infection control and hospital epidemiology february 2007, vol. 28, no. 2

original article

How to Assess Risk of Disease Transmission to Patients When There Is a Failure to Follow Recommended Disinfection and Sterilization Guidelines

William A. Rutala, PhD, MPH; David J. Weber, MD, MPH

background. Disinfection and sterilization are critical components of infection control. Unfortunately, breaches of disinfection and sterilization guidelines are not uncommon. objective. To describe a method for evaluating a potential breach of guidelines for high-level disinfection and sterilization of medical devices. methods. The appropriate scientific literature was reviewed to determine the frequency of failures of compliance. A risk assessment model was constructed. results. A 14-step protocol was constructed to aid infection control professionals in the evaluation of potential disinfection and sterilization failures. In addition, a model is presented for aiding in determining how patients should be notified of the potential adverse event. Sample statements and letters are provided for communicating with the public and individual patients. conclusion. Use of a protocol can guide an institution in managing potential disinfection and sterilization failures. Infect Control Hosp Epidemiol 2007; 28:146-155

In the United States in 1996, there were approximately infection failure on record involved the distribution of an 46,500,000 surgical procedures and a much larger number of inactive lot of glutaraldehyde disinfectant solution that had invasive medical procedures.1 For example, there are about been used by 60 hospitals in Belgium and involved 34,879 5 million gastrointestinal endoscopies per year. Each of these patients. In this incident, 25,589 patients were screened for procedures involves contact between a medical device or sur- infection with hepatitis B virus (HBV) and hepatitis C virus gical instrument and a patient’s sterile tissue or mucous mem- (HCV), and no acute infections were observed.8 It is our branes. A major risk of all such procedures is the introduction experience that the number of incidents that are published of pathogenic microbes, which can lead to infection. Failure or reported in the press represent a small fraction of the to properly disinfect or sterilize equipment has led to person- disinfection and sterilization instrument reprocessing failure to-person transmission via contaminated devices (eg, My- incidents that result in patient notification. These failures cobacterium tuberculosis transmitted via contaminated bron- may result from human error (eg, incorrect setting of the choscopes). Thus, achieving disinfection and sterilization temperature on a steam sterilizer or failure to clean items through the use of disinfectants and sterilization practices is before disinfection),4 equipment or product failure,8-10 or essential for ensuring that medical and surgical instruments systemic problems (ie, an organizational, procedural, or en- do not transmit infectious pathogens to patients. vironmental factor that facilitates the failure, such as the Deficiencies in disinfection and sterilization leading to use of incorrect channel connectors). Equipment failure in- infection have resulted either from failure to adhere to sci- cidents may stem from problems with design, manufacture, entifically based guidelines 2-4 or failures in the disinfection maintenance, or storage, as well as from a lack of user com- or sterilization processes.5 Patient notifications due to im- petence.11 This article presents a scheme for performing an proper reprocessing of semicritical (eg, endoscopes) and evaluation of possible failures of high-level disinfection or critical medical instruments have occurred regularly; these sterilization of patient care items. It will also provide a incidents generally involve single institutions, but may also method for assessing patient risk for adverse events, espe- involve multiple institutions (Table 1).6,7 The largest dis- cially infection.

From the Department of Hospital Epidemiology, University of North Carolina Health Care System, and the Division of Infectious Diseases, University of North Carolina School of Medicine, Chapel Hill (both authors). Received July 21, 2006; accepted November 3, 2006; electronically published January 24, 2007. ᭧ 2007 by The Society for Healthcare Epidemiology of America. All rights reserved. 0899-823X/2007/2802-0006$15.00. assessing risk to patients from disinfection and sterilization failures 147 disinfection and sterilization are those that enter sterile tissue or the vascular system and guidelines must be sterile because any microbial contamination could result in disease transmission. This category includes surgical Sterilization describes a process that destroys or eliminates instruments, cardiac and urinary catheters, implants, and ul- all forms of microbial life and is carried out in healthcare trasound probes used in sterile body cavities. facilities by either physical or chemical methods. Steam under Semicritical items are those that come in contact with mu- pressure, dry heat, ethylene oxide gas, hydrogen peroxide gas cous membranes or nonintact skin. Respiratory therapy and plasma, and liquid chemicals are the principle sterilizing anesthesia equipment, some endoscopes, laryngoscope blades, agents used in healthcare facilities in the United States. When esophageal manometry probes, anorectal manometry cathe- chemicals are used for the purposes of destroying all forms ters, and diaphragm fitting rings are included in this category. of microbiological life, including fungal and bacterial spores, Thus, these medical devices should be free