Pharmacoeconomic Guidelines: South Korea
Country/Region: South Korea
PE Guidelines
Guidelines for the Economic Assessment of Medicines (January 2021)
PDF in Korean
PDF in Korean
PE Guidelines Source:
Health Insurance Review & Assessment Service (HIRA)
http://www.hira.or.kr/
http://www.hira.or.kr/
Additional Information:
Bae EY, Hong J, Bae S, et al. Korean Guidelines for Pharmacoeconomic Evaluations: Updates in the Third Version. Appl Health Econ Health Policy. 2022;20(4):467-477. doi:10.1007/s40258-022-00721-4
Information current as of Tuesday, May 21, 2024
Key Features
| Key Features | |
|---|---|
| Type of Guidelines | PE Guidelines |
| Title and year of the document | Guidelines for the Economic Assessment of Medicines (20-Jan 2021) |
| Affiliation of authors | Health Insurance Review & Assessment Service (HIRA) |
| Purpose of the document | To clarify previously released pharmacoeconomic (PE) guidelines (2006, 2011) regarding ambiguities in interpretation, and to account for evolving methodologies for estimating effectiveness and costs. The revisions will provide the following: 1) clear guidelines to serve as practical standards, 2) reflection of the progress and level of consensus in economic evaluation methodologies globally and what is feasible and accepted domestically, 3) minimization of uncertainty surrounding evaluation results, and 4) support for consistent decision-making by improving transparency and comparability across submissions. |
| Standard reporting format included | Structured reporting format with detail encouraged to facilitate review and to ensure reviewer may replicate the process and reproduce the results. Data, analysis code, and program used in the analysis must be submitted together. |
| Disclosure | Not required |
| Target audience of funding/ author's interests | Data submitters, those performing economic evaluation, and reviewers. |
| Perspective | Healthcare system perspective |
| Indication | Requires the submission of an economic evaluation in the process of listing new drugs in the National Health Insurance. |
| Target population | Study population should be specified, and generalizability to the target population should be discussed with supporting evidence. |
| Subgroup analysis | In the case of subgroup analysis, the characteristics of patients constituting the subgroup are presented in detail, and the rationale for the subgroup is presented and supported by relevant evidence (variability between groups in treatment effects, etc.). Report if the subgroup analysis was pre-planned. |
| Choice of comparator | Comparable drugs available in the market, and of those, the most widely used one amongst them should be used as the comparator. Multiple drugs, other interventions such as surgery, or comparator arms used in clinical trials (e.g., SoC) may be considered. Sufficient rationale for selecting the comparator should be included. Comparator arms from clinical trials may be used only under certain conditions: (1) if they are recommended as standard treatments in medical guidelines, 2) if there is no head-to-head evidence with the drug with the highest market share and the result of an indirect comparison is highly uncertain, or 3) if the drug(s) are used in current practice in South Korea. Regardless of which comparison target is selected, the reason for selection should be sufficiently mentioned in the report. |
| Time horizon | Long enough to capture all meaningful differences in costs and outcomes between alternatives. Sensitivity analyses using the data observed during the clinical trial need to be presented as well. If lifetime horizon is used, the model and input data used should be justified, and the external validity of modeled period needs to be assessed. If extrapolated benefits account for a large part of the overall benefits, the impact of this extrapolation needs to be considered and accounted for. |
| Assumptions required | Yes, required. Model specification and assumptions used for the extrapolation should be detailed on the basis of the model selection. To justify the extrapolation model and assumption, they must be supported by clinical reasons and other external data sources. |
| Preferred analytical technique | CUA is preferred. |
| Costs to be included | Cost items depend on the analysis perspective. Since the revised guidelines take a health care system perspective, only direct health care costs are included in the basic analysis, and the analysis includes transportation costs and time costs for patients and guardians informal care costs, etc., can be presented separately. Productivity costs due to morbidity and death are not included in the basic analysis, and results including these items are presented separately. Unrelated future medical costs not related to the treatment of the disease are also not included in the basic analysis. Costs incurred in clinical trials that do not occur in the actual treatment process or transfer costs are excluded from the analysis. |
| Source of costs | Data sources with a low likelihood of bias should be selected and quality of data should also be included. Acceptable data source for resource usage include clinical studies, claims data, clinical practice guidelines hospital data, and expert opinions. When using the data source, reliability and consistency should be ensured. Since unit costs and utilization patterns differ between countries, values recently calculated which are representative of the Korea population are preferred. Country-specific usage data and prices are recommended, akin to many economic evaluation guidelines. It is also advisable to adjust cost calculation based on time of analysis. There is a publication for costing methods standards [in Korean]: https://www.neca.re.kr/lay1/program/S1T11C145/report/view.do?seq=8 |
