Epidemics and Outbreaks in Ophthalmology
Kirk R. Wilhelmus
Abrupt waves of illness have afflicted mankind since the start of civilization. The first eyewitness account of a plague involving the eyes dates from ancient Greece.1 In the Middle Ages, a diaspora of ocular inflammatory disease spread far and wide during the Crusades. International trade and war globalized ocular infections so that a blinding scourge could flare wherever and whenever people congregated.2 Ophthalmology crystallized as a specialty in the 19th century to confront contagious eye diseases.3 Infection, though quelled as a leading cause of vision loss (Table 44.1), still provokes suffering and chronic disability and can emerge without warning.
BASICS OF EPIDEMICS
An epidemic is the increased frequency of a disease in a group of people. An outbreak is a localized upsurge. Apprehending the context of communicable and transmissible infections is built on fact-finding and problemsolving. Just as an investigative journalist looks at who, where, and when, an epidemiological inquiry begins with the fundamentals of person, place, and time (Fig. 44.1). The ophthalmologist-detective uses these descriptive statistics to figure out how and why an epidemic or outbreak arose and what to do about it.
Person
An epidemic infection is the episodic, often sudden increase in the number of people developing an infectious disease at the same place and time. Epidemics are most easily recognized when many individuals become ill within a short interval.
A common-source epidemic, also called a holomiantic infection, is an outbreak due to a shared intermediary. A point-source epidemic follows simultaneous exposure to a single contaminated vehicle. The intensity of epidemicity is proportional to the number of new cases and the area of distribution. A pandemic is an epidemic that affects many people over a wide region.
An epidemic may erupt visibly within a group of persons. If, on the other hand, the incidence is low and cases are widely distributed, an emerging outbreak can be obscured. Recognizing a relative rise from the noisy background of an infrequent disease requires collaborative effort (Fig. 44.2). Knowing the baseline frequency and distribution of a condition is a prerequisite for being able to discern an epidemic.
An endemic infection is one that persistently endures within a population or region. The intensity of endemicity is classified by its sustained prevalence. A hyperendemic infection disproportionately affects a particular community. A holoendemic condition touches nearly everyone in a region. Less prevalent disease may be described as being mesoendemic or, if erratically present, hypoendemic.
TABLE 44-1 Transitions in the Epidemiology of Infectious Blindness | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
 FIG. 44.1 Communicable and transmissible infections are related to person, place, and time in an epidemic-endemic continuum. |
Characterizing the endemicity level depends on the particular disease. The threshold for defining hyperendemic trachoma may be 20% while hyperendemic onchocerciasis generally has a prevalence over 40%. A disease that affects relatively few people or has minor consequences can be unapparent, but when prevalence and severity rise, eye disease becomes a societal dilemma. In a worst-case scenario, if infection-related blindness approaches 5%, the survival of an entire community is in jeopardy.4
Place
The environment provides multiple opportunities for exposure to infectious agents and can be responsible for regional differences in ocular disease.5 For example, the prevalence of microbial keratitis varies by location,6, 7, 8, 9, 10 and 11 partly because of differences in geography and biometeorology (Table 44.2) that shape socioeconomics. Filamentous fungal keratitis, in particular, is relatively more common in developing countries,12 at tropical latitudes (Fig. 44.3), and with windy, humid, and warmer climates.13, 14
Medical mapping can reveal clustering (Fig. 44.4) and suggest links between infected individuals and their habitat. An atlas that plots infection hotspots gives insight into the ecology of communicable eye diseases and helps to allocate resources to places most in need.15 The application of a geographic information system to public health informatics has the potential to reveal predisposing factors and to track the effect of an intervention.
 FIG. 44.2 Multicenter experience of fungal keratitis shows peak during 2006 linked to a contact lens cleaning solution, United States, 2001-2007. (Reproduced with permission from Gower EW, Keay LJ, Oechsler RA, et al. Trends in fungal keratitis in the United States, 2001 to 2007. Ophthalmology. 2010;117:2263-2267.) |
TABLE 44-2 Population-Based Epidemiology of Ulcerative Keratitis | |||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||
 FIG. 44.3 Correlation of latitude with relative proportion of fungal infection among microbial keratitis. (Reproduced with permission from Leck AK, Thomas PA, Hagan M, et al. Aetiology of suppurative corneal ulcers in Ghana and south India, and epidemiology of fungal keratitis. Br J Ophthalmol. 2002;86:1211-1215.) |
 FIG. 44.4 Trachoma in an African village, showing three clusters involving neighboring households of extended families, Tanzania, 2000-2002. (Reproduced with permission from Polack SR, Solomon AW, Alexander ND, et al. The household distribution of trachoma in a Tanzanian village: an application of GIS to the study of trachoma. Trans R Soc Trop Med Hyg. 2005;99:218-225.) |
Time
The recognition of a secular trend for an infrequent infection is challenging. Changes over time may be related to microbial mutations, altered vectors, or variations in the environment. Several eye infections exhibit a seasonal ebb and flow, reflecting the influence of weather on microbial growth and on human activity. For example, in agricultural communities the harvest period brings an upsurge of corneal injuries and secondary infection.16, 16a Global warming portends an ascending spiral of ocular infectious disease.17
The public health importance of an infectious disease is based on its prevalence and incidence. Prevalence is the proportion of people who have clinical or microbiologic evidence of infection at a given time. Incidence is the rate of transmission or acquisition of infection during a given interval. For acute disease, prevalence is approximately equal to incidence times the duration of an infection.
