Table 4 Statistical analysis and categorisation of outcome for early benefit assessment in Germany (G-BA/IQWIG) and single technology appraisal in England (NICE) – a comparison
From: Requirements for benefit assessment in Germany and England – overview and comparison
Methodological element | Benefit assessment in Germany (G-BA/IQWIG) | Single Technology Appraisal in England (NICE) |
|---|---|---|
Systematic reviews & meta-analyses | Assessment of review quality; assessment of heterogeneity. | Assessment of data quality; assessment of heterogeneity. |
NICE may commission an additional independent review. | ||
Sensitivity analyses | Done with regard to methodological aspects of information retrieval and evaluation. | Done with regard to methodological aspects of information retrieval and evaluation, and uncertainty associated with parameter precision (probabilistic sensitivity analysis). |
Quality of evidence base | ||
Level of evidence | Clearly defined levels of evidence. | Defined levels of evidence. |
Quality of studies (internal validity) | Classification of potential bias. | Quantification of potential bias. |
G-BA may request further evidence as part of the decision. | NICE guidance may be reviewed/up-dated 1–5 years after initial appraisal. | |
Validity of endpoints | Classification of potential bias | Quantification of potential bias |
(blinding, ITT, reporting). | (blinding, ITT, reporting). | |
Extrapolation of results (modeling) | Not done. | Qualitative extrapolation and quantitative modeling of data regarding study duration, patient population, choice of comparator, and type of outcomes. |
Generalisability of study results (external validity) | Descriptive evaluation. | Qualitative description or quantitative extrapolation (modeling). |
Categorisation of outcome | Proof/indication/hint/no proof of (lack of) (additional) benefit (or harm). | Recommended/optimised/only in research/not recommended. |