No abstract

BACKGROUND: South Africa´s diagnostic algorithm for TB diagnosis from 2011 to 2017 employed the Xpert® MTB/RIF assay as the initial screening test for TB diagnosis and rifampicin (RIF) susceptibility, followed by submission of a specimen for GenoType® MTBDRplus. This study aimed to determine the concordance between the two assays in terms of RIF susceptibility and explore reasons for discordance.METHODS: This was a retrospective laboratory-based study that included all MTBDRplus results of tests performed at the Braamfontein Mycobacteriology Referral Laboratory between 1 September 2014 and 31 August 2015. The patient´s Xpert RIF result was linked with the MTBDRplus result.RESULTS: The overall concordance between RIF susceptibility results was 96.4%. There were 68 discordant RIF results. The most common reasons for discordance identified were possible false Xpert RIF-resistant results (22%), mixed infection/heteroresistance (16%), transcription errors (7%) and erroneous manual interpretation of the MTBDRplus strip (7%). Xpert RIF resistance detected using delayed hybridisation was associated with discordance.CONCLUSIONS: The overall concordance between the MTBDRplus and Xpert RIF results were very good. Management of discordance should include repeat specimens for Xpert and MTBDRplus and rpoB sequencing. All variables should then be considered before treatment regimens are altered.

Objective We aimed to provide an analysis of A. baumannii complex (ABC) isolated from blood cultures in South Africa. Materials and methods ABC surveillance was conducted from 1 April 2017 to 30 September 2019 at 19 hospital sites from blood cultures of any age and sex. Organism identification was performed using the MALDI-TOF MS and antimicrobial susceptibility testing (AST), MicroScan Walkaway System. We confirmed colistin resistance with Sensititre, FRCOL panel, and selected for whole-genome sequencing. Results During the study period, we identified 4822 cases of ABC, of which 2152 cases were from 19 enhanced surveillance sites were reported during the enhanced surveillance period (1 August 2018 to 30 September 2019). Males accounted for 54% (2611/4822). Of the cases with known age, 41% (1968/4822) were infants (< 1-year-old). Seventy-eight percent (1688/2152) of cases had a known hospital outcome, of which 36% (602/1688) died. HIV status was known for 69% (1168/1688) of cases, and 14% (238/1688) were positive. Eighty-two percent (1389/1688) received antimicrobial treatment in admission. Three percent (35/1389) of cases received single colistin. Four percent (75/2033) were resistant to colistin. At least 75% of the isolates (1530/2033) can be classified as extensively drug-resistant (XDR), with resistance to most antibiotics except for colistin. The majority, 83% (20/24), of the colistin-resistant isolates were of the sequence type (ST) 1. Resistance genes, both plasmid- and chromosomal- mediated were not observed. Although all isolates had, nine efflux pump genes related to antimicrobial resistance. Conclusion Our surveillance data contributed to a better understanding of the natural course of A. baumannii disease, the patient characteristics among infants, and the level of resistance. At least two-thirds of the isolates were extensively drug-resistant, and four percent of isolates were resistant to colistin.

No abstract

Background: Pre-analytical errors contribute significantly to the reduced quality of laboratory results and may adversely impact patient management. Appropriate completion of the laboratory requistion form (LRF) by clinical staff is an important element of the pre-analytical phase. As a quality improvement initiative, a sample of LRFs submitted to the Charlotte Maxeke Johannesburg Academic Hospital Microbiology Laboratory was audited. Methods: A pilot study reviewing LRFs for the period 1–18 September 2020 for tissue, pus, sterile site fluids and pus swab specimen types was undertaken. An assessment of the completeness and correctness of information in LRFs was performed. Parameters were assigned to four quality indicator (QI) groups, namely patient identifiers, clinician identifiers, test request and clinical details. Results: An audit of completed LRFs for 172 specimens was performed, suggesting wide variability in the completion of parameters. Clinical details (description of the site of specimen collection, diagnosis and medication) were the most poorly completed components. The contact number of the requesting healthcare worker (HCW) was missing in 91% of requests. The most consistently completed and reliable QI was patient’s details. Other mandatory parameters, including the HCW’s name and practice number, were completed in 95% and 99% of LRFs, respectively. Conclusions: The inconsistent completion of key parameters in the LRF is of concern, and larger studies are warranted to determine the broader implications of our findings. Strategies to improve the completion of microbiology LRFs in this setting include educating the medical staff and students, expanding mandatory fields in the LRF and implementing an electronic requisition system.