C. Lucas

Continuous infusion of vancomycin in neonates

A one year, retrospective audit of intermittent vancomycin therapy within the Neonatal Unit at the Royal Hospital for Sick Children, Glasgow, found that only 33% of 984 vancomycin trough levels were within the British National Formulary for children (BNFc) target range of 10–15 mg/l and 25% were <10 mg/l. A detailed, prospective review of 20 courses (15 patients) over one month, while using the same dosing guidelines, highlighted that only 23 of 50 concentration measurements (46%) were within the target range and 20% were <10 mg/l, even though the initial doses used (table 1) were higher than those recommended by the BNFc. Dose adjustments were common and up to 80 mg/kg/day was often required for older infants. Appropriate interpretation of concentration results was compromised …

Improved early diagnosis of Pseudomonas aeruginosa by real-time PCR to prevent chronic colonisation in a paediatric cystic fibrosis population

Early detection of Pseudomonas aeruginosa in children with cystic fibrosis is hampered by the need to process a sub-optimal specimen type, namely cough swabs, which are known to have a lower positive yield than sputa or more invasive samples. This delay in the detection of low levels of P. aeruginosa could potentially result in the loss of an opportunity to initiate early aggressive antibiotic therapy and result in chronic colonisation, with a poorer overall prognosis. Quantitative real-time PCR (qPCR) offers an opportunity to increase the detection rate of P. aeruginosa compared to traditional culture techniques. This study examined 500 cough swabs and 42 sputum samples from paediatric patients and showed that detection of P. aeruginosa could be increased in both sample types by 100% and 45% respectively. Overall the sensitivity was 100% and specificity of 58% when compared to culture as a gold standard. These results although initially promising require careful consideration both from a treatment and infection control standpoint as the significance of detection of very low levels of P. aeruginosa is unclear.

Intermittent versus continuous infusion of vancomycin in neonates

Objective To compare the effectiveness of intermittent and continuous vancomycin infusions in achieving target concentrations in neonates and to assess the implications for clinical practice. Methods The study involved infants who required treatment with vancomycin during their admission to the neonatal intensive care unit at Yorkhill Hospital, Glasgow, between March and June 2011. During the first month, intermittent vancomycin (Ivanc) infusion was started, at a dose of 10, 15 or 20 mg/kg over 1 h every 12 or 8 h, according to creatinine concentration. Thereafter, doses were adjusted to maintain troughs of 12–15 mg/l.1 Ivanc dosing was maintained in month 2 while staff received training on the continuous vancomycin (Cvanc) infusion guidelines, which comprised of a loading dose of 15 mg/kg followed by a continuous infusion of 20–60 mg/kg/day according to creatinine and corrected gestational age (CGA). Doses were adjusted to maintain concentrations of 15-25 mg/L. These guidelines were implemented in month 3. During months 1 and 4, the following data were collected: age; weight; date, time and dose given; vancomycin concentrations; creatinine concentrations at the start and 2 weeks after therapy. Clinical characteristics were compared by t-test and concentrations by χ2 test. Results There were 20 courses (15 patients) of Ivanc infusions and 20 (17 patients) of Cvanc infusions. There were no differences in CGA (mean±SD 37.6±4.8 weeks Ivanc versus 38.7±2.7 weeks Cvanc) or weight (2.39±0.95 kg versus 2.66±0.82 kg) between the groups. Mean±SD vancomycin concentrations were 13.3±3.9 mg/l (Ivanc) and 21.7±4.6 mg/l (Cvanc). The first concentration was within the target range in 45% of the Ivanc and 75% of the Cvanc courses. Although similar proportions were above the target ranges (20% and 25% respectively), low concentrations were only observed in the Ivanc group (35%). Overall, 34% of 50 samples were within the target range in the Ivanc group and 77% of 82 in the Cvanc group (p=0.0001; CI95 0.24 to 0.58). The starting dose was correct in 80% of patients in the Ivanc group and 95% in the Cvanc group. In the Ivanc group, 51 doses (18%) were administered >30 min outside the prescribed time; no administration problems were identified in the Cvanc group. In the Ivanc group, eight patients had at least one dose change and two had >3; in the Cvanc group, 10 patients had at least one change and none had >3. Creatinine concentration did not increase in either group. Conclusions The results suggest that continuous vancomycin infusion in neonates is more likely to maintain concentrations within the target range compared to intermittent vancomycin infusion. In clinical practice, the continuous administration schedule helps to minimise errors with dose times and associated monitoring inaccuracies.

