Abstract
Background
In October 2018, the Hispaniola Project was initiated to build local expertise in infection care and prevention at three pediatric oncology units (POUs) in Haiti and the Dominican Republic. Surveillance of healthcare-associated infections (HAI) was a central aim. Severe and prolonged neutropenia is a frequent risk factor for infections in oncology patients. Among HAIs, bacteremia is one of the most serious; bacteremia requires timely isolation and identification of the offending microorganism and the antimicrobial susceptibility. These diagnostic interventions allow informed therapeutic and prophylactic measures. Here, we report our experience in bacteremia in these 3 POUs.
Methods
We conducted prospective infection surveillance of all patients admitted to three POUs in Hispaniola Island. Blood culture methods followed standard national procedures. We used the 2018 US Centers for Disease Control National Healthcare Safety Network case definitions for primary laboratory-confirmed bloodstream infections (LCBI), and we categorized infections as healthcare-associated or present on admission (POA). We reviewed data collected from January 2019 to December 2020 and used descriptive statistics to report our results.
Results
Our review identified 66 LCBIs with an overall rate of 3.52 infections per 1000 patient-days. Of these, 40 (61%) were healthcare-associated, and 26 were POA. The majority (41, 62%) of patients were undergoing chemotherapy at the time of the infection, with induction being the most common phase (23). The most common oncologic diagnosis was acute lymphoblastic leukemia (43, 65%), followed by solid tumor (12, 18%). Fifty-three (80%) of the infections met the LCBI-1 criteria, with the other 13 categorized as LBCI-2. Of the 53 LCBI-1, 7 (13%) were considered related to mucosal barrier injury (MBI-LCBI 1 definition). The most commonly identified organisms were Klebsiella spp. (13, 19%) and coagulase-negative Staphylococcus (13, 19%). Antibiotic resistance was observed in many of the identified pathogens, with nearly half (25, 44%) of the 57 bacterial isolates having any resistance and a quarter (14, 25%) with resistance to multiple classes, including cephalosporins, fluoroquinolones, and aminoglycosides. Eleven (17%) patients were admitted to the Intensive Care Unit as a result of the LCBI. Thirteen deaths were recorded among the patients with LCBIs, with 6 (46%) associated with the HAI and 7 (54%) related to disease progression.
Conclusions
Our findings demonstrate that resistant pathogens were frequent among the LCBI isolates. Our preliminary results are guiding clinical management to be vigilant in our care of patients at high risk for bacteremia and poor clinical response by initiating more effective antimicrobials sooner. Importantly, reviewing reasons for antimicrobial resistance and implementing best antimicrobial use practices will protect our fragile antibiotic arsenal. Infection surveillance programs, such as ours, and other initiatives which promote infection prevention and control in POU will increase the quality of care for these vulnerable patients.
Tissue culture: Methods and applications: Ed. by Paul F. Kruse, Jr and M. K. Patterson, Jr. 1973. New York and London: Academic Press. Pp. xxvii and 868; 264 text figs. 10.15.