Heated birthing pools as a source of Legionnaires' disease.

In June 2014 Public Health England confirmed a case of Legionnaires' disease (LD) in a neonate following birth at home in a hired birthing pool incorporating a heater and a recirculation pump which had been filled in advance of labour. The case triggered a public health investigation and a microbiological survey of an additional ten heated birthing pools hired or recently hired to the general public across England. The birthing pool used by the parent of the confirmed case was identified as the source of the neonate's infection following detection of Legionella pneumophila ST48 in both patient and environmental samples. Legionella species were detected by quantitative polymerase chain reaction but not culture in a further three pools together with other opportunistic pathogens identified by culture and matrix-assisted laser desorption ionization-time of flight (MALDI-ToF) mass spectrometry. A Patient Safety Alert from NHS England and Public Health England was issued stating that heated birthing pools filled in advance of labour should not be used for home births. This recommendation remains in place. This investigation in conjunction with other recent reports has highlighted a lack of awareness regarding the microbiological safety of heated birthing pools and their potential to be a source of LD and other opportunistic infections. Furthermore, the investigation raised important considerations with regards to microbiological sampling and testing in such incidents. Public health authorities and clinicians should consider LD in the differential diagnosis of severe respiratory infection in neonates within 14 days of a water birth.

Design and validation of a qPCR assay for accurate detection and initial serogrouping of Legionella pneumophila in clinical specimens by the ESCMID Study Group for Legionella Infections (ESGLI).

Prompt detection of Legionella pneumophila is essential for rapid investigation of legionellosis. Furthermore, as the majority of L. pneumophila infections are caused by serogroup 1 (sg1) strains, rapid identification of such strains can be critical in both routine and outbreak scenarios. The ESCMID Study Group for Legionella Infections (ESGLI) was established in 2012 and immediately identified as a priority the validation of a reliable, easy to perform and interpret, cost-effective qPCR assay to standardise the detection of L. pneumophila DNA amongst members. A novel L. pneumophila assay targeting the mip gene was designed and combined with previously published methodologies amplifying the sg1 marker (wzm) and the green fluorescent protein gene (gfp) internal process control. The resulting triplex assay was validated internationally on the three qPCR platforms used by the majority of European Legionella reference laboratories: ABI 7500 (Life Technologies), LightCycler 480 Instrument II (Roche) and Rotor-Gene Q (Qiagen). Clinical and EQA specimens were tested together with a large panel of strains (251 in total) to validate the assay. The assay proved to be 100% specific for L. pneumophila and sg1 DNA both in silico and in vitro. Efficiency values for mip and wzm assays ranged between 91.97 and 97.69%. Limit of detection values estimated with 95% confidence were adopted for mip and wzm assays on all three qPCR platforms. Inhibition was not observed. This study describes a robust assay that could be widely implemented to standardise the molecular detection of L. pneumophila among ESGLI laboratories and beyond.

Design and application of a core genome multilocus sequence typing scheme for investigation of Legionnaires' disease incidents.

Sequence-based typing (SBT) for Legionella pneumophila (Lp) has dramatically improved Legionnaires' disease (LD) cluster investigation. Microbial whole genome sequencing (WGS) is a promising modality for investigation but sequence analysis methods are neither standardised, nor agreed. We sought to develop a WGS-based typing scheme for Lp using de novo assembly and a genome-wide gene-by-gene approach (core genome multilocus sequence typing, cgMLST). We analysed 17 publicly available Lp genomes covering the whole species variation to define a core genome (1,521 gene targets) which was validated using 21 additional published genomes. The genomes of 12 Lp strains implicated in three independent cases of paediatric humidifier-associated LD were subject to cgMLST together with three 'outgroup' strains. cgMLST was able to resolve clustered strains and clearly identify related and unrelated strains. Thus, a cgMLST scheme was readily achievable and provided high-resolution analysis of Lp strains. cgMLST appears to have satisfactory discriminatory power for LD cluster analysis and is advantageous over mapping followed by single nucleotide polymorphism (SNP) calling as it is portable and easier to standardise. cgMLST thus has the potential for becoming a gold standard tool for LD investigation. Humidifiers pose an ongoing risk as vehicles for LD and should be considered in cluster investigation and control efforts.

