Skip to main content
PMC home page
JAMA Network logoLink to JAMA Network
. 2021 Nov 18;140(1):43–49. doi: 10.1001/jamaophthalmol.2021.4852

Association of Public Health Measures During the COVID-19 Pandemic With the Incidence of Infectious Conjunctivitis

Juan M Lavista Ferres 1, Thomas Meirick 2, Whitney Lomazow 2, Cecilia S Lee 2, Aaron Y Lee 2,✉, Michele D Lee 2
PMCID: PMC8603236  PMID: 34792555

Key Points

Question

What were the associations of COVID-19–associated public health measures with the epidemiology of infectious conjunctivitis?

Findings

A model involving publicly available smartphone mobility data was able to show the difference in actual behavior compared with expected trends based on data from previous years and included analysis of noninfectious eye conditions for comparison. The adoption of COVID-19–associated public health measures was associated with a 34% decrease in conjunctivitis-associated search activity and a 37% decrease in emergency department encounters for infectious conjunctivitis.

Meaning

These findings show that search metrics in conjunction with mobility data may provide quantifiable metrics of the associations of public health interventions with transmissible diseases.

This comparative effectiveness study evaluates whether internet search interest and emergency department visits for infectious conjunctivitis were associated with public health interventions adopted during the COVID-19 pandemic.

Abstract

Importance

Infectious conjunctivitis is highly transmissible and a public health concern. While mitigation strategies have been successful on a local level, population-wide decreases in spread are rare.

Objective

To evaluate whether internet search interest and emergency department visits for infectious conjunctivitis were associated with public health interventions adopted during the COVID-19 pandemic.

Design, Setting, and Participants

Internet search data from the US and emergency department data from a single academic center in the US were used in this study. Publicly available smartphone mobility data were temporally aligned to quantify social distancing. Internet search term trends for nonallergic conjunctivitis, corneal abrasions, and posterior vitreous detachments were obtained. Additionally, all patients who presented to a single emergency department from February 2015 to February 2021 were included in a review. Physician notes for emergency department visits at a single academic center with the same diagnoses were extracted. Causal inference was performed using a bayesian structural time-series model. Data were compared from before and after April 2020, when the US Centers for Disease Control and Prevention recommended members of the public wear masks, stay at least 6 feet from others who did not reside in the same home, avoid crowds, and quarantine if experiencing flulike symptoms or exposure to persons with COVID-19 symptoms.

Exposures

Symptoms of or interest in conjunctivitis in the context of the COVID-19 pandemic.

Main Outcome and Measures

The hypothesis was that there would be a decrease in internet search interest and emergency department visits for infectious conjunctivitis after the adaptation of public health measures targeted to curb COVID-19.

Results

A total of 1156 emergency department encounters with a diagnosis of conjunctivitis were noted from January 2015 to February 2021. Emergency department encounters for nonallergic conjunctivitis decreased by 37.3% (95% CI, −12.9% to −60.6%; P < .001). In contrast, encounters for corneal abrasion (1.1% [95% CI, −29.3% to 29.1%]; P = .47) and posterior vitreous detachments (7.9% [95% CI, −46.9% to 66.6%]; P = .39) remained stable after adjusting for total emergency department encounters. Search interest in conjunctivitis decreased by 34.2% (95% CI, −30.6% to −37.6%; P < .001) after widespread implementation of public health interventions to mitigate COVID-19.

Conclusions and Relevance

Public health interventions, such as social distancing, increased emphasis on hygiene, and travel restrictions during the COVID-19 pandemic, were associated with decreased search interest in nonallergic conjunctivitis and conjunctivitis-associated emergency department encounters. Mobility data may provide novel metrics of social distancing. These data provide evidence of a sustained population-wide decrease in infectious conjunctivitis.

Introduction

Acute conjunctivitis, commonly referred to as pink eye, is highly contagious and extremely common.1 It is estimated that conjunctivitis accounts for 1% to 2% of primary care visits in the US and that most cases are viral in causative mechanism.2 Acute viral conjunctivitis is transmitted through close personal contact, fomites, droplets in the air, and poor hand hygiene.3,4 Despite infection control programs and measures, acute viral conjunctivitis continues to be a public health concern and a substantial economic burden.5,6,7 The true incidence of acute conjunctivitis is unknown because many cases may be self-diagnosed and conjunctivitis is not a reportable disease.

