New Tool Can Help Pinpoint Risk for Violence

Deborah Brauser

September 26, 2012

September 26, 2012 — Using structured methods and tools, such as the Historical, Clinical, Risk Management–20 clinical subscale (HCR-20-C), may help psychiatric residents to improve risk assessments for violence in their patients, new research suggests.

In a retrospective study, records were reviewed for 151 violent and 150 nonviolent inpatients who were given violence risk assessments at admission by either psychiatric residents or attending psychiatrists. Interestingly, the attending psychiatrists were found to be only "moderately accurate" in their predictions, whereas the residents' assessments "were no better than chance," report the investigators.

However, once information from the HCR-20-C was added to their assessments, the accuracy of the residents' risk assessments was raised almost to the level of their more experienced peers.

"I think the number 1 takeaway is that there is this accuracy gap that no one has really found thus far between the more inexperienced mental health clinicians and those who are more experienced," lead author Alan R. Teo, MD, now a clinical lecturer in the Department of Psychiatry at the University of Michigan in Ann Arbor, told Medscape Medical News.

Dr. Alan Teo

Dr. Teo, who was a resident at the University of California, San Francisco, during the course of this study, noted that although this "would seem to be a common sense thing," he was surprised to find that the residents "did such a terrible job."

"You might as well have flipped a coin in terms of predicting whether a patient was going to be violent or not. But the other important finding of this study was that this isn't just a hopeless endeavor. You could use a brief tool to overcome that accuracy gap," he said.

The study was published online August 31 in Psychiatric Services.

Clinician Injuries Common

The investigators note that violence within psychiatric inpatient units and emergency departments is common; and it is a frequent contributor to injuries in clinicians.

"Trainees in psychiatry, clinical psychology, and other mental health disciplines often complete rotations in these acute settings and are especially vulnerable to being victims of patient aggression," they write.

Dr. Teo reported that he was witness to these types of incidents.

"In the first year of my residency in psychiatry, some of my co-residents were being assaulted by patients in the psychiatric units they were working on. And the administration was very concerned," he said

"We hoped to use the data from our study to learn how we might prevent future residents and other psychiatric staff from being hurt. And also to promote patient safety," said Dr. Teo.

The researchers write that although several structured risk assessment tools have been developed, these measures "are relatively new and are only beginning to be adopted in many areas of clinical practice."

According to a release, the HCR-20 is a tool that is routinely used in a number of settings in Canada, where it was first developed.

Improved Accuracy

For this study, patient records were assessed for 151 individuals who had assaulted staff while hospitalized involuntarily between January 2003 and December 2008. These data were compared with those of 150 inpatients with no recorded incidents of violence during their stays.

Violent episodes consisted of physical aggression, such as hitting, kicking, or biting.

A total of 38 psychiatry residents (mean age, 30.7 years; mean length of residency, 1.2 years) rated 52 of these patients on a 4-point assault precaution checklist at admission, whereas 41 attending psychiatrists (mean age, 46.8 years; mean length of postresidency experience, 13.7 years) rated the remaining 249 patients.

The HCR-20-C was used by trained research clinicians, blinded as to whether these patients later became violent, to examine the validity of the initial assessments. Area under the curve (AUC) was also used to assess validity of the assessments.

"As a general rule, the predictive validity of AUCs of .80 to .90 is considered excellent, .70 to .79 is acceptable, and .60 to .69 is modest," report the study authors.

Not surprisingly, the attending psychiatrists had significantly higher predictive validity in their estimates of violence risk than did the residents (P = .02).

Although the attendings' assessments were only deemed "moderately accurate" (AUC = .70), the residents' assessments were much worse (AUC = .52).

"Incremental validity analyses showed that addition of information from the HCR-20-C had the potential to improve the accuracy of risk assessments by residents to a level (AUC = .67) close to that of attending psychiatrists," write the investigators.

However, the HCR-20-C "yielded little incremental validity over that of the unstructured clinical risk estimates" for the attending psychiatrists (AUC = .69).

Training Implications

"This study has implications for training and education regarding violence risk assessment," write the researchers.

"Information needed to rate the HCR-20-C is routinely collected in the course of admitting patients..., and prompting trainees to rate this structured measure could help them to attend to valid risk markers from among the vast amount of information that may become available during the course of an admission workup."

The investigators add that improving these assessments could also help in developing antiviolence interventions.

"Similar to a checklist a pilot might use before takeoff, the HRC-20-C has just 5 items that any trained mental health professional can use to assess their patients," said Dr. Teo in a release.

"Basically, the upshot is that this simple tool is something that could help out people who otherwise were not doing a good job in making a risk or threat assessment," he said.

He added to Medscape Medical News that, especially in this age of what seems to be increased episodes of community violence, psychiatry needs to be as evidence-based as possible.

"Given concerns about random, or maybe not-so-random, acts of violence such as mass shootings, we need to think: what can we do to try and prevent those incidents in the future? This is a huge issue," said Dr. Teo.

"I hope that future research will also concentrate on using shorter, practical, user-friendly tools. And that will increase the likelihood that the decision makers will actually use it in routine clinical practice."

The study was supported by a grant from the National Institute of Mental Health, a Minority Fellowship from the American Psychiatric Association and the Substance Abuse and Mental Health Services Administration, and a Clinical and Translational Award from the National Institutes of Health. The study authors have disclosed no relevant financial relationships.

Psychiatr Serv. Published online August 31, 2012. Abstract