Table 3.
Overall direction and quality of evidence from the literature on the impact of ERP at national level.
Source: Synthesis and assessment by the authors based on primary and secondary data collection
| Study endpoints | Issues identified within endpoints on ERP impact | Impact of ERP positive (+) negative (−) or ambiguous (±)a | Quality of empirical evidence on the impact of ERP (where applicable)c | Duration evidence applies to: short-term (S) or long-term (L)d |
|---|---|---|---|---|
| Cost-containment | Generating healthcare savings | + | Very low | S |
| Magnitude of healthcare savings depends on ERP design | ± | Very low | ||
| Prices/price levels | Achieves lower pharmaceutical prices | ± | Very low | S/L |
| Pharmaceutical prices depend on ERP design | ± | Very low | ||
| Pharmaceutical prices depend on market features | ± | Very low | ||
| Drug use | ERP helps contain consumption | ± | Not Available | S |
| ERP improves drug diffusion and use | ± | Not Available | ||
| Availability and launch delays | Possibility of withdrawal from market | + | Very low | S/L |
| ERP causes launch delays, launch sequencing or no launch | + | Very low | ||
| Affordability | ERP leads to pharmaceutical prices in line with the purchasing ability of healthcare systems or patients | ± | Very low | S |
| ERP provides scope for increasing affordability | ± | Very low | ||
| Fairness/social welfare | ERP can lead to social welfare improvement | -b | Not Available | S |
| ERP may neglect country-specific health system priorities | − | Very low | ||
| Microeconomic efficiency | More affordable prices through price revision | +b | Not Available | S |
| Contributes to stable share of pharmaceutical expenditure on total health spend | ± | Not Available | ||
| Contains costs while guaranteeing access to medicines | ±b | Very low | ||
| Industrial policy and innovation | May discourage incremental innovation and investment in (incremental) R&D | −b | Very low | S |
| May influence manufacturing and/or R&D investment decisions | +b | Very low | ||
| May indirectly incentivise innovation | ±b | Very low | ||
| Overall | ± | Very low | S | |
aRegarding the direction of impact of ERP, the “+” sign indicates that ERP contributes to achieving the stated goal(s); the “−” sign indicates that it does not contribute to achieving the stated goals. The sign “±” is used in those cases where the impact of ERP on the relevant endpoint and issue is ambiguous. This is generally observed when the impact of ERP depends on other factors, such as the modalities of ERP methodology or other exogenous factors. In order to arrive at the direction of impact as shown, a simple-vote counting methodology was adopted by counting the number of identified studies providing positive evidence and the number of those providing negative evidence
bInconclusive evidence
cThe overall quality of the identified empirical evidence has been classified as “high”, “moderate”, “low”, “very low” and not available. During vote counting only studies examining each endpoint/issue empirically were considered for quality assessment. As discussed in the Methods section, some studies referencing evidence using a post-only design were classified as “very low” quality, whereas studies performing regression analysis were considered to be of “low” quality. Where quasi-experimental designs or difference-in-difference methodologies were used, the quality of evidence was classified as “high”. Under each endpoint/issue, when different types of empirical studies were considered, the quality of evidence was assessed based on the majority, for example, when empirical evidence under an endpoint was given by three studies using a post-only design and one study using a regression analysis design, then the quality of empirical evidence under this endpoint was considered as “very low”
dThe last column describes the duration of the relevant evidence and indeed whether the evidence provided under each endpoint/issue considered the short or long-term impact of ERP, denoted by “S” or “L”; “S/L” denotes circumstances where both short- and long-term impact are considered