One way to supplement disease surveillance is the use of web-based surveillance. Search query data have been applied to many studies since 2002 with varying levels of success. An early project, Google Flu Trends, initially demonstrated potential but failed to correctly anticipate flu trends in 2011-2013.8 Unfortunately, Google Flu Trends used a purely algorithmic approach that searched among 50 million query terms without guidance from human experts.8,9 More recently, Google search trends have been retrospectively validated in early detection of conjunctivitis outbreaks, with more than 80% of identified outbreaks detected before the official outbreak issuance date.10

On March 11, 2020, the World Health Organization declared the COVID-19 outbreak a worldwide pandemic.11 Around the world, multiple countries went into lockdown, and many states and territories throughout the US instituted sheltering orders, which severely limited travel and extraneous social interaction. On April 3, 2020, the US Centers for Disease Control and Prevention set out guidelines to “[w]ear a mask,” “[s]tay at least 6 feet (about 2 arm lengths) from others who don’t live with you,” and “[a]void crowds.”12 Any person with flulike symptoms or exposure to symptomatic persons was instructed to quarantine for 14 days.

We hypothesized that imposed social distancing, emphasis on hygiene, and travel restrictions during the COVID-19 pandemic altered the dynamics of infectious conjunctivitis. In this study, we used causal inference bayesian models to evaluate the potential association of the COVID-19 pandemic with infectious conjunctivitis in the US.

Methods

This study was conducted in accordance with the Declaration of Helsinki. This database study was approved by the University of Washington Institutional Review Board. Informed consent was not required because this research involved no more than minimal risk to the participants.

Our primary outcome of interest was infectious conjunctivitis. Therefore, we excluded any cases of allergic conjunctivitis from all study data sets, as detailed below.

Google Search and Mobility Data

Google trends data (publicly available at https://trends.google.com/) were aggregated on February 11, 2021, for the period between February 21, 2016, and February 6, 2021, for conjunctivitis, corneal abrasions, floaters, and photopsias. The 3 control terms were used because they represent commonly searched but nontransmissible eye problems not associated with COVID-19, social distancing, or decreased mobility. For conjunctivitis, the top associated queries were “pink eye,” “conjunctivitis,” and “pink eye symptoms.” For corneal abrasions, the top associated queries were “corneal abrasion,” “eye abrasion,” and “cornea abrasion.” For the ocular symptom of floaters, the top associated queries were “floater in eye,” “eye floaters,” and “floater in the eye.” For photopsias as a search topic, the top associated queries were “flashes of light,” “light flashes,” and “seeing stars.” Queries were normalized to compare the frequency of queries on conjunctivitis and other ophthalmic conditions. Because of a known transient increase in searches regarding the potential COVID-19–associated conjunctivitis from a US vice presidential candidate, we decided a priori to exclude the November 2020 data from our analyses.

The community mobility reports were developed by Google as a response to public health officials asking for data to better understand the COVID-19 pandemic (available publicly at https://www.google.com/covid19/mobility/). The data are collected from geolocation data from mobile devices that run Google Maps service and anonymized and aggregated using differential privacy at different levels, including data from city or state and country levels. Previous work demonstrated Google mobility data are associated with the effective reproductive number (the mean number of persons infected by an individual with an infection) for COVID-19.13 Data were downloaded on February 2, 2021, and included data from February 15, 2020, to January 31, 2021. For our study, we used country-level data for US and Australia, which was aggregated at a weekly level. We compared the changes in queries with the mobility data. Given that conjunctivitis has incubation and contagious periods, we then computed Pearson correlations between both mobility and time series at different lags and report the lag that maximized the correlation.

Emergency Department Data

Electronic health record data from the University of Washington were text mined for clinical impression and diagnoses made by the treating physicians for emergency department (ED) encounters during February 2015 to February 2021. The full text of the physician notes was extracted from the clinical enterprise data warehouse and text mined for diagnoses and impressions of interest. We included all diagnoses of “conjunctivitis”; however, “allergic conjunctivitis” was excluded. As a negative control, monthly diagnoses of corneal abrasions and posterior vitreous detachments (PVDs) were extracted. The total number of distinct new patients presenting to the ED each month was calculated.

