Surveillance for nosocomial infections

2. Surveillance for Nosocomial Infections

Introduction

Hospital programmes of infection control (IC) should include surveillance to detect common source outbreaks, identify problem areas, help set priorities for infection control activity, and meet national standards. Surveillance can also provide data to help convince clinicians and managers of the need for improvements in infection control practices. Surveillance must be performed in a systematic way with the aim of reducing rates of hospital infection. Surveillance results should be fed back to clinical and managerial staff and should lead to action.

Surveillance followed by action for improvement can have a significant impact on rates of hospital acquired infection (HAI), called nosocomial infection or health care facility-associated infection in some countries. The Study on the Efficacy of Nosocomial Infection Control (SENIC) [1] found that hospitals that had a programme of surveillance and fed results back to clinical staff had considerably lower infection rates than others. French and colleagues have demonstrated the effectiveness of repeated prevalence surveys [2] and the US National Nosocomial Infections Surveillance (NNIS) system has shown a significant reduction of nosocomial infection rates nationally in the US [3].

Surveillance can be defined as the systematic, active on-going observation of the occurrence and distribution of a disease within a population and of the events that increase or decrease the risk of the disease occurrence. If the incidence, distribution and associations of a disease are known, then resources can be targeted, predisposing factors can be reduced or eliminated, and the incidence of the disease reduced.

The purpose of surveillance of nosocomial infections is to reduce the incidence of HAI and thus to reduce the associated morbidity, mortality, and costs. Before beginning surveillance activities it is essential to develop a clear plan. It should address 1) what questions are being asked, 2) how infections are to be defined, 3) how the data are to be collected, stored, retrieved, summarised and interpreted, 4) how to feed the results back to frontline practitioners, and 5) how to use the information to bring about change.

Surveillance practices are similar to clinical audit, except that for an audit the practice and outcomes of medical care (in this case the prevention and control of HAI) is compared with a standard. By repeated audit cycles, practice is brought closer to the ideal.

Methods of Surveillance of HAI

Surveillance of infectious conditions requires strict definitions. In many cases there are no universally agreed definitions therefore the infection rate will vary with the definition used. For this reason, comparisons can be made between units or institutions only if the same set of definitions is used and applied in exactly the same way. It is often more meaningful and more useful to use surveillance data from a single institution to measure trends over time, either to alert staff to increasing problems or to monitor the effectiveness of interventions.

The definitions used should distinguish between HAI and community-acquired infection (CAI). Hospital-acquired infections can be defined as those that were neither present nor incubating at the time the patient was admitted. Detailed definitions of specific infections have been published by several organisations, including the World Health Organization [4], the US NNIS [5], and the Hospital Infection Society [6].

Some infections may present after the patient has been discharged from hospital. In surveillance for surgical site infection, as many as 70% of infections may present after discharge. This has led to the development of "post-discharge surveillance". However, post-discharge surveillance often poses considerable logistic problems and may add further expense to surveillance activities.

Formal surveillance of infections requires each patient to be assessed, often repeatedly, by trained staff. For this reason, true infection surveillance (and especially incidence surveillance) is very expensive due to the need for staff time. Because of this, surveillance is often done routinely by analysing laboratory reports, or by informal ward visits, or by a combination of the two. However, it must be recognised that these methods are not accurate. Laboratory reports are not always indicative of true infection. Negative reports (or no report) do not always mean infection is absent. Nevertheless, active surveillance (case finding by the Infection Control Nurse [ICN] ) increased detection from approximately 25% of defined infections to more than 85% in some studies [7] . These methods are particularly useful for identifying infections that may require action by the Infection Control Team and for measuring trends over periods when laboratory, medical, and infection control nursing practice remain constant.

Incidence and prevalence of HAI

The prevalence of HAI is the number of cases of active HAI in a defined patient population either during a specified period of time (the period prevalence) or at a specified point in time (point prevalence). The prevalence rate is the proportion of patients in the population who have an active infection at the time of the survey. The incidence of HAI is the number of new cases of disease that occur in the defined patient population during a specified time period. The incidence rate is the number of new cases of HAI that appear in the population at risk during the specified time period.

Alert condition surveillance

Alert condition surveillance means monitoring the incidence of specific clinical conditions, such as infectious diarrhoea or tuberculosis. This is part of the daily work of the Infection Control Team, which is directed towards, for example, the early identification of outbreaks of viral diarrhoea and the implementation of control procedures. Because this activity is usually not performed in a systematic way, it is not an accurate measure of true incidence.

