Table 3 Key findings from primary studies specific to GPs, prescribing and EO included in all systematic reviews

Atkin, Ogle, Finnegan, and Shenfield (1996): RCT (Chhina et al., 2013) [23]

Concurrent medication prescription to older adults

EO visit including education on adverse medication reactions and the importance of hospital-admission prevention in older adults

Although the mean number of medications concurrently prescribed per older adult decreased in the EO group at one year, the number significantly decreased across both groups (df = 3, F = 3.78, p < 0.02)

Avorn and Soumerai (1983): RCT (Chhina et al., 2013) [23]

Excessive prescribing of common medications (cephalexin, propoxyphene, and cerebral and peripheral vasodilators)

Pharmacist-led behavioural theory-based EO visits

The GPs in the EO group reduced the mean number of the common medications they prescribed (5439 mean units to 4174, p = 0.0001) 14% more than the GPs in the no-intervention control (5415 mean units to 4921) did at one year.

Berings, Blondeel, and Habraken (1994): RCT (Smith & Tett, 2010) [19]

Benzodiazepine prescribing

EO visit supplemented with mail-outs

Greater reductions in benzodiazepine prescribing was seen in GPs who received EO and mail-outs (21% between-group difference), as opposed to mail-outs alone (14% between-group difference)

Bernal-Delgado, Galeote-Mayor, Pradas-Arnal, and Peiro-Moreno (2002): RCT (Chhina et al., 2013) [23]

Anti-inflammatory prescribing

Structured EO with printed materials that explained tenoxicam was a less cost-effective option compared to diclofenac

Structured EO (Tenoxicam packages prescribed reduced by 22.5% [95%CI 34.42 to −10.76]) reduced tenoxicam prescribing significantly more than unstructured EO visits (Tenoxicam packages prescribed reduced by 9.78% [95%CI −17.70 to −1.86]).

Clyne (2014): RCT (Clyne et al., 2016) [2]

Inappropriate prescribing

Pharmacist-led EO, a GP-led medication review guided by web-based algorithms and information leaflets

Intervention group reduced inappropriate prescribing rates significantly more than usual care control group at one year (OR = 0.3, 95%CI = 0.1 to 0.7, p = 0.02)

De Burgh (1995): RCT (Chhina et al., 2013 [23]; Smith & Tett, 2010) [19]

Prescribing of benzodiazepines for insomnia and anxiety

Educational visit and supporting materials from a pharmacist or doctor

Intervention group reduced prescribing of benzodiazepines more than the control group, but this difference was not significant.

Gall, Harmer, and Wanstall (2001): before-and-after (Kamarudin et al., 2013) [27]

Inappropriate prescribing of supplements to malnourished patients

Practical and theoretical EO visit on how to use nutritional guidelines, assess for malnutrition and treat nutritional deficits

A significant 15% reduction in total prescribing of supplements was seen at three months (438 patients were prescribed supplements at baseline, which reduced to 372 patients at follow-up).

Graham, Hartzema, Sketris, and Winterstein (2008): cohort (Chhina et al., 2013) [23]

COX-2 prescribing

EO visit on evidence-based osteoarthritis management, emphasising minimising COX-2 prescribing.

General practitioners in intervention group significantly reduced COX-2 prescribing by 0.76 defined daily doses/patient compared to the control at 3 months; however, this effect was not maintained at 12 months.

Ilett et al. (2000): RCT (Chhina et al., 2013) [23]

Antibiotic prescribing

EO visit delivered by a therapeutics advisor involved delivering, and briefly discussing, the best practice guidelines for antibiotic prescription for otitis media, urinary tract infections, and upper and lower respiratory tract infections

The number of non-recommended antibiotic prescriptions (e.g. cefaclor and roxithromycin) per provider increased for GPs who received EO and GPs in the control group at three months. However, prescriptions of non-recommended antibiotics increased more for GPs in the control group, meaning non-recommended antibiotic prescribing decreased 74% more in the EO group.