of all microor- they may be called chemical sterilants. These same germicides ganisms (ie, mycobacteria, fungi, viruses, and bacteria), al- used for shorter exposure periods may also be part of the though small numbers of bacterial spores may be present. disinfection process (ie, for high-level disinfection). Intact mucous membranes, such as those of the respiratory Disinfection describes a process that eliminates many or tract or the gastrointestinal tract, generally are resistant to all pathogenic microorganisms, with the exception of bacterial infection by common bacterial spores but susceptible to other spores, on inanimate objects. Disinfection is usually accom- organisms, such as bacteria, mycobacteria, and viruses. Semi- plished by the use of liquid chemicals or wet pasteurization critical items minimally require high-level disinfection using in healthcare settings. The efficacy of disinfection is affected chemical disinfectants (eg, glutaraldehyde, hydrogen perox- by a number of factors, each of which may nullify or limit ide, ortho-phthalaldehyde, peracetic acid with hydrogen per- the efficacy of the process. Some of the factors that affect oxide, or chlorine). both disinfection and sterilization efficacy are prior cleaning Noncritical items are those that come in contact with intact of the object, the organic and inorganic load present, the type skin but not mucous membranes. Intact skin acts as an ef- and level of microbial contamination, the concentration and fective barrier to most microorganisms; therefore, the sterility duration of exposure to the germicide, the nature of the object of items coming in contact with intact skin is not critical. (eg, presence of crevices, hinges, or lumens), the presence of Examples of noncritical items are bedpans, blood pressure biofilms, the temperature and pH of the disinfection process, cuffs, crutches, bed rails, linens, bedside tables, patient fur- and, in some instances, the relative humidity of the sterili- niture, and floors. In contrast to critical and some semicritical zation process (eg, for ethylene oxide disinfection). items, most noncritical reusable items may be decontami- Cleaning, on the other hand, is the removal of visible soil nated where they are used and do not need to be transported (eg, organic and inorganic material) from objects and sur- to a central processing area. There is virtually no documented faces, and it normally is accomplished by manual or me- risk of transmitting infectious agents to patients via non- chanical means using water and detergents or enzymatic 20 critical items when they are used as noncritical items and products. Thorough cleaning is essential before high-level dis- infection and sterilization, because inorganic and organic ma- terials that remain on the surfaces of instruments interfere table 1. Reprocessing Failures of Semicritical or Critical Medical with the effectiveness of these processes. Instruments Resulting in Patient Notification Over 35 years ago, Earle H. Spaulding12 devised a rational No. of approach for disinfection and sterilization of patient care persons items and equipment. This classification scheme is so clear Location or institution, year Instrument involved exposed and logical that it has been retained, refined, and successfully Sacramento, CA, 2002 Endoscope 750 used by infection control professionals and others when plan- Toronto, ON, 2003 Endoscope 146 ning methods for disinfection or sterilization.13-18 Spaulding12 Seattle, WA, 2004 Endoscope 600 believed that the nature of disinfection could be understood Sacramento, CA, 2004 Endoscope 1,331 more readily if instruments and items for patient care were San Francisco, CA, 2004 Endoscope 2,000 Long Island, NY, 2004 Endoscope 177 divided into 3 categories on the basis of the degree of risk Charleston, NC, 2004 Endoscope 1,383 of infection involved in the use of the items. The 3 categories Toronto, ON, 2003 Prostate biopsy probe 900 he described were critical, semicritical, and noncritical . This Pittsburgh, PA, 2005 Endoscope 200 terminology is employed by the Centers for Disease Control Leesburg, VA 2005 Endoscope 144 and Prevention (CDC) in their guidelines for environmental San Diego, CA, 2006 Endoscope 300 19 18 Augusta, ME, 2006 Prostate biopsy needle 481 infection control and disinfection and sterilization in health- Dept Veterans Affairs, 2006 Prostate biopsy equipment 2,075 care facilities. San Diego, CA, 2006 Surgical instrument 82 Critical items are so categorized because there is a high note. Modified from a presentation by Douglas Nelson, MD, at the 33rd risk of infection if the items are contaminated with any mi- Annual Conference and International Meeting of the Association for Pro- croorganism, including bacterial spores. Thus, critical items fessionals in Infection Control and Epidemiology; Tampa, Florida, 2006. 148 infection control and hospital epidemiology february 2007, vol. 28, no. 2

