Nosocomial Infections: Creating a Roadmap for Surveillance
By Patricia S. Grant, RN, BSN, MS, CIC
An article on nosocomial (hospital-acquired) infections can take one of several major paths, including listing all the statistics and financial aspects involved or reviewing the many published practice guidelines to decrease risk of infection transmission. Since infection control (IC) today is very different from our acute care-based roots and many of the statistics and guidelines are in that vein, this article will explore what makes nosocomial infections unique to each healthcare facility.
We are fortunate to have at our disposal the Study on the Efficacy of Nosocomial Infection Control (SENIC), which clearly outlines the benefits derived from an active IC surveillance program.1 The scientific community continues to provide guidance related to nosocomial infection surveillance. Only a few of these landmark studies can be referenced, however, as the knowledge base has grown too vast to list all relevant contributions.2-9
Whether you work in a traditional hospital setting, ambulatory clinic, nursing home, rehabilitation center, free-standing dialysis center or any kind of facility in between, nosocomial infections will present themselves and demand your attention. Think of yourself as the "conductor," overseeing the well-being of patients, employees, physicians, auxiliary healthcare workers, volunteers and visitors.10 As conductor, you must determine not only when a nosocomial event occurs, but when to investigate, tabulate, intervene and assess again.
It is almost impossible to separate nosocomial infections from surveillance activities, for they are married items. One (nosocomial infections) is an event and the other (surveillance activities) involves the culture in which the event exists. It is not enough to simply acknowledge and tabulate which infections have occurred or determine why an infection may or may not have happened. There must be an overall plan, a nosocomial infection roadmap equipped with signage, directions, destinations and a key for what each symbol means.
The first decision related to surveillance of nosocomial infections involves acknowledging a nosocomial event itself. You must know ahead of time why you are trying to identify a nosocomial event, otherwise there will be little direction to your activities. The determination of why a nosocomial event is first identified -- then later analyzed as a single event or in aggregate format -- is related to the type of healthcare facility at which you work. You must know the population you are serving, what the risk factors for infection are, and what accrediting/licensing organizations to which you are held accountable.
The best place to start is by analyzing two prior years' worth of data. Network with your peers for additional information, and then integrate this data into your purpose-of-nosocomial-infection identification framework. Other questions to ask are: What services are provided at your healthcare facility? What are the ages of your patients? What is the number of employees? Does your facility have overnight or ambulatory populations? Are you studying prior outbreaks, epidemic situations or endemic problems? This detective work will help you to understand why you do what you do related to nosocomial infections, and more importantly, will eliminate unnecessary busy work.
Understanding the occurrence of nosocomial infections will determine how you collect your data and whether or not you will analyze the information by a simple line-listing of events and descriptive statistics, or use a more complicated statistical analysis with benchmarking capabilities.
You should plan for your desired outcome related to nosocomial infections so you do not get to the end of the road only to discover that you have collected the wrong or incomplete data. Data collection without purpose-driven initiatives will waste your time and make your program an exercise in busywork minus measurable outcomes.
If the two years of data you are analyzing is controlled, you may want to focus on the processes involved with a nosocomial event, not just the outcome. Whether your two-year retrospective journey brings you to outcome-driven or process-driven nosocomial-infection prevention efforts remains to be seen; however, you will have ownership and rationale for the prevention efforts you define.
How you use the data collected is related to the purpose of data collection and the realities of nosocomial infection occurrence as a whole. A word of caution about "what-if" data collection: Do not collect data just in case it may be needed later. When you know the purpose of nosocomial infection data, then true outbreak or epidemic situations are obvious.11
Use of data must be directly related to the feedback system you will employ. Will you benchmark against your own internal, historical data or will it be against an outside system such as the Centers for Disease Control and Prevention (CDC)'s National Nosocomial Infections Surveillance (NNIS) System data? Or will you benchmark your findings against a data set provided by a corporate-run healthcare institution?
Determination of data use must be done with the end user in mind. Knowing exactly how -- and more importantly who-- will be using your nosocomial infection data toward process improvement is critical to the success of such programs. Knowing who will use the nosocomial infection information supplied will keep you focused on the projects related specifically to IC and prevention.
