Deferred prescribing in practice: an observational study in community pharmacies in France during 2024–5 and feedback from pharmacists

Abstract

Background

Excessive use of antibiotics promotes antibiotic resistance. Deferred prescribing (prescriptions dispensed according to clinical progress) is a strategy to reduce this overconsumption. Recognized in some countries, this practice is poorly documented and not recommended in France.

Objectives

To determine the rate of deferred outpatient prescriptions, their dispensing rate, and their characteristics.

Methods

This observational study was conducted over 1 year in 20 pharmacies and focused on prescriptions with a handwritten note indicating deferred dispensing. The data collected included patient age and sex, prescriber specialty, drug class, dispensing delay, and population density. Feedback from pharmacists was also collected.

Results

Of 174 deferred prescriptions (0.021% of prescriptions), 38% were dispensed immediately, 20% were dispensed later, and 42% were not dispensed. The majority were prescribed by general practitioners (84%). Dispensing was significantly more immediate in urban areas (P = 0.02). Feedback analysis showed that pharmacists had a positive perception of deferred prescribing, seeing it as a tool for empowerment, antibiotic reduction, and collaboration. However, its implementation was limited by logistical constraints and patients' perception of the prescription as a right to dispensing.

Conclusion

Effectiveness of deferred prescribing in reducing antibiotic consumption appears to be limited, especially in urban areas. Deferred prescription has little effect on antibiotic consumption. It seems more important to learn not to prescribe antibiotics when infections are viral. Delaying the prescription may help, but communication between pharmacists, doctors, and patients needs to be strengthened. Harmonization of practices and patient education need to be improved in this context of cascading uncertainties.

Key messages.

• Deferred prescriptions account for 0.021% of all prescriptions, 84% of which are issued by general practitioners.

• With 38% of prescriptions being dispensed immediately and 20% being dispensed later, deferred prescription has little effect on antibiotic consumption.

• Harmonization of practices and patient education need to be improved in this context of cascading uncertainties.

Introduction

Prescribing medication is a central pillar of patient care in France, occurring in nearly 78% of general medical consultations [1–3]. France ranks among the countries with the highest consumption of medication [4, 5], ranking second in Europe for antibiotic consumption in 2024 [6]. This excessive consumption contributes to the spread of antibiotic resistance, described as a “silent pandemic” by the World Health Organization [7]. It exposes patients with upper respiratory tract infections to unnecessary adverse effects, without any significant impact on complication rates [8–12]. National and international recommendations encourage more judicious prescribing. The “Choosing Wisely” campaign lists antibiotic prescription for acute bronchitis, uncomplicated influenza, serous otitis media, and nasopharyngitis, among the practices to be avoided [13]. The Pediatric Infectious Diseases Group also recommends discussing or delaying antibiotic therapy in cases of maxillary sinusitis lasting more than 10 days [14].

In light of these observations, deferred prescribing has emerged as an alternative strategy aimed at reducing inappropriate antibiotic use [12, 15]. Deferred prescriptions also exist for other drugs, such as painkillers and contraceptives. It consists of giving the patient a prescription and deferring to dispense the treatment for 48–72 h, depending on how the symptoms evolve [16]. This type of prescription should be distinguished from advanced prescribing (advance prescriptions follow a protocol and can be applied to multiple patients in a hospital setting), which generally follows a protocol [17]. In the case of conditional prescribing, dispensing is conditional on a clinical and/or biological parameter with the performance of rapid diagnostic tests (RDTs) (e.g. a sore throat test or urine strip test) [7, 17–19].

