Inappropriate antibiotic prescribing and demand for antibiotics in patients with upper respiratory tract infections is hardly different in female versus male patients as seen in primary care

Abstract

Background: Unnecessary prescribing of antibiotics is a major public health concern. General practitioners (GPs) prescribe most antibiotics, often for upper respiratory tract infections (URTIs), and have in general been shown to prescribe antibiotics more often to women. No studies have examined the influence of patient gender on unnecessary antibiotic prescribing.

Objectives: To study a possible gender difference in unnecessary antibiotic prescriptions for URTIs in general practice; to assess whether a possible difference is explained by patient demand for antibiotics.

Methods: Post-hoc analysis of a cross-sectional study including 15 022 patients with URTI (acute rhinitis, acute otitis media, acute sinusitis, acute pharyngotonsillitis) from Argentina, Denmark, Lithuania, Russia, Spain and Sweden (HAPPY AUDIT Project). The association between gender and unnecessary antibiotic prescriptions, unadjusted and adjusted for treatment demand, was analysed using logistic regression models.

Results: A total of 25% of patients with URTI received antibiotics; in 45% of the cases, antibiotics were unnecessary. Overall, no gender difference for unnecessary prescribing of antibiotics for URTIs was found. Women with acute otitis media received an unnecessary antibiotic twice as often as men (14.4% versus 7.1%). In Danish patients with acute pharyngotonsillitis, there was a gender difference in unnecessary prescriptions for antibiotics (women 29.1% versus men 48.6%). Some 14% of patients receiving unnecessary antibiotics demonstrated a demand for antibiotics, but no gender difference was found in this group.

Conclusion: This study indicated a high rate of unnecessary antibiotic prescribing for URTIs in general practice, but overall found no gender differences in receiving unnecessary antibiotic prescriptions.

Keywords::

• General practice
• antibiotics
• infectious diseases
• upper respiratory tract
• gender

KEY MESSAGE:

• In general practice, there seems to be no difference in unnecessary prescribing of antibiotics for men and women with upper respiratory tract infections.

• Minor differences between genders could not be explained by the differences in patient demand for antibiotics.

INTRODUCTION

Unnecessary use of antibiotics is a major public health problem causing development of antimicrobial resistance (1). Some 50–60% of all antibiotics prescribed in general practice are for respiratory tract infections (2,3), which are among the most common reasons for patient encounter (9–28%) (4,5). Upper respiratory tract infections (URTIs) represent the majority of RTIs worldwide (6) and lead to prescribing of antibiotics in 60% of all consultations (7). However, most URTIs are caused by virus where antibiotics have no beneficial effect (8–10). Studies from general practice have demonstrated high rates of unnecessary antibiotic prescribing for patients with URTIs: 18% of patients with acute otitis media are treated unnecessarily with antibiotics (11), 22% with acute sinusitis, and 71% with acute tonsillitis (12).

In earlier studies of unnecessary antibiotic prescribing many factors linked to the general practitioner (GP) have been associated with increased rates of unnecessary antibiotic prescribing for RTIs, e.g. the GPs’ seniority, high-volume practice and single-handed practice (13,14). However, no studies have yet examined the association between patient gender and unnecessary/inappropriate prescribing of antibiotics in general practice, although female gender in several studies has been associated with a higher likelihood of receiving a prescription for antibiotics in general (2,15–17). This has led us to the hypothesis that women with URTIs may be more likely to receive an unnecessary prescription for antibiotics compared to men. Previous studies have also shown that patient demand for unnecessary antibiotics influences the GPs’ prescribing for RTIs (18,19). Thus, it is interesting to study possible gender differences in patient demand for antibiotics.

The aim of this study was to examine whether women with URTIs had a higher risk of being unnecessarily treated with antibiotics compared to men. Furthermore, we wanted to investigate whether a potential increased risk of unnecessary antibiotic prescribing among women could be explained by an increased demand for antibiotics.

METHODS

Study design

The current study represents a post-hoc analysis— performed in 2013—of the dataset of the HAPPY AUDIT study (health alliance for prudent prescribing, yield and use of antimicrobial drugs in the treatment of respiratory tract infections) (20). Briefly, data were collected from GPs in six countries: Sweden, Denmark, Argentina, Spain, Russia and Lithuania. A total of 618 GPs participated voluntarily in the study, after receiving random invitations. All patients with RTIs were registered during a three-week period in the winter of 2007/2008. The study protocol was submitted to a legally constituted ethics committee and deemed exempt from review (The Scientific Ethics Committee for the County of Vejle and Funen, Denmark).