figure 1. Protocol for exposure investigation after a failure of disinfection and sterilization procedures do not come into contact with nonintact skin or mucous protocol for exposure investigation membranes. Although exposure events due to possible failures of disin- reprocessing of endoscopes fection or sterilization are often unique, one should approach the evaluation of a potential failure using a standardized ap- Physicians use endoscopes to diagnose and treat numerous proach. As with evaluation of outbreaks of microbial infec- medical disorders. Although endoscopes are a valuable di- tion, one must be prepared to assess the unique aspects of agnostic and therapeutic tool in modern medicine and the each possible disinfection or sterilization failure by adapting incidence of infection associated with use has been reported the following recommended approach. to be very low (about 1 case per 1.8 million procedures),21 We propose a sequence of 14 steps that form a general more healthcare-associated outbreaks of infection have been approach to the evaluation of a possible failure of disinfection linked to contaminated endoscopes than to any other medical or sterilization that could result in patient exposure to an device.2-4,22 To prevent the spread of healthcare-associated in- infectious agent (Figure 1). Because failure to disinfect a non- fections, all heat-sensitive endoscopes (eg, gastrointestinal en- critical patient care item (eg, a blood pressure cuff) is very doscopes, bronchoscopes, and nasopharyngoscopes) must be unlikely to result in a patient exposure, reference to disin- properly cleaned and, at a minimum, subjected to high-level fection in the following section refers to high-level disinfec- disinfection following each use. High-level disinfection can tion of semicritical items. be expected to destroy all microorganisms, although if high Step 1. The first step in assessing a possible disinfection numbers of bacterial spores are present, a few spores may or sterilization failure is to confirm whether the suspected survive. Recommendations for the cleaning and disinfection failure did, in fact, occur. To do this, the infection control of endoscopic equipment have been published and should be professional should review the circumstances of the reported strictly followed.18,23 Unfortunately, audits have shown that failure, including the time and date of possible failure(s), the personnel often do not adhere to guidelines on repro- type of sterilization method used, and the evidence of failure, cessing,24-26 and outbreaks of infection continue to occur.9,10,27 including review of process parameters and physical, chem- To ensure that reprocessing personnel are properly trained, ical, and/or biological indicators. Some common failures in- there should be initial and annual competency testing for clude failure to subject the medical item to any disinfection each individual who is involved in reprocessing endoscopic or sterilization after cleaning, failure to expose the item to instruments.18,23,28,29 the proper temperature during the sterilization process, fail- assessing risk to patients from disinfection and sterilization failures 149

figure 2. Sample script to be used when discussing potential exposure with a patient by telephone ure to provide the proper duration of disinfection, failure to current problem(s). It is often better to arrange a face-to- expose the instrument to the disinfectant at the proper con- face conference, to assure complete communication of the centration, or failure to clean the item prior to disinfection. facts with feedback from attendees, than it is to use e-mail If the initial evaluation reveals that no medical items that or telephone consultation. were potentially inadequately processed were used in patient Step 5. A complete and thorough evaluation of the pos- care, then one can limit the evaluation to determining sible disinfection or sterilization failure should be rapidly whether the disinfection process failed and correcting any completed. The infection control professional should review processing error (ie, there is no patient safety issue involved). the exact circumstances of the possible disinfection or ster- All potentially inadequately processed items must, of course, ilization failure, including the dates and results of all process be reprocessed. If a disinfection or sterilization failure is not measures and physical, chemical, and biological indicators confirmed, the investigation may be concluded. obtained in the recent past, going back far enough to assess Step 2. If a possible disinfection or sterilization failure the time and date of the first possible malfunction. Process has occurred, one should immediately impound any medical measures may include the duration and temperature of ster- items that may not have been appropriately disinfected or ilization, as well as the sterilant concentration and pressure. sterilized (ie, do not allow such items to be used in patient Common problems with automated endoscope reprocessors care). All items reprocessed since the last successful processing have included failure to properly clean the item, failure to procedure (as demonstrated by process measures and/or set the correct duration of exposure to the high-level disin- physical, chemical, or biological indicators) should be im- fectant or sterilant, failure to attach the channel connectors pounded. Retrieving all items may require visiting all areas properly, and failure to use the correct channel connectors.5 where the medical or surgical items may be stored or used, Step 6. Once a failure of disinfection or sterilization has including the central processing facility, patient units (in- been documented, it is important to initiate the evaluation cluding community-based clinics), operating rooms, and of potential patient exposures. The first step is to create a storerooms. Impounded items should be retrieved and stored line listing of all patients who may have been exposed to in a place where they cannot be released for use and should potentially contaminated medical or surgical devices. The line be clearly labeled to prevent inadvertent use. listing should include, at a minimum, the following infor- Step 3. The suspect or defective disinfection or sterili- mation: patient name, identification number, date(s) of ex- zation unit should immediately be taken out of use and not posure, contaminated