Cost Savings From a Policy to Diagnose and Prevent Transmission of Adenoviral Conjunctivitis in Employees of a Large Academic Medical Center

Key Points

Question  What is the economic impact of a quality improvement policy to diagnose suspected adenoviral conjunctivitis in medical center employees using polymerase chain reaction (PCR) before furlough?

Findings  In this quality improvement study, a total of 1520 employees underwent conjunctival swabbing for PCR for clinical suspicion of adenoviral conjunctivitis; 130 (8.6%) had positive PCR test results and required a 1- to 2-week furlough. The estimated cost savings from furloughing 130 employees instead of 1520 was more than $3 million during 7 years.

Meaning  These findings suggest that use of accurate diagnostic testing for adenoviral conjunctivitis can result in cost savings from fewer employee furloughs compared with furloughs based on clinical diagnosis, the typical method of diagnosing conjunctivitis.

Importance  Adenoviral conjunctivitis is highly contagious, can be associated with systemic infections, and can cause chronic visual impairment. It accounts for a large proportion of acute conjunctivitis. Outbreaks of epidemic keratoconjunctivitis (EKC) are costly in terms of productivity loss from work furloughs and spread to patients and have resulted in clinic and departmental closures.

Objective  To examine the institutional cost savings of a policy to diagnose adenoviral conjunctivitis and triage and furlough medical center employees with this condition to prevent outbreaks.

Design, Setting, and Participants  This quality improvement study assessed Johns Hopkins Medicine employees with red eye from November 1, 2011, through October 31, 2018, who were triaged at the occupational health clinic whose conditions were diagnosed using polymerase chain reaction (PCR) validated for adenoviral conjunctivitis.

Interventions  Only employees with positive PCR test results were furloughed, with furlough length tailored to subtype (a minimum of 2 weeks for EKC and 1 week otherwise).

Main Outcomes and Measures  Total number of furloughs avoided and cost savings associated with reducing unnecessary furloughs.

Results  Of 2142 employees with red eye, 1520 (71.0%) underwent PCR testing; 130 (8.6%) had positive adenoviral PCR test results, of whom 41 (31.5%) had EKC. Furloughing 130 employees with positive PCR test results vs furloughing all 1520 employees clinically suspected of having adenoviral conjunctivitis represented an estimated annual savings of $442 073, or $3 094 511 during 7 years. The cost of performing PCR on employees suspected of having adenoviral conjunctivitis was 5.0% of the cost associated with furloughing all employees with red eye. No outbreaks occurred.

Conclusions and Relevance  In this quality improvement study, this policy, notable for development and use of PCR for adenoviral conjunctivitis on a large scale, resulted in substantial cost savings from fewer work furloughs compared with the number of employees who would have been furloughed based on clinical diagnosis. These results may provide impetus for policy adoption by other institutions and for development of a rapid, sensitive, and specific diagnostic test for adenoviral conjunctivitis.

Acute conjunctivitis is costly, imposing health, economic, and social burdens. Estimates of the total direct and indirect costs of treating bacterial conjunctivitis in the US in 2005 were $377 million to $857 million in 2007 dollars,¹ but most diagnoses of conjunctivitis are made presumptively. Viral conjunctivitis is estimated to account for 35% to 75% of all cases of acute conjunctivitis.^2,3 Human adenoviruses are believed to account for 65% to 90% of cases of viral conjunctivitis, making them the most common cause of viral conjunctivitis.^3-6

Adenoviral conjunctivitis can range from mild follicular conjunctivitis to highly contagious epidemic keratoconjunctivitis (EKC), which has forced the closure of clinics, hospitals, and care facilities.^7-13 Adenovirus differs from bacterial conjunctivitis: it is highly contagious, EKC can cause chronic vision impairment from corneal scarring and dry eye,¹⁴ no widely accepted diagnostic test exists,¹⁵ and supportive care remains the mainstay given the lack of effective treatment.¹⁶ Although non-EKC genotypes cause a self-limited follicular conjunctivitis without corneal involvement, they can cause severe respiratory infections in vulnerable populations.¹⁷ Screening patients and health care employees for adenoviral conjunctivitis is, therefore, important.^10,18 Issuing unnecessary work furloughs without an accurate diagnosis, however, negatively affects clinical operations and individual employees.

