◆ The ICU is a focus for the emergence of bacterial resistance.
◆ Overuse of antibiotics is the main driver.
◆ Catheter-associated bacteraemia rates are the most frequently used performance indicator.
◆ Methicillin-resistant Staphylococcus aureus (MRSA) screening and topical suppression is effective in preventing spread.
◆ Screening for multiresistant Gram-negative bacteria needs to be considered according to local prevalence.
Of all hospital wards, the intensive care unit (ICU) is the one with the highest risks of acquisition and transmission of hospital-acquired infection. Most patients are given antibiotics, there are numerous invasive procedures, and a large number of staff frequently come into contact with the patient. Infections acquired in the ICU have significant consequences for the rest of the hospital when the patient is discharged to a general ward. Furthermore, colonization or infections with multiresistant organisms can significantly delay discharge of a patient from ICU at considerable cost to the hospital. Surveillance is important to identify emerging outbreaks of multiresistant infections and the overuse of antibiotics that can give rise to them. Continuous collection of data using widely recognized definitions can allow comparisons between units and be used to reduce rates in outliers. Definitions and methods should be based on recognized international systems such as those of Centers for Disease Control (CDC) or European Centre for Disease Control (ECDC). National networks such as The Intensive Care National Audit & Research Centre (ICNARC) should form the basis for extracting denominator numbers and exchange of information between units.
Identifying rates of infection and feeding back the results to the clinical teams is effective in reducing rates. Local and timely investigation of the causes of an infection is the best method of preventing further cases. Root cause analysis is a formal tool that has been used widely to identify specific failings and strategies to avoid them in future. In the USA nosocomial infection is no longer reimbursed by insurance companies and in other countries there is increasing scrutiny of these infections and in some cases financial penalties are applied. As a result there is now much interest in surveillance as a means of improving patient outcome.
The ECDC has used a range of definitions in multinational ICU surveillance in Europe . ICU-acquired infection is defined as occurring later than 48 hours after admission. Device associated health care-associated infection (HAI) is an infection in a patient with a device used within 48 hours before onset of infection. Devices are intubation, central vascular catheter (CVC), and urinary catheter. Elsewhere CDC definitions are widely used, but not specific to ICU .
Antibiotic consumption and resistance levels vary widely between units. Defined daily doses per 1000 occupied bed days is a common comparative measure, but the definition of length of stay must be agreed and antibiotics delivered to ICU may not necessarily be used. Dosage increased due to severity of illness or reduced due to renal failure are sources of error. One study found no relationship between prevalence of multiresistant pathogens and consumption of carbapenems, quinolones, or cephalosporins . Poor hand hygiene or lack of antibiotic control may be an explanation. High rates of resistance in pseudomonas are related to high usage of carbapenems and quinolones.
Catheter-related or associated bacteraemia
Central line-associated bloodstream infections increase the length of stay in ICU and increase the risk of mortality. With appropriate catheter management these infections are avoidable and so catheter-associated bacteraemia is likely to emerge as a standard performance measure Box 281.1. Surveillance can cease 48 hours after the patient leaves the unit . Bloodstream infections are central-line associated when a central line or umbilical catheter was in place (or within 48 hours) and not secondary to an infection at another site. If a line was inserted in the Emergency Department and is diagnosed within 24 hours of arrival in ICU it is still considered ICU associated. If it occurs within 48 hours of transfer out of an ICU it is attributed to that ICU. Pacemaker wires, femoral arterial catheters, intra-aortic balloon pump devices, or devices without lumens are not counted as central lines. The catheter-associated bloodstream infection rate is expressed per 1000 central line days. Surveillance at least at a regional level should be continuous.
Reproduced from CDC Device Associated Module January 2012: Central Line-Associated Bloodstream Infection Event (CLABSI).
One of the most significant examples of the efficacy of surveillance was the successful Michigan program to reduce CVC infections . This project used peer pressure to encourage ICUs to join the program, created strong networking in the group, portraying catheter-related bacteraemia as a social problem, used a bundle of measures to reduce infections and data generated as a disciplinary measure. Four recommendations were applied related to insertion of the catheter: hand washing, using full barrier precautions, cleaning the skin with chlorhexidine, and avoiding the femoral site when possible. The fifth recommendation was to remove unnecessary catheters. In 90 ICUs, the number of catheter-related bloodstream infections was collected over 3 years. The mean and median rates of catheter related bloodstream infection per 1000 catheter days decreased from 7.7 and 2.7 at baseline to 1.3 and 0 at 16–18 months to 1.1 and 0 at 36 months. The project demonstrated that through a bundle of measures and a quality improvement ethos, improvements could be achieved and sustained.
