The Impact of SANE Programs on Adult Sexual Assault Investigation and Prosecution

• Part 1: Overview
• Part 2: SANE Effectiveness Depends on Local Circumstances
• Part 3: Analysis
• Victims Interaction With Police

Overview

Kristina Rose: Welcome to all of you to our JUICE session today, which stands for Just Using Intellectual and Creative Energy. I'm Kristina Rose, and I'm the Acting Director for the National Institute of Justice, a component of the Office of Justice Programs in the U.S. Department of Justice.

And today we have with us Dr. Rebecca Campbell. And Dr. Campbell is a professor of psychology at Michigan State University. And Dr. Campbell has made important contributions to the research in the area of how legal, medical and mental health systems respond to the needs of rape survivors. Her current research, which she'll be discussing today, examines the effectiveness of sexual assault nurse examiner programs in the prosecution of adult sexual assault.

She is the author of "Emotionally Involved: The Impact of Researching Rape," which won the 2002 Distinguished Publication Award from the National Association for Women in Psychology. Dr. Campbell also received the 2008 Distinguished Contributions to Psychology in the Public Interest Award from the American Psychological Association.

Her Ph.D. is from Michigan State University. And we are very grateful to have her with us today.

Becky.

(Applause.)

Rebecca Campbell: Thank you very much, Kris. Good afternoon, everyone. Thanks for coming. It's a pleasure to be here today and to have the opportunity to talk with all of you about my research on the impact of sexual assault nurse examiner programs on the investigation and prosecution of sexual assault cases.

What I'm going to be talking about today is a project that I recently completed that had two primary goals. The first was to conduct a series of research studies, where the goal was to try to figure out whether sexual assault nurse examiner programs, or SANE programs, can have a positive impact on adult sexual assault prosecution. So we wanted to find out whether prosecution rates would increase from before to after the implementation of a SANE program. And then, if so, why? Why is there that change? What is it about a SANE program that helps contribute to increased prosecution rates?

And then, building on those findings, the second goal in our project was to develop an evaluation toolkit for practitioners. The idea was to take science into practice by developing a step by step tool guide, so practitioners could also study whether they are having a positive impact on their criminal justice systems. And if not, what could they be doing to try to increase prosecution rates.

Now, I had a wonderful team at Michigan State that helped me do this project, so I want to take a moment to acknowledge my co investigators and all of my project staff.

And what I'm going to be doing in my presentation today is I'm going to start off with a very brief background of information about sexual assault nurse examiner programs and adult sexual assault prosecution. And then I'm going to give you an overview of the multistudy research project. I'm going to highlight our key methods and the key findings from this study. And then I'm going to wrap up by describing that practitioner toolkit and discuss its utility to practitioners.

So first off, by way of background, what we know about adult sexual assault prosecution is that national data indicate about 17 percent of all reported sexual assaults are going to be successfully prosecuted. Now, this rate is in the ballpark of other felonies, but it's low even among felonies. And this has been a long-standing frustration for survivors, for community practitioners, including law enforcement and prosecutors, as well as for rape victim advocates.

So in response to this problem, communities throughout the United States have been developing multidisciplinary response programs to try to improve reporting rates as well as prosecution rates. And one of these programs that's received the most attention nationally has been sexual assault nurse examiner programs, or SANE programs.

Now, SANE programs were created by the nursing profession. And they were created to provide comprehensive medical care to rape survivors. So they provide a complete examination of the survivor and treat any injuries that might have come from the assault. They also attend to victim's emotional needs. So they focus on the crisis of surviving sexual assault and making sure that the victims' emotional well being is attended to. They also do an excellent job with collecting forensic evidence, documenting the sexual assault crime, as well as any injuries that the victim might have sustained in the assault.

Most SANE programs also provide training and consultation to law enforcement and prosecutors in their communities, so that police and prosecutors know what to do with this forensic evidence and the injury documentation that the nurses have collected, and show them how it can be used to help build a successful case. And should the case go to trial, the nurses can provide expert witness testimony, so that they can explain to judges and juries the nature of the forensic evidence and the injuries.