Barrow-in-Furness: a large community legionellosis outbreak in the UK.

A community outbreak of legionellosis occurred in Barrow-in-Furness, Cumbria, during July and August 2002. A descriptive study and active case-finding were instigated and all known wet cooling systems and other potential sources were investigated. Genotypic and phenotypic analysis, and amplified fragment length polymorphism of clinical human and environmental isolates confirmed the air-conditioning unit of a council-owned arts and leisure centre to be the source of infection. Subsequent sequence-based typing confirmed this link. One hundred and seventy-nine cases, including seven deaths [case fatality rate (CFR) 3·9%] were attributed to the outbreak. Timely recognition and management of the incident very likely led to the low CFR compared to other outbreaks. The outbreak highlights the responsibility associated with managing an aerosol-producing system, with the potential to expose and infect a large proportion of the local population and the consequent legal ramifications and human cost.

Bartonella species as a cause of infective endocarditis in the UK.

Bartonella spp. are increasingly implicated in infectious endocarditis cases in the UK. Herein, we attempted to quantify their role in this syndrome and explored the epidemiology of Bartonella-associated endocarditis in the UK. Between November 2005 and October 2010, samples from 685 endocarditis patients were submitted to the Health Protection Agency for Bartonella serology. Serological evidence of infection was obtained for 57 (8·3%) patients. PCR-based evidence of infection was obtained from 13 out of 14 patients for whom heart valve tissue was available, with Bartonella quintana implicated in 12 cases and B. henselae in one. Six patients with B. quintana endocarditis were recent immigrants into the UK, of whom four lived in poor socioeconomic conditions. These results indicate that Bartonella is a not uncommon cause of endocarditis in the UK and should be considered particularly in patients raised in eastern Europe and/or with a history of homelessness or alcoholism.

Application of Legionella pneumophila-specific quantitative real-time PCR combined with direct amplification and sequence-based typing in the diagnosis and epidemiological investigation of Legionnaires' disease.

The detection of Legionella pneumophila DNA in clinical specimens using quantitative real-time polymerase chain reaction (qPCR) combined with direct sequence-based typing (SBT) offers rapid confirmation and timely intervention in the investigation of cases of Legionnaires' disease (LD). We assessed the utility of a specific L. pneumophila qPCR assay targeting the macrophage infectivity potentiator (mip) gene and internal process control with three clinical specimen types from confirmed LD cases. The assay was completely specific for L. pneumophila, as demonstrated by positive results for 39/39 strains from all subspecies and 16 serogroups. No cross-reaction was observed with any of the 54 Legionella non-pneumophila (0/69 strains) or 21 non-Legionella (0/58 strains). All L. pneumophila culture-positive respiratory samples (81/81) were qPCR-positive. Of 80 culture-negative samples tested, 47 (58.8%) were qPCR-positive and none were inhibitory. PCR was significantly more sensitive than culture for samples taken ≤ 2 days of hospitalisation (94.7% vs. 79.6%), with the difference being even more marked for samples taken between 3 and 14 days (79.3% vs. 47.8%). Overall, the sensitivity of the qPCR was ∼30% greater than that of culture and direct typing on culture-negative PCR-positive samples resulted in full 7-allele profiles from 23/46, 5 to 6 alleles from 8/46 and ≥ 1 allele from 43/46 strains.

Travel-associated Legionnaires disease in residents from England and Wales travelling to Corfu, Greece, August to October 2011.

