INTRODUCTION
Sexual crimes are a major cause of public concern. When new offenses are committed by known offenders, such concerns are increased when it appears that the offense should have been predicted and prevented (Mann, Hanson, and Thornton Reference Mann, Hanson and Thornton2010). Sexual assault is a unique type of contact violence with serious negative consequences for the victims. This type of violence is also considered a significant global issue and public health problem (Hwa et al. Reference Hwa, Chen, Wu, Shun, Liu, Lee and Chen2010). A previous study estimated that 13% of women and 3% of men are victims of sexual assault at some time in their lifetime (Spitzberg Reference Spitzberg1999). However, assessing the risk of sexual offenders is an essential task that provides the basis for many decisions that have profound consequences for both offenders and the public. Thus, providing first-line specialists, such as practitioners, parole officers and probation officers, with highly effective methods that minimize the risk of misclassification is critical (Hempel et al. Reference Hempel, Buck, Cima and van Marle2013).
Sexual assault not only damages or impairs a victim’s body (through disfigurement, loss of the ability to bear children, loss of hearing or vision, paralysis, or human immunodeficiency virus [HIV] infection) but can also cause serious psychological trauma, including an increased fear of further assault, symptoms of posttraumatic stress disorder (PTSD) such as depression, anger, crying alone, substance abuse, nightmares, and feelings of depression, anxiety, fear and powerlessness (Yang et al. Reference Yang, Miller, Zhang, LeHew and Peek-Asa2014). The Annual Report of the National Police Agency in Taiwan (http://www.npa.gov.tw) has estimated that 4,245 and 3,752 sexual assault cases were reported in Taiwan in 2012 and 2013, respectively, with high clearance rates in both years (96.06 and 96.06%, respectively). Nevertheless, many victims do not seek medical care or report the assault to the police. There is an increasing epidemic of sexual assault in Taiwan and in other Asian countries (Hwa et al. Reference Hwa, Chen, Wu, Shun, Liu, Lee and Chen2010).
To the best of our knowledge, however, few evidence-based studies have attempted to determine the risk of incident psychiatric disorders in sexual assault victims in Taiwan, who face the burden of this sexual crime. To identify the incidence of psychiatric disorders in sexual assault victims, the present study explored the potential for condition-related incident psychiatric disorders, because the incidence has been considered to underscore crucial implications for understanding of the overall pathogenesis of psychiatric disorders in this population. As mentioned, because sexual offenses cause harm to the victims’ physical and psychological functions, the purpose of the present study was to explore the context of the incidence of psychiatric disorders among sexual assault victims in Taiwan by conducting a nationwide population-based cohort study.
METHODS
Data Sources
In Taiwan, the National Health Insurance (NHI) program has provided health care for the population since its inception in 1995. This single-payer health care program provides medical coverage for more than 99% of the population of Taiwan. The Bureau of NHI manages the claims and administrative data contained in the National Health Insurance Research Database (NHIRD) and releases it for research purposes. The bureau depersonalizes the data to ensure individual privacy by scrambling any personal information before releasing the data to researchers. In the NHIRD, diseases are defined using International Classification of Diseases 9th Revision Clinical Modification (ICD-9-CM) codes (Chi et al. Reference Chi, Wang, Chen, Wang, Chen and Tung2015).
In this study, we retrieved the data from the Longitudinal Health Insurance Database 2005 (LHID2005), which contains the claims data of 1,000,000 beneficiaries enrolled in the year 2005, randomly sampled from the Registry for Beneficiaries of the NHIRD, which contains the registration data of all beneficiaries (approximately 25.68 million individuals) for the period from January 1, 2005 to December 31, 2005. The LHID2005 has been validated as a representative sample of the Taiwanese population in terms of age, sex and average payroll bracket (Cheng et al. Reference Cheng, Kao, Lin, Lee and Lai2011). The LHID2005 has been extensively used in thousands of published epidemiological studies (Chen et al. Reference Chen, Keller, Kang, Hsieh and Lin2013). Because the National Health Research Institute has addressed any confidentiality assurance problems, all procedures were performed in accordance with the guidelines of our institutional ethics committee and adhered to the tenets of the Declaration of Helsinki. All patient information was anonymous.
Study Subjects
We first recruited only those subjects who received a new diagnosis of sexual assault from 1997 to 2010. Subjects who received a psychiatric disorder diagnosis before the sexual assault were excluded to prevent overestimating the risk of incident psychiatric disorders between January 1997 and December 2004. Figure 1 illustrates the selection process. The final study cohort group comprised 81 sexual assault victims without any co-morbidity of psychiatric disorder and 324 control subjects who were matched at a ratio of 1:4.
