Outbreak Investigation: Study Provides Glimpse into National Experience

Outbreak Investigation: Study Provides Glimpse into National Experience

How often are outbreak investigations being conducted at U.S. hospitals? Fairly frequently, according to a group of researchers who also sought to examine the various triggers for investigations, the types of organisms involved, and the control measures being utilized by healthcare facilities. They also discovered that norovirus caused 18.2 percent of all infection outbreaks and 65 percent of ward closures in U.S. hospitals during a two-year period.

By Kelly M. Pyrek

How often are outbreak investigations being conducted at U.S. hospitals? Fairly frequently, according to a group of researchers who also sought to examine the various triggers for investigations, the types of organisms involved, and the control measures being utilized by healthcare facilities. They also discovered that norovirus caused 18.2 percent of all infection outbreaks and 65 percent of ward closures in U.S. hospitals during a two-year period.

A team of researchers from Chartis, Main Line Health System, Lexington Insurance Company, and APIC Consulting Services conducted a national survey of infection preventionists regarding outbreak investigations at their institutions during 2008 and 2009, collecting responses from 822 members of the Association for Professionals in Infection Control and Epidemiology (APIC). Thirty-five percent of the 822 hospitals reported they had investigated at least one outbreak in the previous two years. Four organisms caused nearly 60 percent of the outbreaks: norovirus (18.2 percent), Staphylococcus aureus (17.5 percent), Acinetobacter spp (13.7 percent), and Clostridium difficile (10.3 percent). These results reflect 386 outbreak investigations reported by 289 hospitals over a 24-month period.

Medical/surgical units were the most common location of outbreak investigations (25.7 percent), followed by surgical units (13.9 percent). Nearly one-third (29.2 percent) of outbreaks were reported in a category that included emergency departments, rehabilitation units, long-term acute care hospitals, psychiatric/behavioral health units, and skilled nursing facilities.

According to the results, the average number of confirmed cases per outbreak was 10.1 and the average duration was 58.4 days. Unit closures were reported in 22.6 percent of the cases, causing an average 16.7 bed closures for 8.3 days. Of reported outbreaks, only 132 (52.2 percent) of investigations were reported to an external agency, with just 71 (28.4 percent) involving assistance in the investigation by an external resource. In most states, reporting to the state health department is required and can provide hospitals with expertise to expedite and expand their outbreak investigations.

Lead study author Emily Rhinehart, RN, MPH, CIC, of Chartis Insurance, says the results were not surprising, at least in regard to Clostridium difficile, Acinetobacter and Staphylococcus aureus. "We all had a sense that these organisms were the most common causes of outbreaks, and I think we had a sense that norovirus was occurring more and more frequently, but this study quantifies it. Perhaps the one slight surprise was where the outbreaks of norovirus were occurring. It was previously thought of as a community-acquired infection, or you could say a cruise ship-acquired infection, but to see it happening in  hospitals -- not within the population at risk for other HAIs -- but in the populations that are in behavioral health and rehab settings, really changes the focus."

Statistics on outbreak investigation can be elusive. As Rhinehart, et al. (2012) write, "Reports in the literature describing outbreaks routinely include the epidemiology of the outbreak with details focusing on the patient population (e.g., newborns, mothers under obstetric care, critical care patients, and others), the causative agents, theories on the cause and mode of transmission, and control measures. Such reports continue to be published from hospitals and other healthcare settings all over the world, adding to the knowledge of the various causes of outbreaks, modes of transmission, or involvement of medical devices and interventions as well as the success (or lack of success) of various control measures. However, because the reporting of outbreaks in the literature is voluntary and depends on the motivation and diligence of the professionals involved, we do not have any comprehensive data on the frequency of outbreak investigations or confirmation of their occurrence. Although many outbreaks or clusters of HAI go undetected, many are successfully investigated and documented internally. Some of this experience is shared through scientific posters and papers at professional society meetings or other presentations, but it is unknown how many outbreaks actually occur and are investigated."
Rhinehart says the study was a breakthrough in terms of adding to and enhancing this existing information. "We have benchmarks on the normal incidence of healthcare-acquired infection -- we know the incidence rates down to the birth weights of neonates and their risk for infection, for example, so we already have that stratification. We also have information on infection rates related to the types of surgery. We have all that, so what we didn't have until this study is the frequency of outbreaks."

