Table 3.

Key obstacles to the implementation of quality-assured clinical bacteriology laboratories. Many of the most difficult challenges are outside the laboratory and its control.

Challenge encountered	Solution attempted or proposed
Overall quality assurance
Achieving a broad-based commitment to quality	• Sustained engagement with clinicians and hospital administration to promote the relevance of the laboratory. • Leveraging key colleagues as “champions” of the laboratory. • Focusing on achievable goals, and avoiding unrealistic expectations of perfection. • Promoting a shared-responsibility quality mindset, and the objective of constant improvement.
Pre-analytical phase
Engaging and generating demand from clinicians	• Series of clinician-focused hospital lectures. • Pharmacist-led antimicrobial stewardship intervention for inpatients receiving antibiotics, described elsewhere (29).
Poor-quality specimen Wrong test ordered	• Communication, education, and persistence.
Inappropriate transport conditions	• Would be facilitated by the existence of a medically qualified microbiologist in the hospital.
Improperly labeled specimens Absent, incomplete, incomprehensible requisitions	• Technologists provided blood culture bottles only after requesting physicians supplied a properly filled requisition. • Verifying labeling on specimens on arrival.
Analytical phase (in laboratory)
Adapting standard operating procedures (SOP)	• Using best-practices for SOP in LRS (20), including straightforward language, short sentences, a legible font and size, and avoiding acronyms. • AVOIDING pictograms, illustrations, or figures as a replacement of plain language. • AVOIDING calendar-based QC schedule. Instead, embedding QC tasks within concrete pre-specified activities (e.g., QC after opening a new reagent box).
Standardizing practice	• Regular and repeated direct observations were the single most useful action available for identifying knowledge and procedural gaps, and for promoting reference to SOPs.
Availability of pyrolidonyl aminopeptidase (PYR) for identification of Enterococcus sp.	• In the absence of PYR reagent, we relied on bile esculin (BE) hydrolysis to differentiate Enterococcus sp. from the majority of alpha-hemolytic Viridans Group Streptococcus sp. other than S. bovis. Optochin susceptibility was determined in order to identify S. pneumoniae among Streptococcus sp. that may hydrolyse BE. • Pathogens potentially misidentified as Enterococcus sp. using this second-line approach include S. mutans, S. salivarius, and S. suis, but these are infrequent blood-borne pathogens and would be treated by regimens directed at Enterococcus sp.
Availability of rabbit plasma for coagulase testing	• We relied on importation of a commercial kit (Pastorex™ Staph-Plus; Bio-Rad, Marnes-la-Coquette, France) to presumptively identify S. aureus. • Alternatively, human plasma discarded from blood banks could be used instead of rabbit plasma or commercial kit, but it needed to pass QC using an appropriate S. aureus strain.
Availability of Mueller-Hinton blood agar for AST of fastidious bacteria	• Use of chocolate agar was permitted as long as QC testing of antimicrobial discs using representative bacterial strains yielded expected results.
Availability of sheep-blood agar	• Hospital administration unwilling to keep two healthy sheep on the hospital grounds for this purpose. • Approached local farmers to bleed their sheep, but could not agree on reimbursement and logistics. • When human blood agar was used for blood-agar preparation, we stressed the importance of testing for catalase, and careful inspection of the hemolysis pattern using magnification and effective lighting. In addition, using human blood agar magnified the importance of routine Gram staining for confirming bacterial morphology, and of media quality control in general.
Complexity of bacterial identification	• Simplified presumptive bacterial identification schemes, based on manual phenotypic testing, were used instead of routine full identification.
Inadequate space	• Few short-term solutions.
Post-analytical phase
No telephone system in hospita No way to reach physician for critical results	• Successfully implemented a policy of demanding the mobile phone number of the ordering physician on the test requisition.
Inadequate laboratory information system	• Implemented WHONET software (27, 28) with support from EPHI, in order to link data between institutional testing and national AMR surveillance. • Unfortunately, WHONET is not capable of interfacing with the regular laboratory information system in the laboratory. • Unfortunately, WHONET is not capable of interfacing with in-hospital electronic medical record systems.
No in-hospital electronic medical record	• We relied on a written ledger that was consulted on a daily basis by ordering physicians. • Efforts to provide printouts of WHONET entries in lieu of this ledger were thwarted by consistent lack of consumables (paper, toner) and controversy about the need for this step.
	• We provided all registered physicians with electronic access to a commercially available online resource on which we could load cumulative antibiogram data as well as tailored clinical guidelines on antimicrobial use, informed by local susceptibility patterns (Sanford Guide with Stewardship Assist™; https://www.sanfordguide.com/stewardship-assist/.
Data analysis responsibility	• Ongoing data analysis for quality assurance remains a challenge without external support.
Power/training differential between technologists and laboratory clients	• The promotion of mutual respect through lectures and open communication was helpful.
Challenges outside the testing cycle
Overcoming local supply and foreign currency exchange issues	• See text.
Engagement from hospital & university community toward importance of lab sector
Attractiveness of microbiology as a career path, given lack of recognized specialty in LRS
Attractiveness of Infectious Diseases as a career given lower revenue compared to procedural specialties.

SOP denotes standard operating procedures; PYR, pyrolidonyl aminopeptidase; AMR, antimicrobial resistance; AST, antimicrobial susceptibility testing; LRS, low resource settings; QC, quality control; BE, bile esculin.