As a result other countries have considered using catheter-associated or catheter-related blood stream infection as a measure of ICU performance in preventing nosocomial infection. ‘Matching Michigan’ was run by the National Patient Safety Agency in the UK. Definitions were based on CDC and ECDC and gained wide participation despite being voluntary and time limited. Results were sufficiently encouraging to consider a national voluntary quality improvement scheme . Other studies have demonstrated that audit and feedback of these infections gradually reduces the number of cases.
In many countries, computerized data collection of morbidity and mortality is becoming more extensive, but infection surveillance is not usually the primary focus. While catheter-related bacteraemia requires additional laboratory resources for diagnosis (semiquantitative analysis of blood samples or roll cultures of catheter tips), catheter-associated bacteraemia is less stringent and more easily applicable. Collection of data on catheter-associated bacteraemia may not require significant extra resource as long as duplicate data entry is avoided. Voluntary surveillance may be under-reported if results are used for performance management or reimbursement or if there are major resource implications. On the other hand mandatory surveillance can be costly and less useful in small units. The main parameter for surveillance should be CVC associated infection expressed as a proportion of CVC patient-days using the ECDC definition. CVC patient-days (number of patients with at least one CVC in situ every 24 hours summed over one month) are needed to calculate device-associated rates. In the longer term, it would be desirable for microbiology departments to develop capacity for routine semi-quantitative cultures to allow use of catheter-related bloodstream infection as the eventual ideal marker.
Ventilator-associated pneumonia shortens patient survival and can arise from cross infection. However definitions are problematic and the use of bronchoscopy for diagnostic purposes varies widely between countries. Rates identified by the treating surgical intensivist and the infection control service in the same hospital can differ widely . Therefore, it cannot be recommended as a subject for long-term surveillance.
Pseudomonas aeruginosa, including multiresistant strains, contaminating taps and other water sources in ICUs has been associated with the development of bacteraemia in patients, particularly neonates . The hands of staff are the likely means of transmission and bathing infants in tap water a probable source. The hand wash station should only be used for hand washing and not for disposal of body fluids. Outbreaks have been reported with detergent diluted with contaminated tap water. Some neonatal units check rectal swabs every week to establish the prevalence of pseudomonal carriage, which can be helpful in monitoring the effect of interventions.
Studies confirm the effectiveness of point of use water filtration in reducing infection in critical care units. In one study, pseudomonas infections fell by 56% and water isolates fell from 97% to 0% when filters were introduced . However filters have to be replaced monthly and are expensive. Sensor taps have been implicated in some outbreaks of Pseudomonas bacteraemia, probably because of biofilm accumulation in the end of the tap due to poor water flow and the thermoregulatory valve providing a carbon source. Surveillance of water samples in augmented care units for Pseudomonas aeruginosa has been advised.
In hospitals where methicillin-resistant Staphylococcus aureus (MRSA) is prevalent, the likelihood of carriage and acquisition is often highest in the ICU. Identification of carriage followed by topical suppression has been successful in reducing the risk of transmission. MRSA screening in many countries is performed on admission to hospital and/or the critical care unit. If a rapid PCR based method is used, the result is available the same day allowing immediate source isolation and more appropriate use of glycopeptides. If not, a patient with unknown carriage status may need to be source isolated or treated with topical suppression pending a result. All carriers are treated with topical chlorhexidine and mupirocin to reduce shedding into the environment and cross infection. Identified carriers should be source isolated in a single room. Units with universal screening report higher rates of MRSA as the result of better ascertainment, but only 20% of cases are likely to have been acquired during ICU stay .
Although topical suppression of all patients was cost effective in the short term, it carries a risk of selection of resistance. Rapid PCR based screening of all patients on admission and weekly thereafter is most cost effective . Isolation or cohorting without topical suppression is effective, but increases costs unless prevalence is high (10%). The use of topical chlorhexidine and mupirocin may have been responsible for the dramatic reduction in MRSA prevalence in some countries, but probably is not cost effective when prevalence is already low.
Extended spectrum β-lactamase-producing Gram-negative bacteria
Extended spectrum β-lactamase producing Gram negative bacteria (ESBL) are rapidly increasing as a cause of invasive infection in ICU. Overuse of antibiotics is a major driver. Screening may be justified in units where there have been previous outbreaks or control of antibiotic use is poor. Surveillance over 11 years in Dutch ICUs showed changes in antimicrobial resistance in Klebsiella pneumoniae. Resistance to ceftazidime rose from 4.2% to 10.8%, ciprofloxacin from 5.8 to 18.5% and cefuroxime from 2.8 to 7.9% . The prevalence of ESBL increased from 2% to 8%. Resistance was significantly greater in ICU than surgical wards.