Alert organism surveillance

Alert organism surveillance is the continuous monitoring of the incidence of specified organisms isolated by the microbiology laboratory. Alert organisms might include methicillin-resistant S. aureus (MRSA), glycopeptide-resistant enterococci, gentamicin-resistant coliforms and Clostridium difficile (by identification of its toxin). The isolation of an organism is not necessarily indicative of infection and the failure to isolate an organism does not prove the absence of infection. Furthermore, care must be taken to avoid the bias produced by duplicates and screening cultures. That is, when tracking an organism it should only be counted once. Nevertheless, this is a useful method of surveillance for infection control purposes. It has the advantages of simplicity and being inexpensive, and in computerised laboratories the surveillance can be automated. In particular, in a given hospital it can show trends in the isolation of specific organisms in different wards over time.

Prevalence surveys

In this method hospital in-patients are surveyed over a short period of time, ideally on a single day. Many hospitals and Infection Control Teams find prevalence surveys to be more practical than incidence surveillance since they can be performed by just a few people - often temporarily recruited from other tasks - once or twice a year. Repeated prevalence surveys are not a complete substitute for incidence surveillance, but they are useful where resources are limited.

Prevalence surveys are useful to indicate the extent of nosocomial infection within a hospital or region, to indicate specific problems requiring more extensive investigation, and to define the changing patterns of HAIs in a single hospital. If prevalence surveys are repeated at regular intervals and the results fed back to medical and nursing staff they can perform some of the same functions as continuous surveillance.

In general, prevalence rates tend to be lower than incidence rates since prevalence studies are less effective in identifying acute or short-lived infections. Repeated prevalence surveys have shown themselves to be useful for monitoring trends in rates of both HAI and CAI. They are practical to perform with relatively limited resources and they produce information on both infected and uninfected patients that can be used to identify independent risk factors. When properly applied, prevalence surveys can also be used to analyse the effectiveness of intervention strategies.

Prevalence studies have shown that, depending on the patient population, the prevalence of HAI averages around 9 -10%. Although most prevalence studies have been applied to the entire hospital, it is probably more effective to target certain areas or services where infection rates are suspected or known to be high.

Incidence surveillance

In this method, all patients are monitored over a period of time for the presence or absence of HAI. This is the best method for producing accurate measures of infection rates, however as with prevalence studies, it requires structured analysis, strict definitions and trained staff to visit all patients repeatedly. Because it is time consuming, incidence surveillance usually cannot be done continuously; rather it is often targeted in areas where problems are known or suspected. It is desirable to get surgical teams to do their own incidence surveillance of (say) clean surgical wound infection, supervised by the Infection Control Team. This means the surgical teams take ownership of the problem and are more likely to take action if rates are high.

Numerator data

The patient's name, identifying number or code, ward or unit, medical service at the time the infection began to develop, and date of admission are the necessary numerator information to collect. The date of onset of the infection, preceding risk factors such as respiratory therapy before new pneumonia, site of infection, significant organisms cultured and their sensitivity patterns help to describe the infection. Additional information may be helpful but should not be collected routinely unless it will be used. (Patient's primary diagnosis, age, sex, a measurement of severity of illness, physician's name, antimicrobial therapy, indirect risk factors such as immunosuppressive diseases or therapies are examples.)

Denominator data (population at risk)

Rates are always calculated with the numerator (number of persons with the infection or condition) divided by the denominator (number of persons at risk for the infection). The more precisely the denominator captures the potentially preventable risk elements the better. For example, nosocomial pneumonia cases among patients who had respiratory therapy divided by number of patients discharged in a month or on a specific care unit provides some estimate of risk. However nosocomial pneumonias among such patients divided by number of patients receiving respiratory therapy yields a much better rate.

What are the standards for rates of HAI?

There are no published standards of HAI rates. The rate of HAI will vary with patient risk, and therefore, there will be different rates in different units. Ayliffe has pointed out that there is an ‘irreducible minimum ‘ rate of HAI due to the inherent risks of underlying disease and medical interventions. Rates will also vary depending on the level of facilities and staffing available in different hospitals of medical systems. In general rates should be compared with peer institutions.

