Rapid detection of methicillin-resistant Staphylococcus aureus directly from clinical samples: methods, effectiveness and cost considerations.

Methicillin-resistant Staphylococcus aureus (MRSA) isolates is a serious public health problem whose ever-increasing rate is commensurate with the pressure it is exerting on the healthcare system. At present, more than 20% of clinical S. aureus isolates in German hospitals are methicillin resistant. Strategies from low-prevalence countries show that this development is not necessarily inevitable. In the Scandinavian countries and the Netherlands, thanks to a rigorous prevention programme, MRSA prevalence has been kept at an acceptably low level (<1-3%). Central to these 'search and destroy' control strategies is an admission screening using several MRSA swabs taken from mucocutaneous colonisation sites of high-risk patients ('MRSA surveillance').It has also been reported that the speed with which MRSA carriage is detected has an important role to play, as it is a key component of any effective strategy to prevent the pathogen from spreading. Since MRSA culturing involves a 2-3 day delay before the final results are available, rapid detection techniques (commonly referred to as 'MRSA rapid tests') using PCR methods and, most recently, rapid culturing methods have been developed. The implementation of rapid tests reduces the time of detection of MRSA carriers from 48-72 to 2-5 h. Clinical evaluation data have shown that MRSA can thus be detected with very high sensitivity. Specificity however is sometimes impaired due to false-positive PCR signals occurring in mixed flora specimens. In order to rule out any false-positive PCR results, a culture screen must always be carried out simultaneously. The data provide preliminary evidence that a PCR assay can reduce nosocomial MRSA transmission in high-risk patients or high-risk areas, whereas an approach that screens all patients admitted to the hospital is probably not effective. Information concerning the cost-effectiveness of rapid MRSA tests is still sparse and thus the issue remains debated.

Introduction

Staphylococcus aureus (S. aureus) is one of most important bacterial pathogens in medicine today, accounting for a high proportion of cases of severe infection in both hospital and outpatient medical care. According to the findings of the German surveillance system of nosocomial infections, one has to assume that 18% of the 60,000 hospital infections that occur each year in intensive care are caused by S. aureus [1]. Of these, methicillin-resistant S. aureus (MRSA) strains account for a significant proportion: given that at least 15% to 20% of the clinical isolates are methicillin resistant [2], more than 2000 hospital MRSA infections have to be expected annually [1]. This is alarming because methicillin resistance in S. aureus not only means limited effectiveness of antibiotic treatment, but also leads to prolonged hospital stay and higher morbidity and mortality rates [3], [4].

In Germany, yet another year of increasing cases of MRSA continues the rising trend that prevailed in previous years. According to the 2007 study of the Paul Ehrlich Society, the average MSRA rate among German clinical S. aureus isolates is now 20.3% [2]. Thus, on a global scale, Germany falls in the middle of the MRSA ‘ranking’ list, while the Netherlands and Denmark show a remarkably low MRSA rate (<3%), which has remained stable throughout recent years [5]. This is largely due to a rigorous and coordinated policy of hygiene measures. Japan and the USA, on the other hand, dominate the top rankings, reporting the highest MRSA rates worldwide, which now average around 50% [6].

On-admission screening cultures for MRSA are one of the mainstays of the successful ‘search and destroy’ infection control policy in the Netherlands. Patients who are to be hospitalised are systematically screened for MRSA carriage, depending on their individual risk profile. Although similar guidelines (published in 2004 by the Commission for Hospital Hygiene and Infection Prevention (KRINKO)) are also available for Germany (Table 1 (Tab. 1)) [7], many institutions have never attempted to implement an active surveillance programme for financial reasons. Instead, clinicians in these hospitals rely on the ‘passive’ acquisition of MRSA information from clinical culture. However, by doing so, 38% to 77% of MRSA carriers remain undetected, or are detected too late, and thus may act as a potential reservoir for MRSA dissemination [8], [9], [10]. Despite the fact that culture-based MRSA screening swabs have proven to be cheap, sensitive and practicable, the delay between sample acquisition and reporting of results remains a significant drawback. Reliable identification and testing results are usually available only 48–96 h after sample collection, and during this time MRSA cross-transmission could occur if patients are not placed under contact precautions (‘precautionary isolation’) [11], [12], [13]. As these measures may be unnecessary or, if not applied, unidentified MRSA-positive individuals may remain a hidden reservoir for cross-transmission, the need for speedier methods to detect MRSA is widely acknowledged.

