Investigation of Outbreaks

Erica S. Shenoy

Hospital epidemiologists and infection preventionists spend a substantial portion of their time identifying and investigating possible transmission events, outbreaks, and, sometimes, pseudo-outbreaks. While a portion of infections identified in hospitalized patients are likely from endogenous patient flora,¹ healthcare environments, medical equipment, healthcare personnel (HCP), other patients, and visitors can be reservoirs of pathogens that can be transmitted to susceptible hosts.²^,³^,⁴ The majority of these transmission events likely go undetected except in circumstances where the organism is unusual either by incidence or by susceptibility pattern, in which case the suspicion of a transmission event may be raised and an investigation initiated. Transmission of resistant bacteria is more likely to be detected than that of susceptible bacteria in acute care settings because of the prevalence of use of antibiotics that would reduce detection and suppress amplification of transmitted bacteria.

An outbreak is defined as an increase in the occurrence of an event above a background rate. This definition assumes that a background rate is known. That is the case for some infections and pathogens, but not for all. Most facilities are limited in the resources available to conduct surveillance on all infections and pathogens and instead focus on a subset of infections (ie, deviceassociated infections, specific surgical site infections) or pathogens (ie, methicillin-resistant Staphylococcus aureus [MRSA], vancomycin-resistant Enterococcus [VRE], multidrug-resistant Gram-negative organisms [MDROs], or Legionella pneumophila). For those infections and organisms that are tracked prospectively, identifying an increase in incidence is straightforward, except in instances where incidence may be spread over several hospital units (as may occur with infections transmitted by portable medical equipment), in which case the increase in cases may not be apparent when assessing on a unit-by-unit basis. For infections or pathogens that are not routinely tracked, determining if an outbreak is occurring requires a retrospective review to generate a baseline rate. For an organism that has never been isolated at a facility, even a single case could be considered above baseline. The costs of outbreaks in healthcare facilities, in terms of the personnel and other resources necessary for the investigation and outbreak control, can be substantial.⁵^,⁶

In the course of outbreak investigations, a pseudooutbreak, the “red herrings” of hospital epidemiology, may be identified. Pseudo-outbreaks may appear to be outbreaks based on an increase in occurrence of infections, but the clinical scenarios are at odds with the organisms identified. While these events do not represent a concern for infection transmission or patient safety, they can often highlight sources of contamination that may change practice in clinical or laboratory environments. They are also important to investigate for clinical reasons as patients may be treated for nonexistent infections by clinicians unaware of the pseudo-outbreak.

Regardless of identification of transmission pathways and causes, investigations are resource-intensive and can have substantial impacts on hospital finances through bed/ward closures and reputational impact beyond the resource allocations to the investigation itself.⁷^,⁸^,⁹^,¹⁰ This chapter provides an overview of approaches to identifying and investigating possible transmission events, outbreaks, and pseudo-outbreaks, with a focus on the essential steps, and provides examples of investigations that illustrate the success of these approaches. An abbreviated glossary of terms used throughout is provided (Table 9-1).

IDENTIFICATION OF POTENTIAL TRANSMISSION EVENTS

Routine surveillance for healthcare-associated infections (HAIs) using accepted definitions forms the basis for identifying increases in incidence of infections that define an outbreak. This approach works well for those organisms and types of infections for which there are standard definitions and resources allocated to surveillance. Identification of a potential outbreak, however, may also come from frontline clinicians or laboratorians who bring a cluster of infections to the attention of an infection preventionist or hospital epidemiologist. At that point, a determination must be made as to whether or not the cluster represents an increase from baseline—and thus the baseline infection rate needs to be established.

Some events are so rare as to make the decision to investigate even a single case straightforward. This sentinel event is true of any case of hospital-onset L pneumophila infection or postoperative group A Streptococcus
infection. Both are subject to routine surveillance and a single case warrants further investigation. For infections that are not routinely tracked, single cases can also warrant investigation: for example, if a pathogen with a novel mechanism of resistance in a single patient is identified for the first time, an investigation may be warranted. More challenging are clusters of infections of common organisms that are not routinely tracked at a particular facility. For example, a cluster of infections with methicillin-susceptible S aureus (MSSA) may represent transmission events but may also represent endogenous infection given the high prevalence of colonization among patients. For any cluster, complete reliance on common time (ie, infections occurring close together in time) or space (ie, infections occurring in patients on a specific hospital unit or procedural area) may lead to missed detections given persistence of many organisms in the built environment over time.