ORIGINS OF AN EPIDEMIC
A cluster of ocular infections encompasses a web of causation that links an infectious agent with the eye. Exogenous infection begins when an infectious agent is transferred from an external source to the eye as its target or portal of entry. An endogenous infection seeds the eye from the bloodstream. An infecting microorganism must avoid, subvert, or circumvent the body’s defenses and then adhere, survive, and invade. Aft er exposure to an infectious agent, an interrelated set of factors promote, potentiate, and precipitate infection (Table 44.3).
Infectiousness
Commensal microorganisms colonize the outer eye without causing infection or an evident immune response. Latency is an asymptomatic condition that can reactivate and cause contagious disease. Pathogens are microbes capable of invasion and of creating illness. Several lines of evidence define a causal relationship between a pathogenic microorganism and a disease.18, 19
Ocular pathogens fall into two broad categories. Primary pathogens regularly cause disease in otherwise healthy eyes after injury or exposure. Opportunistic pathogens cause disease when the eye’s physiologic or immune defenses are compromised. Many ocular microbial isolates fall between these extremes. Incidental pathogens are microorganisms residing in environmental niches or the normal flora that cause disease if directly inoculated in a sufficiently large amount to overcome the eye’s clearance mechanisms.
Microbiologists measure infectiousness by the minimum number of infectious units needed to produce disease. For ocular pathogens, at least 100 to 1,000 colonyforming or plaque-forming units are generally needed to generate experimental ocular infection, but the requisite infectious dose can be 10-fold to 1,000-fold higher depending on the microorganism and the host.
Epidemiologists measure infectiousness by counting the number of new cases arising among susceptible people. Epidemic spread is influenced by an infectious agent’s capability to propagate by serial transfer from one person to the next.
Approximately a fourth of recognized human pathogens are known to infect the eye. Of species considered to be emerging or reemerging, nearly half are ocular pathogens. Some are capable of inducing neoplasms (Table 44.4).
Host immunity determines susceptibility for several communicable diseases. Implementation of an immunization policy can suppress endemic infection and blunt an epidemic. Smallpox, once the foremost cause of infective blindness, was eradicated by vaccination. More recently, mumps essentially disappeared as an ocular disease after widespread immunization (Fig. 44.5), and Haemophilus influenzae preseptal cellulitis declined after Hib vaccine was introduced.20 The immunity of a population also partially explains an upswing of one infectious strain as another wanes (Fig. 44.6).
TABLE 44-3 Risk Factors for Contact Lens-Related Bacterial Keratitis | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
TABLE 44-4 Ocular Neoplasia Linked with Infection | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
 FIG. 44.5 Mumps incidence and reported cases of mumps keratitis, United States, 1922-1992. (Reproduced with permission from Sutphin JE. Mumps keratitis. Ophthalmol Clin North Am. 1994;7:557-566.) |
 FIG. 44.6 Changing epidemiology of adenovirus serotypes, Scotland, 1981-1991. (Reproduced by permission of the BMJ Publishing Group. O’Donnell B, McCruden EAB, Desselberger U. Molecular epidemiology of adenovirus conjunctivitis in Glasgow 1981-1991. Eye. 1993;7:8-14.) |
TABLE 44-5 Routes of Transmission | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Route of Transmission
The theory of contagion, once thought to involve an invisible miasma, was reworked in the late 19th century. Pasteur’s germ theory explained how communicable diseases could pass between people, through the air, on a vehicle, or by a vector (Table 44.5). Following a chain of infection, outbreaks arise when contaminated objects or risky practices are repeatedly shared.
Hands and Eyes
The eye examination involves close contact. Patients shedding viruses and bacteria in respiratory and ocular secretions are potentially infectious.21 Some viruses and prions can enter the body by way of the eye (Table 44.6). An infectious agent making landfall in an ophthalmologist’s office can die out or run rampant (Fig. 44.7). On the other side of the slit-lamp, physicians harbor microorganisms.22 An ophthalmologist with conjunctivitis is a smoldering tinderbox apt to spark an office outbreak.23
TABLE 44-6 Systemic Infections Transferable through the Eye | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Water and Air
Viruses and other microorganisms, including protozoa, that survive in water can, with appropriate exposure, infect the outer eye. Maps of two clusters of Acanthamoeba keratitis, for example, implicated waterborne spread.24, 25 Water-related transmission has also occurred with microsporidial keratitis,26 whose emergence in Asia (Fig. 44.8) seems to be associated with the rainy season.27, 28 Airborne transmission of amoebic cysts and fungal spores may also be responsible for cases of ocular infection.