Table 1

Patient groups that should be screened for MRSA carriage

Rapid MRSA identification

Conventional screening for methicillin-resistant S. aureus generally relies on plate-based culture methods with or without prior broth enrichment. Any growth of S. aureus on primary plates is considered suspect and processed for positive identification and antimicrobial susceptibility testing. Susceptibility testing can be performed by either manual procedures (disk susceptibility testing or agar dilution) or by using one of the current automated microbiology systems. However, as methicillin resistance is difficult to recover from low inoculum or mixed flora samples, traditional methods are labour intensive and time-consuming and may necessitate a further 2 to 3 days to confirm positives [14], [15]. Although culture-based methods conform to the MRSA screening standard, speedier testing is of course desirable in order to resolve (or continue) precautionary infection control measures.

In general, a reduction in diagnostic turnaround time can follow two paths: either by rapid confirmation of methicillin resistance in positive cultures of S. aureus or by rapid molecular or non-molecular detection of MRSA directly from the patient sample. For rapid confirmation of MRSA in pure cultures, several new methods have been developed in recent years, foremost the chromogenic media, the PBP-2a latex agglutination test and the mecA PCR using colonies from overnight cultures [15], [16]. Generally, these methods can speed up the identification of MRSA, but they cannot shorten the incubation steps (24–48 h) required after the sample reaches the laboratory. Thus, the methods mentioned above are useful in terms of speeding up diagnosis, but the maximum saving in time is not more than 1 day and therefore the overall benefit of these methods remains limited. As a result, rapid detection methods have been developed where either the primary culture – the time-limiting step – is no longer necessary (PCR) or the incubation times are much shorter (approximately 5 h) than during conventional procedures (rapid culture techniques). See Table 2 (Tab. 2).

Table 2

MRSA rapid tests

MRSA rapid culture

BacLite MRSA is the first example of a rapid non-molecular MRSA screening test. This new commercial rapid culture-based assay was developed by 3M Company. The procedure does not rely on discernible colonies growing on the primary plates; rather, the presence of bacteria is measured by adenylate kinase (AK) activity. In the assay, AK detection is combined with a selective broth enrichment (which contains cefoxitin, ciprofloxacin and colistin and thus pre-enriches methicillin-resistant staphylococci), magnetic microparticle extraction and selective (lysostaphin) lysis to add target organism specificity [17]. The kit comes complete with the reagents and media needed to run the assay, and analysis occurs in automatic steps inside the BacLite instrument. The test allows negative results to be confidently reported within 5 h [17]. Current evaluation data of the BacLite test originating from clinical or comparative studies are still sparse and so the final verdict is not yet in. However, preliminary results are encouraging: sensitivity (specificity) reached 94.6% (96.9%) in nasal swabs and 95.9% (88.8%) in inguinal swabs [18], [19]. In another study, the new assay was shown to detect without exception all MRSA strains in large collections of strains comprising highly diverse genetic backgrounds [20].

PCR-based detection of MRSA directly from the clinical sample

Most molecular methods for identification of methicillin resistance in S. aureus have been PCR based. The current protocols do not usually include any bacterial culture and thus allow turnaround times in the range of only 1 to 5 h (Table 2 (Tab. 2)). Since methicillin resistance is caused by a mecA gene product, the low-affinity penicillin-binding protein (PBP2'), the first PCR attempts to detect MRSA directly from the patient sample were refined from those mecA protocols that had originally been used for mecA confirmation of a pure S. aureus culture. As clinical samples often contain both coagulase-negative staphylococci (CoNS) and S. aureus, either of which can carry mecA, detection of the mecA gene alone is not sufficient for discriminating between MRSA and methicillin-resistant CoNS in a mixed flora clinical sample. Thus, several multilocus PCRs that simultaneously amplify DNA sequences specific for both the species (e.g., nuc, clfA, fem, or 16S rRNA) and methicillin resistance (mecA) have been proposed. However, if directly used on specimens rather than on cultured bacteria, these assays are again unable to differentiate between methicillin-susceptible S. aureus and methicillin-resistant CoNS in mixed cultures [21]. The only way to reliably detect MRSA is to make sure that the mecA signal found definitely originates from an S. aureus and not from a coexisting S. epidermidis or S. haemolyticus [21].