Search data and ED data were analyzed using a counterfactual control method14 with the objective of building a counterfactual model of what would have happened if there were no lockdowns or changes in behavior. A bayesian structural time series15 is a generalization of difference-in-differences approach to time-series analyses. The methods model the counterfactual of a time series observed before and after the initial social distancing and provide a bayesian time-series estimate of the outcome. The method constructs a counterfactual control combining a set of candidate’s time series in the pre–social distancing period and a seasonal component. The model then computes the posterior distribution of the counterfactual time series after March 11, 2020 (the date of the World Health Organization’s announcement of a global pandemic), and the difference between the observation and the counterfactual time series is a bayesian posterior distribution for the causal outcome. We include the directed acyclic graph of our modeling as eFigure 1 in the Supplement. Pearson correlation coefficients were used to measure correlation. All statistical analyses were performed with Python version 3.8.12 (Python Software Foundation; https://python.org/) and the package CausalImpact, version 1.2.15 Reported P values are based on a 1-sided hypothesis test; P values were not adjusted for multiple comparisons. A P value less than .05 was considered statistically significant. An open-source repository with the analyses used in this manuscript are provided as a GitHub repository (https://github.com/jlavista/conjunctivitis).

Results

The overall proportion of search interests on conjunctivitis were similar between January 2016 to December 2019 (mean [range] search interest, 77 [71-80] total normalized queries per year; possible range, 0-100), while the proportion during 2020 were lower (mean search interest, 59 total normalized queries per year) (Figure 1A). A consistent trend of lower search interests during summer months (July, August, and September; mean [range], 60.5 [58-62] total normalized queries) and higher search interests during winter months (January, February, and March; mean [range], 82.2 [80-84] total normalized queries) was shown during 2016 through 2019. In contrast, the search interest in conjunctivitis abruptly decreased in March 2020 and remained low throughout the year (mean [range], 45 [38-49] total normalized queries; with the exception for the following outlier), except for a peak during November 2020 (peaking at 100 total normalized queries), which coincided with the US vice presidential debate. No substantial differences in the proportion of search interests for terms about corneal abrasion, floaters, or photopsia occurred between 2016 and 2020 (Figure 1B-D).

Figure 1. Seasonal Time Series of Google Search Trends.

Figure 1.

Open in a new tab

Comparison of conjunctivitis search queries (A) with other frequently searched ophthalmic conditions (B-D) during 2016 through 2020. Normalized search interest searched against the week of the year. The shaded area corresponds to the beginning of the COVID-19 pandemic in March 2020.

To examine the temporal association between social distancing and the change in conjunctivitis search interest, we used community cell-phone mobility data to quantify the decrease in mobility of the US population (Figure 2). Both time series were mildly correlated (r = 0.52) without temporal correction. The time interval that maximized the correlation in the drop of search interest with social distancing was 3 weeks (r = 0.92; Figure 3). Using searches for corneal abrasion, floaters, and photopsia, we created a counterfactual control using a seasonally corrected, bayesian time-series model (Figure 4A). After the widespread adoption of social distancing in March 2020, we found a decrease of 34.2% (95% CI, −30.6% to −37.6%; P < .001) in conjunctivitis-associated search activity.

Figure 2. Correlation of Reduction of Mobility and Conjunctivitis.

Figure 2.

Open in a new tab

Smartphone mobility index is plotted from February 15, 2020, to June 15, 2020, with the change in the search interest for conjunctivitis.

Figure 3. Correlation of Search Interest With Varying Lag Intervals.

Figure 3.

Open in a new tab

The correlation of query interest for conjunctivitis with smartphone mobility index was measured at different time lag intervals in weeks.

Figure 4. Bayesian Causal Inference Modeling for the Effect of Social Distancing.

Figure 4.