Fourteen cases of Legionnaires' disease were confirmed in residents from England and Wales with history of travel to Corfu, Greece, in the 14 days before symptom onset. These cases were reported to the Health Protection Agency national surveillance scheme for Legionnaires' disease in residents of England and Wales between August and October 2011. In addition, one case in a Greek national and a case of non-pneumonic legionellosis in a resident from Scotland were also reported. Few cases shared the same accommodation site in Corfu during their incubation period. Epidemiological investigations and microbiological analysis of clinical and environmental samples excluded a single source but rather implicated several accommodation sites as sources of sporadic infection. Control measures have since been implemented at these accommodation sites and no further cases have been reported. This incident highlights the value of epidemiological typing and the importance of effective international response to control and prevent legionella infection.

An outbreak of Legionnaires' disease associated with a display spa pool in retail premises, Stoke-on-Trent, United Kingdom, July 2012.

Twenty-one confirmed cases of Legionnaires' disease (Legionella pneumophila serogroup 1) were identified in the Stoke-on-Trent area of England with onsets since 2 July 2012. Sequence-based typing results are available for nine cases; all are a unique type (ST1268). Initial interviews highlighted a number of possible environmental sources. Inspection of premises of interest revealed an operating spa pool on display, from which the outbreak strain was identified. All cases had visited the retail premise with this spa pool.

Mycoplasma pneumoniae infection in primary care investigated by real-time PCR in England and Wales.

Real-time PCR was employed to detect a conserved region of the P1 cytadhesin gene of Mycoplasma pneumoniae in combined nose and throat swabs collected from patients attending GP surgeries during 2005-2009 with symptoms of respiratory tract infection (RTI). Samples were collected as part of an annual winter epidemiological and virological linked study in England and Wales. A total of 3,987 samples were tested, 65 (1.7%, 95%CI 1.3-2.1) had detectable M. pneumoniae DNA. Positive patients were detected of both gender, aged from 9 months to 78 years, who had clinical signs of upper RTI, fever and/or myalgia, an influenza-like illness to lower RTI. Mixed infections were identified in four cases, two with influenza A H1, one with H3 and one with influenza B. Children aged 5-14 years were more likely to have detectable M. pneumoniae in samples than all other age groups (Fishers p = 0.03), attributed to the 2005-2006 season in which 6.0% (12/200, 95%CI 3.4-10.3) of 5-14 year olds had detectable M. pneumoniae in comparison to 2.2% in 2006-2007 (3/141 95%CI 0.5-6.4), 2.2% in 2007-2008 (2/89 95%CI 0.1-8.3) and 0% in 2008-2009 (0/151 95%CI 0-2.9).

Antimicrobial susceptibility testing of historical and recent clinical isolates of Bordetella pertussis in the United Kingdom using the Etest method.

Reports of the development of antimicrobial resistance by Bordetella pertussis to macrolides in the United States and Taiwan, together with a recent increase in pertussis notifications and laboratory-confirmed cases in England and Wales in 2008, prompted the examination of historical and recent clinical isolates from patients for evidence of such resistance in our collection. Isolates submitted to our laboratory as part of the enhanced surveillance scheme for pertussis, from 2001 to 2009, were tested against three agents, erythromycin, clarithromycin and azithromycin, by the Etest (bioMérieux) method. All isolates (n = 583) were fully susceptible to all three agents tested (minimum inhibitory concentrations [MICs] 256 microg/ml. Although no evidence of resistance was found in the strains tested from the United Kingdom, screening for antimicrobial resistance of B. pertussis may be warranted in cases that are unresponsive to macrolide treatment and to provide early warning of such emergence in the future.

Investigating a cluster of Legionnaires' cases: public health implications.

