Factors associated with suspected drug-facilitated sexual assault

Setting

In 2004, we recruited 7 of the 34 hospital-based treatment centers for sexual assault and domestic violence in Ontario. These centres were selected because they served both rural and urban areas of the province and people representing Ontario's culturally diverse population. Nurses and physicians provided on-call services 24 hours a day, 7 days a week, to victims, who generally presented within 72 hours of being assaulted. Services included crisis intervention, physical assessment and treatment of injuries, collection of medicolegal evidence, prophylactic medication for pregnancy and sexually transmitted infections, and referral to counselling and community agencies for ongoing support.

We received ethics approval for our study from the research ethics board of Sunnybrook and Women's College Health Sciences Centre and from each of the other participating sites.

Definitions

We chose the Delphi method,12 a systematic, iterative means of determining the parameters of a phenomenon, to generate key criteria for suspected drug-facilitated sexual assault. Using the Delphi method, an advisory committee, which comprised 2 toxicologists, a forensic biologist, a physician sexual assault examiner and the program coordinators from the 7 participating sites, aided in identifying 16 symptoms that might be associated with suspected drugging (Table 1). We classified cases of suspected drug-facilitated sexual assault as those in which the person reported a suspicion of having been drugged in combination with at least 1 of the 16 associated symptoms, gave a valid reason for believing that a sexual assault had occurred (Table 1) and presented within about 72 hours of being assaulted.

Data collection

We created a screening form to gather information about the participants. Staff members at an eighth sexual assault and domestic violence treatment centre piloted the form for 3 weeks. We revised the form on the basis of their feedback. The final version of the form included the list of valid reasons that would support a person's belief that she or he had been sexually assaulted and the list of symptoms associated with a suspicion of having been drugged. It also contained 19 other items in 5 brief sections: victim information, assault history, alcohol use, drug use and time frame (Appendix 1, available at www.cmaj.ca/cgi/content/full/180/5/513/DC2).

In May 2005, we trained 5 site coordinators by videoconference using a “train-the-trainer” model. During this session, we presented information about drug-facilitated sexual assault and the goals, procedures and inclusion criteria of the study. We introduced the screening form and explained how to complete it. We asked the 5 site coordinators to train staff members at their own sites. In January 2006, we provided in-person training to all available staff members at the remaining 2 sites.

For each consecutive person who presented to a participating site between June 2005 and March 2007, the attending health care provider completed the screening form. We asked site coordinators to fill in missing information, when it was possible to do so, and to return the forms to the study coordinator (N.R.), who reviewed them for completeness and participant eligibility. We entered the data from the forms into a database on a biweekly basis.

Statistical analysis

We examined variables pertaining to the location of the service; the date and the victim's time of arrival at the centre, as well as her or his sociodemographic characteristics (e.g., age, sex), health history (physical and mental health problems) and voluntary use of substances (over-the-counter medications, prescription medications, street drugs) in the previous 72 hours; and the assault (e.g., type of assault, presence of a weapon) (Table 2).

We first described the sample characteristics using descriptive statistics: frequency counts and percentages for categorical variables and means and standard deviations (SD) for continuous data. Next, we compared victims of suspected drug-facilitated sexual assault and victims of other types of sexual assault using χ² tests with cross-tabulation or the Fisher exact test when cell sizes were small. Finally, to determine what factors, if any, were associated with drug-facilitated sexual assaults, we entered variables that were significant at p < 0.10 in the bivariate analyses into a multivariable model using logistic regression. We entered age as a continuous variable. We classified time from assault to presentation as less than or equal to 24 hours or greater than 24 hours. We reclassified the location of service as large urban centre or other (suburb, smaller city, town, rural area). We examined all variables for multicollinearity. We used the Hosmer–Lemeshow goodness-of-fit statistic to check the fit of the model. We defined p values of 0.05 or below as statistically significant. We report the findings as odds ratios (ORs) with 95% confidence intervals (CIs).

Because of substantial missing data, we did not enter into the multivariable model 3 variables that were significant in the bivariate analyses: kissing or fondling, cunnilingus or fellatio, and presence of a weapon. To assess the effect of missing data on the remaining factors, we ran models with missing values set to “no,” with missing values set to “yes,” and with missing values left out. The results of these analyses were essentially the same, and we present here the model with missing values set to “no.”

Limitations

This study had limitations. First, the results can be generalized only to cases reported to specialized sexual assault services, which may represent only a fraction of the rapes in the community. In a US survey of 3001 women, Kilpatrick and colleagues5 found that only about one-fifth of those who reported having been sexually assaulted had received post-rape medical care. Thus, our study may represent the tip of the iceberg with regard to this issue. Second, it is possible that we did not collect some key information that might have further differentiated victims of suspected drug-facilitated sexual assault from victims of other types of sexual assault (e.g., victim's relationship to the assailant, activity immediately before the assault) or that might have aided in interpreting the results (e.g., whether assailant plied the victim with alcohol, amount of alcohol consumed). Finally, as might be expected, given that many victims reported amnesia of the event, substantial data were missing for some variables (e.g., type of sex acts perpetrated).

Conclusion

Suspected drugging was common in a sample of people attending sexual assault and domestic violence treatment centres. Providing access to hospital-based toxicologic screening services with the sensitivity to detect “date rape drugs” would help to meet the needs of victims of drug-facilitated sexual assault. At the same time, given that suspected drug-facilitated sexual assault is associated with victims' voluntary substance use, it is imperative that we increase awareness of the effects of using alcohol, particularly in combination with prescription, over-the-counter and street drugs, through public awareness campaigns, including posting information at clubs, pubs, bars, university and college campuses, and community health centres.7^,13^,20 We also need to educate men that a woman who is intoxicated is incapable of consenting to sexual contact and that using alcohol or drugs (or both) to incapacitate a woman for the purposes of sex represents sexual assault.

@@ See related commentary by Butler and Welch, page 493