Sourcing quality-of-life weights obtained from previous studies: theory and reality in Korea

Korean Guidelines for Pharmacoeconomic Evaluations: Updates in the Third Version

The first version of the pharmacoeconomic (PE) guidelines was published in South Korea in 2006. Despite its first revision in 2011, there were still ambiguities in its interpretation. Moreover, methodologies for estimating effectiveness and costs have also evolved since then. Under these circumstances, the Health Insurance Review and Assessment Service published the third version in January 2021. This article reviews the revision process and major changes made in the new edition of the PE guidelines. The revision was processed through reviews of the previous 50 PE submissions, international guidelines, academic literature, and surveys and advisory meetings to obtain stakeholders' opinions. The analysis perspective has changed from a limited societal perspective to a healthcare system perspective. In addition to the drug with the highest market share, drugs used in clinical trials can be selected as comparators under certain conditions. The discount rate decreased from 5% to 4.5%. Furthermore, the revised guidelines provide more detailed and specific instructions for items including non-inferiority margin, extrapolation, utility elicitation, and uncertainty. Treatment switch and co-dependent technology guidelines are newly included; the budget impact analysis guideline is deleted. Through this revision, transparency and consistency of decision-making is expected to improve.

A Review of Utility Measurement Methods Used in Pharmacoeconomic Submissions to HIRA in South Korea: Methodological Consistency and Areas for Improvement

Pharmacoeconomic (PE) guidelines, first published in 2006 and later updated in 2011, were developed to guide the preparation and submission of PE data to the Health Insurance Review and Assessment Service (HIRA) for drug reimbursement decision making in South Korea. This study, which was conducted as part of a project for revision of the PE guideline, reviewed utility values used in the PE submissions processed at HIRA during 2014-2018 to identify aspects of the current guidelines that may need to be revisited. A total of 50 PE submissions were processed at HIRA over the 5 years. Of these, 47 submissions that used quality-adjusted life-years as an outcome measure were included in this review. Data were extracted from full copies of the manufacturer's initial submissions and committee documents provided by HIRA. Of the 47 submissions, nearly half (n = 23, 48.9%) used published sources to obtain health state utility values, followed by direct methods using time trade-off (n = 7) or standard gamble (n = 2) and indirect methods with patient-level data using the EQ-5D-3L (n = 4) or the EQ-5D-5L (n = 2). Mapping, using the EQ-5D-3L as a target measure, was also adopted in six submissions, although it was somewhat unfavourably described in the guideline. Notably, 52.2% of the submissions with published sources took utility values from different sources for different health states defined in a single model. In addition, details of utility measurement methods or mapping functions taken from published sources were relatively poorly reported. Moreover, the preferences of the Korean general public, preferred by the guideline, were rarely reflected in the utility values used in submissions relying on published sources (95.7% for foreign values only/mixed) and mapping (66.7%). While most submissions with direct and indirect methods used domestic preference values, the former was occasionally criticised by assessment committees because of health state descriptions. This review highlights a considerable amount of inconsistency in the measurement of utility values used in the PE submissions during 2014-2018, indicating a strong need for methodological standardisation.

A Meta-Regression Analysis of Utility Weights for Breast Cancer: The Power of Patients' Experience

To summarize utility estimates of breast cancer and to assess the relative impacts of study characteristics on predicting breast cancer utilities. We searched Medline, Embase, RISS, and KoreaMed from January 1996 to April 2019 to find literature reporting utilities for breast cancer. Thirty-five articles were identified, reporting 224 utilities. A hierarchical linear model was used to conduct a meta-regression that included disease stages, assessment methods, respondent type, age of the respondents, and scale bounds as explanatory variables. The utility for early and late-stage breast cancer, as estimated by using the time-tradeoff with the scales anchored by death to perfect health with non-patients, were 0.742 and 0.525, respectively. The severity of breast cancer, assessment method, and respondent type were significant predictors of utilities, but the age of the respondents and bounds of the scale were not. Patients who experienced the health states valued 0.142 higher than did non-patients (P <0.001). Besides the disease stage, the respondent type had the highest impact on breast cancer utility.