The red eye policy at Johns Hopkins Medicine (JHM), an ongoing multidisciplinary, quality improvement initiative implemented in November 2011,¹⁸ was devised to promote swift triage, accurate diagnosis, and prevention of adenoviral conjunctivitis transmission from employees to patients and other employees. This policy became necessary when the ophthalmology department closed the “red eye room” to which those with possible infectious conjunctivitis had been triaged and examined quickly. In this study, we estimate the institutional cost savings from implementing the policy.

The JHM Institutional Review Board deemed that this quality improvement initiative did not require approval because it did not constitute human subjects research; therefore, informed consent was not required. All data were deidentified. This quality improvement study followed the Standards for Quality Improvement Reporting Excellence (SQUIRE) reporting guideline.

The goal of the policy is to prevent spread of infection and to lessen the risk of compromising clinical operations from granting unnecessary work furloughs, which is accomplished through rapid, accurate diagnostic testing and employee education (Figure). The policy implements polymerase chain reaction (PCR) developed and validated at JHM for adenovirus in conjunctival specimens.^18,19 Both PCR and viral culture are the diagnostic gold standards for adenoviral conjunctivitis.^19,20 As described previously,^18,19 total nucleic acid is isolated from conjunctival specimens (processed volume, 400 μL), and adenovirus DNA is detected by real-time PCR.²¹ The analytical sensitivity (95% detection rate or limit of detection) is 300 copies/mL, and the assay detects at least 16 adenovirus serotypes. When adenoviruses are detected, serotype is determined by nested PCR of the hexon gene hypervariable regions 1 to 6.²²

Quiz Ref IDA cornea specialist and head of infection control for the Department of Ophthalmology at Johns Hopkins University School of Medicine (I.C.K.) taught nurse practitioners in occupational and environmental medicine at the Johns Hopkins Hospital the history, signs, and symptoms consistent with adenoviral conjunctivitis and how to swab conjunctivae. Employees are swabbed if they have had a red eye for fewer than 7 days and have 1 or more of the following: discharge or tearing, contact with a person with red eye, and/or recent cold, cough, or flu-like symptoms. After the employee’s eye is swabbed, the employee is promptly furloughed; PCR results return within a few hours. Those employees with PCR-negative test results are directed to see their own ophthalmologist or are scheduled for an ophthalmology department appointment within 24 hours. Employees with a recent (3-month) history of ophthalmic surgery or ocular inflammation are excluded from swabbing and urgently referred to the ophthalmology department.

The first version of the red eye policy mandated a 2-week work furlough from date of symptom onset for employees with PCR-confirmed adenoviral conjunctivitis. For quality improvement purposes, early in the third year of the policy (January 2014), the laboratory performed retrospective typing from the first 2 years. Two-thirds of employees with PCR-positive results had non-EKC adenoviral conjunctivitis, almost all of which resolved by 7 days. Typing helps rule out EKC, takes a few days, and may predate EKC signs and symptoms. Accordingly, prospective typing was initiated in April 2014. The policy was modified so that employees with positive PCR test results who had non-EKC adenoviral conjunctivitis receive a 1-week work furlough and those with EKC receive a 2-week work furlough.

All employees must return to the occupational health department for clearance to return to work, even if they have a physician’s note. Criteria allowing release are lack of pain, tearing, discharge, and redness.

At the outset, the study team agreed on the data to record for quality improvement: department and job description; history (signs and symptoms, contact with individuals with red eye, and use of antibiotics); dates of onset, presentation to the occupational health department, and PCR swabbing to ascertain elapsed time; external assessment by nurse practitioners; typing results; ophthalmology consultation if obtained; and date of release to return to work. The number of employees the nurse practitioners saw for any red eye issue, the number swabbed for suspected adenoviral conjunctivitis, the number who tested PCR positive, and the number with EKC were tabulated (Table). History, external findings, typing, and clinical course were examined for correlations and monitored for outbreaks. We defined an outbreak as 2 or more employees from the same department seen within 2 weeks in the occupational health department for conjunctivitis caused by the same adenovirus type. We chose 2 weeks to account for incubation time.²³

We obtained the 2018 mean hourly wage for a JHM employee from the human resources departments of constituent entities (Johns Hopkins Health Systems and the Johns Hopkins University School of Medicine).

Because JHM is a not-for-profit enterprise, we assumed the productivity cost of a furlough is the employee’s base salary plus benefits for the furlough period. To determine these costs, we first calculated the mean annual wage of all JHM employees. Because the mean wage differs for workers of Johns Hopkins Health Systems and the Johns Hopkins University School of Medicine, we weighted the 2 wages by the number of employees in each entity to obtain the mean wage. We increased the wage by 32% to account for fringe benefits, then calculated the weekly wage by dividing by 52. Under these assumptions, the weekly wage plus benefits is $1139. To obtain the cost of a furlough, we multiplied this amount by furlough length: 2 weeks for employees with EKC and 1 week for employees without EKC.