Outbreaks of multiresistant Acinetobacter are very disruptive to the functioning of critical care units and cause bacteraemia and serious respiratory infection. Control of spread can be very difficult even with full infection control measures. Few ICUs screen routinely for the organism except during an outbreak. However, active surveillance (testing and either isolating or decolonizing) is cost saving when the colonization prevalence is greater than 1%. . Even slight increases in length of stay in ICU can give rise considerable additional costs. Swabbing pharynx, wounds, axilla, and groin has a low sensitivity 13–78% compared with a sponge wiped on the upper arm and thigh (89%).
As health systems become more focussed on preventable infections, mandatory surveillance (or least peer pressure) will become more common. Financial incentives or penalties may be applied on the basis of reported numbers of hospital-acquired infections. While considerable reductions in infection rates can be achieved, use of national surveillance networks, standard definitions and methodologies are essential to ensure valid comparisons between units and appropriate actions for outlying performances.
1. European Centre for Disease Prevention and Control (2012). European surveillance of Healthcare-associated Infections in Intensive Care Units. HAIICU Protocol v 1.01: Standard and Light. Available at: http://www.ecdc.europa.eu/en/aboutus/calls/Procurement%20Related%20Documents/5_ECDC_HAIICU_protocol_v1_1.pdf (accessed 28 October 2015).
2. Centers for Disease Control (2012). Bloodstream Infection Event (Central Line-associated Bloodstream Infection and Non-central Line-associated Bloodstream Infection). Available at: http://www.cdc.gov/nhsn/PDFs/pscManual/4PSC_CLABScurrent.pdf (accessed 28 October 2015).
3. Hanberger H, Arman D, Gill H, et al. (2009). Surveillance of microbial resistance in European Intensive care Units: a first report from the Care-ICU programme for improved infection control. Intensive Care Medicine, 35, 91–100.Find this resource:
4. Pronovost PJ, Goeschel CA, Colantuoni E, et al. (2010). Sustaining reductions in catheter related bloodstream infections in Michigan intensive care units: observational study. British Medical Journal, 340, c309.Find this resource:
5. Julian Bion, Annette Richardson, Peter Hibbert et al. (2012) ‘Matching Michigan’: a 2-year stepped interventional programme to minimise central venous catheter-blood stream infections in intensive care units in England. British Medical Journal Quality Safety doi:10.1136/bmjqs-2012-001325Find this resource:
6. Thomas BW, Maxwell RA, Dart BW, et al. (2011). Errors in administrative-reported ventilator-associated pneumonia rates: are never events really so? American Surgeon, 77, 998–1002.Find this resource:
7. Crivaro V, Di Popolo A, Caprio A, et al. (2009). Pseudomonas aeruginosa in a neonatal intensive care unit: molecular epidemiology and infection control measures. Bio Medical Centre Infectious Diseases, 9, 70.Find this resource:
8. Trautmann M, Halder S, Hoegel J, Royer H, and Haller, M. (2008). Point-of-use water filtration reduces endemic Pseudomonas aeruginosa infections on a surgical intensive care unit. American Journal of Infection Control, 36, 421–9.Find this resource:
9. Kohlenberg A, Schwab F, Behnke M, Geffers C, and Gastmeier P. (2011). Screening and control of methicillin-resistant Staphylococcus aureus in 186 intensive care units: different situations and individual solutions. Critical Care, 15, R285.Find this resource:
10. Edgeworth JD. (2011). Has decolonization played a central role in the decline in UK methicillin-resistant Staphylococcus aureus transmission? A focus on evidence from intensive care. Journal of Antimicrobial Chemotherapy, 66, (2), ii41–7.Find this resource:
11. van der Donk CF, Beisser PS, Hoogkamp-Korstanje JA, Bruggeman CA, Stobberingh EE, and the Antibiotic Resistance Surveillance Group. (2011), A 12 year (1998–2009) antibiotic resistance surveillance of Klebsiella pneumoniae collected from intensive care and urology patients in 14 Dutch hospitals. Journal of Antimicrobial Chemotherapy, 66, 855–8.Find this resource:
12. Lee BY, McGlone SM, Doi Y, Bailey RR, and Harrison LH. (2011). Economic value of Acinetobacter baumannii screening in the intensive care unit. Clinical Microbiology and Infection, 17, 1691–7.Find this resource: