Background: Febrile neutropenia (FN) is a common complication to chemotherapy associated with a high burden of morbidity and mortality. Reliable prediction of individual risk based on pre-treatment risk factors allows for stratification of preventive interventions. We aimed to develop such a risk stratification model to predict FN in the 30 days after initiation of chemotherapy.
Methods: We included consecutive treatment-naïve patients with solid cancers and diffuse large B-cell lymphomas at Copenhagen University Hospital, 2010-2015. Data were obtained from the PERSIMUNE repository of electronic health records. FN was defined as neutrophils <= 0.5x10E9/L at the time of either a blood culture sample or death. Time from initiation of chemotherapy to FN was analysed using Fine-Gray models with death as a competing event. Risk factors investigated were: age, sex, body surface area, haemoglobin, albumin, neutrophil-to-lymphocyte ratio, Charlson Comorbidity Index (CCI) and chemotherapy drugs. Parameter estimates were scaled and summed to create the risk score. The scores were grouped into four: low, intermediate, high and very high risk.
Results: Among 8,585 patients, 467 experienced FN, incidence rate/30 person-days 0.05 (95% CI, 0.05-0.06). Age (1 point if > 65 years), albumin (1 point if < 39 g/L), CCI (1 point if > 2) and chemotherapy (range -5 to 6 points/drug) predicted FN. Median score at inclusion was 2 points (range 5 to 9). The cumulative incidence and the incidence rates and hazard ratios of FN are shown in figure 1 and table 1, respectively.
Conclusion: We developed a risk score to predict FN the first month after initiation of chemotherapy. The score is easy to use and provides good differentiation of risk groups; the score needs independent validation before routine use.
G. Daugaard, None
P. Brown, None
H. Sengeløv, None
A. Mocroft, None
J. Lundgren, None
M. Helleberg, None