OBJECTIVES: To describe the multidisciplinary investigation and management of a rapidly increasing number of cases of Legionnaires' disease in the North Shropshire area, UK during August 2006. STUDY DESIGN: Epidemiological and environmental investigation of a cluster of cases of Legionnaires' disease. METHODS: Outbreak investigation included: agreeing case definitions; case finding; epidemiological survey; identification and environmental investigation of potential sources; microbiological analysis of clinical and environmental samples; mapping the location of potential sources; and the movement and residence of cases. RESULTS: Three cases of Legionnaires' disease were admitted to a local hospital between 30 and 31 August 2006. Two of these cases were Shropshire residents, with the third living in Wales. A fourth case was also identified which, it was thought, may have been linked to this cluster as the patient had a history of travel to the same area as the two Shropshire residents. Over the next few weeks, three more cases were identified, two of whom were admitted to hospital. Subsequent detailed environmental, epidemiological and microbiological investigation did not support the hypothesis that any of these cases could be linked to a common source. CONCLUSIONS: The results of this investigation strongly suggest that a single source was not responsible for the cluster, and it was concluded that this incident was a pseudo-outbreak. This investigation serves as a reminder that clusters can and do occur, and that an apparent outbreak may be a collection of sporadic cases distinguishable only by rigorous epidemiological, environmental and microbiological investigation.

Distribution of Legionella pneumophila serogroups, monoclonal antibody subgroups and DNA sequence types in recent clinical and environmental isolates from England and Wales (2000-2008).

Clinical isolates of Legionella pneumophila, obtained from 167 patients, who acquired their illness in the community in England and Wales between January 2000 and March 2008, were compared with 276 environmental isolates of L. pneumophila obtained over the same period as part of the routine sampling of 'managed' water systems. The 443 isolates were typed by monoclonal antibody (mAb) subgrouping and the internationally standardised, seven-gene loci, sequence-based typing (SBT) scheme of the European Working Group for Legionella Infections (EWGLI). Of the clinical isolates, 97.6% were L. pneumophila serogroup (sgp) 1, compared with only 55.8% of environmental isolates (P = 0.0002); 91.6% were subgrouped as mAb3/1+ve, compared with only 8.3% of environmental isolates (P < 0.0001). The isolates were very diverse, with SBT identifying 111 sequence types (STs) (index of diversity [IOD] 0.954). Among the clinical isolates, 42 ST were seen, with one (ST47) accounting for 25.7% and three (ST47, ST37 and ST62) accounting for 46.1% of all isolates. Eighty-two STs were identified among the environmental isolates, with two (ST1 and ST79) accounting for 34.1% of these. Comparison of the STs seen among clinical and environmental isolates showed that there was very little overlap between the two populations (P < 0.0001), with common clinical strains found in the environment very infrequently: 0.4, 0.7 and 0% (ST47, ST37 and ST62, respectively), and common environmental strains rarely causing disease: 4.8 and 1.2% (ST1 and ST79, respectively). Combining phenotypic and genotypic data identified 144 phenons (IOD 0.970); 52 among clinical isolates and 101 among environmental isolates. The most abundant clinical strain, mAb 'Allentown' ST47, accounted for 22.8% of cases, but was only found once in the environment. Conversely, mAb 'Oxford/OLDA' ST1 was the most common environmental strain (17.0%), but only caused two infections. A review of the published data shows that mAb 'Allentown' ST47 is also an important cause of infection in France and possibly in the Netherlands. However, it was not found in a large study of German clinical isolates. This study confirms previous work showing that just a few strains of L. pneumophila cause the majority of community-acquired Legionella infection in England and Wales, and that these clinically significant strains are only rarely found in managed water systems. These data suggest that knowing which particular strain is present in an environment might be at least as important as knowing the quantity in which legionellae are present.

Pertussis outbreak on a neonatal unit: identification of a healthcare worker as the likely source.