Prevalence surveys show average whole hospital rates of HAI of 7-10%. It is thought that about 30% of these may be preventable, depending on the patient population. Surgical site infection rates in clean surgery should probably be less than 5%, and even less than 1% may be achievable.

Feedback

It is pointless to collect masses of data if it is only seen by the Infection Control Team. It is essential that surveillance results are fed back regularly to the front-line clinical staff in order to help them choose actions to reduce infection rates. It has been shown on many occasions that feedback – with educational and practical help from the Infection Control Team – is one of the most effective ways of effecting change in hygienic practice.

Minimal Requirements for Surveillance

Monitor infection patterns (sites, pathogens, risk factors, location within the facility)
Detect changes in the patterns that may indicate an infection problem
Direct the rapid implementation of control measures
Monitor antibiotic use and resistance
Provide the staff with exactly the information they need in order to improve infection prevention practices.

References

Haley RW, Culver DH, White JW, Morgan WM, Emori TG, Munn VP. Hooton TM. The efficacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals. American Journal of Epidemiology 1985; 121:182-205. Back to text
French GL. Repeated prevalence surveys. Ballière's Clinical Infectious Diseases 1996;3:179-195. Back to text
Report. Monitoring Hospital-Acquired Infections to Promote Patient Safety -- United States, 1990-1999. MMWR 2000; 49:149-153. Back to text
WHO Prevention of hospital-acquired infections, A practical guide. 2nd edition, 2002 (WHO/CDS/CSR/EPH/2002.12). Back to text
Horan TC, Emori TG. Definitions of nosocomial infections. In: Abrutyn E, Goldmann DA, Scheckler WE, eds. Saunders Infection Control Reference Service . Philadelphia : W. B. Saunders, 1998:17-22. Back to text
Report. National prevalence survey of hospital-acquired infection: definitions. A preliminary report of the Steering Group of the Second National Prevalence Survey. Journal of Hospital Infection 1993;24:69-76. Back to text
Perl TM. Surveillance, reporting and the use of computers. In Wenzel RP(ed.). Prevention and control of nosocomial infections. Third Edition. Williams & Wilkins, Baltimore 1997. Back to text

Bibliography

Gaynes RP. Surveillance of nosocomial infections. In, Bennett JV, Brachman PS (eds) Hospital Infections, 4 th Edition. Philadelphia: Lippincott-Raven, 1998.