TABLE 9-1 Glossary of Terms

*Term*	*Definition*
Hospital-acquired infection (HAI), also referred to as healthcare-associated infection	An HAI is an infection that is acquired in the hospital, either based on standard epidemiological definitions such as CDC NHSN criteria or based on molecular analyses that determine a particular infection was acquired within the hospital, regardless of the timing with respect to hospital admission or healthcare exposure
Transmission	Transmission occurs when a susceptible individual is exposed to an infectious material and develops either infection or colonization with the organism
Mode of transmission	Infectious diseases are transmitted through three modes: contact (direct and indirect), droplet, and airborne
Sources of infectious agents/reservoirs	Sources of infectious agents include healthcare personnel, patients, visitors, and the healthcare environment, inclusive of portable medical devices
Exposure	Exposures occur when a susceptible individual comes in contact with an infectious material through the various routes of transmission
Outbreak/cluster	A cluster or outbreak describes an increase in infections above a baseline rate
Pseudo-outbreak	Pseudo-outbreaks occur when there is an apparent increase in a particular infection; however, there is no associated clinical disease and no transmission has occurred

For all of these reasons, infection preventionists and healthcare epidemiologists can spend resources investigating outbreaks that turn out to be nonoutbreaks and can also fail to detect ongoing transmission events. It is imperative to minimize both overdetection and underdetection in order to deploy limited resources in an effective manner and increase patient safety.

REASONS TO INVESTIGATE A POTENTIAL OUTBREAK

Outbreak investigations are aimed, first and foremost, at stopping any ongoing transmission. Ideally, the root cause is identified through the course of the investigation, and effective strategies are then deployed to prevent future transmissions. Although in some cases transmission ceases and a root cause cannot be initially identified, an investigation may still be considered to avert future events and advance the field of healthcare epidemiology. There are often public health or regulatory reasons to investigate a potential outbreak. These can relate to conditions that are reportable to public health authorities as well as infections that, while not reportable, may be of interest to local, state, and national public health authorities. Through the course of an investigation, a particular facility may contribute to broader understanding of the epidemiology of an important pathogen that can be disseminated through public health agencies.

INSTITUTION OF APPROPRIATE CONTROL MEASURES

Control measures range from routine interventions (ie, enhanced hand hygiene observations and feedback, education, screening, and isolation) to more disruptive interventions (ie, closure of hospital units, sequestration of a product or device, screening of HCP). The choice of control measures is driven by the balance between risk to patients of transmission, the efficacy of various measures, resources available to implement, and potential adverse effects of interventions on patients. It should be noted that often the first round of control measures will be instituted at the very early stages of an investigation and that interventions are often revised throughout the course of an investigation as more data become available.

STAGES OF INVESTIGATION/ANATOMY OF AN INVESTIGATION

Each investigation may have distinct features. A general outline of steps is provided here, with the understanding that (1) each step may be more or less critical depending on the particular investigation, (2) steps may often occur simultaneously, and (3) investigations often have an iterative component that is a natural consequence of outbreak investigation.

Initial Stages of Investigation

Step 1. Ensure that any clinical or surveillance isolates that may be needed are stored for potential future use. Clinical and surveillance isolates obtained in the routine course of care are often discarded after a brief period of storage, and thus, involving the microbiology laboratory in the earliest stage of an investigation is important to ensure isolates that are still available are stored for possible future analysis.

Step 2. Notify appropriate leadership within the institution (ie, leadership of affected service lines, including physicians and nurses, microbiology, quality and safety, hospital administration), define the lead investigator (eg, designated infection preventionist paired with hospital epidemiologist), and assess resource needs (ie, data extraction, chart review, analytics, statistics, microbiological).

Step 3. Fact gathering. Perform preliminary literature review including search of outbreak databases,¹¹ institutional historical data, and confirmation if an outbreak is occurring or has occurred. Literature reviews can assist in guiding the current investigation as they highlight mechanisms of transmission and often provide lessons learned that can be applied to an active investigation.

Step 4. Develop a case definition, and apply it to relevant time frame to enable case finding and development of a line listing to enable chart review to identify potential common exposures and host factors. Cases may need to be subcategorized as confirmed, probable, and possible. Confirmed cases usually require laboratory confirmation. Probable cases usually have both typical clinical features and an epidemiological link. Possible cases have fewer clinical symptoms and weaker epidemiological links.¹²

Commonly included data elements are demographic variables, comorbidities, hospital locations, duration of hospitalization, invasive procedures, devices, medications and products administered, HCP or services involved in care, onset of infection, and details related to infection course. The relevant aspects of chart review should be determined in advance to the extent possible to minimize work effort of chart review and to allow for, when possible, electronic extraction of relevant fields. In many cases, it is useful for the investigative team to create a proposed line listing with variables of interest and to review them after having completed one or two chart reviews. At that time, additional fields may become relevant and be added so that chart review can be done as efficiently as possible on the remaining cases. When feasible, the use of electronic data extraction tools rather than manual review to increase data gathering efficiency and completeness should be employed. The mode of transmission (contact, droplet, airborne) should be considered when determining the parameters of the investigation as well as patient risk factors (ie, immunocompromised, nonimmune) and environmental risk factors (ie, common physical locations).