Multilocus PCR protocols

A variety of strategies have been attempted to counter the specificity problem. Only two of these have been found to be sufficiently suitable for routine use (Table 2 (Tab. 2)). The multilocus PCR protocols (the ‘first generation’ in MRSA PCR testing) enable one to increase specificity, in that the original mecA and nuc gene loci have been augmented with further loci specific to frequently encountered CoNS (such as S. epidermidis and/or S. haemolyticus). On the basis of the patterns of the PCR products, one can now deduce from which species (S. aureus: nuc; or CoNS: specific marker gene) the mecA gene has probably been detected. There are several commercial tests that exploit this principle: e.g., hyplex StaphyloResist (BAG) or LightCycler Staphylococcus + MRSA Kit (Roche Diagnostics). The only drawback of the multilocus systems appears when the pattern is ambiguous, showing the presence of all three PCR signals (mecA plus nuc plus CoNS maker gene). In this case, it is not possible to determine from which species the mecA gene has been detected, and the only way to resolve this ambiguity is to incubate the sample and test the culture isolates (which cannot be considered rapid). Although the frequency of ambiguous tests resulting from mixed flora specimens seems to be relatively low (<5%) [22], [23], a substantially higher false-positive rate (approximately 20%) has been acknowledged by other authors [24]. Basically, all systems that rely on multilocus PCR are capable of producing quick and accurate results (Table 2 (Tab. 2)) [25], [26], [27], but they can still be impaired by the presence of mecA-positive CoNS in the sample. Sometimes a definitive diagnosis is only possible by culturing the swab [21].

SCCmec-PCR / single-locus PCR

In 2004, a new real-time PCR concept involving the amplification of DNA sequences in the region of the open reading frame orfX, where the staphylococcal cassette chromosome mec (SCCmec) integrates with the S. aureus chromosome, was published [28]. Unlike earlier assays targeting the separate detection of mecA and several different marker genes, this assay yields only one amplification product (mecA-orfX) and is therefore often referred to as ‘single-locus’ PCR. Because the chromosomal orfX is almost always S. aureus specific, an amplification product can only be detected in mecA positive S. aureus, but not with mecA positive CoNS. As with the SCCmec technique, it is possible to reliably prove the presence of MRSA from a mixed flora specimen, without running the risk of a false-positive result due to CoNS. Thus the new methods belong to a ‘new generation’ of MRSA rapid tests [28], [29]. There are any number of publications illustrating that the SCCmec PCR combines the advantages of sensitivity and specificity, thus leading to convincing performance data (Table 2 (Tab. 2)) [28], [29], [30], [31], [32], [33], [34]. At present, the SCCmec assay is commercially available as Xpert MRSA (Genzyme Virotech), BD GeneOhm MRSA (Becton Dickinson), GenoType MRSA Direct and GenoQuick MRSA Direct (both Hain Lifesciences) (Table 2 (Tab. 2)).

There are, however, pitfalls too. Mainly, the SCCmec cassette has proven to be unstable where the chromosomal orfX fragment merges with the mecA gene, resulting in false-negative PCR results. This has only been observed occasionally, but as numerous examples in medical bacteriology have proven, the single cases of today can proliferate out of control, leading to severe diagnostic (and therapeutic) problems in future. False-positive results have also been observed. Possible reasons include an orfX gene in CoNS that is homologue in sequence to that of S. aureus [29], [35], or an SCCmec cassette from which the mecA gene has been deleted [36], [37]. In rare cases, the mecA gene is replaced by a different gene that also produces a false-positive result [29]. Time will tell how problematic these impairments may become. At present, based on basic practical experience, the SCCmec concept is far superior to multilocus testing.

Rapid, bedside MRSA testing

Current PCR formats that rely on laboratory-based protocols inevitably introduce a delay in the production of results, because the sample has to be transferred to the laboratory. In addition, many laboratories collect samples over a period of a few hours (or even a few days) in order to test them in batches. Thus, the maximum speed that can be achieved is limited to 3–5 h for conventional PCR, and 1–2 h for real-time PCR systems (GeneOhm MRSA; Light Cycler Staphylococcus/MRSA Kit) (Table 2 (Tab. 2)). To further reduce the time from screening to notification of test results, the POCT (point-of-care testing) concept has very recently been applied to MRSA testing. One newly developed assay is called ‘Xpert MRSA’ and is performed on a closed, self-contained, fully integrated and automated platform (GeneXpert DX instrument; Genzyme Virotech), which represents a paradigm shift in the automation of molecular analysis, producing accurate molecular results on demand. Specimens do not need to be tested in batches; rather, due to autonomous PCR modules, testing can occur at any time, on any day. The Xpert MRSA assay fully integrates and automates all the steps that are involved in PCR processing (sample preparation, amplification and detection) in one disposable cartridge [38], [39]. The assay is based on the SCCmec concept and each run can be completed in about 1 hour from the time of sample acquisition. As the GeneXpert device requires little operator handling and specialised knowledge, the assay can be installed and performed in the immediate vicinity of the patient (e.g. on the clinical ward or in the emergency room). The advantages of Xpert MRSA have to be weighed against the disadvantages, however: non-laboratory personnel have to be trained in the method in order to ensure testing competency. Adequate quality management may be considered another drawback to Xpert MRSA. If there is insufficient quality control, the risk that erroneous data are seen and acted upon can be high. Lastly, relative to the batch testing protocols, the running costs of the Xpert MRSA assay (= reagents, consumables) are quite high, and performing the procedure means an increased workload for the hospital staff.

