pone.0356480.t007 - <p> </p>

Information on patients’ clinical problems and prescriptions was extracted from practice medical records before and after the intervention and analysed using multilevel generalised linear mixed models (GLMM), including practice as a random intercept, and accounting for trial arm, age, deprivation and co-morbidities. Symptoms recorded on the first and second Profiles were compared. Resolution of clinical problems was described, and participants’ perspectives sought.</p> <p>Results</p><p>We recruited one general practice group, comprising six practices, and 60 patients.

Three patients were lost to the study. Intervention arm patients were more likely to have more than one problem addressed than control arm patients (22/27[82%] vs. 11/30[35%], adjusted odds ratio [aOR] 7.48, 95% confidence interval [CI] 1.99-28.10). All intervention arm patients either benefited or expected to benefit following referrals.

For two patients, the intervention proved critical: detection of B12 deficiency in a patient using metformin and flecainide de-prescribing. Fewer patients reported pain by the study’s end (16/27 vs 24/27). All patients were grateful for ADRe, but clinicians had reservations concerning staff time.</p> <p>Conclusion</p><p>The ADRe Profile benefited patients by reducing risk of suboptimal prescribing and iatrogenic harm.

The project is registered with the Clinicaltrials.gov as NCT04663360 (<a href="clinicaltrials.gov/study/NCT04663360" target="_blank">clinicaltrials.gov/study/NCT04663360</a>).</p></div>