A Space-Time Bayesian Analysis of Non-Medical Prescription Drug Overdoses and Child Harm in California: 2001-2011

Background & Purpose: Characteristics of the alcohol and illicit drug use environment are associated with negative outcomes for children.  It is unknown, however whether prescription drug abuse, which has reached epidemic levels in the US is similarly related to rates of problems like child maltreatment and child unintentional injury over space and time.  We used data from California zip codes between 2001 and 2011 to examine the following: Controlling for alcohol and other drug use, is prescription drug use associated with hospital discharges for child maltreatment or injury? 

Methods: Data were obtained from the California Office of Statewide Health Planning and Development (hospital discharge data) and the US census (environmental characteristics).  We conducted panel analyses of all California zip codes in the years 2001-2011 (n=18,517 space-time units).  These data were used to examine the ecological associations between non-medical prescription drug abuse, child maltreatment, and child unintentional injury, while controlling for indicators of other substance use and zip-code level environmental characteristics.   Due to changes in the size and shape of zip codes that occur over time, we used Bayesian space-time misalignment Poisson models, which control for zip code instability and spatial autocorrelation. 

Results:  The percentage of hospital discharges linked to prescription drug overdoses was positively associated with the number of hospital discharges for child maltreatment [relative rate=1.103 (95% confidence interval 1.010, 1.183] and child injury (relative rate= 1.061, (95% confidence interval 1.018, 1.103) over the ten year period, controlling for other drug or alcohol use and environmental factors. Among the other substance use variables, hospital discharges associated with alcohol abuse or dependence was positively associated with child maltreatment and negatively associated with child unintentional injury, while hospital discharges associated with illicit drug abuse or dependence was positively associated with both maltreatment and unintentional injury.

Conclusions: Contrary to societal beliefs that prescription drugs abuse is generally less harmful than alcohol or illicit drug use, increase in community level prescription drug abuse may contribute to a greater rate of maltreated or injured children.  Higher community rates of adults abusing prescription drugs may limit the amount of adults able to intervene against maltreatment or protect local children.  The geospatial distribution of prescription drug abuse overlapped in several areas with alcohol and illicit drug abuse or dependence, suggesting that some California zip codes may have enhanced risk from misuse of multiple substances.  Policy interventions designed to reduce substance abuse may also lower rates of harm for children.