Vehicles and Fomites
When proprietary eyedrops came into widespread use in the mid-20th century, some contaminated solutions led to corneal infection. In-use contamination of eyedropper bottles by bacteria and viruses still occurs, but common-source outbreaks from regulated products are rare. During the early 21st century, clusters of ocular infections occurred from nonsterile solutions dispensed by compounding pharmacies. In developing countries, traditional medications and folk remedies are liable to contamination and potentially harmful to the injured eye.
Ophthalmic devices offer another route oftransmission. In the early years of production, contact lenses, contaminated by Pseudomonas aeruginosa or filamentous fungi, led to clusters of ocular infection. Outbreaks of atypical infections among contact lens wearers have been superimposed on a background rate of microbial keratitis.29 Analogously, clusters of endophthalmitis after cataract surgery have been linked to contaminated intraocular lenses.
The outpatient clinic, hospital ward, intensive-care unit, and surgical center offer many opportunities for transmitting infections to patients. Inanimate articles, often referred to as fomites, are capable of retaining viable microorganisms. Contact instruments, multiuse eye drops, and surfaces in the waiting room are sources of transmission.30 The regular disinfection of tonometers and other reusable objects that come into contact with ophthalmic patients aims to reduce health care-associated infection.31 Safety concerns motivate a push toward single-use solutions, disposable instruments, and regulatory oversight of ophthalmic products.
Vectors and Animals
Arthropods capable of carrying ocular pathogens include flies that are able to transmit microorganisms directly to the ocular surface.32 Larval nematodes can be transferred by the bite of a fly in areas endemic for onchocerciasis. Other insects are intrinsically toxic: individuals have developed dermatoblepharitis from injured beetles oozing a potent vesicant.
 FIG. 44.7 Outbreak of adenovirus conjunctivitis at a healthcare-facility instigated by a community epidemic, Chicago, 1985-1986. (Reproduced by permission of the University of Chicago Press. Warren D, Nelson KE, Farrar JA, et al. A large outbreak of epidemic keratoconjunctivitis: problems in controlling nosocomial spread. J Infect Dis. 1989;160:938-943.) |
 FIG. 44.8 Emergence of microsporidial keratitis at two eye centers (SNEC and CGH), Singapore, 2004-2007. (Reproduced with permission from Loh RS, Chan CML, Ti SE, et al. Emerging prevalence of microsporidial keratitis in Singapore: epidemiology, clinical features, and management. Ophthalmology. 2009;116:2348-2353.) |
An epizootic is an epidemic in animals. Ophthalmic epizootics include Moraxella bovis conjunctivitis in cattle, Pasteurella multocida conjunctivitis in rabbits, and herpesvirus keratoconjunctivitis in dogs. An epornithic is a bird epidemic, such as Mycoplasma gallisepticum conjunctivitis in finches. A zoonosis is an infection that spreads from vertebrates to humans. Zoonoses that affect the eye include bartonellosis, toxoplasmosis, histoplasmosis, psittacosis, and baylisascariasis.
One epornithic that is a zoonosis of ophthalmic importance is Newcastle disease.33 This paramyxovirus infection affects chicken and turkey flocks where the virus spreads between birds by direct contact, through airborne and waterborne routes, and via fecal-oral transmission. Human conjunctival infection can be acquired by handtoeye autoinoculation and from one family member to another. Clusters of conjunctivitis among poultry workers have also occurred with avian influenza viruses.
PROGRESSION OF AN EPIDEMIC
Incubation Period
The incubation period is the interval between the initial exposure of a sufficient dose of an infective agent and the onset of illness (Fig. 44.9). During the incubation period, the host prepares to react, and the pathogen proliferates—although the person is not necessarily infectious.34 The duration of the incubation period depends on the pathogen (Table 44.7), the inoculum size, and individual susceptibility. In an outbreak, a histogram of patients’ incubation periods usually produces a slightly skewed, bell-shaped curve along the time axis. For viral causes of conjunctivitis, 90% of cases begin between one half and two times the median incubation period.35
Attack Rate
Infectivity is the relative ability of an infectious agent to induce infection. Infectivity depends on the characteristics and dose of the agent, the means of exposure, and the host’s susceptibility. The greater the infectivity, the higher the opportunity would be that an epidemic could begin. The probability of transmission can be estimated by counting how many susceptible contacts of an infectious individual get the infection.