Its use is formalized in several international recommendations (National Institute for Health and Care Excellence, Canadian, and Norwegian recommendations) for common conditions, such as acute otitis media, sinusitis, tonsillitis, and cystitis [20–25]. Deferred prescriptions can then be used depending on the duration of symptoms, certain clinical scores, and the patient's general condition. In France, although there is no regulatory framework for this practice, the use of deferred prescribing is beginning to spread. The ongoing study “Antibiotic Therapy in Acute Otitis Media (ADOMA)” specifically evaluates this practice in patients with acute otitis media [16]. French medical students are interested in this practice, and several theses seemed to show that French general practitioners (GPs) seem to be familiar with and use deferred prescriptions, in part to satisfy patients' opinions and expectation. Compared with immediate prescriptions, deferred prescriptions are only used in 30%–50% of cases, thereby contributing to a reduction in antibiotic consumption. Its use depends heavily on the clinical situation (duration of symptoms, general condition), the patient's profile (who must be deemed reliable), and access to care. The recommendations in France is to control a patient in 48 h when GP has a doubt in a bacterial infection, in a context where the average waiting time for a general medical consultation is 6 days [26, 27], deferred prescribing would be a relevant compromise between the fight against antibiotic resistance and patient safety [12]. Most prescriptions with deferred prescriptions involve symptomatic treatments. As a result, patients frequently visit pharmacies to pick up their symptomatic treatments. This is a particularity in France, and the average number of medications prescribed per visit is 2.4. In 63.4% of cases—or 677 578 cases—at least three medications were prescribed [28].

However, data on the actual frequency and effective dispensing of deferred prescriptions in outpatient medicine remain limited.

The primary objective of this study was to determine the actual frequency of deferred outpatient prescriptions and their effective dispensing. The secondary objectives were to characterize these prescriptions and obtain feedback from the pharmacists.

Materials and methods

Study protocol

This cross-sectional study [29] and exploratory mixed-method was conducted from 1 October 2024 to 30 September 2025 in pharmacies in the Puy-de-Dôme department. The pharmacists collected prescriptions with handwritten notes indicating deferred dispensing. Every 2 months, they hand-delivered anonymized photocopies of the front pages of deferred prescriptions. The pharmacists noted whether the drug had been dispensed and the total number of prescriptions per month. This study was conducted in accordance with the STrengthening the Reporting of OBservational studies in Epidemiology [30]. In the absence of data in the literature on the rate of deferred prescriptions, an arbitrary and feasible decision was made to recruit 20 pharmacies.

Inclusion and exclusion criteria for pharmacies

Pharmacies were identified using a hybrid sampling method, combining stratification and snowballing [31, 32]. Stratification was carried out according to municipal density on the basis of the four-level municipal density grid (as of 1 January 2022) of the National Institute of Statistics and Economic Studies [33]. The Puy-de-Dôme department had 219 retail pharmacies in September 2024 [34], distributed among 30% high-density, 31% medium-density, 35% low-density, and 4% very low-density pharmacies [35]. Five pharmacies were selected from each density level (high, intermediate, low, and very low). High-density and intermediate-density areas were then grouped into urban areas (N = 10), and low-density and very low-density areas were grouped into rural areas (N = 10). There were no criteria for excluding pharmacies.

Each pharmacist completed a questionnaire providing contact details, the number and qualifications of pharmacy staff, average daily footfall, the presence of a supervising pharmacist, and population density (see Supplementary Appendix I).

Prescription inclusion and exclusion criteria

Deferred prescriptions written between 1 October 2024 and 30 September 2025 by doctors of all specialties were included, if a handwritten note on the prescription indicated when the treatment should be taken, for example, “after a predefined period” or “depending on the evolution of symptoms.” Prescriptions for conditional dispensing, contingent upon completion of a RDT in a pharmacy (e.g. urine strip, RDT for tonsillitis, RDT for influenza/COVID-19), were excluded. In cases of doubt regarding the inclusion of a prescription on the basis of the written notes, A.R. and J.C. jointly decided whether to include the prescription.

The number of monthly outpatient prescriptions could not be obtained from the pharmacy software. The calculation of the deferred prescription rate was based on the total number of prescriptions (outpatient and inpatient) according to each pharmacy's software.

Data collection

Continuous variables (such as prescription date, patient age, total monthly number) and categorical variables [dispensation (yes or no), dispensing time (immediate, deferred, no dispensing), geographical area of the pharmacy, patient sex, prescriber specialty, therapeutic class prescribed, and drug name] were collected and compiled in an Excel file in accordance with the recommendations of the Clermont-Ferrand Clinical Research and Innovation Department.

Statistics

The primary endpoints were to determine the proportion of deferred prescriptions among all prescriptions and to evaluate the immediate, deferred, or nondispensing of these prescriptions.