We describe the investigation and containment of an outbreak of pertussis on a neonatal unit. Bacterial culture, polymerase chain reaction (PCR) and serology were used to confirm suspected cases. Two infants with pertussis were identified and a nurse with prolonged cough was traced as the likely source. Control interventions included mass chemoprophylaxis of healthcare workers and patients and exclusion from work of healthcare workers with cough. The use of PCR allowed rapid assessment of the extent of the outbreak. This outbreak highlights the risk to hospitalised infants posed by circulation of Bordetella pertussis in young adults and illustrates the utility of PCR in rapidly assessing the extent of outbreaks. Prevention strategies such as universal vaccination of adolescents, or selective vaccination of healthcare workers, should be considered in the UK.

Comparison of clinical and environmental isolates of Legionella pneumophila obtained in the UK over 19 years.

Between January 1980 and December 1998, 3458 cases of Legionnaires' disease were reported to the national surveillance scheme in England and Wales. Of these, 463 (13.4%) were reported as proven by culture and isolation of Legionella spp., with 96.3% being Legionella pneumophila. Serogroup (Sgp), monoclonal antibody (mAb) subgrouping and restriction fragment length polymorphism (RFLP) analysis data were obtained for 321 (69.3%) of these, of which 284 were classified as being unrelated to any other isolate in the study. Typing data were also available for 117 unrelated environmental isolates of L. pneumophila obtained from England and Wales, giving a total of 401 unrelated isolates in the study. Of the clinical isolates, 88.0% were Sgp1, compared with only 42.7% of environmental isolates (p <0.001); 79.6% of clinical isolates were subgrouped as mAb2+, compared with only 12.8% of environmental isolates (p <0.001). RFLP typing identified 67 types among the 401 isolates, with clinical isolates showing significantly less diversity than environmental isolates (index of diversity (IOD) 0.944 vs. 0.958; p <0.05), with three RFLP types (1, 5 and 14) accounting for 40.0% of all clinical isolates. Combining the phenotypic and genotypic data resulted in 173 distinct phenons, with clinical isolates showing significantly less diversity than environmental isolates (IOD 0.964 vs. 0.996; p <0.01). Three phenons accounted for 30% of all clinical isolates. These data strongly suggest that some strains of L. pneumophila are more likely to cause human infection than would be expected from their distribution in the environment.

Identification of Legionella spp. by 19 European reference laboratories: results of the European Working Group for Legionella Infections External Quality Assessment Scheme using DNA sequencing of the macrophage infectivity potentiator gene and dedicated online tools.

Identification of Legionella spp. can be achieved by DNA sequencing of the macrophage infectivity potentiator (mip) gene. The External Quality Assurance (EQA) scheme described in this report is the first to assess the proficiency of laboratories using this methodology. The results obtained from two EQA distributions sent to European reference laboratories involved in Legionella outbreak control and environmental monitoring are presented. Each distribution contained a panel of ten coded Legionella strains. All strains were from clinical and environmental sources and were considered to be wild-type strains. Participants used dedicated online tools to compare sequence text files against a database of known Legionella spp. The majority of centres (seven of ten, and 11 of 12) correctly identified all strains tested, in the first and second distributions, respectively. Typically, sequence similarity values of 98-100% were obtained when the test strains were compared with sequences contained in the database. In all but one case, lower values indicated a poor quality sequence. The exception was associated with the identification of a putative new species in the first panel. Genotypic identification of Legionella can be achieved by the use of standard protocols, dedicated identification libraries, and online tools. EQA schemes provide an independent measure of performance, and it is recommended that laboratories performing these techniques participate in such schemes, thereby allowing optimisation of and improvements in their performance.

Rapid detection and evolutionary analysis of Legionella pneumophila serogroup 1 sequence type 47.

