Baseline Hematologic Reserve and Laboratory-Defined Leukopenia/Neutropenia During Vancomycin Therapy: A Real-World Risk-Stratification Cohort Study.

BACKGROUND

Laboratory-defined leukopenia/neutropenia may occur during vancomycin therapy, but mild threshold-crossing abnormalities do not necessarily establish definite drug-induced toxicity.

OBJECTIVE

To develop and evaluate a risk-stratification strategy based on baseline hematologic reserve and vancomycin treatment duration.

METHODS

This single-center retrospective cohort included adult inpatients receiving intravenous vancomycin for ≥3 days. Patients with baseline leukopenia/neutropenia, hematologic disease, malignancy or antitumor therapy, non-intravenous use, or unavailable blood counts were excluded. The primary endpoint was laboratory-defined leukopenia/neutropenia, defined as new white blood cell count (WBC) <4.0 × 109/L or absolute neutrophil count (ANC) <2.0 × 109/L during therapy or within 3 days after discontinuation. A 4-component baseline hematologic reserve score was constructed from WBC, ANC, monocyte count, and hemoglobin.

RESULTS

Among 506 patients, 75 developed leukopenia/neutropenia, corresponding to an event rate of 14.8%. The score assigned 1 point each for WBC ≤7.77 × 109/L, ANC ≤6.62 × 109/L, monocyte count ≤0.40 × 109/L, and hemoglobin ≤102 g/L. Event rates increased from 7.5% to 40.9% across scores of 0 to 4 (P for trend < .001), and were 8.6%, 12.2%, and 34.5% in the low-, intermediate-, and high-risk groups, respectively. Compared with the low-risk group, the score-defined high-risk group had higher odds of laboratory-defined leukopenia/neutropenia (adjusted odds ratio [OR] 7.52, 95% CI, 3.95-14.33, P < .001). Vancomycin duration >14 days was also independently associated with the outcome (adjusted OR 5.08, 95% CI, 2.77-9.29, P < .001). The highest risk was observed in high-risk patients treated for >14 days, with an event rate of 75.0%. The baseline score model showed moderate performance (area under the curve [AUC] 0.697), whereas the extended model incorporating the score-defined risk group, treatment duration, and clinical covariates improved discrimination (AUC 0.786).

CONCLUSION AND RELEVANCE: A 4-component baseline hematologic reserve score combined with treatment duration may help identify patients who warrant closer complete blood count monitoring during vancomycin therapy. These findings should be interpreted as laboratory-event risk stratification rather than proof of definite vancomycin-induced toxicity or as a mandatory discontinuation threshold.