The secondary endpoints included the population of patients concerned, the population density of the territory, the specialties of the prescribing doctors, the therapeutic classes prescribed, and the seasonal variations in this practice.

Categorical data are described as numbers and percentages, whereas quantitative data are presented as medians and interquartile ranges given their non-Gaussian statistical distribution. Normality (Gaussian distribution) was evaluated using the Shapiro‒Wilk test. Comparisons between groups for categorical variables were performed using the chi-square test or Fisher's exact test (theoretical number of cases <5) and the nonparametric Kruskal‒Wallis test for quantitative variables. Statistical analyses were performed using Stata software (version 15; StataCorp, College Station, USA). All the statistical tests were performed bilaterally for a 5% risk of Type I error.

Feedback

After the quantitative data were collected, semi-structured interviews were conducted with the participating pharmacists. Two rural pharmacists and one urban pharmacist recruited at random from the list. The interviews were recorded using a dictaphone and manually transcribed into a Word document. The content of the interviews was then analyzed using an artificial intelligence tool. The two researchers, J.C. and A.R., also reviewed and analyzed the interviews by grounded theory analysis.

Results

Pharmacies recruited and deferred prescription rates

Twenty pharmacies were invited to participate, two of which declined due to staff shortages; thus, two new pharmacies were contacted (Fig. 1). A total of 20 pharmacies were included and evenly distributed between urban (N = 10) and rural (N = 10) areas. The majority had a training ground (75%), with no significant difference between urban and rural areas (P = 1). No pharmacies were excluded or lost to follow-up between 1 October 2024 and 30 September 2025. The data in Table 1 show a balanced distribution of pharmacies and professionals between urban and rural areas, with comparable characteristics. A total of 174 deferred prescriptions were recorded. The proportion of deferred prescriptions was 0.021% (∼2 prescriptions per 10 000). Two pharmacies recorded no deferred prescriptions during the study.

Figure 1.

Flow chart of deferred prescription inclusion. DD, deferred dispensing; ND, not dispensed; DI, immediate dispensing.

Table 1.

Descriptive analysis of pharmacies and prescriptions.

	Total	Urban area	Rural area	P
Number of pharmacies	20 (100%)	10 (50%)	10 (50%)
Internship placement				1
Yes	15 (75%)	8 (53%)	7 (47%)
No	5 (25%)	2 (40%)	3 (60%)
Number of professionals in pharmacies	7 (3.5; 10.5)	8 (6; 11)	5 (2; 10)	0.19
Of which pharmacists	3 (1.5; 4)	3 (3; 4)	2 (1; 4)	0.17
Of which pharmacy technicians (N = 18)	3 (1; 6)	3 (1.5; 4.5)	3 (1; 6)	0.82
All prescriptions combined (outpatient and inpatient)
Monthly number	3324 (1892; 4463)	3944 (2715; 4531)	1919 (1407; 4215)	0.23
Annual number	832 949 (100%)	489 835 (59%)	343 114 (41%)
Prescriptions deferred
Total	174 (100%)	109 (63%)	65 (37%)
Percentage	0.021%	0.022	0.018	0.31

The data are presented as numbers (percentages) or median (25th; 75th percentiles). In the case of missing data, the total number is presented in brackets in the first column. All tests were performed bilaterally. Percentages are rounded to the nearest whole number.

Inclusion of deferred prescriptions

The words “dispensed” or “not dispensed” were missing from 65 prescriptions initially (failure to annotate the prescription or describe the treatments dispensed on the back of the prescription); 52 were distributed among 4 pharmacies. Pharmacy No. 1 indicated that the absence of a note was equivalent to immediate dispensing. Pharmacies 2 and 3 searched their software for dispensing data for all the deferred prescription orders collected. In total, dispensing was unknown for 21 prescriptions.

Statistical analyses

Twenty-one prescriptions with unknown dispensing status were excluded.