Open in a new tab

A, Searches over time for conjunctivitis modeled using the counterfactual controls of searches for corneal abrasion, floaters, and photopsias. The top section shows the predicted (orange line, shaded area as 95% CI) with the actual search interest (dark blue line). The vertical line represents the start of the pandemic. The bottom section shows the cumulative effect with 95% CIs over time after the start of the pandemic. Bayesian modeling shown for conjunctivitis (B), posterior vitreous detachment (C), and corneal abrasion (D) visits to the emergency department (ED), controlled for the total emergency visit volume.

The rate of ED encounters for conjunctivitis, PVD, and corneal abrasions was examined by text-mining physician notes. A total of 1156 encounters with a diagnosis of conjunctivitis were noted from January 2015 to February 2021. A mean (SD) of 17.1 (4.9) ED encounters for conjunctivitis per month were noted prior to March 2020, compared with 8.4 (5.2) ED encounters per month after COVID-19–associated public health measures began. While there was a decline in total ED encounters, bayesian time-series modeling showed that even after controlling for the decrease in total ED encounters, conjunctivitis declined 37.3% (95% CI, −12.9% to −60.6%; P < .001; Figure 4B) during the pandemic. Conversely, diagnoses of PVD (increased 7.9% [95% CI, −46.9% to 66.6%]; P = .39) or corneal abrasion (declined 1.12% [95% CI, −29.3% to 29.1%]; P = .47) did not show a decline using the same modeling (Figure 4C and D).

To examine the reproducibility of the association of social distancing with conjunctivitis in a phase-shifted setting, we examined data from Australia as an English-speaking country in the southern hemisphere (eFigure 2 in the Supplement). During 2016 to 2019, a seasonal trend of lower search interests occurred during February and March (mean [range], 46 [42-54] total normalized queries) and higher search interests during August and September (mean [range],73 [64-83] total normalized queries), directly opposite of data from the US. Conjunctivitis search queries in Australia also showed a similar decline of 40.2% after widespread adaptation of COVID-19–associated public health measures (95% CI, −32.6% to −47.8%; P < .001).

Discussion

Shortly after implementation of COVID-19 pandemic–associated infection control measures in March 2020, online searches about conjunctivitis decreased by approximately 34%. This trend was corroborated by a decrease in the number of patients presenting to our ED for treatment of conjunctivitis. However, search data and ED encounters for noncommunicable ophthalmic conditions did not significantly change during the pandemic.

Most infectious conjunctivitis cases are viral and caused by adenovirus.16 Viral conjunctivitis is spread through direct contact with persons who are infected or indirect contact via contaminated fomites, which are materials contaminated with bodily secretions and/or fluids.17 Past initiatives have investigated handwashing for infection control; however, 46% of patients with culture-positive adenovirus conjunctivitis were also found to have virus on their hands, and thus handwashing in itself was ineffective at reliably removing the virus from contaminated hands.4,18 In addition, respiratory droplets can persist for up to several weeks on plastic surfaces.3,17 A successful nosocomial infection prevention strategy used at Moorfields Eye Hospital primarily focused on isolating patients with suspected adenoviral conjunctivitis into designated waiting rooms and minimizing time spent in a clinic by these patients.19 Implementing these distancing measures decreased their nosocomial infection rate from 48.4% of new adenoviral conjunctivitis cases to 22.7% in the first year and 3.4% in the 2 years following. Another successful infection-control strategy was used in Taipei, Taiwan, where text messages sent by the government encouraged parents to keep children with conjunctivitis at home.20 The increase in isolation of individuals with infections led to a relatively shortened epidemic. Similarly, the combination of social distancing, travel restrictions, and increased emphasis on hygiene during the COVID-19 pandemic likely led to a decrease in viral transmission.

The decrease in internet search interest did not occur immediately after COVID-19 restrictions began, but instead approximately 2 to 3 weeks later. This time lag may suggest that the decrease in the incidence of conjunctivitis cases in this study were from viral causes, adenovirus in particular. The incubation period for adenovirus conjunctivitis is typically thought to be 5 to 12 days, although certain types of adenovirus have incubation periods of more than 2 weeks.21 Once a patient becomes symptomatic, they typically remain so for roughly 2 weeks.7 The 2-week to 3-week delay in decreased search interest would coincide with the end of an incubation period for a patient infected just prior to the wide adaptation of COVID-19 mitigation measures. Additionally, this period would correlate with the timing in which we would expect to observe the potential impact of public health measures on the outbreak rate of conjunctivitis.