So in light of the different kinds of resources that SANEs provide both survivors as well as the legal community, it really does raise the question of whether SANE programs could be making a dent on this problem of underprosecution of sexual assault. And what we know from prior research on SANE programs is that there have been several case studies that have suggested that SANE programs do increase prosecution rates once they've been implemented in a community.

Now, many of these are case-study research, so they don't have appropriate methodological controls to be able to determine whether the effect was really attributed to the SANE program. For that, you would need a quasi experimental design. And at the time we began our project, there was only one quasi experimental design that had ever been done. It was a research study, funded by NIJ, conducted by Crandall and Helitzer in New Mexico. And what they did was compare prosecution rates two years before to three years after the implementation of a SANE program in a New Mexico community.

And what they found was very striking. They found statistically significant increases in the number of forensic exams, or kits as they're sometimes called, being collected. The arrest rates went up. The number of charges went up. Convictions went up. Sentences went up. Everything went up. It was a complete out of the ballpark home run.

So what we wanted to do in our research was try to build on this work and examine whether there would be a positive impact of SANE programs in a different state, different context, different community. And we also wanted to use an even stronger quasi experimental design than what this prior study had done.

We wanted to look at a much longer time span. So we looked at five years before to seven years after the implementation of a SANE program. And we really wanted to focus on the comparability of the before and after cases because that's very critical in a quasi experimental design. And I'm going to come back to that point in a few minutes, because it really is quite, quite essential.

And then, building on prior work, we wanted to look at this issue of why there may be change. It would be very nice to see that SANE programs increase prosecution rates, but why is that? Is it as simple as if you introduce a SANE program, prosecution rates go up? Not likely.

So what is it about the program, about the community context, what are the critical ingredients that really need to be in place, for a SANE program to have the opportunity to make a positive impact on the criminal justice system?

Part 2:SANE Effectiveness Depends on Local Circumstances

Rebecca Campbell: And with that overview, let me tell you a little bit about the research studies.

The first question that we wanted to look at was this issue of whether we would replicate our finding that there was an increase in prosecution rates. So what we wanted to do was compare investigation and prosecution rates in a Midwestern county before and after the implementation of the SANE (Sexual Assault Nurse Examiner) program.

So what we did was we culled through the police records in the five largest police departments in this jurisdiction — and that did account for the majority of reported sexual assaults in this county — as well as the prosecutor records. So we could see how many cases were making it how far through the system before the SANE program was implemented and then how many cases were making it how far through the system after the SANE program, to see if we would get a statistically significant increase.

And then we wanted to examine the issue of why: Why would there be an increase if, in fact, we do replicate this?

So to do that, what we decided to do was switch methodologies here and use qualitative methods for this component of the project, where we did in-depth qualitative interviews with law enforcement and prosecutors to understand what their experiences were with the SANE program. And we interviewed police and prosecutors who were there both before and after the SANE program so that they could really speak to what is different in this community in the way it functions.

And then in the next piece of our project, we wanted to take a much closer look at the impact of the SANE program in the early stages of investigation and prosecution. So much of the literature is really focused on prosecution rates. But you can't get to prosecution unless the police send the case from the police to the prosecutors. So let's rewind and see what happens at the stage of the investigation.

So we did some very, very detailed coding from police records from the largest three departments. Again, we had to narrow it down a little bit because of the level of detail that we were coding. And what we wanted to see in these reports was whether there was any evidence that the SANE program was changing the way police were doing their investigation. Was there any evidence to suggest that the police were drawing upon the SANE program as a resource in that very early stage? And would that help predict cases moving on to the prosecutor and then ultimately beyond that?

Now, in the next component of the study, we wanted to understand these issues from the perspective of the survivors themselves. Up to this point I've been talking about the criminal justice system and the criminal justice perspective on this. What about the survivors themselves? So we did in-depth qualitative interviews with 20 survivors who had SANE exams and asked them about their experiences at the program as well as with the criminal justice system.

And then in the last component of the project, we did validation interviews with the forensic nurses. And we asked them about their experiences of providing patient care to the survivors as well as how they work with criminal justice system personnel in their community.