General characteristics of deferred prescriptions

Among the 118 prescriptions with age data, 26% were for children ≤ 12 years old, 51% were for patients aged 13–65 years old, and 23% were for seniors >65 years old (see Table 2). The male/female sex ratio of patients who received a deferred prescription (n = 149) was 0.78. Deferred prescriptions were for antibiotics in 91% of the cases, 79% of which were beta-lactams (Fig. 2). Most deferred prescriptions were from GPs (84%). The 16% of prescriptions come from prescribers other than GPs; there were specialists in pediatrics, dental surgery, anesthesiology, orthopedic surgery, gynecology, dermatology, neurology, rheumatology, internal medicine, and emergency medicine. Between October and March, 76% of deferred prescriptions were recorded.

Table 2.

Characteristics of deferred prescriptions according to delivery time.

	Total	Not dispensed	Delayed dispensing	Immediate misuse	P
Number of prescriptions delayed	153 (100%)	64 (42%)	31 (20%)	58 (38%)
Areas					0.02
Urban	92 (60%)	37 (40%)	13 (14%)	42 (46%)
Rural	61 (40%)	27 (44%)	18 (30%)	16 (26%)
Age (years) (N = 118)					0.21
≤12	31 (26%)	17 (55%)	4 (13%)	10 (32%)
12 > age ≥ 65	60 (51%)	22 (37%)	15 (25%)	23 (38%)
>65 years old	27 (23%)	7 (26%)	7 (26%)	13 (48%)
Sex (N = 149)					0.79
Female	84 (56%)	37 (44%)	16 (19%)	31 (37%)
Male	65 (44%)	25 (38%)	14 (22%)	26 (40%)
Prescribers					0.52
GPs	128 (84%)	55 (43%)	27 (21%)	46 (36%)
Other prescribersa	25 (16%)	9 (36%)	4 (16%)	12 (48%)
Drug classes					0.34
Antibiotics	139 (91%)	60 (43%)	29 (21%)	50 (36%)
Other classesb	14 (9%)	4 (29%)	2 (14%)	8 (57%)
Day of the weekc (N = 148)					0.13
Monday	35 (24%)	14 (40%)	5 (14%)	16 (46%)
Tuesday	31 (21%)	16 (52%)	5 (16%)	10 (32%)
Wednesday	23 (16%)	6 (26%)	10 (44%)	7 (30%)
Thursday	26 (17%)	11 (42%)	2 (8%)	13 (50%)
Friday	33 (22%)	16 (49%)	6 (18%)	11 (33%)
Seasonal
October to March	117 (76%)
April to September	36 (24%)

The data are presented as numbers (percentages). Where data are missing, the total number is shown in brackets in the first column. All tests were performed bilaterally. Percentages are rounded to the nearest whole number. Values in bold indicate P < 0.05.

^aOther prescribers included specialists in pediatrics, dental surgery, anesthesia, orthopedic surgery, gynecology, dermatology, neurology, rheumatology, internal medicine, and emergency medicine.

^bOther drug classes included analgesics, anti-inflammatories, contraceptives, antifungals, antiemetics, and antiepileptics.

^cWorking days of the week, four prescriptions on Saturdays, and one prescription on Sundays were also included.

Figure 2.

Hierarchical distribution of prescribed drug classes. ¹Other drug classes included analgesics, anti-inflammatories, contraceptives, antifungals, antiemetics, and antiepileptics. ²Other antibiotics included aminoglycosides, cyclines, furans, and quinolones. Amox. + Clav., amoxicillin + clavulanic acid; ATB, antibiotics.

Characteristics of deferred prescriptions according to delivery time

Among the 153 prescriptions, 38% were dispensed immediately, 20% were dispensed later, and 42% were not dispensed, meaning that prescriptions were dispensed in at least 58% of the cases (Table 2). Dispensing did not vary significantly according to age (P = 0.21). It affected 45% of children ≤12 years old, 63% of patients aged 13–65 years old, and 74% of seniors >65 years old (Fig. 3).

Figure 3.

Distribution of the types of dispensing by age group.

With respect to population density, the difference observed (P = 0.02) indicated potentially more frequent delayed dispensing in rural areas, whereas immediate dispensing (misuse) seemed to predominate in urban areas. Depending on the time of dispensing some patients may have wanted the treatment for organizational reasons, such as weekends or going on vacation. But no significant differences were observed to the day of the week. Regardless of the method of dispensing, no significant differences were observed according to sex, drug class, and type of prescriber.

