-NRS only available for first 12 weeks

ERG: WI-NRS was the primary outcome in KALM-1 and KALM-2; 5-D itch scale choice for economic model less justified

Committee: EAG accepted model structure but noted uncertainty in using 5-D itch scores for treatment efficacy modelling

ICER impact: uncertain_direction

Company: Original base case: 5% waning per year in comparator arm, no waning in difelikefalin arm. No evidence provided initially

ERG: Waning assumptions uncertain without supporting evidence

Committee: Company explored evidence from economic literature review of 3 NICE HST appraisals with itching outcomes; presented scenario analyses showing higher waning rates than 5% per year

ICER impact: increases

Company: Original base case: 5% waning per year for comparator arm, no waning for difelikefalin. Revised base case: 10% waning per year for comparator arm, no waning for difelikefalin

ERG: Accepted 10% waning per year in comparator arm and updated base case accordingly

Committee: Accepted 10% waning in comparator arm as conservative. Recognised as plausible that difelikefalin would also have waning effect. Welcomed scenario analysis with 5% waning in treatment arm and 10% in control arm

ICER impact: uncertain_direction

Company: Carried out separate primary data collection study across UK dialysis centres to map WI-NRS and 5-D itch scale to EQ-5D-3L

ERG: Not explicitly stated but implied acceptance

Committee: Accepted that mapping study was robust

ICER impact: uncertain_direction