Another reason why our results may be reflective of a decrease in viral conjunctivitis (such as adenovirus) is the seasonality observed in our data set. Adenoviral conjunctivitis has been shown to cause outbreaks in a seasonal manner, with higher incidence typically occurring during winter and spring months.22,23,24,25 Data from online searches and social media have previously reflected this seasonality and been shown to anticipate and identify epidemic outbreaks.23,26,27 The data (both search data and number of ED encounters) demonstrate a similar seasonality in years prior to the pandemic, suggesting that our findings likely reflect true disease incidence. Phase-shifted trends between the US and Australia also support the reliability of search data in reflecting the overall incidence rate of conjunctivitis that varies with seasonality.

A recent report by Deiner et al26 reviewed recent search data with regard to COVID-19–associated ocular signs. In this study, the authors found a similar significant decrease in searches for conjunctivitis and pink eye during the pandemic relative to the 5 years prior and suggested that this decrease in search volume may be secondary to a decrease in non–COVID-19 infectious conjunctivitis. Our study further supports this hypothesis by controlling the data with both internet searches and ED encounters for noncommunicable ophthalmologic diagnoses by using bayesian modeling. Additionally, it suggests that online search metrics have potential in providing a quantifiable measure of social distancing implementation.

Limitations

There are several limitations to our study. While we are able to show a decrease in the number of patients seeking emergency care for conjunctivitis, it is possible patients were less likely to go to the ED because of a fear of contracting COVID-19. We did note a decrease in total ED encounters following the start of COVID-19 restrictions. However, if this were the only contributing factor, we would also expect a similar decrease in other (noncommunicable) disease presentations. Indeed, when we controlled for the total number of ED encounters, we observed a decrease in the number of conjunctivitis-associated encounters, while the number of patients in EDs presenting with corneal abrasions or PVDs remained largely unchanged. In addition, while online search data regarding conjunctivitis decreased considerably, we did not see an equivalent drop in search terms regarding negative control diagnoses.

Additionally, we are unable to attribute the associations we found to any single infectious pathogen. However, the correlation of a decrease in search interest with the incubation period of adenovirus, along with the knowledge that nearly all infectious conjunctivitis cases are secondary to adenovirus, suggest that our results might be driven primarily by adenoviral conjunctivitis. Lastly, it is possible that we were not able to exclude all noninfectious cases of conjunctivitis because of misclassification and/or coding errors; however, these would have biased our results toward the null hypothesis in showing no decrease in cases after the onset of the pandemic.

COVID-19 itself may cause a confounding effect. Infection with COVID-19 has been associated with conjunctivitislike symptoms in 6.6% to 31.6% of patients.28,29 It is possible that this may be underreported because of the reduction of eye care during the pandemic as well as emphasis on more life-threatening clinical signs and symptoms of COVID-19. While the true incidence of ocular symptoms in the setting of infection with COVID-19 appears to be unknown, conjunctivitis searches associated with COVID-19 would bias our results toward the null hypothesis; if anything, there would have been an increase in internet searches. To illustrate this point, there was a transient influx of interest for viral conjunctivitis when former US Vice President Michael Pence appeared to have conjunctival vasodilation on national television on October 7, 2020, leading to online and media speculation that he may have been infected with COVID-19.30 This brief increase in internet searches returned to low levels 1 week later and has remained at a low level for the remainder of 2020.

Conclusions

To our knowledge, this is the first article to illustrate evidence of a sustained population-wide decrease in the rate of infectious conjunctivitis. Public health initiatives may have contributed to a decline in the number of new infections nationally because there was a relative decrease in ED encounters for conjunctivitis during the same period. Search metrics in conjunction with mobility data may provide quantifiable metrics of social distancing implementation in the future and a method of surveillance of infectious conjunctivitis.

