The feasibility of a Bayesian network model to assess the probability of simultaneous symptoms in patients with advanced cancer

Abstract

Although patients with advanced cancer often experience multiple symptoms simultaneously, clinicians usually focus on symptoms that are volunteered by patients during regular history-taking. We aimed to evaluate the feasibility of a Bayesian network (BN) model to predict the presence of simultaneous symptoms, based on the presence of other symptoms. Our goal is to help clinicians prioritize which symptoms to assess. Patient-reported severity of 11 symptoms (scale 0-10) was measured using an adapted Edmonton Symptom Assessment Scale (ESAS) in a national cross-sectional survey among advanced cancer patients. Scores were dichotomized (< 4 and ≥ 4). Using fourfold cross validation, the prediction error of 9 BN algorithms was estimated (Akaike information criterion (AIC). The model with the highest AIC was evaluated. Model predictive performance was assessed per symptom; an area under curve (AUC) of ≥ 0.65 was considered satisfactory. Model calibration compared predicted and observed probabilities; > 10% difference was considered inaccurate. Symptom scores of 532 patients were collected. A symptom score ≥ 4 was most prevalent for fatigue (64.7%). AUCs varied between 0.60 and 0.78, with satisfactory AUCs for 8/11 symptoms. Calibration was accurate for 101/110 predicted conditional probabilities. Whether a patient experienced fatigue was directly associated with experiencing 7 other symptoms. For example, in the absence or presence of fatigue, the model predicted a 8.6% and 33.1% probability of experiencing anxiety, respectively. It is feasible to use BN development for prioritizing symptom assessment. Fatigue seems most eligble to serve as a starting symptom for predicting the probability of experiencing simultaneous symptoms.

Figure 1

Directed acyclic graph of the Baysian network; as constructed during structure learning by the algorithm with the lowest prediction error estimate (Tabu search algorithm). Edges (arrows) between nodes indicate that the model found a direct association between those variables. An edge points from a parent node towards a child node, indicating that the BN structure found that the child node is conditionally dependent on the parent node.

Figure 2

Calibration plots per predicted simultaneous symptom. The mean observed frequencies are plotted against the mean predicted probabilites per decile. Each decile represents a group of patients with a similar conditional probabilty predicted by the model. The black line corresponds to a model with ideal calibration, that is, a model that perfectly predicts the conditional probabilities.

Figure 3

Preliminary symptom prediction system that uses the presence of fatigue as system input. Fatigue is the second most frequently volunteered symptom by patients with advanced cancer, and in our Bayesian network this symptom was identified as the symptom that most frequently was directly associated with other USD-listed symptoms (7 other symptoms). When using fatigue as system input, the conditional probabilities (%) of the patient also experiencing pain, dry mouth, sleeping problems, lack of appetite, nausea, shortness of breath and anxiety can be presented. This may help clinicians to prioritize which symptoms to assess.