To estimate the cost of unnecessary furloughs in the absence of the policy, we assumed that all individuals who received swabbing would have been furloughed for 2 weeks because of high clinical suspicion for adenoviral conjunctivitis. Thus, we multiplied the total number of swabs by the costs of 2 weeks of salary and benefits. Next, we calculated the cost of furloughing only PCR-positive employees. We multiplied the number of PCR-positive employees by the weekly wage and the furlough length. We calculated the number of unnecessary furloughs by subtracting the number of PCR-positive employees from the number of individuals who underwent PCR testing.

Before policy implementation, individuals with suspected adenoviral conjunctivitis did not undergo PCR testing for adenovirus. Thus, PCR testing and typing costs, which the laboratory reported as $114, are incurred as a result of the policy. We multiplied this cost by the number of employees who underwent PCR testing for clinical suspicion of adenoviral conjunctivitis.

We calculated cost savings by subtracting the furlough and PCR testing costs with the policy from the estimated furlough costs without the policy. We calculated the total costs for the 7-year period of November 1, 2011, through October 31, 2018.

Quiz Ref IDDuring 7 years, 2142 individuals with red eye presented to the JHM occupational health clinic (Table). A total of 1520 individuals were suspected of having adenoviral conjunctivitis and underwent swabbing for adenovirus PCR testing. Of these, 130 employees (8.6%) had positive PCR test results for adenoviruses, 41 (31.5%) of whom had EKC. Therefore, 1390 employees with negative PCR test results who would have been furloughed without the policy were able to return to work the next day because they did not have adenoviral conjunctivitis. Most employees with negative PCR results were seen by a JHM comprehensive ophthalmologist within 24 hours. Common conditions diagnosed among the employees with negative PCR results examined at JHM were allergic conjunctivitis, dry eye, hordeolum, and contact lens overuse.

Without the policy, all 1520 individuals suspected of having adenoviral conjunctivitis would have received a 2-week furlough at a cost of $2278 per furlough and a total cost to the institution of almost $3.5 million for the 7-year period. Following the original policy guidelines, only the 130 individuals with positive PCR results would have been furloughed for 2 weeks, at an estimated total institutional cost of $296 140. After subtracting the cost of PCR testing on 1520 individuals suspected of having viral conjunctivitis, the cost savings of the policy with a uniform 2-week furlough for all employees with positive PCR results would be nearly $3.0 million during the 7 years, or a mean of $427 591 per year.

Use of differential furlough lengths based on viral typing would reduce the number of employees who needed a 2-week furlough to 41 individuals, with the remaining 89 employees with positive PCR results requiring a 1-week furlough, for a total tailored furlough cost of $194 769. Basing furlough lengths on typing vs a universal 2-week furlough for all employees with positive PCR test results regardless of EKC status led to a mean additional savings of $15 000 per year. The savings were $3 094 511 during 7 years, or a mean of $442 073 per year.

Quiz Ref IDThe cost of performing PCR on the 1520 employees suspected of having adenoviral conjunctivitis ($173 280) was 5.0% of the cost associated with furloughing all these employees for 2 weeks without this policy ($3.5 million). Since implementation of this red eye policy, no outbreak of adenoviral conjunctivitis has occurred. When the definition of outbreak was extended from 2 to 4 weeks, typing and departments remained disparate, indicating no outbreak.

In this quality improvement study, implementation of a red eye policy at our institution, featuring single-site triage by trained professionals and diagnosis of adenoviral conjunctivitis using PCR validated for adenoviral conjunctivitis, has substantially reduced the number of employees furloughed unnecessarily. During a 7-year period, an estimated 10-fold reduction in work furloughs was noted, representing cost savings to the institution of approximately $3.1 million. Without the policy, 1520 employees with red eye might have received 2-week furloughs. With the policy, however, only 130 employees (8.6%) with diagnostically confirmed adenoviral cases received furloughs. In addition, because of typing and symptoms that confirmed lack of EKC, two-thirds of employees with positive PCR test results for adenoviral conjunctivitis required a 1-week furlough instead of the minimum 2-week furlough for EKC. Institutions that adopt a similar policy may benefit from cost savings as a result of accurate diagnosis and reduction in the number of work furloughs while also preserving clinical operations and maintaining infection control.

