https://orcid.org/0000-0002-0165-8934
Abstract
Background
Factors associated with the appropriateness of antibiotic prescribing in primary care have been poorly explored. In particular, the impact of computerised decision-support systems (CDSS) remains unknown.
Objectives
We aim at investigating the uptake of CDSS and its association with physician characteristics and professional activity.
Methods
Since May 2022, users of a CDSS for antibiotic prescribing in primary care in France have been invited, when registering, to complete three case vignettes assessing clinical situations frequently encountered in general practice and identified as at risk of antibiotic misuse. Appropriateness of antibiotic prescribing was defined as the rate of answers in line with the current guidelines, computed by individuals and by specific questions. Physician’s characteristics associated with individual appropriate antibiotic prescribing (< 50%, 50–75% and > 75% appropriateness) were identified by multivariate ordinal logistic regression.
Results
In June 2023, 60,067 physicians had registered on the CDSS. Among the 13,851 physicians who answered all case vignettes, the median individual appropriateness level of antibiotic prescribing was 77.8% [Interquartile range, 66.7%–88.9%], and was < 50% for 1,353 physicians (10%). In the multivariate analysis, physicians’ characteristics associated with appropriateness were prior use of the CDSS (OR = 1.71, 95% CI 1.56–1.87), being a general practitioner vs. other specialist (OR = 1.34, 95% CI 1.20–1.49), working in primary care (OR = 1.14, 95% CI 1.02–1.27), mentoring students (OR = 1.12, 95% CI 1.04–1.21) age (OR = 0.69 per 10 years increase, 95% CI 0.67–0.71).
Conclusion
Individual appropriateness for antibiotic prescribing was high among CDSS users, with a higher rate in young general practitioners, previously using the system. CDSS could improve antibiotic prescribing in primary care.
KEY MESSAGES
Individual appropriateness for antibiotic prescribing is high among CDSS users.
CDSS use could passively improve antibiotic prescribing in primary care.
Factors associated with appropriateness for antibiotic prescribing for primary care diseases are: prior use of CDSS, general practice speciality vs. other specialities, younger age and mentoring of students.
Introduction
In primary care, the appropriateness rate of antibiotic prescribing is one of the preferred metrics to evaluate the impact of stewardship programs [Citation1–4]. Appropriateness refers to the concordance of antibiotic prescribing with current guidelines and encompasses the need for prescription or not, when appropriate the correct choice of drug, dosage, delivery route and/or duration [Citation2]. In the United Kingdom and Canada, evidence suggests that appropriateness remains low in primary care with antibiotic prescribing rates higher than ‘ideal’ targets established for common infectious diseases [Citation3,Citation5,Citation6].
Computerised Decision Support Systems (CDSS) are a potential solution to improve antibiotic prescribing. However, data are lacking about the potential role of CDSS as a medium for improving the knowledge base and antibiotic prescribing in primary care [Citation3,Citation7–9].
We used clinical case vignettes upon registration in a CDSS dedicated to antibiotic prescribing in primary care, to investigate the appropriateness of physicians’ intended antibiotic prescribing, and to assess associated factors among physician characteristics, type of professional activity, and their use of the system.
Methods
This is a cross-sectional study based on clinical case vignettes, assessing the knowledge and practice attitudes of users registered in the ‘Antibioclic’ CDSS for/in France [Citation10]. From May 2022, a free but mandatory registration to the CDSS was implemented and three case vignettes were submitted to all physicians registering to the system. The ‘Antibioclic’ CDSS (http://www.antibioclic.com) was developed by French academics and released in 2011 and is described elsewhere [Citation10,Citation11]. Access and use are free of charge to any healthcare professional or service user 24 h/day, 7 days/week. It is widely used by French GPs, with more than half of French GPs (51,000/80,000) registered to the system as of December 2023. Users are described in . It relies on a task-network model to translate national guidelines into an easy-to-use system for a large panel of infectious diseases (n = 36) at the point of care. The task-network model allows the physician to tailor a query based on the patient’s key characteristics for antibiotic prescribing (disease, age group, pregnancy/breastfeeding, chronic kidney disease, and other conditions specific to the infectious disease of interest – e.g. urinalysis results for urinary tract infections). The CDSS also allows users to download educational leaflets for parents (e.g. appropriate home management of isolated fever), access e-learning content on urinary tract infections, and to a geolocated directory of referring infectious disease physicians for tele-expertise requests. The CDSS is not integrated into primary care electronic health records and physicians’ prescribing software and operates as a standalone application.
Table 1. Characteristics of responders vs. non-responders.
Upon registration and for the two weeks thereafter, newly registered users received daily reminders asking them to participate in a short evaluation of their antibiotic prescribing habits in theoretical clinical situations identified as being at risk for antibiotic misuse [Citation12].