Feedback analysis

After the quantitative results were obtained, feedback from volunteer pharmacists regarding the study, deferred prescriptions, and explanations for the high percentage of immediate dispensing (50%) was explored. This figure of 50% was extrapolated from the initial results. The interview guide included the following three questions: What did you think of this research? As a pharmacist, what do you think of deferred prescriptions? It seems that 50% of deferred prescriptions are dispensed immediately across all pharmacies. Do you have an explanation for this?

Three pharmacists conducted semi-structured interviews. They were anonymized as follows: P1 (rural area), P2 (urban area), and P3 (rural area). The three pharmacists interviewed expressed a very positive opinion of the research, finding it useful, relevant, and well conducted, particularly because of its practical focus and its aim of promoting multiprofessional coordination. This research highlights the value of collaboration between doctors and pharmacists: “It's very nice; it promotes coordination between doctors and pharmacists; it's multiprofessional” (P1). For some, it allowed them to take a step back and look at their own practices, identifying the doctors who regularly prescribe: “It's interesting, we looked at it and saw which doctors do it in our practice” (P2). Some were surprised by the number of deferred prescriptions received, which they had not suspected: “I didn't think we had so many” (P3). Pharmacists also highlighted the quality of the methodological follow-up, particularly the regular reminders from the investigator, which helped maintain a focus on data collection.

Ambivalence: patient autonomy and empowerment or resignation in the face of symptoms? Cascades of uncertainty (doctor/pharmacist/patient)

Pharmacists saw deferred prescribing as a way of empowering patients and a useful tool for limiting the overuse of antibiotics: “It teaches people to take responsibility for their treatment” (P1); “It's very good; it empowers patients” (P2). It also helped to reduce the number of new consultations, resulting in savings: “The doctor doesn't have to see the patient again, which is good for the National Health Service” (P3). However, they quickly noted the practical limitations, particularly the difficulty for patients to assess the real need for antibiotics: “It's very good… and it's not so good because it makes patients more responsible” (P2); “Some people find it too complicated to know whether they will need antibiotics or not” (P3).

Regarding misuse, external explanations: patients, distance

Pharmacists believed that deferred prescriptions worked better with retired people, who had time to return to the pharmacy. They confirmed this estimate of 50% immediate dispensing and offered several converging explanations.

Patients wanting immediate dispensing were those in work, parents, people living far away, those with limited availability, or those anticipating a worsening of their condition: “I do have the forty-something man who says ‘I'm ill; I want my antibiotics’” (P1), as well as patients in a hurry, seeking quick relief: “They think it will be quicker with it”. (P2); “They consider that they need the treatment.” (P3). Patients mentioned dispensing the medicine without consumption: “Patients say: I'll wait, but I don't want to come back” (P3).

Delivery would be immediate, depending on logistical and organizational constraints, such as the distance between home and pharmacy: “They're not going to travel 30 km to come back” (P1); “It's 20 km to get here; they won't come back” (P3). Prescription times and days were also a factor: “On Friday evenings, they prefer to have the antibiotics” (P1); “If it's a prescription made on Friday, people take it” (P3); “On Sundays, the on-call pharmacy can be very far away” (P3).

The pharmacist's advisory role: minimized, difficult, or forgotten.

Pharmacists could be satisfied with having an advisory role to play, but they downplayed it: “It involves us a little bit; we don't dispense immediately; we also explain things to the patient, the patient comes back, they can ask us for advice if necessary, to find out whether they should take their treatment or not, we have a small active role” (P1); “What we say as pharmacists doesn't necessarily carry much weight” (P2). The prescription (even if delayed) was seen by the patient as a right to immediate withdrawal, which diminished the weight of the pharmacist's advice: “What is written on the prescription is what is due” (P1). The prescription delay was seen as difficult to enforce: “It's difficult for us to say ‘don't take it, the doctor said to wait’; it's difficult to refuse to dispense it; patients say ‘yes, I'll wait, but I don't want to come back in two days’mpa#rdquo;(P3).