 FIG. 44.9 Time course of a communicable ocular infection, showing periods of infectivity and of disease. The latent period is the time between exposure to the infectious agent and the onset of infectiousness. The incubation period is often longer than the latent period. Occasionally, the symptomatic period may be very brief or unapparent. Th e period of recovery may result in structural alteration and functional disability. |
TABLE 44-7 Incubation Periods of Ocular Surface Infections | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
|
The attack rate is the percentage of people developing infection in the initial wave after exposure. The figure is calculated as the number of new cases of a disease occurring within the infection’s incubation period divided by the number of possible contacts who could be exposed during the first generation time. The attack rate describes the prevalence of disease transmission in a population unit such as a household, school, or clinic.
Coxsackievirus A24, enterovirus 70, and adenovirus types 8, 19, and 37 are very contagious and have attack rates that sometimes exceed 20%. In one acute outbreak of adenovirus conjunctivitis, the attack rate was 10% at one childcare center and 26% at another.36 During another outbreak in a nursing home, the attack rate of adenovirus conjunctivitis was 13% among employees and 36% among patients.37 Crowding and frequent contact with infected people or contaminated fomites raise the attack rate.
Individual vulnerability also influences the probability of transmission. Protective immunity, whether naturally acquired or induced by immunization, impedes transmission and dampens epidemic cycles. For example, an epidemic of congenital rubella syndrome was curtailed by a national vaccine policy (Fig. 44.10).
Severity
Pathogenicity is the ability of an infectious agent to induce overt disease through infection. Pathogenicity can be measured by the proportion of infected people who develop sufficiently severe clinical disease to require medical care. Thus, the illness rate may differ from the attack rate. Pathogenicity is related to the invasiveness of the agent and to the host’s proinflammatory response.
Virulence describes the consequences of pathogenicity through the assessment of ocular or visual damage caused by the microbe-host interaction. The gradient of ocular infection can vary from mild redness to sight-threatening complications. Virulence is described by the proportion of those having a severe level of clinical disease among those infected. In ophthalmology harmfulness can be judged by the frequency of a serious endpoint such as perforation, phthisis, or blindness. The outcome assessments of visual loss38 and quality of life39 help to quantify the long-term impact of ocular infectious disease.
 FIG. 44.10 Rubella and congenital rubella syndrome (CRS), United States, 1966-2004. Tick marks indicate official vaccination recommendations. (Reproduced from Centers for Disease Control and Prevention. Achievements in public health: elimination of rubella and congenital rubella syndrome—United States, 1969-2004. MMWR Morb Mortal Wkly Rep. 2005;54:279-282.) |
TABLE 44-8 Bioterrorism Agents and the Eye | ||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
DETECTION OF AN EPIDEMIC
An epidemic is often brought to light when a physician notices an unusual condition or realizes that a series of patients is presenting with the same problem. The index case is the first infected patient to be recognized. Some infections are so rare, such as anthrax of the eyelid, that even a single incident calls attention to a worrisome threat (Table 44.8). A sentinel health event is a condition that signals a possible challenge to public health.
 FIG. 44.11 Bacterial conjunctivitis in a healthcare facility, Buffalo, 1990-1992. Threshold levels were calculated from the previous 2 years. (Reproduced with permission from Mylotte JM. Analysis of infection control surveillance data in a long-term care facility: use of threshold testing. Infect Control Hosp Epidemiol. 1996;17:101-107.) |
An emerging outbreak may be revealed as an uptick in health statistics that merge the collective experience of practicing ophthalmologists. For prevalent conditions, information on a disease’s endemic prevalence allows the epidemiologist to determine the threshold that delimits an epidemic.40 An example of threshold assessment is the monitoring done by a hospital’s infectioncontrol office (Fig. 44.11). Establishing a baseline reveals clustering.
Clinical Surveillance
Surveillance is the ongoing observation of disease in a population. Surveillance distinguishes seasonal pulses and tracks longitudinal trends that help to detect clusters in time or place which might signal a possible outbreak (Fig. 44.12). Surgical centers and eye banks have a vested interest in performing quality-assurance surveillance since consecutive cases of the same postsurgical infection can be a tip-off that there may be a contaminated source.
A registry aims to record all cases of a particular disease within a particular population. National registries have been set up to monitor the outcomes of cataract surgery,41 keratoplasty,42, 43 and 44 and ocular trauma45 as well as for reporting specific diseases such as postoperative endophthalmitis or contact lens-related keratitis.46 Maintaining the quality and momentum of a large database requires the investment and incentive of stakeholders.47
Notification systems enable the oversight of ocular adverse events associated with drugs, devices, and vaccines.48, 49, 50 and 51 Electronic medical data held by governmental bureaus,52, 53 and 54 insurance companies, and health maintenance organizations55 can be mined for patterns of ocular inflammatory disease. Even surrogate measures, such as antibiotic sales56 or hits at a Web site providing health information, offer ways to deduce the changing frequency of disease (Fig. 44.13).