Now, in the interest of time, I'm not going to be able to share with you these last data, the validation interviews from the nurses. But suffice it to say that the information that they provided does validate the findings that I will be presenting to you today.

All right. So let's get to our first question — the big $6 million question of whether there is an increase in prosecution rates from before to after the implementation of the SANE program. Again, we're going to be looking at police records and prosecutor records.

I want to say a little bit about the methods here because it's an important issue. The case selection, we're looking at adult cases. And by "adult," we were constrained by our institution's IRB definition of what constitutes "adults," which means that they were 18 years old or older at the time of the assault.

All of these survivors reported a sexual assault; "criminal sexual conduct" is the term used in this particular state, which is first through fourth degree. The overwhelming majority of these are completed-penetration crimes.

But not all sexual assault cases that occurred in this Midwestern county actually go into the sample that we studied because in quasi experimental research, we have to make sure that the before cases are comparable to the after cases. So we're comparing apples to apples, not apples to oranges or apples to who knows what. So we're going to be doing a process of selecting cases from the before and the after to make sure the kinds of cases we're comparing are comparable in all elements except that these were before and these were after.

So what we did was, the before and after cases had to be the same in the sense that they all — all of the assaults occurred in the same county; all of the survivors had a completed medical forensic exam; all of them filed a police report; all of them had their police department's — had their reports filed with the same five departments. So we're not mixing up or confounding different departmental level context; it's the same context in both situations. And all of them had to have the kit analyzed by the state crime lab for DNA evidence. We want to understand whether this forensic evidence in the DNA that they find makes any difference; it needs to make it to the state crime lab to be analyzed, to know whether it's positive, negative or inconclusive for DNA evidence.

So our before cases met all of those criteria, except that their medical forensic exams were done by emergency department doctors in hospitals prior to the implementation of the SANE program in the five years prior. And then the after cases were similar in all those ways, except their medical forensic exams were done by the forensic nurses in the SANE program in the seven years after the implementation of the SANE program.

I also want to take a quick moment to tell you about how we measured prosecution rates because this is an important issue as well. As I said, traditionally the literature focuses on, Was the case successfully prosecuted, yes or no? But we all know there's a lot of steps that lead up to that process. So what we wanted to do was a much, much more refined coding of what we mean by "prosecution rates" to really understand how cases move through the system.

So what we created here is an ordinal category. And so that as we looked up case outcomes for the before cases and the after cases, we would classify their outcomes as one of these four categories. And the "ordinal," meaning that they build on each other.

So the lowest level, the 1s, the case would be classified as a 1 if the outcome was that the case was not referred by the police on to the prosecutors. Now, you'll recall that to get into the sample, you had to have reported to the police. So that's why the lowest level is you made a report but nothing happened to it. The police never sent it on to the prosecutor. The case, so to speak, died in the police department. So that would be category 1.

Category 2 meant that the case was referred on to the prosecutor. So the police moved it to the prosecutor, but the prosecutor didn't warrant it for prosecution. So it died very early on in the prosecutor's office.

The third possible outcome would be that the case was warranted by the prosecutor — meaning that the prosecutor is moving forward with the case — but the case was later dropped or acquitted at trial.

Now, I can appreciate that those sound like two very different scenarios that kind of got mushed together: something that gets dropped versus something that goes all the way to trial and gets an acquittal. True, they are very different, but they share this in common: They both represent a scenario where the prosecutor made an attempt, pursued prosecution, but in the end there's no conviction. And statistically our results bore out that it was OK to smush those two things together.

And then, finally, the fourth level, the highest level that a case could have been classified as, is that there was a guilty plea or there was a conviction at trial.

So let's take a look at our prosecution rates. What did we find? If we look at our first level, the ordinal level number 1, these are the cases that were not referred by the police on to the prosecutors. Before the SANE program was implemented, 49 percent of the cases ended right there in the police department. After the SANE program, that did decrease a little bit, from 49 down to 43 percent.

So for our second level — now, these reflect the cases that were referred to the prosecutor but the prosecutor dropped it very early on — before the SANE program, 17 percent of the cases died very early on in the prosecutor's office. And that dropped a little bit after the SANE program, down to 15 percent.

