Infectious Disease · Medical Literature
Using a decision support tool to reduce inappropriate antibiotic use after a delayed prescription.
Abstract
Antimicrobial resistance is partly driven by overprescribing antibiotics in primary care. Delayed antibiotic prescriptions, which allow patients to monitor their symptoms before deciding to use antibiotics, are a promising intervention. However, this strategy is undermined when patients initiate antibiotic use immediately. Using a behavioural game measure with behaviour-contingent financial incentives that models delayed prescriptions, we examined whether a decision support leaflet issued by the World Health Organization can, alone or in combination with other evidence-based interventions, reduce inappropriate antibiotic use for viral infections. In a pre-registered online experiment (N = 663 UK adults), participants were randomly assigned to one of five experimental conditions: (i) control condition, (ii) leaflet only, (iii) leaflet with value integration task, (iv) leaflet with active symptom monitoring and (v) leaflet with value integration and active monitoring. Inappropriate antibiotic use was lower in all decision support conditions than in the control condition (p < .001, ηG 2 = .02). Decision support tools also reduced decisional conflict (p < .001, ηG 2 = .02), particularly by improving value clarity and perceived decisional support. Integrating decision support tools into delayed prescription procedures can contribute to a more informed and responsible approach to antibiotic use by providing relevant information and tools.
Clinical relevance
Can a WHO decision support leaflet, alone or combined with a value integration task and/or active symptom monitoring, reduce inappropriate antibiotic use among patients who have been issued a delayed antibiotic prescription for a presumed viral infection? __SNAPSHOT__ {"studyDesign":"Pre-registered online randomized controlled experiment (five-arm parallel-group design) using a behavioral game measure with behavior-contingent financial incentives to simulate the delayed prescription context","setting":"Online; United Kingdom (general adult population)","sampleSize":"n = 663 UK adults (allocation across five arms not specified in abstract)","population":"UK adults recruited online; specific inclusion/exclusion criteria, age distribution, comorbidities, and prior antibiotic use not reported in the abstract","intervention":"WHO decision support leaflet alone (arm ii); WHO leaflet plus value integration task (arm iii); WHO leaflet plus active symptom monitoring (arm iv); WHO leaflet plus value integration task and active symptom monitoring combined (arm v)","comparator":"Control condition: no decision support tool (arm i)","primaryOutcome":"Inappropriate antibiotic use for viral infections, operationalized via a behavioral game measure with behavior-contingent financial incentives modeling the delayed prescription scenario; additional outcomes appear to include decisional outcomes (abstract truncated — full outcome list not reported)","followUp":"Not reported (single-session online experiment implied by design)","fundingSource":"Not reported"}
Key findings
Inappropriate antibiotic use was significantly lower across all four decision support conditions combined compared with the control condition (p < .001; generalized eta-squared [ηG²] = .02). The effect size of ηG² = .02 is conventionally considered small. No pairwise between-arm comparisons, absolute event rates, relative risk reductions, confidence intervals, or NNT/NNH values were reported in the abstract. The abstract was truncated before reporting secondary outcomes (including decisional outcomes) and any differential effects between the four active arms. Adverse events are not applicable to this behavioral intervention study.
Practice implications
This study provides controlled experimental evidence that WHO-endorsed patient decision support materials — with or without behavioral augmentation (value integration task, active symptom monitoring, or both) — can reduce simulated inappropriate antibiotic use in a delayed-prescription context in UK adults. The use of a gamified behavioral paradigm with financial incentives is methodologically innovative but introduces ecological validity concerns: real-world decisions occur under illness burden, patient-clinician relationship dynamics, and access barriers not captured in an online experiment. The study does not establish whether these effects translate to actual antibiotic consumption behavior in clinical settings. The patient population is limited to UK adults recruited online, and the abstract does not report whether subgroups (e.g., older adults, those with prior antibiotic experience, lower health literacy) responded differently. The small effect size (ηG² = .02), while statistically significant, requires careful clinical contextualization — the absolute difference in inappropriate use between arms is not reported, and statistical significance at this sample size does not confirm clinical meaningfulness. Whether combined interventions (arms iii–v) outperform the leaflet alone cannot be determined from the truncated abstract. NOTE: This summary is based on the abstract only; the full manuscript was not available for review, and findings should be interpreted accordingly. __LIMITATIONS__ Abstract truncated — secondary outcomes and between-arm comparisons are not reported, preventing evaluation of the relative contribution of each intervention component.,Online experimental paradigm using a behavioral game with financial incentives may not generalize to real-world primary care antibiotic decision-making under illness.,Effect size is small (ηG² = .02); absolute reductions in inappropriate antibiotic use are not reported, limiting assessment of clinical significance.,No confidence intervals reported for any estimate.,Single-session design precludes evaluation of behavioral persistence or durability of any intervention effect.,Sample restricted to UK adults recruited online, limiting generalizability to non-UK populations, elderly patients, pediatric caregivers, or those with low digital literacy.,Behavioral game outcome is a surrogate measure; actual antibiotic fill/consumption rates were not assessed.,Funding source not reported — potential conflicts of interest cannot be evaluated.,Pre-registration status is stated but pre-registration details (registry, deviations from protocol) are not described in the abstract. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ A WHO decision support leaflet, alone or combined with a value integration task and/or active symptom monitoring, statistically significantly reduced simulated inappropriate antibiotic use compared with no intervention in a UK online experiment (p < .001), though the effect was small (ηG² = .02) and the clinical translation of this behavioral paradigm to real-world prescribing remains unestablished. The findings are hypothesis-generating and support investment in patient-facing educational tools as adjuncts to delayed prescribing strategies, but replication in pragmatic primary care trials measuring actual antibiotic consumption is needed before practice change is warranted.