device used, underlying risk factors for used for disinfection or sterilization of medical or surgical infection, development of any healthcare-associated infection devices until such time as its proper functioning can be as- (including pathogen and body site), and other potential ad- sured. This may entail several cycles of operation with an verse events. To develop a line listing of possibly exposed assessment of process parameters and/or physical, chemical, patients in the event of a possible sterilization or disinfection or biological indicators. Medical engineering personnel or the failure, healthcare facilities should consider having tracking manufacturer’s representative usually perform repairs and methods in place to allow retrospective linkage of specific evaluation of the unit. Repair work may require that the pieces of high-risk equipment with specific patients (eg, rec- device be returned to the manufacturer. ords of which patients were evaluated with a specific endo- Step 4. All key stakeholders, including risk management scope). With the expanded use of bar code–labeled equipment personnel, the medical director and nursing director of the and scanning, the ability to link high-risk equipment to pa- involved medical or surgical units, and personnel involved tients should become more common in the future. in disinfection or sterilization, should be informed of the Step 7. Once a failure of disinfection or sterilization pro- 150 infection control and hospital epidemiology february 2007, vol. 28, no. 2

figure 3. Sample patient letter regarding a sterilization failure. aNystrom34; bRutala et al.35 cesses has been documented and possible patient exposure Step 8. All stakeholders should be kept informed of the to a contaminated item has been identified, it is crucial to progress of the investigation, especially if an increased risk determine whether, in fact, the failure could result in an ad- to patients is possible or documented. Key stakeholders in- verse patient event (eg, infection). For example, in our in- clude risk management personnel, the medical and nursing stitution, we require 4 minutes for flash sterilization. We director of the involved patient units, and personnel involved would consider flash sterilization for 3 minutes a breach of in disinfection or sterilization. Other persons who should be our policy. However, some recommendations would state that informed include staff in the public relations, healthcare ad- 3 minutes provides adequate sterilization of an unwrapped ministration, and legal departments. item. Thus, we would not consider items that have been flash Step 9. One should develop a hypothesis regarding the sterilized for 3 minutes to represent a hazard to patients with potential mechanism(s) of the disinfection or sterilization regard to increasing the risk of healthcare-associated infec- failure. Corrective actions (eg, repairs or improved training) tions. Many sterilization processes (eg, steam sterilization) should be initiated to correct deficiencies in reprocessing. Any have an enormous safety margin, and small deviations from item that may not have been appropriately disinfected or standard practice may not represent a patient hazard.30 As- sterilized must be reprocessed. sessments of risk should always be based on a review of the Step 10. Initiate a more detailed study, if necessary, of scientific literature and/or compliance with national guidelines. possible adverse outcomes in patients. This may entail de- assessing risk to patients from disinfection and sterilization failures 151

figure 4. Sample press release regarding the potential exposure event caused by the sterilization failure signing a prospective cohort study. It may require reviewing notify the patients. Choices include the patient’s local medical medical records and/or examining patients for infections, provider, the medical director of the clinic, the attending chemical reactions (eg, colitis), or other adverse events.31 Spe- physician at the time of disinfection or sterilization failure, cific laboratory tests may be necessary, such as performing risk management personnel, or an infection control profes- cultures and/or testing for source patients and exposed per- sional. One should develop a script to be used for notification, sons to identify bloodborne pathogens, such as human im- to ensure that all patients receive the same information (Fig- munodeficiency virus (HIV) (eg, HIV antibody testing at ure 2). Notification may be accomplished by a face-to-face baseline, 12 weeks, and 24 weeks), HBV (eg, hepatitis B sur- meeting, telephone call, or letter sent by registered mail (Fig- face antigen testing at baseline and at 24 weeks), and HCV ure 3). A press release should be prepared in case it is needed, (eg, HCV antibody testing at baseline, HCV polymerase chain and a spokesperson appointed (Figure 4). The wording used reaction viral detection testing at 4 weeks, and HCV antibody in these communications may need to be simplified to ensure testing at 24 weeks). Access to specific tests and receipt of patients’ understanding. A Spanish version of these letters results should be made as easy as possible for patients. may be useful in areas of the country with a large Spanish- Step 11. In conjunction with the legal department, notify speaking population. Our sample notices (Figures 2, 3, and appropriate state and federal authorities, if required by reg- 4) are provided as examples of how to communicate with ulation or law. patients and the media and should not be seen as establishing Step 12. Consider whether patients should be notified of the level of risk for which notification is required. the disinfection or sterilization failure. If it is determined that Multiple methods should be used simultaneously to ensure the failure could result in adverse patient events, then patients complete notification. The notification should include as should be notified. Fear of litigation, loss of business, and much information as possible, such the following: an assess- damage to the hospital’s reputation should not be deterrents ment of the risk, possible adverse events that may occur (eg, to reporting these events to patients.32 Determine who will wound infection), symptoms and signs of the adverse event, 152 infection control and hospital epidemiology february 2007, vol. 28, no. 2

