Table 2 Characteristics of included studies

Spitaels D, 2019, Belgium, NA [29]

Non-randomized and controlled intervention study

Participated:

Intervention group: 426 GPs

Control group: 798 GPs

Completed the outcome questionnaire:

Intervention group: 73 GPs

Control group: 103 GPs

GPs

A 20-min, face-to-face educational outreach visit, with composed of two-components: face to face meeting about the evidence-based knee osteoarthritis management and a printed leaflet was provided for the GPs at the end of the educational visit.

Not visited by academic detailers.

Six months post-intervention

N: (1) Prescription of physical therapy: no significant change between the control and intervention group;

(2) Quality indicator adherence: no significant change between the control and intervention group.

Presseau J, 2018, England, Diabetes UK [30]

Two-arm cluster RCT

Randomized: Intervention group: 22 primary care practices (153 GPs, nurses, and HCAs)

Control group: 22 primary care practices (172 GPs, nurses, and HCAs)

Completed: Intervention group: 22 primary care practices

Control group: 20 primary care practices

GPs, nurses and healthcare assistants

Implementation intervention used behaviour change theory (e.g., SCT, Health Action Process Approach) and behaviour change techniques to develop the intervention, and involved outreach visits, to allow healthcare professionals to dedicate 90 min together to discuss the targeted healthcare behaviors, provided with materials to pre-identify barriers and solutions, short videos using practice-based examples, common barriers and possible solutions.

No intervention was provided and provided materials to control group at the end of the study.

12 months follow-up

N: Six guideline-recommended behaviors: no significant improvement:

Electronic medical records: blood pressure and glycaemic control prescribing, physical activity and nutrition advice: no significant differences between the two groups were found;

Patient survey: diabetes health education and foot examination

Pinto D, 2018, Portugal, National Health Institute [31]

Parallel, open, superiority, cluster RCT

Intervention group: 19 clusters with 120 participating physicians

Control group: 19 clusters with 119 participating physicians

Family physicians

During a 6-month period, 3 educational outreach visits, 3 guidelines were chosen for dissemination, each educational outreach visit was focused on one guideline, last from 15 to 20 min.

Each visit distributed a point of care summary, and a brochure was utilized as a visual aid.

Passive dissemination: by the publication on website.

18 months after the intervention.

N: Prescribed COX-2 inhibitors and the proportion of omeprazole: no statistically significant differences were found.

Wang H-YJ, 2018, USA, NA [32]

Two-arm cluster RCT

Randomized: Intervention group:13 practices, 246 patients

Control group: 12 practices, 233 patients

Completed:

Intervention group: 13 practices, 195 patients

Control group: 12 practices, 176 patients

Primary care physicians

Used social cognitive theory (SCT) to develop the intervention.

Consisted of three components: (1) a printed communication guide, (2) 2 in-office, structured training with patients, each training lasted approximately 45 min and the second training session was 4 to 6 months later after the first session, and (3) auxiliary materials.

No intervention materials except the local free/low-cost screening information sheet

12-month follow-up

N: Patients’ self-reported receipt of routine colorectal cancer screening: small, non-significant effects.

Trietsch J, 2017, Netherlands, ZonMw [36]

Two-arm cluster RCT

Arm A: 10 LQICs, 39 practices (86 GPs)

Arm B: 11 LQICs, 49 practices (122 GPs)

GPs

Audit and feedback with peer review in LQICs: each GP received performance feedback report, each group planned two meetings for each clinical topic (three topics in total from five different topics), test ordering and prescribing meeting, a total of six meetings.

Each meeting: 90 to 120 min.

Feedback reports were generated from diagnostic tests and prescriptions.

Same with Arm A on different topics (three topics in total from five different topics).