The case vignette stories accounted a total of twelve questions. The first was about a three-year-old child who presented with rhinitis progressing to acute otitis media after three days. The second was about probabilistic and targeted antibiotic therapy for uncomplicated pyelonephritis due to E. coli, in a 35-year-old woman. The third case was about the management of consecutive patients with bronchitis, and acute exacerbations of Gold-II or -III chronic obstructive pulmonary disease (COPD). A full description of these case vignettes, including questions, is provided in the supplementary materials. When responding to the questions presented in the vignettes, users were not authorised by the system to use the CDSS.
We extracted data from the ‘Antibioclic’ database in June 2023, and analysed all physicians’ responses to case vignettes. We did not use data from dentists, midwives, students or patients. We focused on the nine questions specifically designed to assess the ‘appropriateness’ of antibiotic prescriptions: three for first case, two for the second, and four for the last case. We were not able to differentiate the four components of appropriateness: drug, dosage, delivery route and/or duration [Citation2].
The ‘appropriateness rate’ was defined (i) by individual, as the ratio of the number of correct answers to the specific questions answered; (ii) by specific question, as the ratio of the number of respondents with correct answers to the total number of respondents for the specific question. Correct answers were those corresponding to the current (2021) French national guidelines for antibiotic prescribing in primary care [Citation13].
For each of the situations at risk of inappropriate prescribing, we estimated the level of appropriateness (95% confidence interval, 95%CI). For users with complete responses, we also computed the individual level of appropriateness across situations. We used multivariate ordinal logistic regression to assess the association between individual appropriateness (three categories: 50% <; 50–75%; > 75%) and individual characteristics: age (by 10-year increment), gender (male/female), general practitioner (GP) (yes/no), practicing in primary care (yes/no), mentoring of medical students (yes/no), using of the ‘Antibioclic’ CDSS prior mandatory registration (yes/no). Significance was set at 5% bilateral. We did not impute missing data. All analyses were performed using the R statistical software (v4.1.1).
All users provided written informed consent for data analysis, in accordance with the European Union General Data Protection Regulation (EU GDPR).
Results
In June 2023, 60,067 physicians were registered in the ‘Antibioclic’ CDSS, of which 18,940 (31.5%) participated in the case vignettes evaluation, and 13,851 (23.1%) had complete responses. The characteristics of the responders were very similar to those of the non-responders, but the large number of respondents resulted in statistical difference in medical practice, with a higher proportion of GPs practicing in primary care (81.3% vs. 79.8%), .
The lowest specific question appropriateness was found for empirical antibiotic prescribing for uncomplicated pyelonephritis, with 40.4% (95% CI, 39.7%–41.1%), and the highest for uncomplicated bronchitis with 98.0% (95% CI, 97.8%–98.2%), . For urinary tract infections, 57.0% (95% CI, 56.3%–57.7%) of respondents intended to prescribe fluoroquinolones empirically for pyelonephritis, despite fluoroquinolone use in the previous 6 months, which contraindicates their use according to French guidelines [Citation13]. For rhinopharyngitis, up to 16.0% (95% CI, 15.4%–16.6%) of respondents proposed a delayed prescription of amoxicillin ± clavulanic acid, although antibiotics are not recommended in this situation [Citation13].
Table 2. Appropriateness of antibiotic therapy by case-vignette.
The median individual appropriateness of antibiotic prescribing was 66.7% [IQR, 55.6%–88.9%] overall, and 77.8% [IQR, 66.7%–88.9%] among complete responders. Only 1,353 (9.8%) physicians with complete responses had individual appropriateness less than 50%.
In multivariable analysis among physicians, the lowest appropriateness was found in older users, with a 31% decrease per 10-year increase (OR = 0.69, 95% CI 0.67–0.71). GPs showed the highest level of appropriateness compared to other specialists (OR = 1.34, 95% CI 1.20–1.49), especially when working in primary care (OR = 1.14, 95% CI 1.02–1.27), as well as those supervising students (OR = 1.12, 95% CI 1.04–1.21), and those with previous use of the CDSS (OR = 1.71, 95% CI, 1.56–1.87), , supplementary Table 1, supplementary Figure 1.
Table 3. CDSS’users Characteristics associated with level of overall individual appropriateness, multivariate analysis (physicians with complete response).
Discussion
Main findings
Up to three-quarters of CDSS users intended to prescribe appropriate antibiotic therapies for common situations at risk of misuse in primary care, with higher appropriateness among young GPs working in primary care, involved in student mentoring, and experienced in using the system.