Cross-sectional sampling gives a snapshot of an infection’s prevalence at a point in time. For example, a 19thcentury population-based survey suggested regional differences of acute conjunctivitis.57 If an entire population can be listed, a fraction can be chosen systematically, as when random-digit telephone dialing was used to estimate the prevalence of infectious complications in the wearing of contact lenses.58 Because of the difficulty and expense of contacting a lot of patients, surveys of ocular infections often sample ophthalmologists within a country59,60 or a specialty society.61, 62
 FIG. 44.12 Relative increase in Acanthamoeba keratitis in comparison with prior experience, Canada, 1988-2011. (Reproduced with permission from Fraser MN, Wong Q, Shah L, et al. Characteristics of an Acanthamoeba keratitis outbreak in British Columbia between 2003 and 2007. Ophthalmology. 2012;119:1120-1125.) |
 FIG. 44.13 Ophthalmic antibiotic sales during an epidemic of coxsackievirus A24 conjunctivitis, French Guiana, 2002-2003. (Reproduced with permission from Dussart P, Cartet G, Huguet P, et al. Outbreak of acute hemorrhagic conjunctivitis in French Guiana and West Indies caused by coxsackievirus A24 variant: phylogenetic analysis reveals Asian import. J Med Virol. 2005;75:559-565.) |
Laboratory Surveillance
Diagnostic laboratories that process specimens enable surveillance of microbial isolates and help clinicians to fulfill the obligation to report selected infectious diseases to a public health bureau—such as the National Notifi able Diseases Surveillance System or the European Surveillance System.63, 64 Combining data from a network of microbiology laboratories gives a comprehensive picture of ocular infectious diseases that might not be otherwise apparent.65 The pooled surveillance of hospital laboratories suggests that eye infections account for about 0.5% of health care-associated infections, an average of one case per 40,000 hospital discharges.66 Laboratory surveillance also complements clinical reporting when monitoring the course of an epidemic (Fig. 44.14).
Molecular epidemiology aids in the search for a chain of infection.67 An initial step is to distinguish a medley of infections caused by various species from a cluster due to a single strain (Table 44.9). The microbial identification of unusual species may need to be confirmed by a reference center.68
The immunology laboratory affords information on the presence and titer of specific antibodies. Besides revealing an individual’s exposure history, serological data can indicate a population’s level of susceptibility. The level of humoral immunity may reflect a region’s vulnerability to epidemic spread of a communicable ocular disease. A high degree of seroprevalence may decrease the risk while an epidemic might be triggered if herd immunity falls below a critical threshold. Although dependent on population size and infectivity duration, population immunity partially explains the recurring periodicity of some infections.
 FIG. 44.14 Active surveillance (laboratory reports) and passive surveillance (clinician reports) during a national upsurge in Fusarium keratitis, United States, 2004-2006. (Reproduced with permission from Grant GB, Fridkin S, Chang DC, et al. Postrecall surveillance following a multistate Fusarium keratitis outbreak, 2004-2006. JAMA. 2007;298:2867-2868.) |
TABLE 44-9 Microbiological Methods of Molecular Epidemiology | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
DYNAMICS OF AN EPIDEMIC
Epidemic Curve
An epidemic curve plots the number of cases over time (Fig. 44.15) and typically follows a log-normal distribution. Graphing the incubation periods of a common-source outbreak also represents the epidemic curve (Fig. 44.16). A swift upsurge may reflect the exposure of several people to a communal source or shared practice that needs to be recognized and interrupted as soon as possible.
 FIG. 44.15 Epidemic curve of 50,000 cases of acute hemorrhagic conjunctivitis, Puerto Rico, 2003. (Reproduced with permission from Centers for Disease Control and Prevention. Acute hemorrhagic conjunctivitis outbreak caused by coxsackievirus A24—Puerto Rico, 2003. MMWR Morb Mortal Wkly Rep. 2004;53:632-634.) |
The downward slope of the epidemic curve may be skewed by waves of additional cases as the transmission of a communicable infection unfolds. For example, the epidemic curve of an outbreak of adenovirus conjunctivitis often gives the impression of falling gradually when it is actually composed of the superimposition of subsequent peaks and troughs as the virus continues its diffusion into new groups (Fig. 44.17). The graph can be further complicated when an epidemic coexists with sporadic cases. Subsequent generations of infectious outbreaks tend to occur in repeated but overlapping cycles separated by one generation time. The numbers affected in each generation describe the epidemic chain.