Figure 1 Flowchart of selection of the study population. LHID2005, Longitudinal Health Insurance Database 2005
Definition of Variables
ICD-9-CM codes 995.83 (adult victims) and 995.53 (child victims) were used to identify sexual assault victims. The outcomes of our interest were incident psychiatric disorders, namely suicide (E950–E959), schizophrenia (ICD-9-CM: 295.xx) and bipolar disorder (ICD-9-CM: 296.xx).
Statistical Analysis
All statistical analyses were performed by using SAS version 9.4 (SAS Institute, Cary, NC). The χ 2 test and two-sample independent t test were used to compare categorical and continuous variables, respectively. A Poisson regression model was also adopted to estimate the incidence rate ratio (IRR) for incident psychiatric disorders between the sexual assault victims and controls after adjustment for related confounding factors. A two-sided p value of <.05 was considered statistically significant.
RESULTS
The demographics of the matched cohort are illustrated in Table 1. We identified 81 sexual assault victims and 324 matched controls. The mean age was 18.39 (sd 10.23) years for the sexual assault victims and 18.39 (sd 10.18) years for the controls (t test, p=1.00). The table shows that there were no significant differences in sex or residential area between the two groups (χ 2 test, p=1.00).
Table 1 Comparisons of Characteristics among the Matched Cohort
During the follow-up period, two sexual assault victims (2.47%) and one control (.31%) developed a psychiatric disorder. The incidence density (ID) per person-year of psychiatric disorders among the sexual assault victims was significantly higher than the controls (9.2 v. 1.1 per person-year, p=.037) (Table 2). In addition, the proportions of suicide among the sexual assault victims and controls were 1.23 and 0%, respectively. The ID of suicide was significantly higher among the sexual assault victims than the controls (ID 4.5 v. 0 per person-year, p=.043).
Table 2 Incidence Density of Psychiatric Disorder among the Matched Cohort
As shown in Table 3, the sexual assault victims exhibited an increased risk of incident psychiatric disorders (crude IRR 10.63, 95% confidence interval [CI] 2.06–54.81) with borderline significance compared with the controls. After adjustment for age, sex, income, residential area and co-morbidity, sexual assault victims remained at a significantly higher risk of incident psychiatric disorders (adjusted IRR 3.40, 95% CI 1.04–26.41).
Table 3 Incidence Rate Ratio (IRR) for Psychiatric Disorder among the Matched Cohort
a IRR was calculated using Poisson regression.
b Adjustment for age, sex, income, urbanization and co-morbidity.
DISCUSSION
Clinical–Epidemiological Aspects of Psychiatric Disorder and Sexual Assault
In 1997, Taiwan enacted the Sexual Assault Crime Prevention Act, amending it with numerous provisions in 2005. The Sexual Harassment Prevention Act was also enacted in 2005 and amended in January 2006. The passage of these two acts offered a comprehensive safeguard for protecting personal security, physical autonomy and privacy. Sexual assault includes rape, sexual coercion, being forced to penetrate someone else and unwanted sexual experiences without physical contact (McCauley and Casler Reference McCauley and Casler2015). Experts have estimated that unreported incidents (the so-called “dark figure of crime”) outnumber reported cases by 7- to 10-fold. Therefore, it is reasonable to infer that statistics on forcible rape in Taiwan underestimate the overall rate of sexual assault cases. As of the beginning of the twenty-first century, two women on average are forcibly raped every hour in Taiwan. Perceptions of government officials, the academic community and the public are in consensus that sexual assault in Taiwan must be handled prudently (Wu Reference Wu2008).
In this study, the mean age of the sexual assault victims was 18.39 (sd 10.23) years. A previous study indicated that 80% of sexual assault survivors experience their first assault before the age of 24 years and that college campuses are a particularly crucial venue for prevention (Sinozich and Langton Reference Sinozich and Langton2014). For college campuses, efforts aimed at reducing risky alcohol use, whether in the form of campus policies and programs or brief interventions in college health centers, would benefit victims by incorporating universal messages about healthy relationships and healthy sexuality; such efforts would reduce or prevent sexual violence (McCauley and Casler Reference McCauley and Casler2015).
Sexual assault differs from other crimes. It is highly private and commonly lacks third-party witnesses, requiring victims to describe the event themselves and to collect biological evidence from their bodies as soon as possible after the incident (Lou, Huang, and Lee Reference Lou, Huang and Lee2009). In addition, while the physical injury typically includes relatively minor injuries such as bruises, scratches, bites and sprains, the psychological damage and fear resulting from an assault is deep (Crowley Reference Crowley1999). In this study, sexual assault was found to be an independent risk factor for incident psychiatric disorders after adjustment for potential confounding factors. People who have been sexually assaulted are more likely to report depressive symptoms compared with people who have not been sexually assaulted (Bailey et al. Reference Bailey, Baker, McElduff and Kavanagh2016). The prevalence rates of PTSD and depression in sexually assaulted women are both estimated to be more than 30% (Zinzow et al. Reference Zinzow, Resnick, McCauley, Amstadter, Ruggiero and Kilpatrick2012). Exposure to lifetime sexual violence often results in an elevated likelihood of PTSD and major depressive disorder (Bailey et al. Reference Bailey, Baker, McElduff and Kavanagh2016). Sexual assault victims most commonly report distress with the physical effects of the assault as well as heightened fear and anxiety when in public places (Fuller Reference Fuller2015).