The definitions of all data collected must be standardized. This rule does not just apply to the event (nosocomial infection) numerator data, but is critical to the rate-driven comparison data (denominators) involved. This rule is important whether or not you are comparing your healthcare facility's nosocomial infection data to external benchmarks.12-13 The definitions must take into consideration any risk-stratification systems needed to factor in extrinsic or intrinsic risk factors toward infection development. End users must be confident that the nosocomial infection information you are providing is useful to them, and part of that usefulness is factoring in the reality that not all infections are created equal. The definitions chosen will have everything to do with the initial work you did when determining the purpose of your nosocomial infection activities. You must customize your definitions to the population you serve, whether it's acute care, long-term care or ambulatory/out patient.
Your use of data is dictated by how the nosocomial infections must be defined: by a descriptive line-listing of nosocomial events: a rate percent by 100 or a rate by 1,000 device or patient days. These decisions must be made during the planning phases of nosocomial infection analysis because this portion is critical to the entire nosocomial infection-identification and transmission-disruption process.
ICPs increasingly face a situation where individuals outside the healthcare facility request nosocomial infection data. Sometimes this data is not retrievable because it was not part of the internal purpose and was never collected. Too often ICPs are told, "You will do this," even when their training has told them the specific request is scientifically invalid, if not misleading. A classic example is when an insurance company wanting your hernia-repair surgical site infection rate when your internal assessment clearly indicated this was not a hot spot requiring extensive interventions.
We owe it to ourselves and to the public we serve to not provide erroneous and potentially dangerous nosocomial infection information. Training in IC, and the epidemiologic framework the discipline is based upon, must be shared whenever an opportunity presents itself. The following is a portion of a letter forwarded to a colleague who was being pressured by her own organization to provide "physician-specific nosocomial infection rates. Through networking she knew I had been successful in not providing this specific erroneous and misleading data set. "... As discussed this morning 'overall physician nosocomial infection rates' have not been a recommended surveillance practice, at least not since I've been trained by APIC beginning in 1990. As you are aware, there are many inherent errors with producing such a rate, with the top few being:
- A true rate must be risk-stratified, and this is not possible when using 'total physician admissions' as the denominator, and 'nosocomial infections' as the numerator. Physicians with higher risk patients (oncology, geriatric, etc.) will have higher rates than those with less acute/non-infectious patients (family practice, ophthalmic surgeons, etc.)
- Even if we were able to produce a 'risk-stratified' nosocomial infection rate for all physicians, what would it reflect? What practice(s) would it identify that the physician could alter? Stated another way, the physician is not the healthcare worker that inserts/maintains most intravenous access devices (primary bacteremia), inserts the foley (UTI), or maintains the ventilator circuits/treatments (ventilator-associated pneumonia).
- Therefore, this particular rate, although asked for by third party payers and/or internal sources, does not reflect true physician practices or outcomes; the rate literally is a number that does not reflect the performance of anybody.
- The main exception to this rule are surgeon-specific surgical site infection rates, and then, these must be risk-stratified either by surgical wound class and/or NNIS risk index, as appropriate. Even this advanced type of IC surveillance has confounding variables that must be considered when used for 'credentialing' purposes. Who were the scrub personnel? Was there sterilizer failure? Were the pre-op antibiotics given in a timely fashion? Was the temperature/humidity in the surgical area correct? Was product failure involved? These are all factors that play into surgeon-specific surgical site infection rates, and as we both know, cannot be accounted for; the labor would be too intensive, with very little in return."
In summary, nosocomial infections equate to extra pain and suffering for those affected. Nosocomial infection events are not limited to patients, nor are all infections preventable; however, the goal should theoretically be zero tolerance. As the ICP responsible for defining the purpose, use, definitions and expression of nosocomial infection happenings, you must remain steadfast in your commitment to providing information that serves the greater good, and that which is based in science and epidemiologic principles.
Patricia S. Grant, RN, BSN, MS, CIC, is director of infection control for RHD Memorial Medical Center and Trinity Medical Center in Dallas.