Pink Eye: Do Antibiotics Matter?

SYNOPSIS: Acute infectious conjunctivitis, commonly referred to as pink eye, is common in children and is caused by bacteria more often than by viruses. Nonetheless, neither the clinical course of uncomplicated cases nor the spread of infection to peers is significantly altered by treatment with topical antibiotics or by exclusion of infected children from daycare and school settings.

SOURCE: Frost HM, Jenkins TC, Meece JC, et al. Etiology and outcomes of acute infectious conjunctivitis in children. J Pediatr 2024;Oct 18:114368. doi: 10.1016/j.jpeds.2024.114368. [Online ahead of print].

The term “pink eye” refers to acute infectious conjunctivitis. It is a common pediatric condition, affecting approximately 12% of children each year. In adults, the cause of pink eye is usually viral, especially adenovirus. Bacterial causes seem more common in children, particularly Haemophilus influenzae. Tests to detect pathogens in children with conjunctivitis usually have involved cultures, with polymerase chain reaction (PCR) being used more recently. The associations between symptoms and chronically carried colonizing bacteria are unclear. In addition, the microbiologic epidemiology of pediatric conjunctivitis likely is evolving with widespread immunization against organisms such as Streptococcus pneumoniae that previously were more common.

Thus, researchers in Colorado and Wisconsin conducted a prospective multi-centered case-control study of acute pediatric conjunctivitis from 2019 through 2023. PCR analysis was done with conjunctival swabs in 194 acutely symptomatic children aged 6 months to 17 years and in 196 age-matched asymptomatic control children. Children with underlying eye disease, allergic conjunctivitis, foreign bodies or ocular trauma, vision loss, nasolacrimal duct obstruction, or recent orbital or periorbital cellulitis were excluded from the study. Children with a history of antibiotic use within the preceding 30 days also were excluded from the study. Control study subjects were enrolled within 30 days of enrollment of the associated case subjects due to possible seasonal variations in colonizing microbes.

PCR testing was done on conjunctival swabs for nine families of viruses and four bacterial pathogens. Standard statistical tests were used to test associations between various groups of data.

The median age of study subjects was 5 years; most were white and Hispanic. Of subjects, 70% were in daycare or school. Enrollment was mostly after the onset of the COVID-19 pandemic, and 46% of subjects were enrolled during the winter season.

Most patients with conjunctivitis (76%) and most control subjects (57%) had bacteria identified on conjunctival swab samples. However, most bacterial species were similarly detected in cases and controls; only H. influenzae was more likely in cases (62%) than in controls (29%), with an odds ratio of 4.59 (95% confidence interval, 2.86-7.37). Viruses were identified in only 5% of children with conjunctivitis. Overall, 23% of study subjects had no virus and no bacteria detected.

Clinical symptoms that were significantly predictive of H. influenzae being detected included purulent eye discharge, runny nose, and nasal congestion. Clear eye drainage was predictive of not having H. influenzae; purulent discharge was noted in 77% of children with H. influenzae and 53% of children without H. influenzae.

Treatment with a topical ophthalmic antibiotic was prescribed for 54% of children with conjunctivitis (68% in Wisconsin, 52% in Colorado; prior to the study, there had been an intervention in Colorado supporting more judicious use of antibiotics for conjunctivitis). A non-fluoroquinolone antimicrobial agent was used for five to seven days in 90% of cases.

By day 5 after diagnosis, 96% of children showed symptomatic improvement, and 85% showed clinical resolution. Rates of resolution and degree of missed childcare/school did not vary based on whether bacteria was identified or which organism was identified. Neither improvement and resolution of symptoms nor missed school/childcare days (two to three) was related to whether an antibiotic was provided. Parents of 20% of children who received an antibiotic reported an adverse reaction. (Details of the adverse reactions were not specified, but 8% of children treated with an antibiotic required an additional medical care visit because of an adverse reaction, as compared to less than 5% of all children requiring a follow-up visit due to symptoms of conjunctivitis.)

The authors noted that no constellation of symptoms was adequate to either predict the presence of bacterial vs. viral conjunctivitis or to differentiate those with and without H. influenzae infection. They also highlighted the benign nature of acute conjunctivitis and the lack of alteration of the clinical course with topical antibiotic therapy. Specifically, they noted that non-quinolone antibiotics provide no clinical benefit and might cause harm when used in the management of acute childhood conjunctivitis.