Comparison with existing literature
The use of the CDSS could improve the knowledge base and antibiotic prescribing. This observation needs to be confirmed in a clinical trial [Citation10]. The reported intended level of appropriateness was close to the ideal targets for the United Kingdoms, France and Canada [Citation3,Citation5,Citation6]. It was higher than that observed in the literature among physicians not using a CDSS [Citation1,Citation3,Citation5,Citation6]. Using claims data from French northern territories, Simon et al. reported an appropriateness of 73–75% for urinary tract infections, compared to 41% among our CDSS users for uncomplicated pyelonephritis. Such differences may be explained by variations in data source, case definition, and accuracy of appropriateness definition. In fact, Simon et al. did not assess adherence to antibiotic prescribing guidelines, but rather prescribing rates against ideal targets, with less granularity on complexity of infection [Citation3].
Factors associated to individual appropriateness in our study were different, yet complementary to those already reported [Citation3,Citation7–9]. The appropriateness was higher in younger physicians, possibly because of recent completion of medical training allowing better compliance to new guidelines focusing on the use of narrow spectrum antibiotics with the shortest possible duration [Citation7]. Young physicians is also the population using the most the CDSS possibly because we promoted its use among medicine faculty students and residents over the last decade [Citation10]. Compared to other specialists, physicians trained as GPs and practicing in primary care may had higher appropriateness rate; possibly because they have a greater experience and knowledge in the management of infections encountered in primary care.
Although CDSS users did not intend to prescribe antibiotics for likely viral infections, we observed that one sixth choose delayed antibiotic prescribing for rhinopharyngitis. Delayed antibiotic prescribing has been shown to be an acceptable trade-off with immediate prescribing for respiratory tract infections [Citation14], but there is no evidence for rhinopharyngitis or other viral infections, for which antibiotics are not recommended [Citation1,Citation5]. Physicians using the CDSS also intend to prescribe fluoroquinolones despite recent exposure. It is therefore important to incorporate fluoroquinolone exposure into decision models to guide prescribing, as has been done in our CDSS since 2018 [Citation11].
Implication for clinical practice and policy
After more than a decade of activity, encompassing physician training and interaction with key stakeholders, we believe that CDSS offering guidance for antibiotic prescribing in primary care is instrumental in raising awareness and knowledge towards antimicrobial stewardship. Effective end-to-end communication between stakeholders, CDSS administrators and users, is also an opportunity for health authorities to maximise public information and health response and public-oriented information. To overcome differences in learning and uptake of CDSS, the integration of these systems into physicians’ prescribing software will be necessary. Ultimately, CDSS should aim for full interoperability to link electronic health record data with laboratory reports and guidelines, to achieve the most individualised recommendation [Citation15]. This could also be the opportunity to deliver contextual alerts and reminders, as well as educational and training messages based on the physician’s prescribing habits (e.g. possible side effects and impact on antimicrobial resistance).
Strengths and limitations
Our study has several limitations. We report theoretical appropriateness based on non-mandatory case-vignette responses, with a limited number of respondents, which preclude the generalisation of our findings to real-world antibiotic prescribing behaviour. This study may primarily assess users’ competency in using the CDSS rather than capturing the nuanced decision-making process involved in actual antibiotic prescribing. As of June 2023, more than 48,000 French GPs are registered in the CDSS, while ∼80,000 are established in primary care. The demographic characteristics of the respondents were similar to those of the non-respondents. The appropriateness rate between complete and incomplete respondents differed by ∼10%, and one may suggest that appropriateness rate might be lower among those who did not respond to all the case vignettes. We lack information on the characteristics of the patients seen by physicians, and on the level of antibiotic prescribing by these physicians in real-life. Moreover, the characteristics of GPs using the CDSS differ from those of the French GPs population as a whole, being younger, more likely to be female (60.9% vs. 46.9%) and working in primary care (78.7% vs. 63.9%). Defining appropriateness in primary care is challenging [Citation1,Citation2,Citation4], and we were not able to differentiate the four components of appropriateness, but we used a robust definition that is reproducible in other settings or countries, in line with local guidelines [Citation2,Citation3]. Finally, it is also possible that the attitudes of physician respondents may be different in real life than they are in the survey.
Conclusion
In conclusion, the appropriateness of intended antibiotic prescribing for theoretical clinical situations in primary care, was high among physicians registered in a CDSS that was created 10 years ago and has been widely used by the primary-care community, especially among young GPs practicing in primary care. There may be a familiarisation and learning process, suggesting that CDSS could passively improve antibiotic prescribing in primary care, and deserves further studies in real-world settings.
Author contributions
TD, JLB, PYB and FT initiated the study, and obtained funding. TD and JLB supervised the study; TD and AM designed the experimental plan; AM managed data and performed the analyses; TD, AM, JLB and PJ can take responsibility for the integrity of the data and the accuracy of the data analysis, AM is the guarantor; TD, AM, JLB and PJ prepared the first draft of the manuscript; All authors contributed to interpretation of the data, critically revised the manuscript, and approved the final version.