figure 5. Description of a hypothetical sterilization failure investigated by infection control personnel. EPA, US Environmental Pro- tection Agency; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus. aMcQuillan et al36; bIppolito et al.37; cCenters for Disease Control and Prevention38; dResnick et al.39; eZou et al.40; fHeymann.41 time range for the adverse event, risk to other contacts, pos- initiated, it is crucial to assess whether these interventions sible prophylactic therapy (including benefits and risks), and have eliminated the problem over the long term. This may recommended medical follow-up. The healthcare facility require long-term surveillance, changes in current policies or must decide who will provide these services and whether the procedures, development of new policies or procedures, or facility will cover the cost of care. In general, we recommend evaluation of current equipment. that if the facility was responsible for the failure, then it should Step 14. Finally, a report of the event should be prepared provide these services at no charge to the patient. However, for presentation to the appropriate hospital or healthcare sys- if the exposure resulted from failures outside the institution tem committees. Consideration should be given to publishing (eg, receipt by the facility of inappropriately obtained or in- the evaluation if the lessons learned from the failure can help adequately prepared tissue used for implants, or a nonsterile prevent the same mistakes at other institutions. device), then the facility may want to offer the services, but provide them at the patient’s expense or at the product or assessing risk instrument manufacturer’s expense. Step 13. Once the problem leading to the disinfection or Assessing the probable risk to a patient after a potential ex- sterilization failure has been identified and corrective action posure is crucial to deciding whether disclosure is warranted. assessing risk to patients from disinfection and sterilization failures 153

Further, if patient disclosure is warranted, proper counseling assessments of chemicals (eg, in foods), air pollutants, and of the patient requires assessing the risk. In many cases, the medications. For example, the US Environmental Protection probable risk may be determined by a careful review of the Agency generally deems a lifetime cancer risk from hazardous literature and construction of an algorithm that determines air pollutant emissions (eg, benzene) greater than 1 in 1 mil- the independent probabilities of disease transmission (Figure lion to be unacceptable. This cancer risk benchmark has been 5). Disease transmission resulting in infection requires a chain used in their publications on the public health significance of events and circumstances including a microorganism path- of hazardous air pollutants. A lifetime risk of health effects ogenic for humans, the presence of the pathogen in the en- of 1 in 1 million or lower for chemical exposure is usually vironment, survival of the pathogen, presence of a portal of considered acceptable by regulatory agencies, either because entry into the potential host, a sufficiently large inoculum, it is small enough to be insignificant compared to other risks and failure of the immune system to prevent infection (ie, a or because it is extremely difficult, if not impossible, to mea- susceptible host). Each of these must occur in the proper sure.42 Similarly, a guideline from the European Medicines sequence, and each represents an independent step in the Agency (CPMP/SWP/5199/02) provides a “threshold of tox- chain. If any step in the sequence fails to occur, this prevents icological concern (TTC)” for the assessment of acceptable infection (eg, if there is no portal of entry into the potential limits of genotoxic impurities in pharmaceutical preparations, host). In many cases, actual probabilities for each of these which corresponds to a 1 in 100,000 risk of cancer.42-46 events (or a range of probabilities) can be determined from Risk is inherent in the practice of medicine, whether it is the scientific literature. The risk of infection is calculated by associated with medication, surgical or radiological proce- adding the independent probabilities, with risk displayed us- dures, medical diagnostic procedures, or diagnosis. Safe prac- ing a logarithmic scale (Figure 5). tice requires that risks be minimized, or ideally, prevented.11 If a calculated risk of disease transmission is extremely Infection control professionals should understand the causes small, such as 8 in 100 trillion, there may be no legal im- of disinfection and sterilization failures in instrument repro- perative to notify patients as there is, in effect, no clinically cessing and implement procedures that minimize or prevent significant health risk. Although there is no fixed or accepted patient exposures to improperly reprocessed medical and sur- risk frequency that necessitates risk disclosure, the legal staff gical instruments. at