Although there are other viral causes of infectious conjunctivitis,²⁴ adenovirus is the most common. It is highly contagious, and EKC can cause chronic visual symptoms from subepithelial corneal infiltrates, symblepharon, or membranes,¹⁴ whereas most viral causes resolve without sequelae. Multiple large EKC outbreaks have occurred in health care settings. In Chicago, 1 outbreak lasted 6 months and involved more than 400 patients, staff, and physicians.¹⁰ In France, an outbreak that involved 41 persons in a long-term care facility cost $29 527, of which $17 184 was for lost productivity because 12 employees became infected and lost 155 productivity-days.⁸ For these reasons, the focus of our policy was adenovirus detection.

To have the greatest impact, diagnostic testing for this policy had to be much more accurate than clinical examination and feature a short turnaround time. Unfortunately, there is no commercially available, highly sensitive diagnostic test for adenoviral conjunctivitis. A point-of-care antigen-based test for adenoviral conjunctivitis is available and produces a result within 10 to 15 minutes. However, reported sensitivity has ranged from 33% (95% CI, 4%-78%) to 90% (no CI reported).^15,25,26 Therefore, despite rapidity, test sensitivity probably would require subsequent PCR tests for many employees with negative test results. Moreover, because it offers no viral typing, epidemiologic information is lacking. In contrast, PCR is highly sensitive (able to detect fragments of adenoviral DNA) and specific. Viral culture and PCR are the gold standards for diagnosis of adenoviral conjunctivitis, but whereas viral culture results take 1 to 2 weeks, PCR results are available in a few hours. Lack of access to a specialized microbiology laboratory and long turnaround time are 2 reasons viral cultures are seldom performed in community or academic practice. For the latter reason, our institution conducted validation tests for PCR for adenoviral conjunctivitis,¹⁷ and this testing serves as a cornerstone of the new red eye policy. At JHM, PCR for adenoviral conjunctivitis is not restricted to employees; patients may also be tested. To our knowledge, PCR for adenoviral conjunctivitis currently is not used at any other institution on a large scale.

Infection control staff monitor the data regularly and would focus efforts if adenoviral cases increase in a department. Moreover, the ongoing nature of this initiative allows modifications when trends are noted. In the policy’s third year, we determined that a 2-week furlough was excessive for most employees. Differential furloughs yielded a further one-third reduction in furlough costs. Because furloughed employees use paid time off that would otherwise go toward vacation or medical leave, most prefer the 1-week furlough.

Although substantial savings in work productivity have been associated with implementing this policy, the original goal was outbreak prevention. In the early 1990s, a large EKC outbreak almost closed JHM’s ophthalmology department. This event led to establishment of a “red eye room” in the ophthalmologic emergency department in which individuals with red eye were sequestered. This room, staffed by ophthalmology residents and an attending ophthalmologist, was a key component of infection control for nearly 20 years. Knowledge of the impact of EKC outbreaks and closure of the ophthalmologic emergency department in 2008 for cost reasons necessitated a new policy. Since inception of the new policy in 2011, no EKC outbreak has occurred at JHM. Although we cannot prove that the policy was preventive, evaluation of health care workers at a single site by trained professionals maximizes adherence to EKC prevention policies and likely decreases health care–associated spread. From billing and coding data, 10% of all visits to our department’s ophthalmologic emergency department were for suspected adenoviral conjunctivitis,¹⁹ comparable with 9.6% of eye-related employee visits to the occupational health clinic for which PCR was performed. Therefore, the same savings likely would be realized if residents used PCR in a red eye room setting.

By one definition, “A disease outbreak is the occurrence of disease cases in excess of normal expectancy.”²⁷ The definition of adenoviral conjunctivitis outbreak used in this study, which incorporated typing by Sanger sequencing, is novel. Polymerase chain reaction followed by whole genome sequencing can determine epidemiologic linkage if 100% identity is shared, as demonstrated in a fatal adenoviral outbreak¹⁷ and an outbreak of coronavirus disease 2019 (COVID-19).²⁸ Without vigilance and outbreak parameters that include department in which the employee works,²³ EKC outbreaks have lasted 1.5 to 6 months.^9,10