Supplemental Material
Download MS Word (118.6 KB)Acknowledgements
Antibioclic is a non-profit organization and is not linked to pharmaceutical companies for the contents of the CDSS or for funding. The steering committee members are volunteers and funding is obtained through competitive calls for tenders from universities, the French health authorities or learned societies. The Antibioclic steering committee (alphabetical order): Elisabeth Bouvet, Tristan Delory, Sylvie Lariven, Josselin Le Bel, François-Xavier Lescure, Pauline Jeanmougin, Nathan Peiffer-Smadja.
The sponsor and the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
These results were partially presented at the 24th ‘Journées Nationales d’Infectiologie’ (oral communication, Grenoble, France).
Data sharing statement
According to informed consent signed by the users of the CDSS platform, the authors cannot publicly release the data used in this analysis. However, any person or structure, public or private, for- profit or non-profit, can request access to the data within the year following publication, upon request to the corresponding author.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Funding
References
- Smith DRM, Dolk FCK, Pouwels KB, et al. Defining the appropriateness and inappropriateness of antibiotic prescribing in primary care. J Antimicrob Chemother. 2018;73(suppl_2):ii11–ii18. doi:10.1093/jac/dkx503.
- Leung V, Langford BJ, Ha R, et al. Metrics for evaluating antibiotic use and prescribing in outpatient settings. JAC Antimicrob Resist. 2021;3(3):dlab098.
- Simon M, Thilly N, Pereira O, et al. Factors associated with the appropriateness of antibiotics prescribed in French general practice: a cross-sectional study using reimbursement databases. Clin Microbiol Infect. 2022;28(4):609.e1–609.e6.
- Thilly N, Pereira O, Schouten J, et al. Proxy indicators to estimate appropriateness of antibiotic prescriptions by general practitioners: a proof-of-concept cross-sectional study based on reimbursement data, North-Eastern France 2017. Euro Surveill. 2020;25(27):1900468.
- Pouwels KB, Dolk FCK, Smith DRM, et al. Actual versus “ideal” antibiotic prescribing for common conditions in English primary care. J Antimicrob Chemother. 2018;73(suppl_2):19–26. doi:10.1093/jac/dkx502.
- Wu JHC, Langford B, Ha R, et al. Defining appropriate antibiotic prescribing in primary care: a modified Delphi panel approach. J Assoc Med Microbiol Infect Dis Can. 2020;5(2):61–69.
- Cadieux G, Tamblyn R, Dauphinee D, et al. Predictors of inappropriate antibiotic prescribing among primary care physicians. CMAJ. 2007;177(8):877–883.
- Nowakowska M, van Staa T, Mölter A, et al. Antibiotic choice in UK general practice: rates and drivers of potentially inappropriate antibiotic prescribing. J Antimicrob Chemother. 2019;74(11):3371–3378. doi:10.1093/jac/dkz345.
- Degnan KO, Cluzet V, David MZ, et al. Development and validation of antibiotic stewardship metrics for outpatient respiratory tract diagnoses and association of provider characteristics with inappropriate prescribing. Infect Control Hosp Epidemiol. 2022;43(1):56–63.
- Delory T, Jeanmougin P, Lariven S, et al. A computerized decision support system (CDSS) for antibiotic prescription in primary care-Antibioclic: implementation, adoption and sustainable use in the era of extended antimicrobial resistance. J Antimicrob Chemother. 2020;75(8):2353–2362.
- Delory T, Le Bel J, Lariven S, et al. Computerized decision support system (CDSS) use for surveillance of antimicrobial resistance in urinary tract infections in primary care. J Antimicrob Chemother. 2022;77(2):524–530. doi:10.1093/jac/dkab392.
- Ferrat E, Le Breton J, Guéry E, et al. Effects 4.5 years after an interactive GP educational seminar on antibiotic therapy for respiratory tract infections: a randomized controlled trial. Fam Pract. 2016;33(2):192–199.
- Gauzit R, Castan B, Bonnet E, et al. Anti-infectious treatment duration: the SPILF and GPIP French guidelines and recommendations. Infect Dis Now. 2021;51(2):114–139.
- Stuart B, Hounkpatin H, Becque T, et al. Delayed antibiotic prescribing for respiratory tract infections: individual patient data meta-analysis. Br Med J. 2021;373:n808.
- Sutton RT, Pincock D, Baumgart DC, et al. An overview of clinical decision support systems: benefits, risks, and strategies for success. NPJ Digit Med. 2020;3(1):17. doi:10.1038/s41746-020-0221-y.