Skip to main content

Table 4 Antimicrobial stewardship strategies in remote primary health care settings

From: Antimicrobial stewardship in rural and remote primary health care: a narrative review

Reference #

First author

Year

Study type

JBI checklist used (score/total)

Country

Participants (number, sites)

Intervention description

Outcome measures

Main findings

Patient and provider education

Rural and remote

[70]

Belongia

2001

Non-randomised, controlled trial

Quasi-experimental Study (8/9)

United States

Parents and 151 primary care clinicians who provide paediatric care in a rural area

Clinician and community education via educational meetings and printed educational materials

Number of solid and liquid prescriptions per clinician, retail antibiotic sales, nasopharyngeal carriage of penicillin-nonsusceptible Streptococcus pneumoniae

Median number of solid antibiotic prescriptions per clinician decreased 19% in the intervention region and 8% in the control region. Median number of liquid antibiotic prescriptions per clinician decreased 11% in the intervention region but increased 12% in the control region. Retail antibiotic sales dropped in the intervention region but not in the control region

[62]

Chiswell

2019

Retrospective pretest–posttest study

Quasi-experimental Study (7/9)

United States

207 ‘walk-in’ patients in a rural primary care practice diagnosed with a respiratory tract infection

One-year patient education intervention programme involving repeated exposure to posters and handouts containing relevant health information

The number of antibiotics prescribed for respiratory tract infections

Antibiotic prescription rate decreased from 56.3% in the preintervention group to 28.8% in the postintervention group (x2 = 15.97, P < 0.001). The number of immediate antibiotic prescriptions dropped from 31.1% in the preintervention group to 13.5% in the postintervention group (x2 = 9.28, P < 0.05)

[65]

Wei

2017

Cluster randomised controlled trial

Randomized Controlled Trial (11/13)

China

Children aged 2–14 years given a prescription following a primary diagnosis of an upper respiratory tract infection in 25 primary care township hospitals across 2 rural counties

Clinician guidelines and training on appropriate prescribing, monthly prescribing peer-review meetings, and brief patient/caregiver education

Primary outcome: Antibiotic prescription rate

Secondary outcomes: Rates of prescribing multiple antibiotics, broad-spectrum antibiotics and intravenous antibiotics, proportion of prescriptions containing nonantibiotic medicines, cost

Antibiotic prescription rate decreased from 82 to 40% in the intervention group, and from 75 to 70% in the control group, yielding an absolute risk reduction in antibiotic prescribing of − 29% (95% CI − 42 to − 16; P = 0.0002)

[66]

Wei (subgroup follow-up analysis of above study)

2019

Cluster randomised controlled trial

Randomized Controlled Trial (11/13)

China

Children aged 2–14 years given a prescription following a primary diagnosis of an upper respiratory tract infection in 14 primary care township hospitals across 1 rural county

Clinician guidelines and training on appropriate prescribing, monthly prescribing peer-review meetings, and brief patient/caregiver education

Primary outcome: Antibiotic prescription rate

Secondary outcomes: Factors in sustaining intervention, rates of prescribing multiple antibiotics, broad-spectrum antibiotics and intravenous antibiotics, proportion of prescriptions containing nonantibiotic medicines, cost

The APR difference in the intervention arm at 6 months is − 49% (95% CI − 63 to − 35; P < 0.0001) compared to baseline. Compared to baseline, the APR difference in the intervention arm at 18 months is − 36% (95% CI − 55 to − 17; P < 0.0001). Compared to that at 6 months, the difference at 18 months represented no change in the APR. Factors sustaining reductions included doctors’ improved knowledge and communication skills and focused prescription review meetings, whereas lack of supervision and monitoring may be associated with relapse

[68]

Cummings

2020

Quasi-experimental

Quasi-experimental Study (6/9)

United States

Rural urgent care centres

Three behavioural interventions: (1) physician and patient education via lectures, presentations, media and distributable materials, (2) public commitment from the Medical Director of Urgent Care, and (3) peer comparison via individual feedback and blinded ranking emails

Proportion of acute respiratory tract infection diagnosis visits that received an inappropriate antibiotic

Percentage of inappropriate prescribing decreased 14.9%, from 72.6 to 57.7% (95% CI − 20.30% to − 9.05%; t, 5.44; P < 0.0001). Interrupted time series analysis showed a significantly lowered rate of antibiotic-inappropriate prescribing (95% CI − 4.59 to − 0.59; P = 0.014)

[74]

Zhang

2018

Cluster randomised controlled trial

Randomized Controlled Trial (11/13)

China

Children aged 2–14 years given a prescription following a primary diagnosis of an upper respiratory tract infection in 25 primary care township hospitals across 2 rural counties

