Temporal Association between Influenza Vaccination Coverage and Ambulatory Antibiotic Use in Children

Abstract

In this population-based retrospective cohort study, increased influenza vaccination coverage was temporally associated with a reduction in ambulatory antibiotic use in children. After adjustment for yearly vaccine effectiveness, antibiotic prescription rate declined by 3/1000 person-months for each 1% increase in influenza vaccination coverage between 2012 and 2017 in South Carolina (P < 0.001).

Excessive ambulatory antibiotic use is a major contributor to antibiotic resistance and Clostridioides difficile infection in the community.¹ Prior studies have demonstrated an increase in ambulatory antibiotic prescription rates by over 20% during the winter season which coincides with the peak of influenza virus activity.² Moreover, >60% of the antibiotics are prescribed for acute respiratory tract infections (ARTI) in children.² Several recent studies have demonstrated a significant decline in ambulatory antibiotic prescription in children.^2,3 Although few factors have been speculated to explain this pattern, little is known about the impact of influenza vaccination coverage on antibiotic prescription rates in children.⁴ The purpose of this population-based retrospective cohort study is to examine the temporal association between influenza vaccination coverage and ambulatory antibiotic prescription rates in children in South Carolina.

METHODS

Medicaid and State Employee Health Plan pharmacy claims for outpatient oral antibiotics were utilized for estimation of community antibiotic prescription fill rates in South Carolina population 6 months to 17 years of age from 2012 to 2017, which represents approximately 60% of the South Carolina population in this age group. Since the peak of the influenza virus activity often occurs during the first 4 months of the year in Southeastern USA, antibiotic prescription fill rates were examined in January through April 2012–2017.⁵ State-level influenza vaccination coverage rates for the 2011–2012 to 2016–2017 influenza seasons for children 6 months–17 years old were obtained from the Centers for Disease Control and Prevention (CDC) FluVaxView database.⁶ Linear regression was used to examine the association between antibiotic prescription fill rates in January through April 2012–2017 and influenza vaccination coverage in the corresponding influenza season after adjustments for influenza vaccine effectiveness in that season as obtained from the CDC.⁶

RESULTS

During the 6-year study period, the annualized antimicrobial prescription rate in children in South Carolina declined from 103 to 79 per 1000 person-months for the months January through April indicating a 23% decline (P < 0. 001) (Table 1). Prescription rates declined for all antibiotics, although the decline varied by antibiotic class with the greatest reductions in macrolides (44%). Influenza vaccination coverage also increased from 51% during the first influenza season of the study and peaked at 61% in the 2014–2015 influenza season (Fig. 1). Influenza vaccine effectiveness ranged from 40% to 52% during the study period except for 2014–2015 when it dipped to 19%. After adjusting for vaccine effectiveness, there was a temporal association between the decline in the antibiotic prescribing rate and the increase in influenza vaccination coverage. Antibiotic prescription rates declined by 3 per 1000 person-months for each 1% increase in influenza vaccination coverage in children (P < 0.001).

TABLE 1. - Annualized Antibiotic Prescription Fill Rates per 1000 Person-Months in Children (6 months–17 years of age), January Through April 2012–2017

Year	Penicillins	Cephalosporins	Macrolides	Overall antibiotics*	Influenza vaccination coverage (%)	Vaccine effectiveness (%)
2012	48	18	24	103	50.6	47
2013	49	18	24	102	52.2	49
2014	37	13	14	74	57.1	52
2015	43	15	15	82	60.7	19
2016	43	16	15	83	59.2	48
2017	43	15	13	79	55.5	40
Percent change†	−12%	−21%	−44%	−23%	+10%	−15%

^*Includes other less commonly used antibiotic classes in children that are not shown in the table.

†Absolute change from 2012 to 2017.

FIGURE 1.: Influenza Vaccine Coverage and Annualized Antibiotic Prescription Fill Rate (Per 1000 Person-Months) in Children (6 Months–17 Years of Age), January Through April 2012–2017. 