Looking at our third ordinal category, these are the cases that were warranted — there was an investment by the prosecutor in the case — but in the end there was no successful conviction. Ten percent of the cases ended that way before, and that slightly increased after the SANE program to 13 percent.

And then finally, looking at our highest ordinal level — these are the cases that were pled guilty or were convicted at trial — before the SANE program, 24 percent of the cases had that outcome. And that increased to 29 percent after the implementation of the SANE program.

Now, the $6 million question is, Is this statistically significant? Is this pattern significant?

So to test this, we used a technique called multilevel modeling because we needed to be able to account for the fact that the cases are coming from five different departments. Even though it's the same five departments before and after, we do have some contextual variability that we need to account for. So we used multilevel modeling with an ordinal regression framework so that we could predict case outcome in this four level variable.

So again, what we have here at the bottom represents our outcome variable: increased case progression through this criminal justice system. That's our dependent variable of how far the cases made it through the system.

So we wanted to see what factors predict whether a case would end up as a 1, a 2, a 3 or a 4. The first thing that we found was that the police department that examined the case, that was handling the case, was significant. So we have a type 1 effect right off the bat. What that means is it made a very big difference which law enforcement agency was handling the case.

So what you can see in this graph, we've got the five departments that we worked with, and you'll see the green bars represent the post SANE referral rates, standardized referral rates. And you can see the green bars all higher than the blue bars, so you're getting a feel that we very likely are going to have some statistically significant results. But what I want to highlight is department A.

If a survivor reported to department A, it was much less likely that the case was going to go forward. So we have a very striking department effect here. That all other things taken into account, which department the survivor happened to live in, what jurisdiction she lived in, where she — the neighborhood, if she went to department A, it didn't matter what the case was like, that department did not refer many cases on. The other ones, in varying degrees, did as well, though.

So after accounting for a police department as our type 1 effect, then we could look at some other contextual variables to see how they predict how far a case would move through the system. And we found two contextual factors that also decreased the likelihood of a case moving forward.

The first was a seasonal effect, what we call the "December effect." Cases that were being processed in the month of December were significantly less likely to make it to those higher stages of case outcomes. Why? Well, what happens in December? People are on vacation. So there's less organizational resources in the law enforcement agencies to attend to the cases that are reported in December. And that in and of itself was enough to make it less likely that a case was going to be prosecuted.

We also found a similar resource effect at the prosecutor's office. In the event of a prosecutor transition, which is a euphemism for prosecution re-election year, we saw an effect such that when the organizational attention, shall we say, of the prosecutor's office was divided, that it also made cases less likely to be prosecuted. So a case that was being processed in the year 2004 was much less likely to make it through the system in and of itself.

Now, how about some good news? DNA evidence, a huge predictor. It was a very statistically significant effect with very large odds ratios; it's around 3. So cases that had positive DNA evidence were approximately three times more likely to make it to those higher levels of case outcomes.

And then, after taking into account all of these things — the department effects; the negative, bad news; the DNA — the SANE program still had a statistically significant effect. So cases that were being processed after the SANE program was implemented, those after cases were much more likely to be prosecuted, make it further through the system, than the before cases.

So this type of multilevel modeling allows us to conclude — with pretty good certainty for a quasi experimental design — that we do see a statistically significant increase in prosecution rates after the implementation, relative to before the SANE program. And we're reasonably confident that that increase is attributable to the SANE program itself and not to other kinds of community context factors that occurred over those 12 years.

Analysis

Rebecca Campbell: So what we wanted to do next, then, was try to understand why. Why is there an increase? Is it just that you put a SANE program in place and prosecution rates go up? Why do rates go up?

So again, what we are doing here is using a variety of mixed methodologies to try to get at this issue. And the first stop on this journey was to talk with police and prosecutors in this community, who had been around both before and after the implementation of the SANE program, to get a feel from them, what's different, what has changed since this program was created.