Commentary

Acute infectious conjunctivitis is common in children. Treatment with topical antibiotics has been widespread with the thought of reducing the spread of infection to others and shortening symptomatic illness. These new data suggest that antibiotic treatment does not significantly alter the clinical course and is associated with bothersome side effects in a non-negligible number of children. Whether antibiotics alter spread of the infection was not a topic of this new study. However, in another recent study of children with conjunctivitis, the infection spread to a household member in 12% of cases regardless of whether the index patient received a topical antibiotic; this transmission rate is lower than that of other common childhood infections that do not prompt exclusion from daycare and school.¹

In a straightforward way, these new data suggest that it would be sensible to withhold antibiotic treatment from children with acute conjunctivitis who do not have serious systemic illness. At the same time, we should remember that this study did not deal with neonates in whom Gonococcus and Chlamydia can cause conjunctivitis and for whom antimicrobial therapy can be very helpful. In addition, this new study did not deal with allergic conjunctivitis, which can occur co-incidentally to or separate from infectious conjunctivitis.

While most adult conjunctivitis is caused by viruses, this study shows that H. influenzae is the most commonly identified infectious cause of childhood conjunctivitis.² At the same time, though, these same germs can colonize eyes without causing symptoms. Further, despite an extensive search for the etiologies of conjunctivitis in this study, nearly one-fourth of patients with conjunctivitis did not carry an identifiable bacterial or viral pathogen.

Some of the past fear of pink eye related to adenoviral disease with epidemic keratoconjunctivitis. This is a severe form of conjunctivitis with flaming red eyes that can transiently impede normal vision and which can remain symptomatic for up to three weeks. This is a more severe disease than the common viral or bacterial conjunctivitis that causes mildly pink eye injection and usually lasts for less than a week. For the severe forms, of course, care to reduce exposures and spread is warranted.

Practically, a real issue is that conjunctivitis often is used as a reason to exclude children from daycare and school. Some doctors feel pressured to provide a topical antibiotic in hopes of getting the child back to regular daycare and school activities more quickly. However, as shown in this new study, antibiotic treatment had no effect on symptoms or on academic activity.

These new data show that while purulent discharge is more common with bacterial than viral conjunctivitis, the presence/absence of purulent discharge is neither sensitive nor specific enough to confirm an etiologic diagnosis. Similarly, while palpable pre-auricular lymph nodes are more common with viral than bacterial conjunctivitis, this finding is not pathognomonic. PCR testing can show the presence of potential pathogens but does not prove that the identified pathogens actually are causing the symptoms.

While a precise etiologic diagnosis of uncomplicated conjunctivitis is not always feasible in clinical settings, the etiologic diagnosis also is not necessary for clinical care. As shown in this new paper, antibiotics likely cause more side effects and lead to potentially more medical care than does mere supportive care, without any additional benefit. In another recent report involving tens of thousands of pediatric patients with acute conjunctivitis in the United States, the use of topical antibiotics did not result in any change in subsequent medical care visits for the ocular symptoms.³

These new data are provocative but also reassuring. It is not usually necessary to prove an etiologic diagnosis of uncomplicated conjunctivitis in children, and antimicrobial therapy is unlikely to alter the clinical course. Rather, effort could be made to allow children to attend daycare and school if they have simple conjunctivitis without systemic symptoms.

Perhaps it would be reasonable to allow children with uncomplicated conjunctivitis who are not on specific treatment to continue routine daycare participation and school attendance. Pink, runny eyes in otherwise well children should not necessarily require any more medical care or alteration of daily activities than does pink nasal mucosa with a runny nose in an otherwise well child. In fact, a recent modeling study related to children with non-severe conjunctivitis showed that withholding antibiotics and allowing continued daycare/school attendance would save more than $750 million and avoid 1,600,000 ophthalmic antibiotic courses per year in the United States.⁴

Philip R. Fischer, MD, DTM&H, is Professor of Pediatrics, Department of Pediatric and Adolescent Medicine, Mayo Clinic, Rochester, MN.

References

1. Kiernan ME, Keith A, Stein AB, et al. Secondary household transmission of conjunctivitis in children. J AAPOS 2024;28:103953.
2. Johnson D, Liu D, Simel D. Does this patient with acute infectious conjunctivitis have a bacterial infection?: The rational clinical examination systematic review. JAMA 2022;327:2231-2237.
3. Shapiro DJ, Geanacopoulos AT, Subramanian SV, et al. Antibiotic treatment and health care use in children and adolescents with conjunctivitis. JAMA Ophthalmol 2024;142:779-780.
4. Frost HM, Stein AB, Keith A, Jenkins TC. Cost-effectiveness of pediatric conjunctivitis management and return to childcare and school strategies: A comparative study. J Pediatric Infect Dis Soc 2024;13:341-348.