Experiencing sexual violence is also related to heightened risks of suicide and suicide attempts (Sumner et al. Reference Sumner, Mercy, Dahlberg, Hillis, Klevens and Houry2015). Adult female sexual assault victims are more likely to consider and attempt suicide than their non-sexually abused counterparts (Behnken et al. Reference Behnken, Le, Temple and Berenson2010). However, the association does not appear to be direct, because some sexual assault victims do not exhibit suicidal tendencies (Cloitre, Scarvalone, and Difede Reference Cloitre, Scarvalone and Difede1997). A possible reason is that some risk and protective factors affect the probability of a sexual assault victim becoming suicidal (Swahn and Bossarte Reference Swahn and Bossarte2007). In the present study, we did not achieve sufficient statistical power to effectively evaluate the presence of incident suicide between the sexual assault victims and controls and then to subsequently adjust for confounding factors given the relatively small sample. From a clinical perspective, many health care or clinical approaches to sexual violence prevention remain in early development and, in general, there is limited experimental evidence of the efficacy of such approaches (Sumner et al. Reference Sumner, Mercy, Dahlberg, Hillis, Klevens and Houry2015). Although the mechanism by which sexual violence and suicide are related still remains to be elucidated, these findings indicate that physicians should routinely enquire not only about physical domestic violence but also about emotional and sexual abuse – particularity given the increasing evidence that emotional abuse may have a higher health impact than physical violence (Jewkes Reference Jewkes2010).
Perceived Limitations
One of the major limitations of this study is that the NHIRD does not contain detailed information regarding the socio-economic status and family history of systemic diseases, which may be risk factors for psychiatric disorders. Second, evidence derived from a retrospective cohort study design is generally lower in statistical quality because of potential biases related to adjustments for confounding variables. Third, a meticulous study design and control measures for confounding factors were used; however, bias resulting from unknown confounders may have influenced the results. Fourth, all data in the NHIRD are anonymous; consequently, relevant clinical variables, such as serum information, imaging and pathology findings, were unavailable. Fifth, although the codes for sexual assault and psychiatric disorder have been used in previous studies, none of the cited studies is a validation study that has determined the accuracy of the coding. Our estimation may have suffered from some level of misclassification bias. Finally, the study investigated a specific population (Han Chinese); therefore, as with other population-specific studies, the findings might not be generalizable.
Conclusion
In the present study, the incidence of psychiatric disorders was significantly higher in the sexual assault victims than in the controls. Assessment of psychiatric disorders should be implemented in the integrative care of this population. In addition, physicians providing clinical care to the sexual assault victims should receive more all-round training regarding the features and management of this type of violence.
Tao-Hsin Tung received the PhD degree in Public Health from National Yang-Ming University, Taipei, Taiwan in 1996 and 2005, and was a PhD candidate in the Department of Crime Prevention and Correction, Central Police University, Taoyuan, Taiwan. In 2004, he joined the Department of Medical Research and Education, Cheng-Hsin General Hospital, as an Associated Researcher, and in 2008 became a Researcher. His current research interests include clinical epidemiology, biostatistics, disease screening, medical law and criminology.
Sheng-Ang Shen obtained his PhD degree from Emory University where he studied psychology. He is now a professor at the Department of Crime Prevention and Correction, Central Police University, Taoyuan, Taiwan. He is a licensed clinical psychologist in Taiwan, and a fellow of the Taiwan Association of Clinical Psychology. He has been involved in the treatment and evaluation of sexual offenders in the community and has also taken on the supervising of therapists in prison. His current research interests are etiology of sexual offending and analysis of deception or lying words from linguistic styles.
Chih-Hao Chen obtained his master’s degree from National Taiwan University where he studied psychology. He is a licensed clinical psychologist in Taiwan, and a fellow of the Taiwan Association of Clinical Psychology.
Chien Huang obtained his PhD in 2014 from National Taiwan University where he studied dual cognitive process analysis of sexual offending and violent assault. He is an assistant professor at the Department of Clinical Psychology at FU JEN Catholic University. He is a licensed clinical psychologist in Taiwan, and a fellow of the Taiwan Association of Clinical Psychology. He has been involved in the treatment and evaluation of sexual offenders in the community and has also taken on the supervising of therapists in prison. His current research interests are etiology of sexual offending and analysis of deception or lying words from linguistic styles.