Sample collection

An important consideration when implementing an active surveillance programme is the question of what sites should be cultured to sufficiently detect MRSA colonisation. The most common carriage site for MRSA is the anterior nares. Culturing additional sites such as the throat, groin, axilla, wounds, non-intact skin surfaces or other sites (depending on the patient’s risk profile) will increase the sensitivity of the screens, but may be inappropriate in terms of cost, time and resources. Thus, most guidelines recommend a combination of nose, throat and skin lesion (wound), yielding the highest sensitivities [7], [25], [40]. There is significant evidence that a similar sampling scheme is appropriate for molecular testing methods as well [31]. Unfortunately, some test systems are licensed only for nose swabs, probably because the PCR methods sometimes fail to screen sites other than mucocutaneous colonisation sites due to the presence of inhibitors (e.g., mucus, pus or blood) that will lead to a false-negative result. In fact, some authors report inhibitory rates of, in part (depending on the reagent batch), 11% [31]. Better extraction protocols are being developed in order to overcome the inhibitory effect of clinical samples for PCR testing.

Effectiveness of MRSA rapid tests

Many clinical studies and models have shown that MRSA screening has a generally positive effect in terms of prevention of new infection and reduction of MRSA transmission [41]. This assessment has been generally accepted and both the Robert Koch Institute (Table 1 (Tab. 1)) and other national institutions have issued a recommendation to this effect [7], [25]. The introduction of an active screening programme has been shown in many studies to have significantly reduced the rate of nosocomial MRSA transmission [42], [43]. For example, a study carried out in the Berlin Vivantes Clinic in Friedrichshain proved significantly that 48% of hospital MRSA cross-infections can be prevented reliably using an active screening programme [43].

Most of the publications to date use data obtained from culture-based screening strategies. PCR data are rare, and the question still remains whether or not the PCR advantage in turnaround times will actually reduce MRSA cross-infection (Table 3 (Tab. 3)) [13], [31], [44], [45], [46], [47], [48], [49]. One has to bear in mind that in most studies active MRSA screening is only one element in a broad range of measures (e.g., isolation policy) to drive down infection rates [50]. It is therefore not surprising that, depending on the study protocol and the underlying hygiene strategy, the assessment of the effectiveness of MRSA rapid tests still remains difficult and leads to inconsistent data (Table 3 (Tab. 3)). Increasingly, it is becoming apparent that using MRSA rapid tests can realise a reduction in transmission and infection rates when it is targeted to patients who are colonised to a large degree with MRSA or who undergo elective procedures with a high risk of MRSA infection (Table 3 (Tab. 3)). Two extremes prove this point: Cunningham et al. found a substantial reduction in transmission rates in patients at intensive care unit admission [48], whereas a team from Geneva University was unable to reduce the frequency of nosocomial infection using a widespread rapid screening on admission compared with standard MRSA control alone [45].

Table 3

Studies evaluating the clinical efficacy (reduction in MRSA transmission) and cost-efficiency of PCR-based MRSA screening

Cost considerations

Because of pressures to keep costs low, new techniques are always received sceptically by most cost bearers of healthcare facilities. This is the case with elaborate culturing and in particular the introduction of molecular testing. Culturing for MRSA costs between 3 and 5 euros (negative result) and 5 to 10 euros (positive result); if, however, testing is augmented by PCR, this increases the cost by at least 15 to 20 euros. Precautionary isolation of a patient while awaiting the final results of screening also incurs incremental costs and ties up personnel and organisational resources. These additional charges must be balanced against the costs incurred by those cases of MRSA infection that are detected too late or not at all. Such costs include prolonged hospital stay and alternative antibiotics. Under the terms of the German diagnosis-related groups (DRG) payment system, the average total loss per patient with MRSA infection has been estimated at approx. 5700 euros; this corresponds to approx. 600 euros per day. These figures do not include the intangible costs (costs which cannot be calculated as such), e.g. lost revenue resulting from a drop in referrals to clinics where MRSA is highly prevalent or the burden on the national health system resulting from ever increasing MRSA resistance.

