Interventions to reduce non-prescription antimicrobial sales in community pharmacies

Antimicrobial resistance (AMR) is a major global health concern. One of the most important causes of AMR is the excessive and inappropriate use of antimicrobial drugs in healthcare and community settings. Most countries have policies that require antimicrobial drugs to be obtained from a pharmacy by prescription. The term 'non-prescription antimicrobial sale' refers to the dispensing and selling of antimicrobial drugs without a prescription in countries where the pharmaceutical policy does not permit the sale of antimicrobial drugs without a prescription. Pharmacies, drugstores, and other medicine outlets are major sources of non-prescription antimicrobial sales in the community setting.

To assess the effects of interventions for reducing non-prescription antimicrobial sales by pharmacists and non-pharmacists in community pharmacies, drugstores, and other medicine outlets. To assess whether the effects of interventions differ according to types of interventions (single or multicomponent), community pharmacy personnel (pharmacists or non-pharmacists), and countries (low to lower-middle-income and upper-middle to high income).

We searched five databases, including CENTRAL, MEDLINE, and Embase, and two trial registers to 26 September 2022. We also conducted reference checking and citation searches.

We included randomized trials, cluster-randomized trials, and quasi-randomized trials evaluating interventions targeted at pharmacists and non-pharmacists in community pharmacies, drugstores, and other medicine outlets. Our primary outcomes were non-prescription antimicrobial sales, symptomatic or asymptomatic infections caused by antimicrobial-resistant pathogens among pharmacy clients or community residents, and adverse events associated with non-prescription antimicrobial drug use in pharmacy clients. Our secondary outcomes were history taking and provision of advice to pharmacy clients, and knowledge of pharmacists and non-pharmacists.

We followed standard Cochrane methods.

We included four trials conducted in pharmacies and drugstores. Three studies were published between 2000 and 2010, and the fourth in 2016. In total, 942 community pharmacies and drugstores participated, including both pharmacists and non-pharmacists. One study conducted in Scotland was a four-arm trial that included educational outreach visits, continuing professional education, and a combination of both as interventions, in comparison to a control group supplied with guideline materials only. Two studies conducted in Portugal and Uganda compared the combination of training and distribution of written materials with a control of no intervention. One study conducted in Thailand and Vietnam compared a sequence of three interventions (regulatory enforcement, education, and peer-monitoring) with a control of no intervention. Only non-prescription antimicrobial sales, history taking and provision of advice to pharmacy clients, and knowledge of pharmacists and non-pharmacists were assessed in the included studies; no study assessed other outcomes. We judged the overall risk of bias for non-prescription antimicrobial sales to be high risk and for the other two outcomes to be some concerns. One study conducted in Scotland assessed single-component interventions (in two intervention arms) versus a control of written guidelines, and reported one primary outcome and one secondary outcome. Non-prescription antimicrobial sales: the study reported the sale or non-sale of antifungal drugs according to simulated patient scenarios. It did not report any differences between the intervention and control groups for the scenarios in which antifungal drugs should not have been sold without a prescription. The certainty of evidence for this outcome was very low. Knowledge of pharmacists and non-pharmacists: the study reported knowledge scores at baseline and follow-up, but did not compare the scores between two single-component intervention arms and the control arm at both time points. Four studies assessed multicomponent interventions versus a control of written guidelines or no intervention, and reported one primary outcome and two secondary outcomes. Non-prescription antimicrobial sales:two studies conducted in Uganda, Thailand, and Vietnam involving 337 randomized community pharmacies and drugstores reported the proportion of non-prescription antimicrobial sales to the total number of requests using a simulated client method. Based on vote counting, both studies favored the intervention, that is multicomponent interventions reduced the sales of non-prescription antimicrobial drugs by pharmacists and non-pharmacists. The information in the other two studies was inadequate to address this primary outcome. One study in Portugal reported that total antibiotic consumption at the municipal level (comprising both prescribed and non-prescribed antibiotic sales) was reduced after a multicomponent intervention. The study conducted in Scotland assessed a multicomponent intervention (in one intervention arm) and reported the sale or non-sale of antifungal drugs according to simulated patient scenarios. It did not report any differences between the intervention and control groups for the scenarios in which antifungal drugs should not have been sold without a prescription. The certainty of evidence for this outcome was very low. History taking and provision of advice to pharmacy clients:two studies conducted in Uganda, Thailand, and Vietnam reported this outcome. In Thailand and Vietnam, there was an improvement in the practices of pharmacists and non-pharmacists in the intervention groups, while Uganda reported a paradoxical decline. Knowledge of pharmacists and non-pharmacists:the study conducted in Scotland reported knowledge scores at baseline and follow-up, but did not compare the scores between the multicomponent intervention arm and control arm at both time points.

No firm conclusions can be drawn about the effects of single-component interventions due to limited evidence. Multicomponent interventions may not reduce the sales of non-prescription antimicrobial drugs in community pharmacies, drugstores, and other medicine outlets; however, the evidence is of very low certainty. Further studies on this topic are needed, particularly to assess the effects of important single interventions such as improving pharmaceutical policies.