Predictors of Young Children's Medical Neglect: A Multinomial Logistic Regression Analysis

Background: Child neglect is a heterogeneous phenomenon and can be better understood using subtypes of neglect (e.g., Mennen, Kim, Sang & Trickett, 2010). One of the subtypes, medical neglect (MN), might significantly compromise children’s physical or mental health. Past studies have shown that children experiencing MN exhibit externalizing behaviors, impaired socialization, and impaired daily living skills (Dubowitz et. al., 2005). However, little is known about the risk factors of MN for young children. Even less is known about the predictors of MN that co-occurs with other forms of maltreatment. This study aims to explore and compare the predictors of MN alone and MN co-occurring with other forms of maltreatment for young children.

Metohds: A total of 625 children aged 0 to 4 from the Longitudinal Studies of Child Abuse and Neglect (LONGSCAN) were included in the study’s sample  (N = 625). Predictors were various factors at child, family, and community levels including child’s socio-demographics characteristics, child’s health, caregiver’s socio-demographics, caregiver’s health and social support, and neighborhood factors. The dependent variable MN was represented by three categories, including no MN, MN alone, and MN co-occurring with other forms of maltreatment. ANOVA and chi-square tests were used for bivariate analysis to detect significant associations. Multinomial logistic regression followed to explore the relations between those significant variables and the outcome, controlling for the effects of other covariates.

Results: More than half of the children (67.7%) had no MN history; some of them (12.6%) experienced MN only, and nearly one-fifth of them (19.7%) experienced MN that co-occurred with other forms of maltreatment. Results of the multinomial logistic regression indicate that older children (Relative Risk Ratio [RRR] = .71, p < .05) and children of caregivers with fewer depressive symptoms (RRR = .96, p < .05) are less likely to experience MN compared with no MN whereas children of caregivers with religious belief (RRR = 2.46, p < .05) are more likely to experience MN compared with no MN. In addition, older children (RRR = .74, p < .000) and children of caregivers with higher levels of social support (RRR = .97, p <.05) are less likely to  experience MN co-occurring with other forms of maltreatment than no MN, whereas children of married caregivers (RRR = 1.94, p <.05) are more likely to have MN co-occurring with other forms of maltreatment compared with no MN.

Conclusions and implications: It is common for neglected children to have co-occurring types of maltreatment compared to having MN only. Consistent with previous studies (Asser & Swan, 1998), caregiver’s religious belief is a significant risk factor for MN. However, it is only associated with MN alone and not related with MN that co-occurs with other forms of maltreatment. Findings provide implications for child welfare workers to provide differentiated services to children with MN. For the MN alone case, caseworkers might need to resolve conflicts between caregiver’s faith and children’s medical care needs; whereas for the co-occurring cases, providing social support interventions might be helpful.