OBJECTIVES: Legionella pneumophila serogroup 1 (Lp1) sequence type 47 is the leading cause of legionellosis in north-western Europe, but, surprisingly, it is rarely isolated from environmental samples. Comparative genomics was applied to develop a PCR assay and to better understand the evolution of this strain. METHODS: Comparative analysis of 36 genomes representative of the Lp species was used to identify specific PCR targets, which were then evaluated in silico on 545 sequenced genomes and in vitro on 436 Legionella strains, 106 respiratory samples, and three environmental samples from proven ST47 sources. Phylogenetic analyses were performed to understand the evolution of ST47. RESULTS: The gene LPO_1073 was characterized as being 100% conserved in all 129 ST47 genomes analysed. A real-time PCR designed to detect LPO_1073 was positive for all 110 ST47 strains tested and agreed with culture and typing results previously obtained for 106 respiratory samples. The three environmental samples were also positive. Surprisingly, 26 of the 44 ST109 strains tested among 342 non-ST47 strains scored positive for LPO_1073. SNP-based phylogenetic analysis was undertaken to understand this result: the PCR-positive ST109 genomes were almost identical to ST47 genomes, with the exception of a recombined region probably acquired by ST47 from a ST62(-like) strain. CONCLUSION: The genomic analysis allowed the design of a highly specific PCR assay for rapid detection of ST47 strains. Furthermore, it allowed us to uncover the evolution of ST47 strains from ST109 by homologous recombination with ST62. We hypothesize that this recombination generated the leading cause of legionellosis in north-western Europe.

Legionnaires' disease: when an 'outbreak' is not an outbreak.

During August 2006, there was a large increase in non-travel related legionella cases throughout England and in the Netherlands.

Assessment of fluorescent amplified fragment length polymorphism analysis for epidemiological genotyping of Legionella pneumophila serogroup 1.

This study assessed the reproducibility and epidemiological concordance of double-enzyme fluorescent amplified fragment length polymorphism (fAFLP) analysis for genotyping of Legionella pneumophila serogroup (sg) 1. fAFLP fragment analysis was performed on three different sequencing platforms (one gel- and two capillary-based) in different laboratories with a well-characterised set of 50 strains of L. pneumophila sg 1. fAFLP data were analysed with the Pearson correlation similarity coefficient, using a range of parameters, and dendrogram outputs were converted to arbitrary types after selection of a specified percentage similarity threshold. The results obtained were compared with those obtained by the standard non-fluorescent AFLP method and were found to be broadly concordant. Using optimised settings for each fAFLP method to analyse the panel of 50 strains, epidemiological concordance (E) and reproducibility (R) values of 1.00 were obtained, and the number of types ranged from nine to 15, compared with E=1.00 and R=1.00, with 16 types, for the non-fluorescent AFLP protocol. The study demonstrated the potential of fAFLP for typing strains of L. pneumophila sg 1 on all three platforms; however, inter-platform comparison of fAFLP data was not achieved. fAFLP analysis may have a role in the fingerprinting of multiple isolates during Legionella outbreak investigations, but further work is required before type designations and identification libraries can be developed.

Diagnosis of Legionella pneumophila infections by means of formolised yolk sac antigens.

Formolised yolk sac antigens of Legionella pneumophila serogroups 1-6 were used to test 1792 serum specimens from 1431 patients with respiratory illness of serological evidence of Legionnaires' disease (LD). Thirty-five patients showed titres against the serogroup 1 antigen diagnostic for LD. Only two further cases were considered to have non-serogroup I infections (both serogroup 4) indicating that such infections are rare. Titres of greater than 1/16 against the serogroup 1 antigen occur in only 3% of subjects without LD and thus the demonstration of such a titre in patients with pneumonia during the early phase of illness can alert the clinician to the likelihood of LD. The supply of serogroup 1 antigen from the Division of Microbiological Reagents and Quality Control to routine diagnostic laboratories will be continued and monovalent serogroup 2-6 antigens will continue to be made available to reference laboratories.

A rapid microagglutination test for the diagnosis of Legionella pneumophila (serogroup 1) infection.

A rapid microagglutination test has been developed which can be performed in 30 minutes. Ninety-seven percent of 96 patients diagnosed as having Legionella pneumophila (serogroup 1) infection by indirect immunofluorescence were also detected by the rapid microagglutination test.