Even without considering intangible or ‘societal’ costs, the financial effects of PCR-based MRSA screening are difficult to estimate, as the findings are strongly influenced by a plethora of competing determinants and confounding factors [49]. The issue is further complicated by different structures, organisational arrangements and hygiene policies across German hospitals [49]. According to the aforementioned study carried out in the Berlin Vivantes Clinic in Friedrichshain (tertiary care hospital, 668 beds, 30,000 admissions/year), it is justifiable to recommend an active screening programme, as it reduces the rate of MRSA transmission and infection, leading to savings of direct medical costs of 110,000 euros/year [51]. One other study has also been published that sought to quantify the cost of the Netherlands’ MRSA policy. The following topics were considered: personnel expenditures, material costs, treatment costs, decontamination measures, lost revenue and lack of staff. The financial consequences were compared to those in a hypothetical situation without the search-and-destroy policy. The authors conclude that without the strict search-and-destroy strategy, the costs associated with the use of alternative antibiotics would be at least twice as high as the costs expended in the actual situation [52].

However, the few investigations into the cost of S. aureus screening have focused mainly on culture-based MRSA detection and were performed at tertiary care hospitals with high MRSA rates. The question of whether the findings can be transferred to PCR techniques and a different epidemiological context is still being debated (Table 3 (Tab. 3)). Even German university hospitals disagree as to whether they should screen possible MRSA carriers by PCR. In Heidelberg, the monetary benefit deriving from PCR-based testing is considered to be approx. 5 times higher than the extra costs (even when counting the expenses for precautionary isolation) [31], whereas in Regensburg cost-effectiveness could not be proved [49].

Summary

In summary, rapid MRSA tests are medically reasonable tools for the timely detection of MRSA carriers, which may be particularly useful in the screening of some high-risk groups of patients. In these patients, rapid MRSA findings can be of twofold value: not only are they in the interests of patients possibly infected with MRSA (in order to start adequate treatment as early as possible), but they also serve to protect other patients from spreading the pathogen. Focusing on the cost/benefit ratio, there is still uncertainty about whether the rapid technologies will lead to overall cost savings. Regarding low-risk patients, it seems that the molecular techniques are neither medically nor economically justified. Thus it is recommended that any hospital that wishes to establish an active surveillance programme should carry out an assessment of the medical and economical value on the basis of its own conditions (percentage of patients at risk, local MRSA rate, organisational arrangements, hygiene policy etc.) [49], [53].

The current MRSA rapid tests fall into one of two categories: PCR-based methods (five systems commercially available, one even suitable for point-of-care testing) and rapid culturing techniques (one system commercially available). All tests can detect MRSA directly from the clinical sample within a few hours and with good sensitivity and specificity. Whereas rapid testing seems to be a reliable way of demonstrating MRSA absence, positive PCR results always need confirmation with culture in order to exclude false-positive results or to gain isolates for further testing (susceptibility, virulence factors).

Glossary

Thermocycler: instrument that repeatedly cycles through various temperatures, thus automating the polymerase chain reaction

Diagnosis-related groups, DRG: German hospital reimbursement system

Incidence: the rate at which new cases occur in a population during a specified period

mecA: gene encoding for methicillin resistance in staphylococci, which is carried on a mobile genetic element termed the staphylococcal chromosome cassette mec (SCCmec)

Methicillin resistance: resistance phenotype (caused by the mecA determinant), conferring cross-resistance to most currently available beta-lactam antibiotics

MRSA: methicillin-resistant Staphylococcus aureus

Negative predictive value, NPV: the proportion of patients with negative test results who are correctly diagnosed

PCR: polymerase chain reaction

Prevalence: the proportion of a population that is affected by the disease at a specific time

Positive predictive value, PPV: the proportion of patients with positive test results who are correctly diagnosed

Point-of-care testing, POCT: diagnostic testing that is performed near to or at the site of patient care with the result leading to a possible change in the care of the patient

Real-time PCR: the PCR signal increases in direct proportion to the amount of PCR product produced and can be monitored at each cycle

RKI: Robert Koch Institute, Berlin

Screening: systematic examination or assessment, performed especially to detect hidden MRSA carriers

Surveillance: active screening programme, performed especially in those patients prone to carry MRSA.

Notes

Acknowledgements

The author thanks Mrs Rachel Murphy for her helpful assistance in writing the English version of the manuscript and Dr Alexander Zitzer for helpful discussions on the topic and for his contribution in putting together the data which are presented in Table 2.

Conflicts of interest

There are no conflicting interests to declare.