 FIG. 44.16 Graph of incubation periods in an office outbreak of adenovirus keratoconjunctivitis, Houston, 1985. The incubation period was estimated as the time between eye examination and the symptomatic onset of conjunctivitis. (Unpublished data.) |
Epidemic Modeling
Infectious diseases can be mathematically modeled based on the SIR model: Susceptible (S) → Infected (I) → Recovered (R). The Kermack-McKendrick equations describe the rate at which these proportions change over time, assuming that everyone in a group comes into contact with everybody else and that recovered individuals are resistant to reinfection.
The rate of new cases at the start of an epidemic can be estimated by β • C • S • I, where β is the transmission probability, C is the contact rate, S is the proportion susceptible, and I is the proportion infected and infectious. For example, in a school of 250 susceptible students where 10 children present with adenovirus conjunctivitis, the attack rate is 25%, and students have 100 contacts per day then (0.25)(100)(0.96)(0.04) = 0.96. Thus, one new case per day would occur at the start of this outbreak. The number of infected children would be predicted to rise day by day unless something were done. This model can be extended to those eye diseases in which infected individuals recover but become susceptible to reinfection.69
 FIG. 44.17 Epidemic spread of adenovirus conjunctivitis from campers to household contacts and the community, Washington, DC, 1954. (Reproduced with permission from the American Medical Association. Bell JA, Rowe WP, Engler JI, et al. Pharyngoconjunctival fever: epidemiological studies of a recently recognized disease entity. JAMA. 1955;157:1083-1092.) |
Reproductive Rate
The average number of people within a susceptible population who are infected by an infectious patient is called the basic reproductive rate (R0). R0 is calculated by β • C • D, where β is the attack rate, C is the average number of new contacts per unit time, and D is the average duration of infectiousness. For example, if a viral strain had a risk of transmission of 15%, an infected person had 12 close contacts per week, and viral conjunctivitis was infectious for 1 week, then R0 would be (0.15)(12)(1) = 1.8. Thus, at the beginning of this epidemic, every case would infect, on average, nearly two others.
The epidemic threshold is the number of susceptible persons necessary to be infected in order for an epidemic to arise from an index case. In general, if R0 < 1, the infection will not be sustained. If R0 = 1, then the infection remains endemic. But if R0 > 1, an epidemic is likely. Thus, a model of an epidemic can predict the number of cases, guide needed resources, and evaluate the effect of control measures.70, 71
Infection control aims to reduce the rate of infection and to lower the reproductive rate below the epidemic threshold. Transmissibility (β) might be decreased by enforced handwashing, no-touch examination techniques, or prophylactic antibiotics. The rate of contact (C) could be reduced by segregating patients or temporarily suspending infected employees. The duration of infectivity (D) might be shortened by early, effective therapy.
The rate of transmission would be reduced if some individuals are not susceptible because of previous exposure or immunization. Herd immunity can be expressed as (R0 – 1)/R0. The effective reproductive number (R′) would be R0 – pR0, where p is the proportion of individuals who are immune.
INVESTIGATION OF AN EPIDEMIC OR OUTBREAK
A cluster of apparently related infections justifies a forensic investigation (Table 44.10). The guiding objective is to identify contributing causes so that efficacious control measures are implemented quickly and capably.72
TABLE 44-10 Checklist of an Outbreak Inquiry | ||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Assessment
The beginning of an outbreak may be masked by community-acquired cases or mistakenly thought to be a random event. On the other side of the coin, overinterpreting the chance occurrence of one or more people having similar signs as a true cluster can lead to an unnecessary and burdensome inquiry. The first steps are to establish a case definition and to apply clinical and microbiological criteria to stratify cases as possible, probable, or confirmed.
Defining an infected case takes advantage of laboratory and clinical evidence filtered through knowledge and experience. When sampling takes place through the normal flora of the tear film, the standard for establishing a pathogen might include the detection of the same isolate from sequential specimens, on separate culture media, or by different techniques. But more than diagnostic tests are required. For example, the number of eyes that become infected after cataract extraction is much less than the number transiently harboring intracameral microorganisms.
Surveillance statistics and census data help to distinguish an epidemic that may be superimposed on endemic cases. The threshold for recognizing a cluster depends on the expected background incidence. Prior knowledge of an infection’s rate—from the literature, a national registry, or institutional records—helps to determine the likelihood of a possible outbreak.73 Postmarketing data and adverse-event reports can aid in detecting a nascent problem.74
Periodic oversight of a center’s pattern of disease can spot nosocomial clusters in the midst of communityacquired cases (Fig. 44.18). Clinics and surgical centers should have an incident management system.75, 76 and 77 The person assigned this responsibility regularly tabulates infectious and other critical events.78 Information can be made available to a disease-specific registry79 or centralized database.80 For uncommon infections for which a baseline incidence can be estimated, statistical charts can indicate when the number of cases exceeds that expected by chance.40, 81, 82
An epidemic aggregates in time, but ascertaining a true cluster of apparently similar cases calls for several criteria.83 An infectious outbreak typically:
Shows a distinctive clustering pattern of infected persons in time and place;
Has a shared route of transmission;
Produces infection selectively among those exposed;
Involves one particular microorganism; and
Is halted by removing the microbial source.