Supplement.

eFigure 1. Directed acyclic graph

eFigure 2. Effects of public health measures in Australia

Click here for additional data file. (50.3KB, pdf)

References

  • 1.Ford E, Nelson KE, Warren D. Epidemiology of epidemic keratoconjunctivitis. Epidemiol Rev. 1987;9:244-261. doi: 10.1093/oxfordjournals.epirev.a036304 [DOI] [PubMed] [Google Scholar]
  • 2.Shields T, Sloane PD. A comparison of eye problems in primary care and ophthalmology practices. Fam Med. 1991;23(7):544-546. [PubMed] [Google Scholar]
  • 3.Nauheim RC, Romanowski EG, Araullo-Cruz T, et al. Prolonged recoverability of desiccated adenovirus type 19 from various surfaces. Ophthalmology. 1990;97(11):1450-1453. doi: 10.1016/S0161-6420(90)32389-8 [DOI] [PubMed] [Google Scholar]
  • 4.Azar MJ, Dhaliwal DK, Bower KS, Kowalski RP, Gordon YJ. Possible consequences of shaking hands with your patients with epidemic keratoconjunctivitis. Am J Ophthalmol. 1996;121(6):711-712. doi: 10.1016/S0002-9394(14)70640-3 [DOI] [PubMed] [Google Scholar]
  • 5.Gottsch JD. Surveillance and control of epidemic keratoconjunctivitis. Trans Am Ophthalmol Soc. 1996;94:539-587. [PMC free article] [PubMed] [Google Scholar]
  • 6.Shekhawat NS, Shtein RM, Blachley TS, Stein JD. Antibiotic prescription fills for acute conjunctivitis among enrollees in a large united states managed care network. Ophthalmology. 2017;124(8):1099-1107. doi: 10.1016/j.ophtha.2017.04.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Lee CS, Lee AY, Akileswaran L, et al. ; BAYnovation Study Group . Determinants of outcomes of adenoviral keratoconjunctivitis. Ophthalmology. 2018;125(9):1344-1353. doi: 10.1016/j.ophtha.2018.02.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Lazer D, Kennedy R, King G, Vespignani A. Big data. the parable of Google Flu: traps in big data analysis. Science. 2014;343(6176):1203-1205. doi: 10.1126/science.1248506 [DOI] [PubMed] [Google Scholar]
  • 9.Wojcik S, Bijral AS, Johnston R, et al. Survey data and human computation for improved flu tracking. Nat Commun. 2021;12(1):194. doi: 10.1038/s41467-020-20206-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Deiner MS, McLeod SD, Wong J, Chodosh J, Lietman TM, Porco TC. Google searches and detection of conjunctivitis epidemics worldwide. Ophthalmology. 2019;126(9):1219-1229. doi: 10.1016/j.ophtha.2019.04.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Cucinotta d, vanelli m. WHO declares COVID-19 a pandemic. Acta Biomed. 2020;91(1):157-160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.CDC . How to protect yourself & others. Published February 5, 2021. Accessed February 8, 2021. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/prevention.html
  • 13.Noland RB. Mobility and the effective reproduction rate of COVID-19. J Transp Health. 2021;20:101016. doi: 10.1016/j.jth.2021.101016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Abadie A, Diamond A, Hainmueller J. Synthetic control methods for comparative case studies: estimating the effect of California’s tobacco control program. J Am Stat Assoc 2010; 105: 493-505. doi: 10.1198/jasa.2009.ap08746 [DOI] [Google Scholar]
  • 15.Brodersen KH, Gallusser F, Koehler J, Remy N, Scott SL. Inferring causal impact using Bayesian structural time-series models. Ann Appl Stat 2015; 9: 247-74. doi: 10.1214/14-AOAS788 [DOI] [Google Scholar]
  • 16.Chintakuntlawar AV, Chodosh J. Cellular and tissue architecture of conjunctival membranes in epidemic keratoconjunctivitis. Ocul Immunol Inflamm. 2010;18(5):341-345. doi: 10.3109/09273948.2010.498658 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ganime AC, Carvalho-Costa FA, Santos M, Costa Filho R, Leite JPG, Miagostovich MP. Viability of human adenovirus from hospital fomites. J Med Virol. 2014;86(12):2065-2069. doi: 10.1002/jmv.23907 [DOI] [PubMed] [Google Scholar]