DISCUSSION

Antibiotics are among the most frequently prescribed classes of medications for children.⁷ It has been estimated that >150 million ambulatory visits result in an antibiotic prescription annually, including >30 million prescriptions for children in the United States.⁸ However, in many of these instances, antibiotics are prescribed for conditions of viral origin, for which they provide no clinical benefit.⁴ The overall rate of antibiotic prescribing in ambulatory settings is declining in children as evidenced by recent studies.^2,3,9 Some of this national decline is attributable to reductions in prescribing for acute respiratory conditions of viral origin through educational campaigns of the general public and health care providers, CDC’s Get Smart campaign, stricter diagnostic criteria for ARTI, and watchful waiting for nonsevere otitis media in selected children.^4,10 There have been other studies assessing educational interventions combined with quarterly individualized antibiotic prescribing feedback resulting in reduced broad-spectrum antibiotic prescribing for patients with pneumonia and sinusitis, although the effect has not been consistent and antibiotic prescribing improvements reversed immediately after the study conclusion.¹¹ In addition, prescribing may have changed due to concerns about adverse drug events and US Food and Drug Administration safety communications.¹²

The current study demonstrates a decline in antibiotic prescription fill rates during the peak of the cold and influenza season (January through April) in children between 2012 and 2017. The lowest overall antibiotic prescription rate was in 2014 when the vaccine coverage and effectiveness were higher which likely translated to less ARTI and related office and emergency room visits and hence a lower antibiotic prescription rate.

Although the current study design does not allow examination of all the factors contributing to the decline in antibiotic prescription fill rates in children, it offers 1 potential explanation that warrants further examination in future studies. The temporal association between the increase in vaccination coverage against influenza virus and decrease in ambulatory antibiotic use in children in South Carolina suggests that vaccination may be an important antimicrobial stewardship tool. Effective influenza vaccination may reduce the burden of ARTI and related office and emergency room visits and hence a decrease in antibiotic prescription rates. Roll out of childhood vaccinations against Haemophilus influenzae type b (Hib) and Streptococcus pneumoniae have contributed to a substantial reduction in the burden of Hib-related disease and a significant decline in progression to complex otitis media after the rollout of conjugated pneumococcal vaccines, PCV7 and PCV13,^13,14 and associated antibiotic use for these conditions.

The current results confirm recent findings of reduced antibiotic use in association with increased influenza vaccine coverage in all age groups in the USA.¹⁵ These results encourage existing efforts to improve influenza vaccine coverage, particularly in children, and further emphasize the use of vaccination as an antimicrobial stewardship tool in the community.

This study has several limitations. Antibiotic prescription is a surrogate for antibiotic use because prescription filled does not necessarily mean the medication was administered. Although most antibiotics used during the winter season in children are likely prescribed for ARTI, the study did not assess indications or appropriateness of antibiotics. Also, with the current design of the study, it is not possible to assess antibiotic prescription in vaccinated children and compare it with their unvaccinated peers, lack of comparison could be a potential distorter of an association between the influenza vaccination and the reduced antibiotic prescription. The observed decline in antibiotic use could be resulting from the changes in prescribing practices in this population rather than vaccination alone as discussed above though it remains a reasonable explanation and warrants further studies to confirm an association. Finally, the study examined the temporal association between influenza vaccination coverage in children and antibiotic prescription rates in the same population rather than direct patient-level antibiotic use data in the same children. While this ecological approach has limitations, it allows measurements of the indirect benefits of influenza vaccination to unvaccinated children due to potentially reduced disease prevalence and transmission.

In summary, there is a temporal association between the increase in influenza vaccination coverage and the decline in ambulatory antibiotic prescription rates in children in South Carolina. Achieving the CDC’s set target influenza vaccination coverage of 70% of the population may be associated with a greater decline in ambulatory antibiotic prescription rates in children in the future.

ACKNOWLEDGEMENTS

The authors thank advisory board members of the Antimicrobial Stewardship Collaborative of South Carolina for facilitating the conduct of this study.