A pseudo-outbreak is a run of apparently similar but unrelated occurrences. A sequence might be due to chance or to misdiagnosis, either by false-positive interpretation of ocular specimens84 or from coincidentally overlapping characteristics.85
Once a cluster of infections is recognized, an investigative team is assembled that optimally includes an epidemiologist, clinician, microbiologist, nurse, and administrator. At regular meetings, the team assigns responsibilities, gathers data, and makes decisions. This multidisciplinary group examines case records and laboratory reports, interviews staff, and inspects facilities in an effort to seek reasons and risk factors for infection.
 FIG. 44.18 Nosocomial and community cases of adenovirus conjunctivitis, Baltimore, 1991. (Reproduced with permission from Gottsch JD. Surveillance and control of epidemic keratoconjunctivitis. Trans Am Ophthalmol Soc. 1996;94:539-587.) |
Reconnaissance
Active investigation of an outbreak begins by documenting the basic variables of an epidemic: person, place, and time. A spreadsheet is constructed of all confirmed and suspected cases. This table catalogues the demographic characteristics of each case, the location where each may have been exposed, and the dates of symptomatic onset and of diagnosis. The database itemizes information on fomites and other vehicles that could have transferred infection. At a health care facility, this includes the equipment, devices, and solutions with serial or batch numbers, as well as the personnel who made patient contact.
An essential task is to characterize the disease. Laboratory criteria should be reviewed to differentiate infection from inadvertent contamination.86 Distinguishing an infectious from a noninfectious etiology may not be straightforward, even when hundreds are affected.87
Identifying the causative microorganism is fundamental to the investigation of an infectious disease. Successful pathogens usually replicate as a clone, so a cluster of ocular infections is recognizable by the phenotypic and genotypic similarity of microorganisms recovered from different patients. Even when the presence of different strains or multiple species excludes a common source, the investigation might still uncover routine procedures that need attention.
Goals of the exploratory phase are to determine the sequence of events and the route of transmission. The investigative team should observe procedures as currently practiced, including patient entry and exit routines. A sociogram or contact matrix can depict the path of spread in the network of transmission.
Analysis
Graphics help to communicate the attributes of person (e.g., age distribution), place (e.g., dot map), and time (e.g., epidemic curve). Plotting the locations where exposure could have occurred, whether operating rooms or patient residences, may disclose environmental links. Comparing genetic analysis of ocular isolates with the characteristics and geographic distribution of infected patients could also reveal routes of dissemination during widespread outbreaks.88
Environmental assessment is particularly relevant for clusters of infection occurring after surgery. The obvious places to look are the preparation area, operating room, and recovery suite. Construction projects in other parts of the hospital and the ventilation system should also be considered as a source of airborne microorganisms.89, 90 Airflow can be tested, and a faulty air-conditioning system should be fixed.91 The use of the facility by other services should be examined and correlated with the cases.
Technical staff should demonstrate that devices used for ophthalmic examination and surgery meet sterility requirements. Equipment maintenance and instrument cleaning should be verified. The time of replacement of reused supplies is checked, and the proper disposal of single-use instruments and solutions evaluated.
Samples can be taken from selected materials and personnel. Instruments and devices, such as a phacoemulsification machine, may need to be separately cultured. Tubing is flushed for microbiological processing because cannulas can harbor microorganisms and allow an apparatus to be persistently contaminated.90, 92
Infection Control
A root-cause analysis should prioritize the critical control points in the clinical process and pinpoint where contamination and exposure could happen.93, 94 By itemizing the various stages of patient flow, apparent problems in routine practice might be found. Even when infections cannot be traced to an identifiable vehicle or route, a systematic investigation may reveal opportunities for improving regular procedures. For example, epidemics of fungal and protozoal infections among contact lens wearers led to the withdrawal of contact lens solutions although a point source of contamination was not established.95
The objective of outbreak control is to avert further cases. Immediate steps of infection control include routine hand washing, regular cleaning and disinfection of contact instruments, and disposal of opened medications and solutions.96 Patients who are possibly infectious may be segregated into a different waiting area and given separate appointment times. Gloves and other personal protective equipment for all examiners may be advised.97
A lack of symptoms does not exclude a person from harboring an ocular pathogen.98 Staff who are infected or colonized with the causative microbe should be furloughed. Operating rooms must adhere to standards of ventilation, disinfection, and sterilization.99 Each procedure should be aseptically separate, among patients and between eyes of the same patient. Prevention begins with hygiene and antisepsis but extends to modifying behavior.