The researchers report in their study (published in the February 2012 issue of AJIC) that within the final sample, there were 289 hospitals that had initiated 386 outbreak investigations within the previous 24 months. Although some hospitals reported no outbreak investigations, the occurrence in those that did initiate investigations was 1.3 investigations per facility for 24 months. Most hospitals reported one investigation; 70 had two or more, with one hospital reporting the initiation of six investigations, which was the highest frequency, according to Rhinehart, et al (2012).

One of the key take-home messages of the study is preparedness. "I think the most important thing people can do when they read our study is say to themselves, 'This could be us, with our own outbreak next week or next month,' and really become better prepared for a potential outbreak," Rhinehart says, "And they must be prepared not necessarily for an outbreak that is organism specific, because they should always be working to reduce the incidence of organisms such as C. diff. Instead, healthcare facilities must get a handle on antibiotic usage so they don't have C. diff in the first place. C. diff outbreaks go on for days, weeks and months; whereas norovirus outbreaks seem to be able to be controlled and contained in several days or several weeks, C. diff is very elusive. I advise infection preventionists to bring our study to healthcare facility leadership as well as to the pharmacy/therapeutics committee and the infection prevention committee and emphasize that they don't want to be faced with this and they better get control of their antibiotic use."

Outbreak investigations occurred most frequently in medical/surgical intensive care units (ICU) (25.7 percent), with surgical ICUs the focus in 15.9 percent. Specific triggers reported for 344 of the 386 outbreak investigations indicate that identification of an unusual organism was the most frequent trigger (38.1 percent); rate above baseline for a specific HAI site and rate above baseline for a specific unit were also frequent triggers at 27.6 percent and 23.8 percent, respectively.  

"One of the challenges to preparedness is that the definition of an outbreak is not criteria-driven, so it's almost as if you know it when you see it -- it's anything above the usual or unusual," says Rhinehart. "The most common trigger we found was an unusual organism. So I think we would define that as not a bug no one ever saw before, but an organism with an unusual susceptibility pattern. So for example, there were probably resistant organisms about which somebody said, 'Oh, this Acinetobacter is resistant to most of the antibiotics we see that it's sensitive to.'"

The APIC Text of Infection Control and Epidemiology defines an outbreak as "an increase over the expected occurrence of an event," and Rhinehart, et al. (2012) note, "Given that the definition of an outbreak is not clear cut, an outbreak could involve many cases of a well-known illness or could be only a few cases of infection caused by an unusual organism. Thus, the decision to initiate an outbreak investigation can be difficult. When considering a circumstance that may suggest an outbreak is possible, IPs and their teams should increase their sensitivity or index of suspicion for outbreaks, especially when new high-risk procedures have been introduced or an unusual organism has been recently described or identified by the hospital laboratory."

Control measures most often implemented were expanded or enhanced precautions and enhanced environmental cleaning, while other frequent control measures included active surveillance cultures of patients at risk, environmental culturing and patient/staff cohorting. As Rhinehart, et al. (20120) explain, "Although control measures were similar, the frequency of implementation was not. The U.S. hospitals in our study more often applied enhanced precautions and enhanced environmental cleaning, whereas the reports in the literature more often applied patient screening/surveillance and personnel screening/surveillance, with isolation and cohorting applied in 32 percent of the outbreaks."

"The control measures we saw align with what is recommended in the MDRO guidelines, so that was good to see, and quite appropriate," Rhinehart says.

Another important weapon in the arsenal was surveillance. According to the researchers, 99.4 percent of hospitals responding to the survey had surveillance programs, including surveillance for central line-associated bloodstream infection (88.6 percent), ventilator-associated pneumonia (79.7 percent), targeted multidrug-resistant organisms (87.7 percent), and catheter-associated urinary tract infections (80.4 percent). Almost half (49.8 percent) performed surveillance on all inpatient surgeries, with 61.4 percent responding that they performed surveillance of surgical site infection of selected surgical procedures. In 73.6 percent of cases, HAI rates were calculated monthly, and 64.5 percent of respondents utilized control charts in some manner.

"With regard to surveillance, I think outside the spectrum of outbreaks, and now within the context of public reporting, that's another reason that the infection preventionist, first of all, is on the frontline with  reporting of CLABSI and VAP," Rhinehart says. "The healthcare facility's risk manager definitely should be involved because if there is fall-out and the facility administration is either losing reimbursement or public reporting mechanisms indicate they have a record that falls below performance within the U.S. benchmark, not only is the IP going to be called to task to say why we have a rate higher than the benchmark but the risk manager will as well. That speaks to the fact that they are all in this together in an effort called patient safety."