| Modeling | Modeling is accepted. Model parameters should be systematically identified, assessed, and described, and each step needs to be reported in a transparent and detailed manner which must follow good model practice. Internal and external validity of the model should follow good modeling practices, and the use of the AdViSHE tool is encouraged as part of validity assessment. AdViSHE tool (Verner et al., 2016): https://pubmed.ncbi.nlm.nih.gov/26660529/ |
| Systematic review of evidences | Yes, conducted using a detailed search strategy, pre-planned protocol and transparent reporting. Recommend using bias assessment and a systematic literature review guideline. |
| Preference for effectiveness over efficacy | An evidence-based head-to-head trial is preferred, but the estimate derived from the indirect comparison can also be used if it is not available. Simple naïve indirect comparisons are discouraged if not appropriately adjusted using a valid method. |
| Preferred outcome measure | QALYs. Final outcome (ultimate therapeutic effect expected from using the pharmaceutical) should be used. |
| Preferred method to derive utility | The hierarchy of the preferred method and details to be reported are as follows. Indirect method with patient-level data collected from clinical trials using a generic preference-based measures (not condition-specific measures) is the most preferred. Choice-based method is preferred. A representative sample of the Korean general public is recommended. Direct methods, mapping, and using published sources are acceptable when data for indirect methods are limited or inappropriate. In this case, justification for its use, together with complete details and sensitivity analyses, is required for informed decision-making. Use of multiple sources should be avoided, if possible, when using published sources. A complete list of items to be reported for each case is included. |
| Equity issues stated | Any assumptions made regarding equity during the analysis process and the group (demographics and socioeconomic) that will benefit from the decision to cover the applied drug are to be described. The recommended method is to treat everyone's health benefits equally, meaning QALY or 1 life year of life extension is evaluated equally for all, regardless of disease status, age, or socioeconomic status. This approach may change depending on the analysis method. |
| Discounting costs | 4.5% (sensitivity analyses with 0% and 3%) When the time horizon is longer than 30 years, if it is deemed necessary, a 3.5% discount rate can be applied to both costs and benefits accrued after 30 years in a sensitivity analysis. |
| Discounting outcomes | 4.5% (sensitivity analyses with 0% and 3%) When the time horizon is longer than 30 years, if it is deemed necessary, a 3.5% discount rate can be applied to both costs and benefits accrued after 30 years in a sensitivity analysis. |
| Sensitivity analysis-parameters and range | If the treatment significantly impacts sectors other than healthcare, the relevant costs or benefits shall be separately measured and presented, or they shall be incorporated into the results and presented as a sensitivity analysis. Additional sensitivity analysis has to be conducted considering the various distribution assumptions and scenarios. |
| Sensitivity analysis-methods | One-way sensitivity analyses are required for the full range of uncertainties, Multivariate analysis is recommended for critical parameter(s). Probabilistic sensitivity analysis (PSA) is encouraged but not required. PSA is recommended with proper distribution, considering the correlation of the parameters. |
| Presenting results | Final results should be presented as an ICER, with total cost and total effect (or utility) of each alternative. |
| Incremental analysis | ICER |
| Total costs vs effectiveness (cost/effectiveness ratio) | ICER |
| Portability of results (Generalizability) | Even if study results conducted outside Korea are utilized, unit costs and resource usage should be based on domestic data. Regarding clinical trial results, the feasibility of their application domestically should be considered and documented using factors such as domestic epidemiology, clinical practice patterns, and genetic background. If significant disparities exist in domestic prevalence or treatment patterns compared to other countries, and applying the results domestically would be problematic, the analysis should rely on domestic data and be presented alongside the clinical trial data analysis results. In cases where multi-national trial results are accessible, both domestic and all subjects' data analyses should be included, provided the domestic sample size is sufficient for drawing statistically meaningful conclusions. However, if this is not possible, sensitivity analyses should reflect the clinical trial results of domestic subjects. |
| Financial impact analysis | Required for submission, but removed from this guideline. Still considered in the HIRA’s decision-making and subsequent price negotiation process. |
| Mandatory or recommended or voluntary | Depends on the advisory level: 1) "Must or shall" means mandatory, 2) "recommended" means should follow but if do not follow the recommended method, must provide justification for the method being used, 3) "preferred" means this is preferred by policymakers, which is theoretically desirable but does not necessarily need to be followed, 4) "can do" means that you may follow the suggested method, and you must provide a legitimate basis why you selected this method |
Acknowledgement:
| Lydia Y Lee, PharmD, MS |