Pharmacists could feel excluded from the advisory role: “The tricky part is how to get people to actually follow the advice” (P2). The cascade of opinions (visiting the chemist and then the doctor's surgery) could justify an increase in prescriptions, in a logic of seeking solutions, according to one pharmacist: “Patients feel they have already done the basics, and if they go to the doctor, they want to take it to the next level. Sometimes the doctor manages to reassure them, but sometimes they feel they are not being taken care of when all they are given is Doliprane, whereas we have already seen them for syrup, saline solution, etc. Comfort is important to people. If they've been coughing for three weeks and they go to the doctor because they can't sleep or work, etc., they're exhausted and need solutions” (P3).

Discussion

Main results

The rate of deferred prescriptions was 0.021%. Among the 153 prescriptions collected, 58% of deferred prescriptions were dispensed, including 38% for misuse (immediate dispensing) and 20% for deferred dispensing. Only one statistically significant difference was observed: misuse appeared to be more frequent in urban areas (P = 0.02). Dispensing appeared to be more common for adults and elderly individuals. Pharmacists attributed this misuse to patient demand and their logistical constraints.

Strengths and weaknesses

A large number of prescriptions were analyzed over an extended 12-month period, with no loss of follow-up among the participating pharmacies and regular and frequent data collection. To our knowledge, this is the only study based on real data rather than patient or doctor reports. This multiprofessional study, conducted by doctors and pharmacists, allowed for a comparison of perceptions.

The number of subjects included could not be calculated, limiting the power of the statistical analyses. The sampling does not reflect the actual distribution of the 219 pharmacies in Puy-de-Dôme, with rural areas being overrepresented. This bias limits the generalizability of the results to the entire department and potentially reduces the dispensing rate (which is lower in rural areas). Measurement bias due to the omission of deferred prescriptions and an underestimation of their frequency is possible. Pharmacists may have modified their dispensing practices because of the Hawthorne effect. Desirability bias may have influenced the responses given by the pharmacists to the investigator, who was a doctor. The exclusion of prescriptions with unclear instructions is a weakness. The qualitative analysis has a small number of interviews.

Comparison with the literature

Improved compliance among children, overdispensing in urban areas, and no variation across days of the week

It appears to be a factor in overprescribing, both in our study and in the literature: children are less likely than adults to collect deferred prescriptions. Compared with children, working adults and people over 65 seem more inclined to collect antibiotics (with no significant difference in our study). There is disconnection between pharmacists' thinking and the figures concerning elderly subjects. Deferred prescribing seems to be more effective in children, whereas pharmacists expressed the opposite view. Differences in understanding between patients and healthcare professionals are common, and healthcare professionals often hold misconceptions [36]. Observations from pharmacists confirm this: working adults are more likely to collect antibiotics, often because of confusion with symptomatic treatments or the belief that they will help them to remain fit for work.

Dispensing was more frequent in urban areas, which is consistent with the literature, whereas consultations in rural areas favor nonprescription [3]. In contrast to what some pharmacists suggest, the distance between home and community pharmacies is not a factor in overdispensing, perhaps the vulnerability of urban dwellers explains this results. Furthermore, no significant difference was observed between days of the week; Fridays do not favor overdispensing.

General practitioners who prescribe beta-lactams

Beta-lactams accounted for most prescriptions. Indeed, amoxicillin and amoxicillin/clavulanic acid account for more than half of antibiotic prescriptions [6]. GPs were responsible for 84% of deferred prescriptions, in line with data from Santé Publique France [37]. There are no national recommendations regarding deferred prescriptions. The Haute Autorité de Santé recommends no antibiotic treatment for nasopharyngitis and bronchitis [38, 39] or reassessment if there is no improvement within 48 h for sinusitis [40].

Delayed prescription versus immediate prescription or no prescription: a cascade of uncertainties

Ninety percent of deferred prescriptions were for antibiotics. Most studies on prescribing have been conducted in comparison with immediate prescribing. However, deferred prescribing could be compared with nonprescribing. In the Cochrane review, only a few studies performed this comparison; antibiotic consumption is lower when antibiotics are not prescribed [12]. This article evaluated the risk of complications and patient satisfaction according to the strategy used: there was no significant difference between delayed prescribing and nonprescribing. In conclusion, for suspected viral respiratory infections, nonprescribing should be preferred.

