Factors influencing the difference between forecasted and actual drug sales volumes under the price–volume agreement in South Korea

• Despite the importance of forecasting the drug-use volume, relevant studies are scarce.
• Factors affecting the gap between forecasted and actual drug sales were identified.
• This study provided the evidence to accurately forecast the drug sales volume.

This study analyzed factors contributing to increases in the actual sales volumes relative to forecasted volumes of drugs under price–volume agreement (PVA) policy in South Korea.Sales volumes of newly listed drugs on the national formulary are monitored under PVA policy. When actual sales volume exceeds the pre-agreed forecasted volume by 30% or more, the drug is subject to price-reduction. Logistic regression assessed the factors related to whether drugs were the PVA price-reduction drugs. A generalized linear model with gamma distribution and log-link assessed the factors influencing the increase in actual volumes compared to forecasted volume in the PVA price-reduction drugs.Of 186 PVA monitored drugs, 34.9% were price-reduction drugs. Drugs marketed by pharmaceutical companies with previous-occupation in the therapeutic markets were more likely to be PVA price-reduction drugs than drugs marketed by firms with no previous-occupation. Drugs of multinational pharmaceutical companies were more likely to be PVA price-reduction drugs than those of domestic companies. Having more alternative existing drugs was significantly associated with higher odds of being PVA price-reduction drugs. Among the PVA price-reduction drugs, the increasing rate of actual volume compared to forecasted volume was significantly higher in drugs with clinical usefulness.By focusing the negotiation efforts on those target drugs, PVA policy can be administered more efficiently with the improved predictability of the drug sales volumes.