Clinician guidelines and training on appropriate prescribing, monthly prescribing peer-review meetings, and brief patient/caregiver education

Cost per percentage point decrease in the antibiotic prescription rate

Incremental cost of US$0.03 per percentage point reduction in antibiotic prescribing

Mixed urban and rural

[67]

Varonen

2007

Randomised controlled trial

Randomized Controlled Trial (8/13)

Finland

30 rural and urban health centres

Nationwide guidelines implementation programme involving education based on a PBL or AD method facilitated by local general practitioners

Compliance with acute maxillary sinusitis management in national Current Care guidelines

Slight increase in the use of the first-line drug amoxicillin (39–48% in AD centres, 33–45% in PBL centres, controls 40%). Proportion of antibiotic courses with recommended duration increased (34–40% in AD centres, 32–47% in PBL centres, controls 43%)

[73]

Little

2001

Randomised controlled trial

Randomized Controlled Trial (9/13)

England

315 children presenting with acute otitis media in 42 general practices (33% mixed urban and rural settings)

Two treatment strategies – immediate antibiotics or delayed antibiotics – supported by standardised advice sheets

Resolution of symptoms, absence from school or nursery, paracetamol consumption

Children prescribed antibiotics immediately had shorter illness [− 1.1 days (95% CI − 0.54 to − 1.48)], fewer nights disturbed (− 0.72 (95% CI − 0.30 to − 1.13)], and slightly less paracetamol consumption [− 0.52 spoons/day (95% CI: − 0.26 to − 0.79)], but had higher incidence of diarrhoea (14/150 (9%) v 25/135 (19%), x2 = 5.2, P = 0.02), compared to delayed prescription

[75]

Haenssgen

2018

Quasi-experimental qualitative study

Quasi-experimental Study (7/9)

Laos

1130 peri-urban villagers

A one-off educational activity comprising of six sections—a mapping exercise, a medicine matching game, a resistance game, a role-play activity, a healthy-wealthy game and a feedback session

Attitudes and knowledge on antibiotics, treatment-seeking behaviour, and social networks

Awareness and understanding of antibiotic resistance improved, but effects on attitudes were minor. Mixed impact on behavioural changes. Activity-related communication spread within groups of greater privilege

Physician support systems

Rural and remote

[64]

Samore

2005

Cluster randomised trial

Randomized Controlled Trial (7/13)

United States

407,460 inhabitants and 334 primary care clinicians in 12 rural communities

6 communities received a community intervention alone and 6 communities received community intervention plus CDSS targeted toward primary care clinicians

Community-wide and diagnosis-specific antimicrobial usage

Prescribing rate decreased from 84.1 to 75.3 per 100 person-years in the CDSS arm vs 84.3 to 85.2 in community intervention alone (P = 0.03). Antimicrobial prescribing for visits in the antibiotics “never-indicated” category during the postintervention period decreased 32% in CDSS communities and 5% in community intervention-alone communities (P = 0.03). Macrolide use decreased significantly in CDSS communities (P = 0.001) but not in community intervention–alone communities

[69]

Gonzales

2013

Cluster randomised controlled trial

Randomized Controlled Trial (9/13)

United States

33 primary care practices in a rural region

Simple clinical algorithm implemented via a traditional printed decision support (PDS) or a computer-assisted decision support (CDS) strategy integrated into the workflow of an electronic health record

Antibiotic prescription rates for uncomplicated acute bronchitis

Percentage of antibiotic prescription decreased compared to baseline at the PDS intervention sites (from 80.0 to 68.3%) and at the CDS intervention sites (from 74.0 to 60.7%) but increased slightly at the control sites (from 72.5 to 74.3%). Differences due to interventions were statistically significant from the control sites (P = 0.003 and P = 0.01 for PDS and CDS, respectively) but not between themselves (P = 0.67)

[72]

Rubin (Extension analysis of study described in Reference #64)

2006

Observational quantitative study

Randomized Controlled Trial (7/13)

United States

99 primary care providers serving rural communities

A standalone personal digital assistant-based CDSS tool for the diagnosis and management of acute respiratory tract infections

Usage patterns and acceptability of the tool

Adherences with CDSS recommendations for the five most common diagnoses and for antibiotic choice were 82% and 76%, respectively. Logistic regression models indicate that provider adherence improved with each ten cases entered into the system (P = 0.001). Respondents believed the CDSS was easy to use, and most (44/65; 68%) reported that patient encounters were either the same duration or slightly faster, when using the CDSS tool compared with their usual practice

[76]

Madaras-Kelly

2006

Experimental quantitative cohort study

Quasi-experimental Study (8/9)

United States

192 patients visiting 2 rural community pharmacies for broad-spectrum antibiotics