So what we did in this study was in depth interviews with nine detectives or detective supervisors and all six prosecutors in the county sex crimes unit. And we asked them whether they perceived that there has been a change in the criminal justice response after the SANE program. We knew quantitatively that there was, but we were interested in whether the qualitative methods would validate our quantitative findings and whether they perceived a change in their normal operating procedure. And then we wanted to collect some information from them about what their experiences had been like, working with the SANE program.

So what I'm going to do now — again, these are qualitative data — so I'm going to share with you two exemplar quotes, one from law enforcement and one from a prosecutor, that really sort of highlight some of the key findings from this study. And then I'm going to show you a visual representation, or a map if you will, of what the full qualitative analysis suggested.

So first off, this is a quote from one of the prosecutors that we interviewed. We asked her, "Do you think the prosecution rates have gone up?" And she said, "Oh, absolutely." And then she went on to say:

"I think it has everything to do with the SANE program, I really do. When I get a case on my desk, I'm like, oh, look, they reported right away, a SANE report. Look, there's injury. I'm sitting here thinking to myself, I've got a good case. I've got corroboration. I've got medical. I've got a good case, and I will use that when talking to the defense attorney."

So what we're hearing from this prosecutor is that yes, something is very different. What arrives on her desk is now very different. She's got forensic evidence. She's got injuries. She's got something to work with to take to a defense attorney and get a guilty plea.

And similarly, when we talked to law enforcement, we also got a very similar picture from them. When asked if they found the SANE program useful, this particular detective supervisor said yes. And went on to say:

"SANE is much more organized than the hospital was. And the SANE information is received rather quickly. If not the same day, the following day, you have the information. With the hospital, it took time to go through their records to obtain any additional reports, which made it difficult for us to proceed with our investigation. SANE, certainly it saves us time. A search warrant takes a lot of time. You go to the hospital. If they don't readily give us the records we are requesting, well, then we have to look through all their files. And once you show them the search warrant, they usually comply. But it's just another step and another. And time is being burned up, when it could be used for other things that are more beneficial to the investigation."

So what this quote is doing is showing us the way in which the SANE program helps out in the early stages of the investigation. The information — meaning the forensic exam report and the injury documentation — the police get that right away. And they can use it right away in their investigation. Before, they had to get a search warrant from the hospital to get the information from the hospital, and that could take time. They only have so much time to spend on any one case, and they were burning up their time trying to get information out of the hospital, whereas now, after the SANE program, they get it right away.

So, in gluing these findings together, what we found from all of these qualitative interviews was that the police and prosecutors emphasize that the medical forensic evidence that is being collected by the SANEs, first off, is just far better quality, in evidence and the documentation, than what they have had. What they're working with is very different. They have better forensics than they ever had before. And that, in and of itself, is a key factor in creating a stronger case that's more likely to be prosecuted.

But what they also said was that the forensic evidence collected by the nurses, and the way in which they get it to them, saves the police time. And time is a very precious commodity to them because then they could use their time to put more investigational effort into the case. So they're not spending all their time trying to get the medical records. What they're doing is that they can now take that information from the nurses and move very quickly into other aspects of the investigation.

Now, the other thing that was highlighted in these qualitative data is that the police and prosecutors emphasized that the training that the SANE program provides for them about forensic evidence, about injuries, was also a very helpful resource to them, as was ongoing case consultation. Meaning that the police and prosecutors knew they could always turn to the SANE nurses and ask them a question: "Hey, what does this mean? What does this mean? Is this consistent with this?"

So they had a go to always available to them as they were building their case. So those two factors provided them with more information and resources. That also contributes to them putting in more investigational effort into the case. And again, that was instrumental in creating a stronger, more complete, more thoroughly investigated case.

So, in the next piece of the project, we're going to switch methods here. We're going to go back to quantitative for a second. The qualitative data there started highlighting that the SANE program might be having a key influence early in the process in the investigation. So we wanted to pursue this a little bit more by taking a really close look at the police records from the three largest departments.

So we did actual coding of actual police reports from the three largest departments. And what we did was code whether or not and in what degree the SANE was involved in the case. So if we read through the report, do we see any evidence that they used this information from the SANEs in a way that influenced their investigation? So, was there a victim exam? Did they ask for a suspect exam? I'll explain what that is in a moment. Was there evidence of this kind of case consultation that I just mentioned?