2. Surveillance for Nosocomial Infections

Introduction Hospital programmes of infection control (IC) should include surveillance to detect common source outbreaks, identify problem areas, help set priorities for infection control activity, and meet national standards. Surveillance can also provide data to help convince clinicians and managers of the need for improvements in infection control practices. Surveillance must be performed in a systematic way with the aim of reducing rates of hospital infection. Surveillance results should be fed back to clinical and managerial staff and should lead to action. Surveillance followed by action for improvement can have a significant impact on rates of hospital acquired infection (HAI), called nosocomial infection or health care facility-associated infection in some countries. The Study on the Efficacy of Nosocomial Infection Control (SENIC) [1] found that hospitals that had a programme of surveillance and fed results back to clinical staff had considerably lower infection rates than others. French and colleagues have demonstrated the effectiveness of repeated prevalence surveys [2] and the US National Nosocomial Infections Surveillance (NNIS) system has shown a significant reduction of nosocomial infection rates nationally in the US [3]. Surveillance can be defined as the systematic, active on-going observation of the occurrence and distribution of a disease within a population and of the events that increase or decrease the risk of the disease occurrence. If the incidence, distribution and associations of a disease are known, then resources can be targeted, predisposing factors can be reduced or eliminated, and the incidence of the disease reduced. The purpose of surveillance of nosocomial infections is to reduce the incidence of HAI and thus to reduce the associated morbidity, mortality, and costs. Before beginning surveillance activities it is essential to develop a clear plan. It should address 1) what questions are being asked, 2) how infections are to be defined, 3) how the data are to be collected, stored, retrieved, summarised and interpreted, 4) how to feed the results back to frontline practitioners, and 5) how to use the information to bring about change. Surveillance practices are similar to clinical audit, except that for an audit the practice and outcomes of medical care (in this case the prevention and control of HAI) is compared with a standard. By repeated audit cycles, practice is brought closer to the ideal. Methods of Surveillance of HAI Surveillance of infectious conditions requires strict definitions. In many cases there are no universally agreed definitions therefore the infection rate will vary with the definition used. For this reason, comparisons can be made between units or institutions only if the same set of definitions is used and applied in exactly the same way. It is often more meaningful and more useful to use surveillance data from a single institution to measure trends over time, either to alert staff to increasing problems or to monitor the effectiveness of interventions. The definitions used should distinguish between HAI and community-acquired infection (CAI). Hospital-acquired infections can be defined as those that were neither present nor incubating at the time the patient was admitted. Detailed definitions of specific infections have been published by several organisations, including the World Health Organization [4], the US NNIS [5], and the Hospital Infection Society [6]. Some infections may present after the patient has been discharged from hospital. In surveillance for surgical site infection, as many as 70% of infections may present after discharge. This has led to the development of "post-discharge surveillance". However, post-discharge surveillance often poses considerable logistic problems and may add further expense to surveillance activities. Formal surveillance of infections requires each patient to be assessed, often repeatedly, by trained staff. For this reason, true infection surveillance (and especially incidence surveillance) is very expensive due to the need for staff time. Because of this, surveillance is often done routinely by analysing laboratory reports, or by informal ward visits, or by a combination of the two. However, it must be recognised that these methods are not accurate. Laboratory reports are not always indicative of true infection. Negative reports (or no report) do not always mean infection is absent. Nevertheless, active surveillance (case finding by the Infection Control Nurse [ICN] ) increased detection from approximately 25% of defined infections to more than 85% in some studies [7] . These methods are particularly useful for identifying infections that may require action by the Infection Control Team and for measuring trends over periods when laboratory, medical, and infection control nursing practice remain constant. Incidence and prevalence of HAI The prevalence of HAI is the number of cases of active HAI in a defined patient population either during a specified period of time (the period prevalence) or at a specified point in time (point prevalence). The prevalence rate is the proportion of patients in the population who have an active infection at the time of the survey. The incidence of HAI is the number of new cases of disease that occur in the defined patient population during a specified time period. The incidence rate is the number of new cases of HAI that appear in the population at risk during the specified time period. Alert condition surveillance Alert condition surveillance means monitoring the incidence of specific clinical conditions, such as infectious diarrhoea or tuberculosis. This is part of the daily work of the Infection Control Team, which is directed towards, for example, the early identification of outbreaks of viral diarrhoea and the implementation of control procedures. Because this activity is usually not performed in a systematic way, it is not an accurate measure of true incidence. Alert organism surveillance Alert organism surveillance is the continuous monitoring of the incidence of specified organisms isolated by the microbiology laboratory. Alert organisms might include methicillin-resistant S. aureus (MRSA), glycopeptide-resistant enterococci, gentamicin-resistant coliforms and Clostridium difficile (by identification of its toxin). The isolation of an organism is not necessarily indicative of infection and the failure to isolate an organism does not prove the absence of infection. Furthermore, care must be taken to avoid the bias produced by duplicates and screening cultures. That is, when tracking an organism it should only be counted once. Nevertheless, this is a useful method of surveillance for infection control purposes. It has the advantages of simplicity and being inexpensive, and in computerised laboratories the surveillance can be automated. In particular, in a given hospital it can show trends in the isolation of specific organisms in different