Outbreak preparedness dictates that a plan be in place before the unthinkable happens. In this study, 531 survey respondents (64.6 percent) reported they had a written policy including triggers for the initiation of an outbreak investigation, and Rhinehart, et al. (2012) note, "It is clear that outbreaks of healthcare-associated infections occur with some frequency in hospitals as well as non-acute settings. An infection prevention and control program and its staff should be prepared for all aspects of an outbreak investigation through written policies and procedures as well as communication with internal and external partners. The study authors add, "An infection prevention and control program and its staff should be prepared for all aspects of an outbreak investigation through written policies and procedures as well as communication with internal and external partners. A written policy and procedure should include triggers for an outbreak investigation, roles for investigation (i.e., hospital epidemiologist, IP, laboratory), authority for the implementation of control measures, internal and external communication plans, and requirements for documentation and retention of records and organisms."

Rhinehart, et al. (2012) add further, "Working with internal, and perhaps assisted by external, experts, it is essential to discuss, determine, and document your organizations triggers for outbreak investigation. This is especially true for organisms most commonly associated with outbreaks such as norovirus and Acinetobacter spp and also the target multidrug-resistant organisms recognized by the organization for ongoing monitoring. Triggers should be set for units at highest risk, such as ICUs, but should also be considered for populations not typically included in our surveillance activities (LTAC, rehabilitation and behavioral health units). IPs working as solo professionals without an experienced, dedicated hospital epidemiologist might benefit most from proactive discussions with internal (and external) experts to set triggers and prepare teams for quick appropriate response when outbreaks occur."

"Infection preventionists need the healthcare facility's leadership to ensure that the right stakeholders are at the table to discuss outbreak preparedness," Rhinehart says. "The facility also must make sure they have a written policy on what they would do in case of an outbreak, because there is nothing worse than making it up on a Friday afternoon at 3 p.m. -- I've been there, done that. So you need to have the structure in place, complete with specific to-do lists and a communication plan.  Get organized around who will be responsible for the various tasks. And then  if they find themselves in an outbreak situation, they should be prepared for when the local media finds out. That response has to be very well organized."

Rhinehart, et al. (2012) note, "As this policy is developed, taking the time to educate risk management and patient safety staff in your organization might provide a worthwhile return on investment when outbreaks occur. Risk management should be notified early on when there is the suspicion of an outbreak. In addition, working closely with the marketing and communications function on policy development is also helpful, should an outbreak occur that elicits public interest and media attention. Determining preferred language and terms prior to an outbreak can be helpful in assuring clear internal and external communications. For example, your organization may avoid the word epidemic in favor of terms such as 'cluster' or 'outbreak.' Whereas lay reporters may refer to 'quarantine,' your organization may prefer to refer to 'precautions' as outlined in the CDC Guidelines. The organizations spokesperson should be aware of the pitfalls of dealing with the media and be prepared to utilize the appropriate terms. Those internal staff dealing with patients and their families should also be supported in using appropriate terms for disclosure of infections in an outbreak situation. Finally, the policy on outbreak investigation should include notification of appropriate external agencies such as the state health department. In most states, such reporting is required. IPs should research the requirements of their state and incorporate the requirements into the policy for outbreak investigation. Although some individual hospital epidemiologists and IPs are resistant to including the state health department in an outbreak investigation, state agencies often bring investigative expertise that many IPs may not have resulting in not only an expedited and more thorough investigation but a great learning experience for the IP, physicians and staff. The use of external infection prevention consultants should also be considered. Use of such consultants provides support for the IP as well as a potential learning experience."

Rhinehart says that infection preventionists must continually employ their detective skills and use surveillance tools to identify clusters and outbreaks early in order to intervene and control them effectively, and that open communication with clinicians and laboratory staff is also key to early identification of potential outbreaks.  Rhinehart encourages infection preventionists to be familiar with guidelines and recommendations before an outbreak occurs. "There are plenty of resources to guide people in how to investigate an outbreak," she says. "Consult the papers on particular organisms in the literature, and review the recently updated guideline for the prevention and control of norovirus from the Centers for Disease Control and Prevention (CDC). With regard to other frequent organisms, APIC has a series of excellent elimination guides on C. diff, the elimination guides. Do your routine investigation, and always remember to reach out for assistance, especially if you are faced with something unusual."

Reference: Rhinehart E, Walker S, Murphy D, O'Reilly K and Leeman P. Frequency of outbreak investigations in US hospitals: Results of a national survey of infection preventionists. Am J Infect Control. February 2012.


Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.