Step 5. With a case definition established, create an epidemiological curve plotting counts of cases over time, and conduct data analysis to present data in ways that can support hypothesis generation. In reality, epidemiological curves may not be needed if there are very few cases, but they can be useful to provide insights into the mechanism of transmission (Fig. 9-1).

Step 6. Based on the available information to date, develop hypotheses to explain findings. In many cases, actions in the form of control measures will need to be taken before the studies to test these hypotheses are able to be conducted. For investigations involving a small number of cases, formal hypothesis testing is not needed or conducted. In these instances, based on the chart review and line-listing data, common links between cases may be readily identified and interventions implemented to end transmission. If an outbreak is larger than a few cases, if the common links are not readily apparent, or if transmission continues despite control measures, formal hypothesis testing with either case-control or retrospective cohort studies can help to identify the source of infection and risk factors. In case-control studies, the cases are defined by the case definition, and controls are identified through a retrospective review to identify at-risk but unaffected subjects. Such studies are generally straightforward to implement, however suffer from both selection and recall bias. Alternatively, a retrospective cohort study can be conducted in which subjects are identified by the exposure (ie, to a particular device) and then followed through some endpoint to determine whether or not the event/disease occurs. Retrospective cohort studies require that all controls can be identified which is usually true in hospitals and is the preferred design as it is less likely to have selection bias.

FIGURE 9-1 A-C. Sample epidemiological curves. (Adapted from CDC Using an Epi Curve to Determine a Mode of Spread.)

Step 7. Data analysis. This step begins earlier with the epidemiological curve but continues throughout the investigation. At any phase, control measures may be instituted
or modified, and revisions in case definition can result in having to revisit steps in the process.

Follow-up Stages of Investigation

At various time points in the investigation, updates in the case definition may occur, which may require revisiting the various steps outlined above. Ongoing surveillance may be required if not already being conducted, and at some point, a decision will need to be made regarding how long after the suspected end of transmission events such surveillance should continue. Periodic review of control measures for efficacy and continued need should take place. If transmission is ongoing in the face of implemented control measures, more drastic measures, such as ward closure, screening of HCP, and other interventions, may be required. Likewise, if transmission is ongoing and a specific physical space, device, or product is suspected but not confirmed, environmental or product sampling may be entertained. Prior to embarking on such sampling, several aspects need to be determined, including how sampling will be done (with preference for following a validated protocol); where samples will be analyzed as many clinical microbiology laboratories are not equipped to process nonclinical isolates; what will be the policy with respect to sequestration of devices or closing of physical spaces while isolates are being analyzed, if appropriate; and plans for remediation if environmental or product sampling identifies a potential source for the outbreak. Those leading an investigation in which such sampling is contemplated should be aware of these considerations before proceeding.

MICROBIOLOGY LABORATORY ASPECTS OF THE INVESTIGATION

The role of the microbiology laboratory and molecular epidemiology is reviewed elsewhere (see Chapter 5); however, the specific considerations related to the aspects of outbreak investigation are reviewed here. Most importantly, the microbiology laboratory is a partner in outbreak investigations, providing initial species identification and antimicrobial susceptibility testing (AST) of clinical and surveillance isolates that often trigger suspicion for transmission. The microbiology lab also serves an important role in preservation of isolates that may need further analysis to determine the relatedness of organisms as part of an investigation. Those analyses may not be possible in the microbiology laboratory, but the microbiology laboratory is often the conduit between the facility and public health authorities or other external partners who can assist if further isolate analyses, such as whole genome sequencing (WGS), are important to pursue.

ENVIRONMENTAL INVESTIGATION

While the environment should be considered in almost all investigations as a potential reservoir leading to transmission, environmental sampling should only be undertaken in specific situations and should be directed by epidemiologic considerations to increase the likelihood of identifying a source and providing results that are useful in the course of an investigation. Examples of situations where environmental or product sampling is indicated include outbreaks with plausible common product source or outbreaks that are plausibly linked based on epidemiological investigation to air handling equipment or water reservoirs. Sampling of hospital surfaces (eg, work surfaces, mattresses, etc.) should be directed and only performed if the results could plausibly identify an important source of ongoing transmission as these surfaces are commonly contaminated, making the interpretation of results challenging. When available, standard protocols should be used.¹³ Most microbiology laboratories are not resourced to process nonpatient samples, and thus facilities in most cases will need to identify external laboratories with the necessary expertise.

DOCUMENTATION AND REPORTING OUTBREAK INVESTIGATIONS

At the conclusion of each investigation, a summary of the activities conducted, findings, and outcomes should be created. This information can serve as the basis for content to be provided to any relevant public health authorities as well as to provide a historical record for the facility. Documentation related to outbreak investigations can be aided by establishing a standardized investigative template, line-listing template, and protocols related to patient and environmental sampling, which can be employed during the course of the investigation. Reporting of outbreaks to public health authorities is governed by applicable rules and regulations.

Only gold members can continue reading. Log In or Register to continue