Doctors resort to deferred prescribing in situations where there is a high risk of immediate inappropriate prescribing. They consider deferred prescribing to be less evil than immediate prescribing. The PAAIR 2 study identified other effective strategies for reducing inappropriate prescribing [41]. Doctors, faced with diagnostic uncertainty and patient anxiety, also face organizational constraints—a lack of time and patients' unavailability for reassessment. The pharmacist, in turn, finds themselves in a situation of uncertainty, hesitating about the appropriateness of dispensing prescriptions and the patient's symptoms. The patient may find himself uncertain about the appropriateness of antibiotic therapy: a cascade of uncertainties.

Misuse could also reflect the illusion of unique invulnerability [42]. Healthcare professionals attribute inappropriate practices to external factors—patients or organizational constraints—rather than their own practices.

Dispensing and consumption: pharmacy loyalty

These observations suggest that the consumption rate is lower than the dispensing rate. Although 82% of patients report returning their drugs to the pharmacy at least once a year [43]. The drugs being dispensed may be kept at home and reused later, particularly for self-medication.

Among patients who received a deferred prescription, 64 did not collect it from their usual pharmacy. National data confirm strong patient loyalty to pharmacies [44, 45]: 79% of the general population uses only one pharmacy, with even greater loyalty among those over 65 (91%). This limits the risk of dispensing at another pharmacy and suggests that the absence of dispensing often corresponds to the absence of consumption.

Deferred prescription: an optimal strategy?

Changes in recommendations, particularly in Belgium, show that deferred prescribing is not always considered the optimal strategy, with some countries now favoring no initial prescribing [46, 47].

Legal framework, harmonization of practices, patient education, and pharmacists' roles

There is no legal or protocol framework for deferred prescribing. In this somewhat unclear context, the harmonization of practices and collaboration between professionals is paramount. Patient participation and partnership in this system also appear to be essential.

Pharmacists have many responsibilities in addition to rapid testing. In particular, they are called upon to provide prevention and renewal of chronic treatments as corresponding pharmacists in local healthcare communities [48, 49].

Conclusion

This observational study, which was conducted in 20 community pharmacies over 12 months, revealed that deferred prescriptions accounted for 0.021% of the total prescriptions dispensed, mainly antibiotics (91%). When presented at the pharmacy, 38% were dispensed immediately, indicating misuse, and 20% were dispensed on a deferred basis. A significant difference was observed according to population density and delivery time (P = 0.02), with immediate delivery appearing to be more common in urban areas. Delivery also appeared to be more common among adults and elderly individuals, although this difference was not statistically significant.

The effectiveness of deferred prescribing remains limited: at least 58% of patients obtains treatment. Compared with nonprescribing, consumption is higher. The difficulty experienced by doctors in not prescribing is also experienced by pharmacists in not dispensing and by patients in assessing their condition, illustrating a veritable cascade of uncertainties. Deferred prescribing does not seem optimal for limiting antibiotic consumption in its current form. This multiprofessional research suggests that collaboration between pharmacists and GPs and the harmonization of practices need to be improved or formalized in the context of deferred prescribing. When antibiotics are unnecessary, it is probably best not to prescribe them.

Supplementary material

Supplementary material is available at Family Practice online.

Funding

None declared.

Data availability

The data underlying this article are available from the corresponding author or from its online Supplementary material.

Ethics approval

This study, conducted outside the Jardé law, was carried out in accordance with MR-004. Authorization was obtained from the French Data Protection Authority (Commission Nationale de l'Informatique et des Libertés) on 18 April 2024. The protocol was approved by the Ethics Committee of Clermont Auvergne University (IRB00011540-2024-28) in 2024, in accordance with the Public Health Code. A file was given to each pharmacy, including a document describing the study objectives, inclusion criteria and instructions for annotating prescriptions (see Supplementary Appendix II) and an information poster for patients, specifying their right to refuse to participate.

Registration and protocol

This study did not require registration, and no study protocol was prepared.