And what we wanted to know was whether this kind of involvement with the SANE program would predict whether the case was referred from the police to the prosecutors. Because again, you can't get a case to be prosecuted if it doesn't get to the prosecutor. So here we're focusing on that very critical juncture of trying to predict: Does the case make it from the police on to the prosecutors?

So our analytic approach here, again, our dependent variable is now logistic regression; we're looking at this decision of, does it refer from the police to the prosecutors? And again, we're going to be using multilevel modeling.

I'm going to go through this a little bit more visually than mathematically in the interest of time.

What we found was that in those police reports, if there was a victim medical exam documented in the police report, the police were more likely to go collect other kinds of evidence. So once they had a medical forensic exam, they were much more likely to actually go to the crime scene and collect other evidence there that might substantiate, corroborate, what they found in the medical forensic exam.

The other thing that we found was that if there was a SANE suspect exam — now, not all SANE programs do this particular service; the one that I studied did — this is an interesting service. This is where the SANE nurse does a complete forensic exam of the suspect's body. It's not done at the SANE program, in order to protect the safety and the confidentiality of the survivor. It's done usually at the law enforcement agency. So the nurse comes in and does a complete head to toe assessment of the suspect to see if there's evidence that might corroborate what they found in the victim exam.

And what we found was that in those cases where there was a suspect exam, law enforcement was also more likely to go collect other kinds of evidence. So, go to the crime scene. Maybe go to his house and look for other types of evidence that might substantiate.

We also found that this meant that the police were more likely to interview the suspects. And I know that sounds kind of strange. Well, if you're doing a suspect exam, of course you're going to interview them. We were actually surprised the number of times when they had an identified suspect, or a known suspect, or even a suspect that was temporarily in custody, where they actually didn't do an interview. So the SANE suspect exam was very instrumental in predicting that there was going to be an interview.

And then finally, we saw a trend effect — that's why it's in dotted lines there — that if there was a SANE suspect exam, the victim was somewhat less likely to withdraw from the process or drop out of the process of prosecution. And I'm going to get to the survivor data in a moment, but I'll preview it here because it's relevant.

What the survivors told us when we asked them about this — because we wondered, why would a SANE suspect exam influence her behavior — because she interpreted this as saying: "Oh, they're taking this seriously. If he has to get the same kind of exam I went through, that means they're taking it seriously. And even though I'm upset, and I want to stick my head in the sand, I don't ever want to think about this ever again." They take this as a sign that the system is interested in their cases, is invested in their cases. And no matter how upset or traumatized they might be, they're going to hang on. They're not going to drop out because they've seen investment from the criminal justice system in their case.

So let's build on that a little bit and see what the survivors had to say in our interviews with them. Again, what we did was in depth qualitative interviews with 20 rape survivors who had exams at the SANE program. And we asked them about their experiences with the forensic nurses and the rape victim advocates at that exam. And then we also asked them about their experiences with the police, the prosecutors and their case outcome.

And again, what I'm going to do is share with you a couple of sample quotes and then show you a visual representation of the model that emerged from the full analysis.

So when we asked survivors about their experiences at the SANE program, the first thing that they always said to us, always, all 20 interviews, had nothing to do with police or prosecution, nothing. Had nothing to do with anything legal. When we asked them about their SANE experiences, where they always started was something like this. This is what one of the survivors told us.

She said: "It was very, at least in my case, it was a very supportive experience. It made me feel right away, the same night it happened, I'm not a victim, and I can get over this. You realize you're not the only one, because that's how you feel. I think SANE is very good as far as the human process is concerned because it doesn't matter whether you prosecute or not, you still have to heal."

So what we're hearing from the survivor and the other survivors is that the SANE program didn't focus on the issues of prosecution. They focused on them, the people, the survivor who has been hurt and traumatized. And they addressed the human elements first. Prosecution, eh, can come, can go. Because it doesn't matter whether you prosecute or not, you still have to heal. And the SANE program was focused on their healing first and foremost.

