A nomogram for predicting complications in patients with solid tumours and seemingly stable febrile neutropenia

Abstract

Background: We sought to develop and externally validate a nomogram and web-based calculator to individually predict the development of serious complications in seemingly stable adult patients with solid tumours and episodes of febrile neutropenia (FN).

Patients and methods: The data from the FINITE study (n=1133) and University of Salamanca Hospital (USH) FN registry (n=296) were used to develop and validate this tool. The main eligibility criterion was the presence of apparent clinical stability, defined as events without acute organ dysfunction, abnormal vital signs, or major infections. Discriminatory ability was measured as the concordance index and stratification into risk groups.

Results: The rate of infection-related complications in the FINITE and USH series was 13.4% and 18.6%, respectively. The nomogram used the following covariates: Eastern Cooperative Group (ECOG) Performance Status ⩾2, chronic obstructive pulmonary disease, chronic cardiovascular disease, mucositis of grade ⩾2 (National Cancer Institute Common Toxicity Criteria), monocytes <200/mm(3), and stress-induced hyperglycaemia. The nomogram predictions appeared to be well calibrated in both data sets (Hosmer-Lemeshow test, P>0.1). The concordance index was 0.855 and 0.831 in each series. Risk group stratification revealed a significant distinction in the proportion of complications. With a ⩾116-point cutoff, the nomogram yielded the following prognostic indices in the USH registry validation series: 66% sensitivity, 83% specificity, 3.88 positive likelihood ratio, 48% positive predictive value, and 91% negative predictive value.

Conclusions: We have developed and externally validated a nomogram and web calculator to predict serious complications that can potentially impact decision-making in patients with seemingly stable FN.

Figure 1

Flow diagram of the FINITE and USH series.

Figure 2

The CISNE nomogram. COPD=chronic obstructive pulmonary disease; ECOG PS=Eastern Cooperative Oncology Group Performance Status; NCI=National Cancer Institute.

Figure 3

Calibration plot showing predicted probability vs actual (observed) frequencies of serious complications in clinically stable patients from the FINITE study (n=1133). (Bootstrap procedure with 2000 repetitions; mean absolute error=0.008; mean squared error=0.00021; 0.9 quantile of absolute error=0.014. Distribution of predicted risks of patients is shown at top of the figure.)

Figure 4

Calibration plot showing predicted probability vs actual (observed) proportions of serious complications in clinically stable patients from the USH registry (n=296). (Bootstrap procedure with 2000 repetitions; mean absolute error=0.036; mean squared error=0.00291; 0.9 quantile of absolute error=0.111. Distribution of predicted risks of patients is shown at top of the figure.)