The effect of sodium thiosulfate on the recovery of Mycobacterium chimaera from heater-cooler unit water samples.

BACKGROUND: Heater-cooler units (HCUs) have been implicated in the recent global outbreak of invasive Mycobacterium chimaera infection among patients following cardiothoracic surgery. Because infected patients tend to remain asymptomatic for extended periods, detection of M. chimaera from HCUs in real time is essential to halting the ongoing M. chimaera HCU-associated outbreak. Sample collection protocols to evaluate the presence of M. chimaera offer conflicting recommendations regarding the addition of sodium thiosulfate (NaT) during the collection process. AIM: To study the effect of NaT on M. chimaera recovery and culture contamination. METHODS: Seventy-six paired HCU water samples (with and without NaT) were collected, processed and cultured simultaneously into Lowenstein-Jensen slants, Middlebrook 7H10 agar plates, and mycobacterial growth indicator tubes (MGITs), and incubated at 37°C. A subset of 31 paired samples was additionally cultured on MGITs and incubated at 30°C. FINDINGS: Of 76 samples incubated at 37°C in each of the three media, with and without NaT, M. chimaera was identified in at least one aliquot of 21 samples. CONCLUSION: The presence of NaT did not significantly increase the probability of recovering M. chimaera in a multi-variable conditional logistic model and culture contamination rates were similar between aliquots with and without NaT. In the subset of samples cultured on MGITs at both 30°C and 37°C, the presence of NaT again was not associated with M. chimaera recovery, but was significantly associated with reduced culture contamination.

A prevalence study of errors in opioid prescribing in a large teaching hospital.

AIM: In 2007, the National Patient Safety Agency performed a study demonstrating that errors in prescribing led to nearly 12,000 adverse clinical incidents a year. The following year, they issued a rapid response report entitled 'Reducing Dosing Errors with Opioid Medicines' designed to be implemented by all NHS trusts. We performed a prevalence study to assess opioid prescribing errors in our large multi-speciality teaching hospital prior to implementation of these recommendations. METHODOLOGY: We conducted a 1 day snapshot of opioid prescriptions on inpatient drug charts. For every chart, all opioid information was entered into the study proforma. All data were reviewed by consensus group and errors categorised by quality and whether they were potentially lethal, serious, significant or minor. RESULTS: A total of 330/722 (46%) charts were found to have opioid prescriptions. On the study day, there were 74 charts with errors and on expert review another 16 erroneous charts were found giving a total of 90/330 (27.2%). The largest quality statement error group was 'unclear prescription, missing information'. There were 4 potentially lethal, 26 serious, 38 significant and 22 minor errors. DISCUSSION: Previous studies have reported opioid prescription error rates of 51.2-70%. Compared with the opioid literature, our trust fares well with an error rate of 27%- four of these errors being potentially lethal. This study has identified where there are weaknesses in our hospital opioid prescribing practice and has aided us in rewriting our acute and chronic pain guidelines with the explicit inclusion of the National Patient Safety Agency recommendations. We have also disseminated the study results at the Trust academic meeting and developed an opioid e-learning package which will be mandatory for all new staff.