The findings and recommendations of an outbreak investigation should be clearly summarized and disseminated. Premature discontinuation of infection-control measures can allow an outbreak to start anew (Fig. 44.19). Proposals need to be incorporated into an operational policy that integrates practice guidelines.100, 101 Sometimes, a revision of the overall care pathway might be proposed.
 FIG. 44.19 Two consecutive outbreaks of nosocomial adenovirus infection. One ophthalmologist had respiratory infection (1), and another had conjunctivitis (2). Control measures included segregation (3) and gloving (4). When the outbreak was thought to be over (5), gloves were discontinued (6), and eye examinations were revised (7). After a subsequent outbreak began, contact instruments were sterilized (8), and gloves were reinstituted (9). (Reproduced with permission from Faden H, Wynn RJ, Campagna L, et al. Outbreak of adenovirus type 30 in a neonatal intensive-care unit. J Pediatr. 2005;146:523-527.) |
PREVALENT ENDEMIC INFECTIONS
Trachoma
Ocular strains of Chlamydia trachomatis dispersed long ago to become endemic.102 Epidemic spread through Europe and Asia during the 19th century contributed to the creation of eye hospitals. By the 21st century, prevalence declined with improved socioeconomic conditions and sanitation (Fig. 44.20). However, trachoma persists in Africa, Asia, the Middle East, and some regions of Australia and Latin America.102a Twenty-one million people are estimated to have active trachoma, 7 million have trachomatous trichiasis, and 2.2 million are visually impaired.
 FIG. 44.20 Worldwide trend in the prevalence of trachoma (dashed line) and trachomatous trichiasis (solid line). (Reproduced with permission from World Health Organization. Global WHO alliance for the elimination of blinding trachoma by 2020. Wkly Epidemiol Rec. 2012;87:161-168.) |
Onchocerciasis
River blindness clusters near flowing waterways where the insect vector breeds. Most of the 18 million people estimated to have onchocerciasis live in Africa, but endemic pockets exist in the eastern Mediterranean and Latin America. Approximately 750,000 people are visually disabled from postinflammatory scarring.
 FIG. 44.21 Trend of new HIV infections. (Reproduced with permission from UN Joint Programme on HIV/AIDS. Global report: UNAIDS report on the global AIDS epidemic 2010. Geneva: World Health Organization; 2010:16.) |
HIV/AIDS
The acquired immunodeficiency syndrome emerged in the 20th century, and a global pandemic peaked at the end of the millennium (Fig. 44.21). The human immunodeficiency virus currently infects 34 million people. The risk of ocular opportunistic infection is 50% or more among those with a low CD4+cell count. Between 400,000 and 2.4 million people are blind from cytomegalovirus retinitis.103
 FIG. 44.22 School outbreak of acute conjunctivitis, London, 1804. (Data extracted from Macgregor P. An account of an ophthalmia which prevailed in the Royal Military Asylum, in 1804. Trans Soc Improv Med Chir Knowledge. 1812;3:30-64.) |
EPIDEMICS OF COMMUNICABLE EYE INFECTION
Adenovirus Conjunctivitis
Several 19th-century epidemics of ocular inflammation in cities of Europe and Asia had epidemiologic characteristics of viral conjunctivitis (Fig. 44.22). One outbreak in central Europe brought about the initial description of superficial punctate keratoconjunctivitis in 1889. Additional European and Asian outbreaks arose during the early 20th century. Epidemic conjunctivitis reached the western United States in 1939 and then quickly spread eastward.104, 105 Hundreds were affected at naval shipyards, and acute keratoconjunctivitis began to strike military posts and to alarm the general population (Fig. 44.23).
 FIG. 44.23 Extremely large outbreaks of epidemic keratoconjunctivitis. (Data extracted from Ford E, Nelson KE, Warren D. Epidemiology of epidemic keratoconjunctivitis. Epidemiol Rev. 1987;9:244-261.) |
The disease, long known to erupt in large clusters, was termed epidemic keratoconjunctivitis in 1942, the same year a virus was first detected. By 1955, adenovirus type 8 was established as a cause. Type 19 was discovered in a 1973 outbreak that swept westward across North America.106 Adenovirus type 37 caused European outbreaks in 1976 and spread to the United States the following year. Several adenovirus strains cause sporadic episodes of conjunctivitis, but many large outbreaks of epidemic keratoconjunctivitis are still due to species D serotypes.107
Attack rates are high within families108 and wherever people come into close contact (Table 44.11). Thus, epidemic keratoconjunctivitis targets settings where peer groups congregate: refugee settlements, summer camps, swimming pools, child-care centers, schools, factories, nursing homes, intensive-care units, hospital wards, emergency rooms, and doctors’ offices. The incidence can reach 25% or higher when a common source is involved.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