Midlov, Bondesson, Eriksson, Nerbrand, and Hoglund (2006): RCT (Chhina et al., 2013 [23]; Kamarudin et al., 2013 [27]; Smith & Tett, 2010) [19]

Benzodiazepines prescribed to older adults

Two EO visits outlining the effects of long and medium acting benzodiazepines in older adults

General practitioners in the intervention group prescribed total (26.63%), and long and medium-acting benzodiazepines (25.80%) significantly less after one year compared to GPs in a wait-listed control group.

Peterson, Bergin, Nelson, and Stanton (1996): cohort (Chhina et al., 2013) [23]

Anti-inflammatory prescribing

EO program that emphasised reducing NSAID prescriptions mainly due to their negative side effects, and increasing use of other medications such as paracetamol, for people with rheumatic disease

Anti-inflammatory prescribing reduced by GPs in both intervention and control groups.

Ray et al. (2001): RCT (Chhina et al., 2013) [23]

Anti-inflammatory prescribing

EO program that emphasised reducing NSAID prescriptions mainly due to their negative side effects, and increasing use of other medications such as paracetamol, for people with osteoarthritis

EO, together with prompts to review NSAID prescription in patient files, significantly reduced the number of days patients had NSAIDs dispensed each year (between-group difference 7% [95%CI 3 to 11%]) by GPs. However, reductions in prescribing were seen in both groups.

Rognstad, Brekke, Fetveit, Dalen, and Straand (2013): RCT (Clyne et al., 2016) [2]

Inappropriate prescribing

GP-led EO program plus audit and feedback

The GPs in the intervention group (n = 250) reduced their inappropriate prescribing practices (measured using Beer’s criteria) by 12.1% (95%CI 16.8 to 6.9%) per 100 patients compared to GPs in the control group (education on antibiotic prescribing for respiratory infections)

Simon et al. (2006): RCT (Clyne et al., 2016) [2]

Inappropriate prescribing

Group EO program designed to increase acceptance of evidence-based computer alerts and was delivered alongside the integration of age-specific medication alerts that appear when potentially inappropriate medications (e.g. long-acting benzodiazepine) were entered by a GP into the patient’s medical record.

Adding EO to alerts did not enhance the efficacy of the alerts (which were also received by control group) at reducing inappropriate prescribing by GPs, as inappropriate prescriptions per 10,000 older adults decreased similarly for both groups (p = 0.52)

Tomson, Hasselström, Tomson, and Åberg (1997): RCT (Chhina et al., 2013) [23]

Prescribing of beta-2-agonists for asthma management

Tailored EO delivered twice per year and including oral and written information about evidence-based management of asthma

General practitioners in the intervention group significantly reduced their prescribing of beta-2-agonists and increased the prescribing of inhaled steroids but the between-group findings were not statistically significant

van Eijk, Avorn, Porsius, and de Boer (2001): RCT (Chhina et al., 2013) [23]

Prescribing anticholinergics to older adults

EO visits (individual vs. group) on the difficulties of managing anticholinergic side effects in older adults

The amount of highly anticholinergic antidepressants prescribed to older adults reduced by 26% (95% CI: - 4 to 48%) in the individual EO group and by 45% (95% CI: 8 to 67%) in the group EO group, compared to control groups.

Witt, Knudsen, Ditlevsen, and Hollnagel (2004): RCT (Chhina et al., 2013) [23]

Prescribing of beta-2-agonists for asthma management

One EO visit that involved discussing an evidence-based asthma guideline and supporting GPs to use it

General practitioners in the intervention group did not significantly reduce beta-2-agonist prescribing or increase the prescribing of inhaled steroids. Although, the reduction in beta-2-agonist and increase in inhaled steroid prescription, was 2 and 7% greater, respectively, for the intervention group compared to the control group.

Zwar, Wolk, Gordon, and Sanson-Fisher (2000): RCT (Chhina et al., 2013 [23]; Smith & Tett, 2010) [19]

Benzodiazepine prescribing

A 20-min EO visit about benzodiazepine prescribing

General practitioners in the intervention group reduced their prescribing rate (per 100 patient encounters) of benzodiazepines for all indications, including sleep problems and anxiety, from 2.3 to 1.7; however, this reduction was like that seen in the control group (a change of 2.2 to 1.6) who received an intervention on an unrelated topic