Estimation of Personal Symptom Networks Using the Ising Model for Adult Survivors of Childhood Cancer: A Simulation Study with Real-World Data Application

Figures & data

Figure 1 Identification of coefficients in the simulated personal symptom networks.

Abbreviations: TPR, true positive rate; FPR, false positive rate; MCC, Matthew’s correlation coefficient.

Figure 2 Correlation of centrality metrics (strength, betweenness, and closeness) between the original sample and subsets with different case-dropping proportions. The definition of the stable and unstable network structure can be found in the Supplementary Materials.

Figure 3 Type I error of testing the differences of edges and centralities for the estimated personal symptom networks.

Table 1 Sociodemographic Characteristics, Cancer Diagnoses, Treatments, and the Presence of 10 Symptom Domains Among 1,708 Adult Survivors of Childhood Cancer Participating in the SJLIFE Study Between 2007 and 2020. Mean (Standard Deviation (SD)) Was Reported for the Numeric Variable and Number (%) Was Reported for Categorical Variables

Variable		Value
Individual characteristics
Age at survey (years)		30.9 (8.6)
Sex	Male	898 (52.6%)
	Female	810 (47.4%)
Race/ethnicity	Non-Hispanic White	1360 (79.6%)
	Other	348 (20.4%)
Attained education	Below college/post-graduate level	1123 (65.7%)
	Above college/post-graduate level	585 (34.3%)
Personal income	< $20,000	969 (56.7%)
	≥ $20,000	739 (43.3%)
Marital status	Ever married or lived as married	993 (58.1%)
	Never married or lived as married	715 (41.9%)
Cancer diagnosis and treatments
Cancer diagnosis	Leukemia	602 (35.2%)
	Lymphoma	328 (19.2%)
	Central nervous system (CNS) tumor	222 (13.0%)
	Solid tumor	541 (31.7%)
	Non-malignancy and histiocytosis	15 (0.9%)
Chemotherapy	No	264 (15.5%)
	Yes	1444 (84.5%)
Radiation	No	747 (43.7%)
	Yes	961 (56.3%)
Symptoms
Cardiac symptoms	Yes	249 (14.6%)
	No	1459 (85.4%)
Pulmonary symptoms	Yes	176 (10.3%)
	No	1532 (89.7%)
Sensation abnormality	Yes	604 (35.4%)
	No	1104 (64.6%)
Nausea	Yes	257 (15.0%)
	No	1451 (85.0%)
Movement problems	Yes	291 (17.0%)
	No	1417 (83.0%)
Pain	Yes	449 (26.3%)
	No	1259 (73.7%)
Memory problems	Yes	470 (27.5%)
	No	1238 (72.5%)
Fatigue	Yes	313 (18.3%)
	No	1395 (81.7%)
Anxiety	Yes	528 (30.9%)
	No	1180 (69.1%)
Depression	Yes	510 (29.9%)
	No	1198 (70.1%)

Figure 4 The identified pairwise associations between symptoms of the personal symptom network estimated based on the adult survivors of childhood cancer from SJLIFE. The edge name “Y₁-Y₂” represents the identified association between symptoms Y₁ and Y₂ independent of individual characteristics, and the edge name “Y₁-Y₂:X” represents the identified association between symptoms Y₁ and Y₂ influenced by individual characteristic X. The black dots represent the average point estimates, and the lines represent the corresponding 95% CIs.

Figure 5 The estimated personal symptom network for two individual cancer survivors: (a) a White non-Hispanic male, aged 30.9 years (overall mean age) at the survey with below college/post-graduate education, ever married or lived as married, and never received chemotherapy and/or radiation, and (b) a racial/ethnic minority female, aged 39.5 years (one SD above the mean) at the survey with above college/post-graduate education, never married or lived as married, and received chemotherapy and radiation. Edges pointed to by arrows with covariates indicate the influence of covariates on the pairwise associations between symptoms.

Figure 6 Values of centrality indices (strength, betweenness, and closeness) in the representative personal symptom network for person 1: a White non-Hispanic male, aged 30.9 years at the survey with below college/post-graduate education, ever married or lived as married, and never received chemotherapy and/or radiation, and person 2: a racial/ethnic minority female, aged 39.5 years at survey with above college/post-graduate education, never married or lived as married, and received chemotherapy and radiation.

Figure 7 Correlations of centrality metrics in case-dropping bootstrap testing for the personal symptom network estimated based on the adult survivors of childhood cancer from SJLIFE.

Data Sharing Statement

The deidentified data will be publicly available on Zenodo.org.