Community pharmacists conducted guided interviews regarding patient symptoms and intercepted inappropriate prescriptions through communication with the ordering clinician to decrease broad-spectrum antibiotic use in upper respiratory infections

Number of patients agreeable for interview, pharmacist time, primary care provider acceptance of the recommendations, and patient opinion data regarding the pharmacy intervention

3% of the patients who were approached declined to discuss their symptoms and treatment with the pharmacist. 7% (n = 4) of patients permitted the community pharmacist to contact the prescriber to discuss first-line therapeutic alternatives. 2 of 3 clinicians contacted by pharmacists were receptive to altering the broad-spectrum antimicrobial to first-line antimicrobial therapy

Surveillance

Rural and remote

[45]

Cuningham

2020

Retrospective data analysis

Prevalence Study (9/9)

Northern Australia

15 remote primary health care clinics

The General Practice version of the National Antimicrobial Prescribing Survey (GP NAPS) tool modified for remote primary health care clinics

Antimicrobials used, indications and the treating health professional to yield similarities and differences in prescribing patterns, appropriateness of antimicrobial use and functionality of the GP NAPS tool

The adapted GP NAPS tool demonstrated potential as an audit tool of antimicrobial use for the remote primary health care setting in Australia. Compared with other Australian settings, narrow spectrum antimicrobials were more commonly prescribed with high appropriateness of use (WA: 91%; NT: 82%; QLD: 65%). The dominant treatment indications were skin and soft tissue infections (WA: 35%; NT: 29%; QLD: 40%)

[77]

Hui

2015

Mathematical model

Diagnostic Test Accuracy Test (7/10)

Australia

Simulated remote indigenous community

Individual-based mathematical model to determine the impact of molecular testing on AMR surveillance of gonorrhoea

Time delay between first importation and the first confirmation that the prevalence of gonorrhoea AMR has breached the 5% threshold (when a change in antibiotic should occur)

In the best-case scenario, the alert would be triggered within 3–6 months of the resistance proportion exceeding the 5% threshold, at least 8 months earlier than using culture alone

Mixed urban and rural

[78]

Schwartz

2019

Database validation

Diagnostic Test Accuracy Test (6/10)

Canada

9272 physicians prescribing antibiotics to patients ≥ 65 years in urban (90.3%) and rural (9.7%) locations of practice

IQVIA Xponent database of dispensed antibiotic prescription counts aggregated at the physician prescriber-level

Agreement and correlation between Xponent and Ontario Drug Benefit database, performance characteristics for Xponent to accurately identify high prescribing physicians

The Xponent database has a specificity of 92.4% (95% CI 92.0–92.8%) and PPV of 77.2% (95% CI 76.0–78.4%) for correctly identifying the top 25th percentile of physicians by antibiotic volume. In the sensitivity analysis, 94% of the top 25th percentile physicians in Xponent were within the top 40th percentile in the reference database. The mean number of antibiotic prescriptions per physician were similar, but the error was greater in rural areas

National policies

Mixed urban and rural

[63]

Hammond

2020

Ecological retrospective database analysis

Prevalence Study (6/9)

United Kingdom

163 urban (80.61%) and rural (14.12%) primary care practices

Incentivising reduced primary care prescribing of co-amoxiclav, cephalosporins and quinolones for any infection

Primary care antibiotic dispensing and antibiotic resistance in community-acquired urinary Escherichia coli

Overall antibiotic dispensing per 1000 registered patients decreased 11%. Antibiotic reductions were associated with reduced within quarter antibiotic resistance to amoxicillin, ciprofloxacin and trimethoprim, reduced subsequent quarter resistance to trimethoprim and amoxicillin, and increased within and subsequent quarter resistance to cefalexin and co-amoxiclav

[68]

Yin

2018

Retrospective database analysis

Prevalence Study (6/9)

China

500 secondary and tertiary hospitals, 600 urban PHC centres and 1600 rural PHC centres

Zero mark-up policies and national policy to improve the rational use of antibiotics in primary health care centres

Data on total and specific antibiotic consumption

Overall antibiotic consumption increased from 12.859 DID in 2012 to 15.802 DID in 2014. When national policies were introduced, this decreased to 13.802 DID in 2016. After an upward trend for 3 years, oral and parenteral antibiotic consumption decreased in rural PHC centres by 12% and 33% from 2014 to 2016

  1. AD academic detailing, AMR antimicrobial resistance, APR antibiotic prescription rate, CDS computer-assisted decision support, CDSS clinical decision-support system, CI confidence interval, DID DDD per 1000 inhabitants per day, GP NAPS General Practice version of the National Antimicrobial Prescribing Survey, NT Northern Territory, PBL problem-based learning, PDS printed decision support, PHC primary health care, PPV positive predictive value, QLD Queensland, WA Western Australia