our institution has reviewed several informed consent cases that cite a 1%-3% frequency as the lower limit for notifying discussion patients about risks associated with medical procedures, such Potential exposure events due to possible failures of disin- as surgery. North Carolina’s informed consent statute refers fection or sterilization are not unusual in healthcare insti- to “usual and most frequent risks” as those that should be tutions (Table 1). Healthcare institutions must take such po- mentioned in an informed consent discussion for a medical tential failures seriously, as multiple outbreaks have been 33 procedure. Although preprocedure informed consent is dif- reported in the literature that resulted from failure to comply ferent than notification after a potential exposure, the orders with recommended practices. Such failures may occur as a of magnitude by which risks differ are instructive. Again, the result of human error, equipment malfunction, or system risk frequencies may be so small (eg, 8 in 100 trillion, as failure. Healthcare institutions must have a well-thought-out displayed in Figure 5) that they are effectively, and legally, of plan for dealing with such events. We have provided a blue- no concern or are far lower than the risk frequencies asso- print for evaluating and correcting potential failures of dis- ciated with many other daily exposures we all encounter (B. Gilbert, JD, MPH, personal communication, January 2005). table 2. Lifetime Odds of Death Due to Selected Types of Even though with risks of this magnitude there may be no Injury, United States, 2002 legal obligation to notify patients, healthcare facilities may Type of injury or event Lifetime odds of death decide in favor of disclosure. Transportation accident 1 in 77 Pedestrian 1 in 612 interpreting risk Car occupant 1 in 228 Drowning 1 in 1,081 The risks encountered in everyday life have been well de- Fall 1 in 229 scribed. For example, the risk of dying in a given year of Exposure to smoke, fire, flames 1 in 1,179 selected causes is shown in Table 2. In assessing everyday Venomous snake or lizard bite 1 in 1,241,661 activities (eg, driving, golfing, or flying), the most common Accidental poisoning 1 in 212 approach we all use is to ask what level of risk we are willing Lightning 1 in 56,439 Flood 1 in 413,887 to tolerate given the sometimes considerable benefits that the Intentional self-harm 1 in 118 activity will provide.42 Similarly, physicians, when providing note. National Safety Council estimates based on data from the Na- medications or recommending procedures, seek to assess the tional Center for Health Statistics and the U.S. Census Bureau. To de- risk-benefit ratio. termine the odds per year, multiply by 77.3 years (eg, the annual odds Regulatory agencies are beginning to develop risk-based of dying as a result of injury caused by lightning are 1 in 4,362,735). 154 infection control and hospital epidemiology february 2007, vol. 28, no. 2 infection or sterilization. The key aspects of the evaluation and training for newly purchased equipment should be ob- are that it ought to be done in an organized fashion and in tained. For example, the manufacturer’s representative or su- a rapid and timely manner. Maintaining communication pervisors should train reprocessing technicians regarding the among key stakeholders is crucial to the process. Although presence of auxiliary channels and in the appropriate repro- we have described the evaluation in a linear fashion, multiple cessing of each new endoscope model. Fourth, proper mon- steps are usually done simultaneously (eg, evaluation of the itoring of equipment should be performed, following the mechanism of sterilization failure and evaluation of patients recommended schedule (eg, for automated endoscope repro- for adverse outcomes). cessors) to ensure the equipment is functioning according to We have provided a general guideline in this article. How- the manufacturer’s specifications. Periodically ensure that the ever, each potential exposure event is unique, and one must high-level disinfectant exposure time for automated endo- be flexible in adapting our recommendations to a specific scope reprocessors is set properly and that the valves and situation. The intent of an evaluation should be to discover connectors are delivering the proper amount of disinfectant the factor(s) that led to the potential exposure and to protect and rinse water. Fifth, proper documentation of equipment patients from adverse events, as appropriate, and not to assign use for specific devices must be maintained. Sixth, regular blame to a particular person or persons. These are very stress- infection control rounds of areas where disinfection or ster- ful events for patients, and sensitivity and empathy toward ilization is carried out should be performed to assure that patients should be paramount. These events are also anxiety actual practice is consistent with policy and procedures. Sev- provoking for healthcare staff associated with the failure and enth, the scientific literature must be routinely reviewed by the staff involved with the corrective measures. In hospitals, infection control professionals with regard to potential out- staff may fail to report incidents because of time constraints, breaks and sources of potential exposures, and lessons must fear of punishment, and a lack of perceived benefit.32 As has be learned from previous failure incidents involving medical been the case with pilots and aviation safety officials, re- devices. Staff must understand what caused these failures and porting of these failure events should be encouraged, and develop procedures that prevent