wards over time. Prevalence surveys In this method hospital in-patients are surveyed over a short period of time, ideally on a single day. Many hospitals and Infection Control Teams find prevalence surveys to be more practical than incidence surveillance since they can be performed by just a few people - often temporarily recruited from other tasks - once or twice a year. Repeated prevalence surveys are not a complete substitute for incidence surveillance, but they are useful where resources are limited. Prevalence surveys are useful to indicate the extent of nosocomial infection within a hospital or region, to indicate specific problems requiring more extensive investigation, and to define the changing patterns of HAIs in a single hospital. If prevalence surveys are repeated at regular intervals and the results fed back to medical and nursing staff they can perform some of the same functions as continuous surveillance. In general, prevalence rates tend to be lower than incidence rates since prevalence studies are less effective in identifying acute or short-lived infections. Repeated prevalence surveys have shown themselves to be useful for monitoring trends in rates of both HAI and CAI. They are practical to perform with relatively limited resources and they produce information on both infected and uninfected patients that can be used to identify independent risk factors. When properly applied, prevalence surveys can also be used to analyse the effectiveness of intervention strategies. Prevalence studies have shown that, depending on the patient population, the prevalence of HAI averages around 9 -10%. Although most prevalence studies have been applied to the entire hospital, it is probably more effective to target certain areas or services where infection rates are suspected or known to be high. Incidence surveillance In this method, all patients are monitored over a period of time for the presence or absence of HAI. This is the best method for producing accurate measures of infection rates, however as with prevalence studies, it requires structured analysis, strict definitions and trained staff to visit all patients repeatedly. Because it is time consuming, incidence surveillance usually cannot be done continuously; rather it is often targeted in areas where problems are known or suspected. It is desirable to get surgical teams to do their own incidence surveillance of (say) clean surgical wound infection, supervised by the Infection Control Team. This means the surgical teams take ownership of the problem and are more likely to take action if rates are high. Numerator data The patient's name, identifying number or code, ward or unit, medical service at the time the infection began to develop, and date of admission are the necessary numerator information to collect. The date of onset of the infection, preceding risk factors such as respiratory therapy before new pneumonia, site of infection, significant organisms cultured and their sensitivity patterns help to describe the infection. Additional information may be helpful but should not be collected routinely unless it will be used. (Patient's primary diagnosis, age, sex, a measurement of severity of illness, physician's name, antimicrobial therapy, indirect risk factors such as immunosuppressive diseases or therapies are examples.) Denominator data (population at risk) Rates are always calculated with the numerator (number of persons with the infection or condition) divided by the denominator (number of persons at risk for the infection). The more precisely the denominator captures the potentially preventable risk elements the better. For example, nosocomial pneumonia cases among patients who had respiratory therapy divided by number of patients discharged in a month or on a specific care unit provides some estimate of risk. However nosocomial pneumonias among such patients divided by number of patients receiving respiratory therapy yields a much better rate. What are the standards for rates of HAI? There are no published standards of HAI rates. The rate of HAI will vary with patient risk, and therefore, there will be different rates in different units. Ayliffe has pointed out that there is an ‘irreducible minimum ‘ rate of HAI due to the inherent risks of underlying disease and medical interventions. Rates will also vary depending on the level of facilities and staffing available in different hospitals of medical systems. In general rates should be compared with peer institutions. Prevalence surveys show average whole hospital rates of HAI of 7-10%. It is thought that about 30% of these may be preventable, depending on the patient population. Surgical site infection rates in clean surgery should probably be less than 5%, and even less than 1% may be achievable. Feedback It is pointless to collect masses of data if it is only seen by the Infection Control Team. It is essential that surveillance results are fed back regularly to the front-line clinical staff in order to help them choose actions to reduce infection rates. It has been shown on many occasions that feedback – with educational and practical help from the Infection Control Team – is one of the most effective ways of effecting change in hygienic practice. Minimal Requirements for Surveillance Monitor infection patterns (sites, pathogens, risk factors, location within the facility) Detect changes in the patterns that may indicate an infection problem Direct the rapid implementation of control measures Monitor antibiotic use and resistance Provide the staff with exactly the information they need in order to improve infection prevention practices. References Haley RW, Culver DH, White JW, Morgan WM, Emori TG, Munn VP. Hooton TM. The efficacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals. American Journal of Epidemiology 1985; 121:182-205. Back to text French GL. Repeated prevalence surveys. Ballière's Clinical Infectious Diseases 1996;3:179-195. Back to text Report. Monitoring Hospital-Acquired Infections to Promote Patient Safety -- United States, 1990-1999. MMWR 2000; 49:149-153. Back to text WHO Prevention of hospital-acquired infections, A practical guide. 2nd edition, 2002 (WHO/CDS/CSR/EPH/2002.12). Back to text Horan TC, Emori TG. Definitions of nosocomial infections. In: Abrutyn E, Goldmann DA, Scheckler WE, eds. Saunders Infection Control Reference Service . Philadelphia : W. B. Saunders, 1998:17-22. Back to text Report. National prevalence survey of hospital-acquired infection: definitions. A preliminary report of the Steering Group of the Second National Prevalence Survey. Journal of Hospital Infection 1993;24:69-76. Back to text Perl TM. Surveillance, reporting and the use of computers. In Wenzel RP(ed.). Prevention and control of nosocomial infections. Third Edition. Williams & Wilkins, Baltimore 1997. Back to text Bibliography Gaynes RP. Surveillance of nosocomial infections. In, Bennett JV, Brachman PS (eds) Hospital Infections, 4 th Edition. Philadelphia: Lippincott-Raven, 1998.