Data collection

Registration was performed during the consultation on a registration chart developed according to the Audit Project Odense (APO) method (21). Variables recorded were: patient gender, age, duration of symptoms, patient symptoms and signs, diagnostic tests performed, assumed aetiology, diagnosis of consultation, choice of treatment and possible patient demand for antibiotics.

URTIs were classified in accordance with the International Classification of Primary Care (ICPC): acute rhinitis (R74), acute otitis media (H71, H72), acute sinusitis (R75), and acute pharyngotonsillitis (R72, R74).

All antibiotic prescriptions were classified as necessary or unnecessary (Table 1). Antibiotic prescriptions following recommendations for antibiotic prescribing according to international guidelines were classified as necessary prescriptions. Prescriptions in conflict with recommendations were classified as unnecessary. We used widely accepted international guidelines for treatment of URTIs: NICE, EPOS and HAPPY AUDIT (22–24).

Table 1. Criteria for classification of the appropriateness of antibiotic prescribing (necessary and unnecessary prescribing) in patients with upper respiratory tract infections (RTIs).

	Necessary prescribing^a	Unnecessary prescribing^b
Acute rhinitis	Never	Always
Acute otitis media	Symptoms ≥ 3 days or	Symptoms < 3 days and absence
	purulent ear secretion	of purulent ear secretion
Acute sinusitis	Symptoms ≥ 5 days or	Symptoms < 5 days and fever
	fever ≥ 38.5°C	< 38.5°C
Acute pharyngotonsillitis	≥ 3 Centor criteria^c or ≥ 2 Centor criteria and a positive strep A test^d	< 3 Centor criteria^c or a negative strep A test^d

^aFollowing international guidelines on antibiotic prescribing.

^bAntibiotic prescription in conflict with international guidelines.

^cPresence of tonsillar exudate, tender anterior cervical lymphadenopathy or lymphadenitis, history of fever and absence of cough.

^dStreptococcus antigen test

Study population

Patients with URTI accounted for 21 206 (63.7%) of the 33 273 patients registered with RTI. Patients aged < 16 years, patients referred to specialist or hospital, and patients of unreported gender were excluded (Figure 1). An additional 105 patients were not enrolled due to insufficient registration.

Figure 1. Flow chart showing the inclusion and exclusion of patients in the study population.

Statistical analyses

The data was analysed in Statistical Analysis System (SAS), version 9.2 (SAS Institute, Cary, North Carolina). Multivariable logistic regression was used, performed with SAS PROC GENMOD, using generalized estimating equations (GEE) to account for clustering at GP level. We analysed unnecessary antibiotic prescribing in men versus women and determined confounding or modification of a gender difference by patient demand for antibiotics. Results are presented as the estimated incidence with 95% confidence interval (CI) of unnecessary prescribing for the corresponding gender. Gender differences are expressed in odds ratios (OR; with 95% CI). The statistical significance of the gender difference was tested by a likelihood ratio test for this OR. A P-value < 0.05 was considered statistically significant.

RESULTS

Antibiotic prescribing

A total of 15 022 patients met the inclusion criteria (70.8% of all patients with URTI). Antibiotics were prescribed in 25.4% of all consultations. There was no difference between male and female patients (Table 2). The lowest antibiotic prescribing rate was found in patients with acute rhinitis (2–4%), and highest in patients with acute otitis media (74–97%) and acute sinusitis (78–93%).

Table 2. Number of office visits, antibiotic prescriptions and proportion of unnecessary prescriptions for antibiotics for upper respiratory tract infections (RTIs) among men and women visiting general practices participating in HAPPY AUDIT^a.