Another benefit of this policy is reducing unnecessary prescribing of antibiotics, indiscriminate use of which can lead to change in ocular surface flora and antibiotic resistance.^29,30 Between red eye room closure and this policy implementation, JHM employees sought care from primary care physicians and urgent care centers. Research has indicated that more than 80% of individuals with acute conjunctivitis enrolled in a national managed care network were seen by nonophthalmologists,³¹ whose ophthalmic acumen may not be accurate. Odds of antibiotic prescription fills for acute conjunctivitis are much higher if first diagnosed by an optometrist, internist, or urgent care physician compared with an ophthalmologist, with odds ratios ranging from 1.3 to 3.3.³¹ The cost of antibiotic eye drops prescribed for conjunctivitis currently ranges from $10 to $34 for generic formulations. In this study, we focused specifically on costs associated with lost productivity and thus did not account for infection-related costs incurred by the employee. Inclusion of antibiotic and downstream infection costs likely would increase the positive financial impact of the policy.

This study has limitations. A limitation of the policy is that it requires PCR or tests that require similar expertise, which may limit implementation to academic centers. As academic centers expand beyond their historically urban locations and absorb outlying hospitals, it must be recognized that community hospitals have their own employee health protocols and their laboratory expertise may be limited. Institutions not equipped to perform PCR would need to decide whether to hire such employees or collaborate with other institutions to develop a cost-efficient approach. The JHM-affiliated hospitals not obligated to adhere to the policy have followed the guidelines, with some employees driving more than 50 miles to the Johns Hopkins Hospital for conjunctival swabbing. The results of this analysis are also limited to one institution’s experience. Future research assessing the impact of a similar policy at other institutions and settings would be beneficial.

Quiz Ref IDAt the time this policy was developed, the decision was made to use PCR, but without a US Food and Drug Administration–cleared nucleic acid amplification assay for detecting adenoviruses in conjunctival specimens, centers must develop tests that use noncommercial primers or adapt tests cleared for use with nasopharyngeal samples. Whole genome sequencing has been able to identify new recombinant types and correctly identify existing types³² but is not widely available for conjunctival specimens. Diagnostic testing for adenoviral conjunctivitis may become important given that the viruses that cause COVID-19 and severe acute respiratory syndrome can cause follicular conjunctivitis that resembles adenoviral conjunctivitis,^33,34 with all demonstrating ocular tropism.³⁵ Five-year data indicate that 10 departments accounted for 60% of employee cases of adenoviral conjunctivitis and 75% of EKC cases.³⁶ One-third of employees with adenoviral conjunctivitis did not work in patient care areas. Because such employees are working from home during the COVID-19 pandemic, cost savings would be reduced by one-third during the present time.

The JHM red eye policy prioritizes triage by trained practitioners at one location, accurate diagnosis of adenoviral conjunctivitis using PCR, and infection control while minimizing the number and cost of unnecessary employee work furloughs. No outbreak has occurred since policy inception, and the cost savings in terms of reduction of unnecessary furloughs have been substantial. Algorithms that use objective criteria and that feature continued multidisciplinary stewardship, such as that at JHM, may be adopted elsewhere. Since the 1970s, the US military has vaccinated personnel 17 to 50 years of age against adenoviral types that cause a mild conjunctivitis and lethal respiratory infection; this vaccine is not available to the general public. Because unnecessary work furloughs are costly and adenoviral conjunctivitis can be associated with severe adenoviral respiratory infections,^37-40 this quality improvement initiative may provide impetus for development of a commercially available, affordable, accurate diagnostic test with quick turnaround not requiring specialized laboratories or individual validation.

Accepted for Publication: January 11, 2021.

Published Online: April 1, 2021. doi:10.1001/jamaophthalmol.2021.0150

Corresponding Author: Irene C. Kuo, MD, Wilmer Eye Institute, Department of Ophthalmology, Johns Hopkins University School of Medicine, 4924 Campbell Blvd, Ste 100, Baltimore, MD 21236 (ickuo@jhmi.edu).

Author Contributions: Dr Kuo had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Kuo.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: All authors.

Administrative, technical, or material support: Kuo.

Supervision: Kuo.

Conflict of Interest Disclosures: None reported.

Additional Contributions: Colleen Espinosa, CRNP, Minnie Sanders, CRNP, and other nurse practitioners in the Division of Occupational and Environmental Medicine, Johns Hopkins University School of Medicine, implemented the red eye policy, and Michael Forman, MS, Department of Pathology, Division of Medical Microbiology, Johns Hopkins University School of Medicine, performed all polymerase chain reaction testing. All these individuals are salaried employees and were not additionally compensated for this work.