Incidence of hospitalization due to child maltreatment in Taiwan, 1996-2007: A nationwide population-based study

Objectives: Little is known regarding the epidemiology of child maltreatment in Asian countries. This study aimed to examine the incidence of hospitalization coded as due to child maltreatment in Taiwan. Methods: We used inpatient claims data of the National Health Insurance for the years 1996 through 2007 for estimation. Hospitalization of children aged 0-12 years with an ICD-9-CM code 995.5 " maltreatment syndrome" or E960-E969 " assault" in the main or secondary discharge diagnosis in claims data was defined as hospitalization due to child maltreatment. Poisson regression was used to examine the trends. Results: A total of 1,726 children with 1,780 hospitalizations due to child maltreatment were identified during 1996-2007. The averaged annual incidence over 1996-2007 was 13.2/100,000 for infants, 3.5/100,000 for children aged 1-3 years, 2.1/100,000 for children aged 4-6 years and 3.3/100,000 for children aged 7-12 years. The incidence increased most among the infant group, from 4.8/100,000 in 1996-1997 to 25.4/100,000 in 2006-2007, a greater than 5-fold increase. The magnitude of the increase in incidence decreased with age, and no change in incidence was noted among children aged 7-12 years. The drastic increase in incidence in infants since 2003 corresponds with the introduction of use of ICD-9-CM code 995.55 " shaken infant syndrome" in Taiwan since 2003. Head injury was the main type of injury in all age groups. Conclusions: The incidence of hospitalization coded as due to child maltreatment in Taiwan differed by age group. Infants had the highest and the largest increase in incidence during the past decade. The increase might be due to improvement in awareness, competence and willingness to diagnose child maltreatment of medical staff in hospitals during the past decade after the revision of Child Welfare Law in 1993 and the new ICD-9-CM code for " shaken infant syndrome" being used since 2003. Practice implications: Physicians and nursing staff should be aware of and provide the detailed and specific information needed for proper discharge diagnosis coding, from which valid estimation of the incidence of hospitalization due to child maltreatment can be made. Administrative data should include both the N-code and E-code for each case of child maltreatment hospitalization to provide more complete information.