  • 18.Jernigan JA, Lowry BS, Hayden FG, et al. Adenovirus type 8 epidemic keratoconjunctivitis in an eye clinic: risk factors and control. J Infect Dis. 1993;167(6):1307-1313. doi: 10.1093/infdis/167.6.1307 [DOI] [PubMed] [Google Scholar]
  • 19.Dart JKG, El-Amir AN, Maddison T, et al. Identification and control of nosocomial adenovirus keratoconjunctivitis in an ophthalmic department. Br J Ophthalmol. 2009;93(1):18-20. doi: 10.1136/bjo.2007.130112 [DOI] [PubMed] [Google Scholar]
  • 20.Yen M-Y, Wu T-SJ, Chiu AW-H, et al. Taipei’s use of a multi-channel mass risk communication program to rapidly reverse an epidemic of highly communicable disease. PLoS One. 2009;4(11):e7962. doi: 10.1371/journal.pone.0007962 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kimura R, Migita H, Kadonosono K, Uchio E. Is it possible to detect the presence of adenovirus in conjunctiva before the onset of conjunctivitis? Acta Ophthalmol. 2009;87(1):44-47. doi: 10.1111/j.1755-3768.2007.01148.x [DOI] [PubMed] [Google Scholar]
  • 22.Deiner MS, Lietman TM, McLeod SD, Chodosh J, Porco TC. Surveillance tools emerging from search engines and social media data for determining eye disease patterns. JAMA Ophthalmol. 2016;134(9):1024-1030. doi: 10.1001/jamaophthalmol.2016.2267 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Deiner MS, McLeod SD, Chodosh J, et al. Clinical age-specific seasonal conjunctivitis patterns and their online detection in twitter, blog, forum, and comment social media posts. Invest Ophthalmol Vis Sci. 2018;59(2):910-920. doi: 10.1167/iovs.17-22818 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Channa R, Zafar SN, Canner JK, Haring RS, Schneider EB, Friedman DS. Epidemiology of eye-related emergency department visits. JAMA Ophthalmol. 2016;134(3):312-319. doi: 10.1001/jamaophthalmol.2015.5778 [DOI] [PubMed] [Google Scholar]
  • 25.Ramirez DA, Porco TC, Lietman TM, Keenan JD. Epidemiology of conjunctivitis in US emergency departments. JAMA Ophthalmol. 2017;135(10):1119-1121. doi: 10.1001/jamaophthalmol.2017.3319 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Deiner MS, Seitzman GD, McLeod SD, et al. Ocular signs of COVID-19 suggested by internet search term patterns worldwide. Ophthalmology. 2021;128(1):167-169. doi: 10.1016/j.ophtha.2020.06.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Leffler CT, Davenport B, Chan D. Frequency and seasonal variation of ophthalmology-related internet searches. Can J Ophthalmol. 2010;45(3):274-279. doi: 10.3129/i10-022 [DOI] [PubMed] [Google Scholar]
  • 28.Wu P, Duan F, Luo C, et al. Characteristics of ocular findings of patients with coronavirus disease 2019 (COVID-19) in Hubei province, China. JAMA Ophthalmol. 2020;138(5):575-578. doi: 10.1001/jamaophthalmol.2020.1291 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Zhou Y, Duan C, Zeng Y, et al. Ocular Findings and Proportion with Conjunctival SARS-COV-2 in COVID-19 Patients. Ophthalmology. 2020;127(7):982-983. doi: 10.1016/j.ophtha.2020.04.028 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Lee BY. Does Vice President Pence have pink eye? here’s what you may have seen during the debate. Published October 8, 2020. Accessed February 10, 2021. https://www.forbes.com/sites/brucelee/2020/10/08/does-vice-president-pence-have-pink-eye-heres-what-you-may-have-seen-during-the-debate/?sh=53ee47f32fe6

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement.

eFigure 1. Directed acyclic graph

eFigure 2. Effects of public health measures in Australia

Click here for additional data file. (50.3KB, pdf)

Articles from JAMA Ophthalmology are provided here courtesy of American Medical Association

RESOURCES