9 months after meetings

N: Tests ordered volumes, and drugs prescribed/practice/1000 patients/6 months: no statistically significant differences were found.

van der Velden AW, 2017, Netherlands, ZonMw [33]

Open, pragmatic, cluster RCT

Randomized:

Antibiotic intervention group: 45 practices

Control group: 41 practices

Completed (second year):

Antibiotic intervention group: 44 practices

Control group: 40 practices

GPs

Multifaceted program aims to improve antibiotic use for RTIs was consisted of GP education, audit/feedback and patient information, two 4-week registration of RITs;

Educational meeting: A meeting discussed the antibiotic prescribing guidelines and antibiotic-related problems with all GPs working in that primary care center in one session (60 to 90 min);

An improvement plan was defined on optimize antibiotic prescribing after the educational meeting.

Patient booklets: symptomatic treatment, natural course and alarm symptoms.

Feedback: all antibiotics dispensed in the year after the GPs meeting.

Usual practices

Two years after the program

P: Overprescribing, non-first choice prescribing and underprescribing for RTIs: sustainably improved antibiotic prescribing, significant differences between the control and intervention group.

Noto H, 2016, Japan, NA [34]

Open cluster-RCT

Randomized:

Intervention group: 22 PCPs

Control group: 20 PCPs

Completed:

Intervention group: 21 PCPs/230 patients

Control group: 15 PCPs/181 patients

PCPs

Received a copy of the Diabetes Treatment Guide, a copy of ‘The Manual and a 30-min seminar regarding ‘The Manual” and the updated copies were disseminated later.

Received a copy of the Diabetes Treatment Guide.

1-year follow-up period

M: Retinopathy evaluation (once annually), and measurements of urinary albumin excretion (every half year, significantly higher in the intervention group compared with the control group, p = 0.016) and serum creatinine (every half year).

Gerber JS, 2014, USA, Pfizer [23]

Cluster RCT

180 healthcare professionals

Intervention group: 9 practices

Control group: 9 practices

Pediatric primary care practices

(1) healthcare professional education, a 1-h review of current guidelines in prescribing, and (2) audit and feedback of antibiotic prescribing.

No intervention

18 months follow up

P: Broad-spectrum antibiotic prescribing: sustainably improved antibiotic prescribing.

Martín-Madrazo C, 2012, Spain, NA [35]

Cluster, parallel RCT

Randomized: Intervention group: 104 healthcare professionals (5 centers)

Control group: 110 healthcare professionals (5 centers)

Completed follow up: Intervention group: 84 healthcare professionals ((5 centers)

Control group: 86 healthcare professionals (6 centers)

auxiliary nurses, dental hygienists, GPs, nurses, pediatricians, midwives and odontostomatologists.

Teaching sessions (Training of healthcare workers) were provided by two nurses within 1 month: 4 sessions of 50 min each for each primary care center on implementation of hydroalcoholic preparations, a video demonstrated the hand hygiene technique (6 steps), each consultation office placed hydroalcoholic solutions, and reminder poster on the walls were placed at key locations (e.g., consultation office, emergency room and waiting room).

No intervention

6 months follow up

P: Hand hygiene compliance level: statistically significant differences were found, p = 0.001.

Enriquez-Puga A, 2009, England, Eli Lilly [22]

RCT

Antidepressant prescribing group: 14 general practices

Antibiotic prescribing group: 14 general practices

General practices

Antidepressant prescribing: Educational outreach visits: first visit: lasted 20 to 40 min, group interactive discussion on the appropriate prescribing, barriers to change, and best solution to overcome them.

A second visit: feedback on prescribing and clarify outstanding issues

Antibiotic prescribing: same with the antibiotic prescribing group.

Two years after the first educational outreach visit

M: Number of items prescribed for amoxicillin with clavulanic acid (co-amoxiclav); average daily quantities for lofepramine; fluoxetine antidepressants and quinolone antibiotics: prescribing of lofepramine (a tricyclic antidepressant) had increased, p < 0.001.

Cates CJ 2009, England, NA [21]

Before–after study

One practice (Manor View) and a nearby control practice (Attenborough)

General practices

Evidence-based printed handout for parents and provide a deferred antibiotic prescription (with not to offer the antibiotics immediately advice).

Usual practices

Three years

M: Prescribing levels of amoxicillin suspension: fell significantly more than national levels.