	Total no. of office visits men/women	Men No. of antibiotic prescriptions (%)	No. of unnecessary prescriptions	% of unnecessary prescriptions (95% CI)^b	Women No. of antibiotic prescriptions (%)	No. of unnecessary prescriptions	% of unnecessary prescriptions (95% CI)^b	P-value^b
All upper RTIs	5 946/9 076	1 495 (25.1%)	677	45.3 (41.7; 48.9)	2 320 (25.6%)	1 032	44.5 (41.0; 48.0)	0.77
Acute rhinitis	3 522/5 120	92 (2.6%)	92	100%	127 (2.5%)	127	100%	–
Acute otitis media	192/292	170 (88.5%)	12	7.1 (3.8; 12.8)	257 (88.0%)	37	14.4 (10.6; 19.3)	0.02
Acute sinusitis	318/649	266 (83.6%)	64	24.1 (18.9; 30.1)	562 (86.6%)	123	21.9 (18.1; 26.2)	0.46
Acute pharyngotonsillitis	1 914/3015	967 (50.5%)	509	52.6 (48.4; 56.8)	1 374 (45.6%)	745	54.2 (49.7; 58.7)	0.47

^aHealth alliance for prudent prescribing, yield and use of antimicrobial drugs in the treatment of respiratory tract infections.

^bThe presented proportions of unnecessary prescriptions, confidence intervals and P-value are adjusted for clustering at doctor level

CI, confidence interval.

Unnecessary antibiotic prescriptions

For both genders, 45% of antibiotic prescriptions were classified as unnecessary (Table 2). The highest proportion of unnecessary prescriptions was found in patients who were prescribed an antibiotic for acute rhinitis (100%), followed by acute pharyngotonsillitis (53.6%). For the whole group of URTIs, no gender difference was found in the proportion of unnecessary antibiotic prescriptions (OR = 1.02; 95% CI: 0.90–1.15). A gender difference was only found in antibiotic prescriptions for acute otitis media, where prescriptions for antibiotics for women were significantly more unnecessary compared to those for men (OR = 2.03; 95% CI: 1.05–3.93).

Table 3 (Appendix, web only) shows the distribution of unnecessary antibiotic prescriptions for male and female patients by country and type of URTI. Sweden had the lowest rates of unnecessary antibiotic prescribing. In four of the six countries, most antibiotic prescriptions for acute pharyngotonsillitis were classified as unnecessary. Only in Denmark could a gender difference in unnecessary antibiotic prescription be found (P-value 0.02). This was mostly accounted for by more unnecessary antibiotic prescriptions for acute pharyngotonsillitis in Danish men as compared to women (OR = 2.16; 95% CI: 1.32–3.56).

Table 3. Rate of unnecessary antibiotic prescribing (% of consultations followed by a prescription) in men and women by country and diagnosis.

Unnecessary prescription of antibiotics		Sweden		Denmark		Argentina		Spain		Russia		Lithuania
		n (%)^a	95% CI^b	n (%)	95% CI	n (%)	95% CI	n (%)	95% CI	n (%)	95% CI	n (%)	95% CI
Total number, ♀+♂	47 (19.3)	114 (27.3)	189 (44.2)	1 038 (48.6)	200 (52.2)	121 (59.0)
All upper RTIs	♀	30 (19.1)	13.7–26.0	68 (24.1)	19.2–29.9	124 (47.3)	37.3–57.5	612 (48.6)	44.1–53.2	144 (55.0)	41.4–67.8	54 (54.5)	43.3–65.3
	♂	17 (19.5)	12.0–30.3	46 (34.1)	26.0–43.1	65 (39.2)	28.9–50.5	426 (48.5)	43.7–53.2	56 (46.3)	33.5–59.5	67 (63.2)	52.9–72.4
Acute rhinitis	♀	5 (100.0)	−	10 (100.0)	−	6 (100.0)	−	92 (100.0)	−	11 (100.0)	−	3 (100.0)	−
	♂	5 (100.0)	−	2 (100.0)	−	9 (100.0)	−	63 (100.0)	−	7 (100.0)	−	6 (100.0)	−
Acute otitis media	♀	2 (18.2)	4.4–51.5	4 (22.2)	8.5–46.9	3 (10.0)	3.9–23.1	26 (15.3)	10.5–21.8	1 (5.3)	0.9–24.7	1 (11.1)	1.4–52.1
	♂	0 (0)	−	1 (14.3)	2.0–58.1	1 (4.2)	0.5–25.7	9 (8.3)	4.0–16.3	0 (0)	−	1 (14.3)	1.9–59.3
Acute sinusitis	♀	7 (9.2)	4.7–17.4	20 (14.6)	9.4–21.9	16 (33.3)	20.0–50.0	59 (24.8)	19.1–31.5	17 (37.8)	28.1–48.5	4 (22.2)	7.4–50.5
	♂	5 (21.7)	9.6–42.0	9 (16.1)	8.5–28.3	5 (26.3)	12.0–48.4	33 (24.8)	17.8–33.5	6 (46.2)	25.3–68.4	6 (27.3)	9.0–58.6
Acute pharyngo-	♀	16 (24.6)	15.4–36.9	34 (29.1)	21.1–38.5^c	99 (55.6)	42.8–67.8	435 (57.4)	52.1–62.5	115 (61.5)	42.9–77.2	46 (66.7)	53.0–78.0
tonsillitis	♂	7 (14.6)	7.1–27.7	34 (48.6)	35.4–61.9	50 (43.9)	32.8–55.6	321 (55.9)	50.5–61.2	43 (47.8)	32.8–63.2	54 (76.1)	64.9–84.5