And similarly, this is what another survivor said. She said: "Up until that time, it would have just been formalities — collecting evidence, the police doing that — and it was the first time, at SANE, where I felt like human, after going through such a horrendous experience and made to feel like I was a bitch in heat or a pig being led to the slaughter. It's right away that it starts the healing."

So it's the healing that's most important to the survivors, that they're not being dehumanized anymore. They're not being objectified anymore. They can become human again at the SANE program. And from that, other things will naturally follow.

So what we found from the other qualitative interviews, gluing it all together is that the overwhelming majority of the survivors had a very positive experience with the SANE program, that really focused on promoting their healing and their psychological recovery from the rape. And then, from that, then they were more willing and able to participate in the criminal justice system. So it's that sort of irony of, don't focus on the prosecution elements right away. Focus on the survivor and her well being. And in so doing, you actually will get increased participation with the criminal justice system because you tended to her first and foremost.

So then we asked the survivors about their experiences with the criminal justice system. And again, even though overall we saw a statistically significant effect, that more cases were moving through the criminal justice system, in this particular sample, we had about a 50 50 split, where 11 of the survivors described very negative outcomes and nine described very positive outcomes. So I'm going to share with you one quote from each of these findings.

So with respect to the negative experiences, this was a very, very common situation that the survivor explained. She said: "There was like no reaction from the detectives, no 'How are you doing with this? Are you okay?' They were just victim blaming. They were looking at me like, 'You had control of this situation; you should have did this,' or, 'you shouldn't have done that.' If they hadn't been cold or unapproachable, it would have been easier for me to tell them what happened."

So what this survivor is telling us is that in her first experiences with the police, because they didn't attend to the humanness, there was no "are you okay"; there was just no shine of compassion toward her; she shuts down, and she's actually not giving information. It would have been easier for me to tell what happened.

If you go further into this interview transcript, the survivor talks about how she didn't give information to her detective because she didn't trust him. Because she didn't feel like it was okay to reveal the full story. That's not going to help a case. But she couldn't do it because she felt no trust from him.

Now, on the flip side, this is what the survivors who had positive experiences tend to say:

"The police were consoling. They were careful, you know. They didn't bombard me. A man had just assaulted me, and I felt calmness right away. I felt safe. They weren't like question, question, question. They made sure I was okay and safe, felt safe there. They let me sit down, be away from the building, and they didn't come at me right away, wanting to know this and this and this. They gave me my time and my space."

So I asked, "Why is that important to you?"

"Well, you need somebody. You need people to be careful with you and to be careful of the way they talk to you and treat you and approach you. Because the way I look at it now, I don't know this man. I don't trust anyone."

So what this survivor was saying is that the experience of being interviewed after an assault can be retraumatizing, revictimizing. They've just been assaulted by a man and had their space violated. And if you come at them question, question, question, they're not going to get much information. That the experience of, "Are you okay? Here, it's okay. Let's get away from the building where this happened; give you a moment," she was able to open up and actually work with the detective and give more information that ultimately was helpful to the case.

Victims Interaction With Police

Rebecca Campbell: So in bringing all of these findings together, what we see from this issue of how and why SANE programs have an effect is that there are two very distinct, mutually reinforcing but distinct tracks: one of which is the work that SANE programs do with the patients — the survivors themselves — and then there's the work they do with law enforcement.

With respect to patients, they focus on providing very high quality patient care — the human element. They do not pressure reporting. They do not encourage reporting. They don't talk about reporting until the very, very, very, very, very end. It's about the survivor and her well-being. They link the patients to advocacy and counseling services in the community. They do provide information about the criminal justice system. It's a very complicated process; the survivors need information. They provide that. They educate them about injuries and forensic evidence as well.

This helps facilitate survivors' healing, and it gives them some hope that something might actually happen with their case. And when you have that, you have a victim who is more willing and able to participate in the criminal justice system.

And then we also have the SANE programs' work with law enforcement. What they provide law enforcement is very good quality medical forensic evidence. Everybody was very clear: The quality of the evidence is far superior to what they were getting from hospital emergency departments. But they also provide education and training on how to use that evidence. It's not just enough to have it; what do you do with it? They address that issue through education and training and ongoing case consultation. And that helps police put more investigational effort into the cases. And that helps them create a more complete, stronger case.

