Discussion

During our study at the SGHUMC in Lebanon, the average BCC rate was elevated at 7%, more than twice the acceptable international benchmark of 3%. This represents approximately 399 contaminated blood cultures annually, contributing to increases in the patient’s length of stay at the hospital and added cost on the healthcare system.

Moreover, it is well established that blood cultures are a major tool for BSI diagnosis. ^{Reference Lamy, Sundqvist and Idelevich17} Yet, BCC remains inevitable ^{Reference Alahmadi, Aldeyab and McElnay18} . Multiple evidence-based strategies have helped us mitigate this risk, but their role is limited. For example, skin decontamination using 2% alcoholic chlorhexidine significantly decreases the percentage of contaminated blood culture, ^{Reference Suwanpimolkul, Pongkumpai and Suankratay19} but the contamination risk cannot be minimized to zero because 20% of the skin contaminants are located in the deeper epidermis. ^{Reference Rupp, Cavalieri, Marolf and Lyden20} In addition, the use of initial specimen diversion devices reduces the contamination rate to 0.22% by diverting the first 2 mL of blood assumed to contain contaminants. ^{Reference Rupp, Cavalieri, Marolf and Lyden20} Other strategies include phlebotomists training, hand hygiene, the use of sterile gloves to palpate the venipuncture site after disinfection and blood culture bottle disinfection with 70 percent isopropyl alcohol. ^{Reference Doern, Carroll and Diekema21}

Despite the above, we still witness high levels of BCC globally that vary between 0.6% and 12.5% among institutions ^{Reference Snyder, Favoretto and Baetz22} . Although some have reported rates as low as 0.2%. ^{Reference Rupp, Cavalieri, Marolf and Lyden20} , the recommended BCC benchmark is at 3%. ^{Reference Wilson23}

In patients with COVID-19, we witnessed an increase in BCU rate of 185.9/1000PD, corresponding to an additional 100 blood cultures/1000CPD, which was accompanied by an elevated BCC rate of 9%. Similar rates have also been reported in other studies. ^{Reference Ohki, Fukui, Morishita and Iwata24} During the study period involving the COVID-19 pandemic, very strict infection prevention and control methods were applied, which implies stringent hand hygiene and skin antisepsis protocols. In other words, traditional measures to mitigate risks of BCC were applied. This was met by a stable and relatively low rate of BSIs, indicating an overuse of blood cultures likely due to a sicker population but not reflective of a population where the likelihood of a BSI is elevated. Would the presence of different blood culture indications have affected the outcome?

In this retrospective study, we identified a high rate of BCC at our institution. It becomes apparent that reducing the rates of BCC starts with reducing our BCU rates by setting indications for blood cultures. The proper selection of patients with a high pretest probability of bacteremia and avoiding sampling in clinical conditions where blood cultures have lower yields than cultures from primary sources is important. ^{Reference Fabre, Sharara, Salinas, Carroll, Desai and Cosgrove25}

Finally, this is a large retrospective, observational single center study spanning over 13 years that contributes raw, observed data. However, we are aware of its limitations that did not allow us to firmly conclude on the etiology behind BCC and its respective impact on morbidity, hospital stay, and cost. It is possible that high BCU where no specific indications guided blood culture drawing, as well as the super vigilant approach of excessive blood cultures early in the COVID-19 pandemic. We also consider the difficulty of factoring in events such as lack of skilled manpower and challenges that may affect financial and technical aspects in our healthcare system. In low-middle-income countries (LMICs), attempting to reduce these unnecessary medical and socioeconomic burdens to provide a more patient centered quality care is essential. If we take a closer look at the year 2019, we notice a rate of BCC of 4%, a value very close to the acceptable international benchmark with a more or less acceptable BCU rate of 105.6/1000 PD. Does this seem like a good balance between BCU and rate of BCC? This is possible because in 2019, we appointed a task force at SGHUMC to tackle blood culture sampling techniques including staff training sessions and the introduction of the vacutainer safety lock blood collection set (The BD Vacutainer® Safety-Lok™ blood collection set, BD medical technology). Minor changes were in progress but, unfortunately, due to several unexpected factors, we were unable to fully implement these changes. These factors included the collapsing economy in Lebanon, the Beirut port August 4, 2020, explosion, the COVID-19 pandemic, and the subsequent, unfortunate emigration of skilled healthcare professionals. This is not unusual for an LMIC. Therefore, it is crucial to adapt the issue of BCC from this angle.

Moreover, in light of limited and challenging healthcare settings, we suggest considering the possibility of establishing institution-based benchmarks or the modification of recommended guidelines and practice while catering for such challenging healthcare settings yet still ensures best safe practice. Strict adherence to aseptic techniques, revision of hospital policies and procedures, and staff training sessions are as important as the continuous surveillance of key quality performance healthcare indicators. Should BCU rate become a new key quality performance indicator? Future studies that explore the rising field of diagnostic stewardship and BCU and contamination especially in LMICs are needed to improve patient outcome and serve as boosters and support multidisciplinary antimicrobial stewardship efforts.