Should screening programs for carbapenemase-producing strains be implemented in ICU patients?

In the last 10 years, Carbapenemase-producing Enterobacteriaceae (CPE), mainly Klebsiella pneumoniae, have broken out in most of the world's hospitals.^1,2 This outbreak poses a great challenge in the management of patients colonised or infected by CPE, which significantly affects the daily work of health professionals and even the structural organisation of hospital centres. The severe therapeutic limitations that usually occur in CPE infections lead to a delay in the initiation of adequate treatment, and often the impossibility of recommending an optimal treatment.¹ In general, infections caused by bacteria resistant to multiple antibiotics generate more prolonged hospitalisations, more frequent complications, including death, and a higher costs than infections caused by sensitive bacteria.³ Invasive infections by strains of carbapenemase-producing K. pneumoniae are associated with higher mortality rates than those caused by sensitive strains.^4–6 Although the attributable mortality rate is difficult to establish, as CPE often cause infections in patients with significant underlying disease, crude mortality rates ranging from 20% to 60% have been reported.^4–7

In addition to being one of the bacterial families that cause greater pathology in humans, enterobacteria, mainly Escherichia coli and K. pneumoniae, are part of our normal microbiota, which means that colonisation by CPE is not rare. The presence of colonised patients is one of the main routes for dissemination of CPE, as well as one of the causes of the persistence of outbreaks and endemics produced by this type of bacteria.^8,9

Enterobacterial infections are usually autoinfections caused by the same strains that colonise the human body, although there is little published evidence of CPE infections in previously colonised patients.¹⁰ Colonisation by multi-resistant bacteria may be a risk factor for poorer patient progress, as previously demonstrated with Acinetobacter baumannii and Pseudomonas aeruginosa.^11,12 According to a recent study carried out in two Greek intensive care units (ICU),¹³ patients colonised by CPE have a 1.79 times greater risk of dying than non-colonised patients.

Early detection of CPE carriers therefore has a dual purpose: epidemiological, to establish effective and early measures that limit rapid and widespread dissemination^9,14; and clinical, to prevent and control the possible downward progression of each patient.

Management of colonisation by CPE becomes even more complex in special risk units such as ICUs. The existence of increased antibiotic-related selective pressure, the presence of critical patients with multiple underlying pathologies particularly susceptible to infection, and an elevated risk of transmission between patients increase the impact of CPE in these units.¹⁵ In a study conducted in 2014 in 11 Spanish hospitals, the average incidence of colonisation during ICU stays was 1.35%, although 36% of the participating hospitals presented an incidence higher than 3%.¹⁶ These figures were obtained in a nation-wide epidemiological context of “regional CPE dissemination”,^16,17 a situation that has worsened to an “endemic” epidemiological context (unpublished personal data) in the last 3 years, which increases the risk of patients colonised by CPE being admitted to the ICU.

In this issue of Enfermedades Infecciosas y Microbiología Clínica, Maseda et al.¹⁸ publish a study conducted in a surgical ICU from January 2012 to December 2013.¹⁸ Out of the 254 patients over the age of 14 admitted to that ICU in the study period, 16.1% were CPE carriers.¹⁸ According to the authors, factors associated with colonisation by this type of bacteria included: previous treatment with broad-spectrum cephalosporins and to a lesser extent with β-lactams/β-lactam inhibitors, abdominal surgery and previous digestive or biliary endoscopy.¹⁸

In light of the above data, active screening for carbapenemase-producing strains in patients admitted to the ICU is a desirable and beneficial measure to reduce person-to-person transmission, prevent outbreaks and improve the prognosis of individual patients.^10,13,19,20 However, in practice, the decision to establish universal active screening in all patients entering the ICU should be considered in terms of another set of closely-linked factors. The epidemiological situation of the hospital and its health area can influence the utility and cost-benefit ratio of this intervention, which would be highly recommended in epidemic situations with active outbreaks caused by CPE.^9,10 However, selective screening for patients with certain risk factors for carrying a CPE could be considered as an alternative in situations of low prevalence.^10,15 These risk factors include, but are not limited to, previous hospital admissions in the last 6–12 months, previous hospitalisation in risk units (such as ICUs), coming from geographical areas with a high prevalence of CPE, having been previously colonised by CPE, prolonged use of antibiotics in the last 3 months or suffering from chronic pathologies susceptible to colonisation.^9,10,15 In addition, following the emergence of a first case of CPE in an ICU, it is recommended to conduct studies to detect colonisation in all hospitalised patients on a weekly basis up to 4 weeks after the last detected case.^19,20

Proper screening for CPE results in a significant workload at different levels, especially in the microbiology laboratories, which require an essential infrastructure and allocation of resources for this purpose. In addition, screening is only the first step in a chain of actions that lead to the control of CPE dissemination. It is a fundamental and necessary first step, whose effectiveness is entirely dependent on the degree of completion of interventions that should occur as a result of said screening.¹⁰ Therefore, before designing and implementing a CPE screening program in an ICU, each health centre must assess its actual capability for response in light of the information generated by the program. Furthermore, channels must be established for the early implementation of appropriate measures to become a reality. At times, the general, functional and even architectural structures of health systems are not specially prepared to address situations where there is a high prevalence of CPE.

The active monitoring of carriers has been successfully used in multifactorial strategies to control the spread of CPE,^13,21,22 although the specific contribution of this measure is difficult to assess given the largely “activating” nature of the other actions.

ICUs are units with specific characteristics that require individualised consideration. However, they are also integrated in a hospital setting with which they have important and inevitable interactions, so any activities aimed at controlling the dissemination of CPE require adequate coordination throughout the hospital as a whole.

The biological and transmission characteristics of CPE as well as the great clinical impact they can generate, especially in high-risk units such as ICUs, make the early detection of colonised patients highly recommendable. Therefore, even though the individual impact of active screening is difficult to establish since it is closely linked to the development of other measures, its use for the detection of CPE in ICUs is widely supported.^{9,10,14,15,19,20} Ideally, although universal screening of all patients admitted to the ICU would be advisable in a situation where the prevalence of CPE is high, as is currently the situation in Spain, the target population included in such screening could be adjusted according to epidemiological circumstances, care and structural aspects of both the ICU and hospital.