recurrence. staff should be immune from disciplinary action if they re- A key component of actual exposure events is assessing the ported the failure promptly.32 Of course, if gross negligence potential risk of disease transmission. Once a risk has been (eg, alcohol- or drug-impaired performance) is uncovered, determined, individual institutions must decide whether the the hospital’s human resources policies should be followed. magnitude of the risk warrants patient notification. Factors Employee performance and the ability to follow hospital pro- in addition to risk that may need to be considered include cedures should be evaluated periodically and, if deficient, local policies, legal recommendations, state or federal regu- employees should be retrained or the institution should fol- lations, availability of postexposure prophylaxis, potential low the appropriate human resources policies. If personnel consequences of infection, time frame for infection, and com- feel that reporting potential exposures will lead to negative municability of the disease. The final decision about whether personal consequences, they are unlikely to report potential to notify patients should be made by key stakeholders, partly events, to the detriment of both the institution and patients. on the basis of risk information provided by infection control It is important to rapidly identify potential exposures and professionals. If notification is planned, it should be done in institute corrective action for the disinfection or sterilization an organized way, and multiple avenues should be used to procedures. assure that all potentially exposed persons are notified. As with everything we do in infection control, prevention is preferable to correction of a deficiency in procedures that can result in potential exposure. The keys to preventing fail- Address reprint requests to William A. Rutala, PhD, MPH, Bioinformatics Building, 130 Mason Farm Road, Division of Infectious Diseases, University ures of disinfection and sterilization include the following. of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7030 (brutala@ First, adhere to authoritative guidelines. Specific guidelines unch.unc.edu). are available from various US agencies (eg, the Centers for Disease Control and Prevention and the Food and Drug Ad- acknowledgments ministration) and professional organizations (eg, the Asso- Financial support for this report was provided by the University of North ciation for Professionals in Infection Control and Epide- Carolina Health Care System. miology and the Society of Gastrointestinal Nurses and Associates). Manufacturers may also have reprocessing rec- references ommendations that should be reviewed. Second, personnel who perform disinfection and sterilization should be properly 1. Centers for Disease Control and Prevention. Ambulatory and inpatient trained (at the commencement of employment and at least procedures in the United States, 1996. Atlanta: Centers for Disease Con- annually after that) and supervised to ensure they consistently trol and Prevention; 1998:1-39. 2. Weber DJ, Rutala WA. Lessons from outbreaks associated with bron- follow the facility’s specific procedures. Competency testing choscopy. Infect Control Hosp Epidemiol 2001; 22:403-408. of personnel responsible for endoscope reprocessing is rec- 3. Weber DJ, Rutala WA, DiMarino AJ Jr. The prevention of infection ommended. Third, appropriate equipment should be used, following gastrointestinal endoscopy: the importance of prophylaxis and assessing risk to patients from disinfection and sterilization failures 155

reprocessing. In: DiMarino AJ Jr, Benjamin SB, eds. Gastrointestinal Dis- vey of digestive endoscopy disinfection procedures. Endoscopy 1997; eases: An Endoscopic Approach. Thorofare, NJ: Slack, 2002:87-106. 29:732-738. 4. Nelson DB. Infectious disease complications of GI endoscopy: part II, 26. Honeybourne D, Neumann CS. An audit of bronchoscopy practice in exogenous infections. Gastrointest Endosc 2003; 57:695-711. the United Kingdom: a survey of adherence to national guidelines. Tho- 5. Centers for Disease Control and Prevention. Bronchoscopy-related in- rax 1997; 52:709-713. fections and pseudoinfections—New York, 1996 and 1998. MMWR Morb 27. Cetse JC, Vanhems P. Outbreak of infection associated with broncho- Mortal Wkly Rep 1999; 48:557-560. scopes. N Engl J Med 2003;348:2039-2040. 6. Vanhems P, Gayet-Ageron A, Ponchon T, et al. Follow-up and manage- 28. Gerding DN, Peterson LR, Vennes JA. Cleaning and disinfection of fi- ment of patients exposed to a flawed automated endoscope washer- beroptic endoscopes: evaluation of glutaraldehyde exposure time and disinfector in a digestive diseases unit. Infect Control Hosp Epidemiol forced-air drying. Gastroenterology 1982; 83:613-8. 2006; 27:89-92. 29. Food and Drug Administration, Centers for Disease Control and Pre- 7. Roy KM, Ahmed S, Cameron SO, Shaw L, Yirrell D, Goldberg D. Patient vention. FDA and CDC public health advisory: infections from endo- notification exercise following a dentist’s admission of the periodic use scopes inadequately reprocessed by an automated endoscope repro- of unsterilized equipment. J Hosp Infect 2005; 60:163-168. cessing system. Rockville, MD: Food and Drug Administration; 1999. 8. Carsauw H, Debacker N. Recall of patients after the use of an inactive 30. Favero MS. Sterility assurance: concepts for patient safety. In: Rutala WA, lot of Cidex disinfection solution in Belgian hospitals. In: Program and ed. Disinfection, Sterilization and Antisepsis: Principles and Practices in abstracts of the 5th International Conference of the Hospital Infection Healthcare Facilities. Washington, DC: Association for Professionals in Society, Edinburgh, Scotland, September 15-18, 2002. Abstract P7.16. Infection Control and Epidemiology, 2001:110-119. 9. Srinivasan A, Wolfenden LL, Song X, et al. An outbreak of Pseudomonas 31. Duffy RE, Brown SE, Caldwell KL, et al. An epidemic of corneal de- aeruginosa infections associated with flexible bronchoscopes. N Engl J struction caused by plasma gas sterilization. Arch Ophthalmol 2000; 118: Med 2003; 348:221-227. 1167-1176. 10. Kirschke DL, Jones TF, Craig AS, et al. Pseudomonas aeruginosa and 32. Leape LL. Reporting of adverse events. N Engl J Med 2002; 347:1633-1638. Serratia marcescens contamination associated with a manufacturing de- 33. North Carolina General Statute §90-21.2. fect in bronchoscopes. N Engl J Med 2003; 348:214-220. 34. Nystrom B. Disinfection of surgical instruments. J Hosp Infect 1981; 2: 11. Amoore J, Ingram P. Learning from adverse incidents involving medical 363-368. devices. Br Med J 2002; 325:272-275. 35. Rutala WA, Gergen MF, Jones JF, Weber DJ. Levels of microbial con- 12. Spaulding EH. Chemical disinfection of medical and surgical materials. tamination on surgical instruments. Am J Infect Control 1998; 26:143-145. In: Lawrence C, Block SS, eds. Disinfection, Sterilization, and Preservation. 36. McQuillan GM, Kruszon-Moran D, Kottiri BJ, et al. Prevalence of HIV Philadelphia: Lea & Febiger, 1968:517-531. in the US household population: the national health and nutrition ex- 13. Favero MS, Bond WW. Chemical disinfection of medical and surgical amination surveys, 1988 to 2002. J Acquir Immune Defic Syndr 2006; 41: materials. In: Block SS, ed. Disinfection, Sterilization, and Preservation. 651-656. Philadelphia: Lippincott Williams & Wilkins, 2001:881-917. 37. Ippolito G, Puro V, De Carli G, Italian Study Group on Occupational 14. Simmons BP. CDC guidelines for the prevention and control of noso- Risk of HIV Infection. The risk of occupational human immunodefi- comial infections. Guideline for hospital environmental control. Am J ciency virus in health care workers. Arch Int Med 1993; 153:1451-1458. Infect Control 1983; 11:97-120. 38. Centers for Disease Control and Prevention. Updated U.S. Public Health 15. Garner JS, Favero MS. CDC Guideline for handwashing and hospital Service guidelines for the management of occupational exposures to environmental control, 1985. Infect Control 1986; 7:231-43. HBV, HCV, and HIV and recommendations for postexposure prophy- 16. Rutala WA. APIC guideline for selection and use of disinfectants. Am J laxis. MMWR Recomm Rep 2001; 50(RR-11):1-42. Infect Control 1990; 18:99-117. 39. Resnick L, Veren K, Salahuddin SZ, Tondreau S, Markham PD. Stability 17. Rutala WA, 1994, 1995, and 1996 APIC Guidelines Committee. APIC and inactivation of HTLV-III/LAV under clinical and laboratory envi- guideline for selection and use of disinfectants. Association for Profes- ronments. JAMA 1986; 255:1887-1891. sionals in Infection Control and Epidemiology. Am J Infect Control 1996; 40. Zou S, Notari EP IV, Stramer SL, Wahab F, Musavi F, Dodd RY. Patterns 24:313-342. of age- and sex-specific prevalence of major blood-borne infections in 18. Rutala WA, Weber DJ, Healthcare Infection Control Practices Advisory United States blood donors, 1995 to 2002: American Red Cross blood Committee. Guideline for disinfection and sterilization in healthcare donor study. Transfusion 2004; 44:1640-1647. facilities: recommendations of CDC. MMWR In press. 19. Sehulster L, Chinn RYW, Healthcare Infection Control Practices Advisory 41. Heymann DL. Control of Communicable Diseases Manual. Washington Committee. Guidelines for environmental infection control in health- DC: American Public Health Association, 2004. care facilities. MMWR Morb Mortal Wkly Rep 2003; 52:1-44. 42. McLaughlin RA, Genter MB, Cook MG, Zublena JP. Pollutants in ground 20. Weber DJ, Rutala WA. Environmental issues and nosocomial infections. water: risk assessment. Raleigh, NC: North Carolina Cooperative Exten- In: Wenzel RP, ed. Prevention and Control of Nosocomial Infections. Bal- sion Service; 1994. Available at: http://www.soil.ncsu.edu/publications/ timore: Williams and Wilkins, 1997:491-514. Soilfacts/AG-439-08/AG-439-8.pdf. Accessed July 2006. 21. Schembre DB. Infectious complications associated with gastrointestinal 43. Van Aerts LAGJM, Kasper P, Queckenberg O, Mueller L, Cartwright A. endoscopy. Gastrointest Endosc Clin N Am 2000; 10:215-232. International workshop on qualification and control of impurities. Drug 22. Spach DH, Silverstein FE, Stamm WE. Transmission of infection by Inf J 2006; 40:155-163. gastrointestinal endoscopy and bronchoscopy. Ann Intern Med 1993; 44. Kroes R, Kleiner J, Renwick A. The threshold of toxicological concern 118:117-128. concept in risk assessment. Toxicol Sci 2005; 86:226-230. 23. Nelson DB, Jarvis WR, Rutala WA, et al. Multi-society guideline for 45. European Medicines Agency, Committee for Medicinal Products for Hu- reprocessing flexible gastrointestinal endoscopes. Infect Control Hosp Ep- man Use. Guideline on the limits of genotoxic impurities, 2004. Available idemiol 2003; 24:532-537. at: http://www.emea.eu.int/pdfs/human/swp/519902en.pdf. Accessed July 24. Jackson FW, Ball MD. Correction of deficiencies in flexible fiberoptic 2006. sigmoidoscope cleaning and disinfection technique in family practice 46. Department of Transportation. Most Americans live in elevated cancer and internal medicine offices. Arch Fam Med 1997; 6:578-582. risk areas, EPA study shows. Available at: http://knowledge.fhwa.dot.gov/ 25. Orsi GB, Filocamo A, Di Stefano L, Tittobello A. Italian national sur- cops/italladdsup.nsf. Accessed June 2006.