So then, why did we get a statistically significant increase in prosecution rates after the SANE program? Well, if you take increased victim participation, and you add it together with a case that's been more thoroughly investigated, it's not a hard stretch to figure out, then, why we see more cases moving through the criminal justice system.

The SANE program works on two levels: preparing the survivors for the criminal justice system and providing resources to the law enforcement community to help create a stronger and more thoroughly investigated case.

So in the second part of the project, what we wanted to do was build on these findings, to take the research findings and start to bring them into practice. And to do that we created a practitioner toolkit. And the idea behind this toolkit is that SANE programs can have a positive impact on the criminal justice system, but not necessarily so.

We just saw, there are some real critical ingredients, in terms of patient care philosophy, resources that you provide to the criminal justice system. Those things need to be in place for you to see a positive impact occurring.

So what that tells us then is that we need to be able to find ways for SANE programs to monitor their program processes and community outcomes. SANE programs need to be studying themselves. They need to be able to have the evaluation resources to see if they are doing the things that would lead to positive outcomes and to assess what kind of outcomes they're getting.

So what we did was we developed a practitioner evaluation toolkit that teaches evaluation skills and builds capacity for SANE programs. So what we did was as we were doing our research, we developed a step by step process that practitioners could do. Now, granted, they're not going to do a very large scale, quasi experimental design with multilevel modeling and all of that. But can you teach them how to do a basic before and after design? Yes.

Can you teach them about comparability of samples? Yes.

Can you link them to different resources to help them do some simple analyses? Absolutely.

And that's what we were trying to do in this toolkit.

So what we did was we took the process of collecting evaluation data and broke it down into a very straightforward seven step process. So first off, we taught them about evaluation questions. And we acknowledged that there are many different ways to evaluate the effectiveness of a SANE program, one of which is legal. And that is the focus of this toolkit, but it is not the only way you can measure the effectiveness of a SANE program.

We teach them a little bit about evaluation design. We tell them what a logic model is, and we provide them sample logic models that show how and why SANE programs might affect case outcomes. And we explain the logic and the steps of doing a pre post design.

Then we provide instruction on relationship building. Now, they very likely have some relationships with law enforcement and prosecutors for services. It's a very different kind of relationship when you're asking for access to their records. So it's a very different kind of relationship building for evaluation, and we provide some guidance on that process.

We teach them about sampling. Again, very important in quasi experimental research, particularly in field based, program initiated evaluations, that they get the sampling right. So we teach them how to do that.

And then we tell them the step by step process for doing data collection. And we provide sample forms, so they can just copy them right out of the toolkit, start filling out the data as they collect it in their community.

We even teach them how to do data analyses. Again, we're not teaching multilevel modeling, but we teach them how to work with Excel for some frequencies, percentages. And then we link them to some online calculators that will help them do some simple nonparametric tests to see if there has been an increase in their community.

And then, finally, this toolkit has a very strong emphasis on utilization — using the data to create change in their communities. So if they found they did not have a positive impact on their community, we provide information and step by step directions for doing focus groups in their communities, to bring together the key stakeholders, try to identify where the problems are, and then work together to create an action plan where they might start to see an increase in prosecution rates.

And for those that did find positive effects, we provide information resources on ways of sustaining those positive changes in their community.

So we think that this toolkit has a lot of utility to practitioners. We see it as a very useful resource for closing the science practice gap by giving practitioners the tools they need to evaluate their own programs. There are about 500 SANE programs in existence. I can count on one hand the number of SANE programs that have been studied with any kind of methodological rigor. We have a bit of a problem here. So we need to find ways of getting resources to SANE practitioners to help address this gap.

So this resource provides help for programs if they're not having a positive impact. We know that this intervention model can. So if it's not, what can we be doing, what can these programs be doing, to try to create that change? And right now my team and I are examining ways that we can get this toolkit ready for national dissemination to share with other SANE programs throughout the country.

So I thank you very much for your attention.

(Applause.)