Audit of ventriculoperitoneal shunt infections in paediatric patients, 2006–2013

Background Infection following ventriculoperitoneal shunt placement remains a significant complication with an incidence of 3–27% cited in literature. Infections cause significant morbidity, and it is important that empirical antibiotic therapy for management is guided by accurate knowledge of prevailing aetiologies and local antibiotic sensitivity patterns. Aims To establish the incidence of shunt infections in our paediatric population, to identify the causative micro-organisms, and to determine the antibiotic resistance patterns of the responsible micro-organisms. Methods Retrospective data collection utilising existing databases in the Royal Hospital for Sick Children, Yorkhill, between 1 January 2006 and 30 September 2013. Results Total number of shunt operations was 308 with 28 episodes of infection involving 27 patients (male = 12, female = 15). The incidence of infection was 9%. In all, 79% of the episodes involved a single pathogen with 21% being mixed pathogens. Coagulase-negative staphylococci were the most common cause of infection (44%). Gram-positive sensitivity to flucloxacillin and gentamicin was noticeably low at 22% and 14%, respectively with 87% of coagulase-negative staphylococci resistant to gentamicin and 81% resistant to flucloxacillin. Conclusions The changing spectrum of Gram-positive organisms has impacted on antibiotic sensitivity patterns, and our local prescribing policy has been adapted in order to manage shunt infections most effectively.

47 Diagnostic dilemmas for CF microbiology labs using MALDI-TOF

Objective The identification of certain bacterial species from the sputum of patients with CF carries prognostic significance so accurate identification is essential. MALDI-TOF technology promises rapid and accurate identification of the sometimes difficult bacteria Pseudomonas aeruginosa, members of the Burkholderia cepacia and other non fermenting organisms. We have compared results from bioMerieux MALDI-TOF and Bruker MALDI-TOF for the Burkholderia cepacia complex (Bcc). Methods A range of organisms from patients attending a Paediatric CF Centre were identified using 2 MALDI-TOF systems and in-house molecular methodology with referral of isolates to a Reference Laboratory if appropriate. Results Five isolates categorised as P. aeruginosa on the bioMerieux system were identified by in-house 16s ribosomal DNA sequencing or Reference Laboratory testing as P. nitroreducens. On retesting on the Bruker MALDI-TOF all 5 strains were identified as P. nitroreducens. Ninety-four PCR-confirmed Bcc strains were tested in parallel and both MALDI-TOF machines categorised them within the cepacia complex. At Genomovar level, B. multivorans and B. vietnamiensis gave concordant results on the 2 systems. All of 28 B. cepacia identified on the bioMerieux MALDI-TOF were B. cenocepacia according to Bruker. Conclusion Misidentification of P. aeruginosa and B. cenocepacia is unacceptable as incorrect segregation and antibiotic management of patients would ensue. Refinement and compatibility of MALDI-TOF databases is essential before this technology can replace current DNA based identification methods and needs to be supported by relevant and robust EQA schemes.

55 Is whole genome sequencing necessary to exclude cross infection with Mycobacterium abscessus ST26 in paediatric cystic fibrosis patients

Objective Patients with Cystic Fibrosis (CF) can be colonized or infected with M. abscessus. Our centre along with others in the UK has seen an increase in patients with this organism. We have reviewed M. abcessus cases since 2011 to attempt to ascertain whether there is a genuine increase in the incidence of this infection and whether there is any possibility of cross infection within our patient group. Methods Retrospective data on paediatric M. abscessus isolates were retrieved from the microbiology laboratory computer system and historical CF databases. Isolates of M. abscessus were sent to PHE, Colindale, for genetic analysis. Results There has been a genuine increase in M. abscessus colonization amongst our paediatric CF population over time. Sampling patterns have varied but isolation rates were independent of numbers of sputum samples and patients tested. Five of 13 paediatric patients colonized with M. abscessus harboured strains belonging to clonal lineage ST26. Of these 5, no patients had any recorded opportunity for direct contact within the hospital and we could find no epidemiological evidence of any periods when cross infection could have occurred. Remaining patient isolates were distinct. The isolates from the ST26 patients were sequenced and whole genome analysis was performed. Conclusion The control of cross infection is an important aspect of CF care. Since diverse ST26 strains have been reported in the context of CF, further genetic analysis including whole genome sequencing may be required to determine any relatedness between isolates and exclude the possibility of hospital acquired M. abscessus infection.