Literature Review
We systematically searched PubMed, Embase, and the Cochrane Library, with two keywords vancomycin AND thrombocytopenia to start the search. Clinical studies that evaluated thrombocytopenia caused by vancomycin, without limitation to patients or controlled drugs. Papers written in languages other English were excluded. Published reviews, case report, conference abstracts and non-clinical studies were also excluded. The reviewer (WGR) assessed publications for eligibility. Any paper with uncertainty in regard to inclusion or exclusion was discussed between 3 authors (WGR, LXX and ZRS) until a consensus of opinion was reached. Newcastle-Ottawa Scale (NOS) quality evaluation scale and JBI checklist for analytical cross sectional studies were used to evaluate the quality of included studies. 19,20 For each study, both the process of data extraction and quality evaluation were performed by one author (WGR) using predesigned tables, followed by the check of another author (LXX). Any disagreement was resolved by discussion among all authors (WGR, LXX and ZRS).
Overall, 600 articles were retrieved. After reviewing the titles, abstracts and full texts, the reviewers agreed on a final selection of 6 studies for inclusion. The screening flow chart is presented in Figure 2, with the basic information of the six studies included shown in table 3. Of these, 5 studies were identified as retrospective cohort with a controlled drugs as linezolid.13-17 One was a cross-section study with a single-arm.18 The patients were mainly adults with bacterial infection, and the duration treatment (vancomycin or linezolid) was more than 7 days. Regarding quality assessment, 5 cohort studies were identified as high quality (score 7-8) based on NOS tool13-17, 1 cross section study as medium quality (score 5) based on JBI tool.18 All studies were therefore acceptable in this review.
We reviewed the frequency, severity and occurrence of bleeding events due to thrombocytopenia caused by vancomycin or linezolid (Table 4 and Figure 2). Only one study showed that the frequency of thrombocytopenia (defined as ≥ 30 % decrease in platelets) caused by linezolid in patients was significantly higher than that caused by vancomycin (OR = 3.38, 95 % CI: 1.87–6.08).17 Five studies showed that vancomycin and linezolid had no significant difference in thrombocytopenia (either defined as platelets<150 x 109/L or<100 x 109/L). As shown in Figure 3, the meta-analysis of the two studies showed that vancomycin and linezolid had no significant difference in thrombocytopenia (defined as platelets<150 x 109/L, RR=1.01, 95%CI=0.64-1.58).14-15 The cross-section study showed that the incidence of vancomycin-induced thrombocytopenia (defined as platelets<100 x 109/L) was 7.1 %.18 Although no life-threatening bleeding events was reported to occur, vancomycin-induced thrombocytopenia can be as low as less than half of the baseline platelets of patients. The above results suggested that the incidence and severity of thrombocytopenia caused by vancomycin should be paid attention.