^aNumber of unnecessary prescriptions for antibiotics (% of total antibiotic prescriptions).

^bThe presented proportions of unnecessary prescriptions, confidence intervals and P–value are adjusted for clustering at doctor level.

^cP-value = 0.02.

♀, female; ♂, male; RTIs, respiratory tract infections; CI, confidence interval.

Patient demand for antibiotics

Of all patients receiving unnecessary antibiotics, 120 (14%) demanded antibiotics. Overall, no differences between male and female demand for antibiotics were found among patients receiving unnecessary antibiotics: Rhinitis (28.5% versus 19.7%, P = 0.56), acute otitis media (0% versus 2.7%), acute sinusitis (0% versus 3.3%), and acute pharyngotonsillitis (4.5% versus 6.7%, P = 0.16).

DISCUSSION

Main findings

Overall, the results could not support the hypothesis that there is a gender difference in the proportion of unnecessary antibiotic prescriptions for URTI. A general gender difference was found for acute otitis media, where women had twice as many unnecessary antibiotic prescriptions than men (14.4% versus 7.1%). In Denmark, men with acute pharyngotonsillitis more often received an unnecessary antibiotic prescription. There were no differences in demand for antibiotics between men and women.

Strengths and limitations

The strength of this study is the large number of participating GPs and patients from different countries. The registration was performed in a real life setting with usual time schedule and no prior information about the patients. It is, therefore, likely that the data reflect the actual conditions in the six countries studied and can be extrapolated to areas with similar practice settings. In addition, our classification of antibiotic prescriptions as necessary or unnecessary was based on well approved international guidelines (22–24).

A possible weakness is that GPs participated on a voluntary basis and may represent a group of doctors more interested in rational antibiotic use. Our results may, therefore, underestimate the rate of unnecessary antibiotic prescribing worldwide. Additionally, participating in an audit may have influenced the prescribing pattern: the GPs might be tempted to register a diagnosis that justified the treatment instead of vice versa, which could lead to diagnostic misclassification bias. However, such a potential bias would not be expected to influence the effect of gender on unnecessary antibiotic prescribing. Furthermore, the fact that data collection took place in 2007 and the analysis in 2013 could potentially lead to bias, but since there was no change in clinical guidelines in that period, this is not expected to influence the outcome.

Interpretation

The high rates of unnecessary antibiotic prescriptions for URTIs in general practice are in accordance with previous studies (7,11,12,16).

Few gender differences in the rate of unnecessary antibiotic prescribing were found. The results point in different directions and do not support our prior hypothesis that women had a higher risk of receiving unnecessary antibiotic prescriptions for URTIs than men.

We found no gender differences in patient demand. Previous studies have concluded that patient demand for treatment was the most important non-clinical factor for prescription of non-indicated antibiotics for RTIs, irrespective of gender (18,19).

Implication for practice and future research

Women received an unnecessary antibiotic prescription for acute otitis media almost twice as often as men. This finding should be taken into account when clinical practice guidelines for URTIs are being updated (25,26). Detailed insights into the decision-making process of antibiotic prescribing by the individual GP, e.g. by qualitative studies, could further elucidate where efforts could best be directed in the continuous campaign to reduce unnecessary antibiotic prescribing.

Conclusion

Our results could not support the hypothesis that there is a gender difference in the proportion of unnecessary antibiotic prescriptions for URTI. There were no differences in demand for antibiotics between men and women.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.