Brief mental health screening for corrections intake.

Authors' note: Points of view expressed in this article do not represent the official position or policies of the U.S. Department of Justice.

Correctional administrators need brief, cost-effective, easy-to-administer and reliable mental health screens to initially identify mentally ill detainees, who can become disruptive and a threat to themselves and others. Current mental health screening at corrections intake varies greatly--from one or two questions to a full-scale clinical analysis. Available instruments are often costly and time-consuming, making them impractical for daily screening of a large number of inmates at intake. As a result, even though most prisons and jails screen inmates for mental illness during booking, (1) research has shown that they miss the majority of inmates with mental health problems, particularly those with less obvious symptoms. (2)

Researchers, through funding by the National Institute of Justice (NIJ), have now developed and validated two brief, free mental health screening tools that proved effective in identifying various levels of mentally ill detainees at intake: the Correctional Mental Health Screen (CMHS) (3) and the Brief Jail Mental Health Screen (BJMHS). (4) The screens use standard one-page questionnaires that correctional officers with modest training can administer in three to five minutes and score simply by adding up "yes" answers.

Both screens proved valid when compared with far longer and more detailed screens administered by trained clinical assessors. The CMHS screens were effective in identifying nine categories of mental disorder in both male and female inmates. The BJMHS was effective in identifying male inmates with mental disorders and is being refined to increase its effectiveness in identifying female detainees with anxiety- and stress-related mental illness.

Using the Screening Instruments

CMHS. The CMHS uses separate questionnaires for men and women: the Correctional Mental Health Screen-Male (CMHS-M) asks 12 yes/no questions, and the Correctional Mental Health Screen-Female (CMHS-F) asks eight yes/no questions about current and lifetime indications of serious mental disorder. Both screens take about three to five minutes to administer. Six questions regarding symptoms and history of mental illness appear on both questionnaires, including whether the inmate ever has been hospitalized for nonmedical, including psychiatric, reasons. The remaining questions on each test focus on types of mental disorders more prevalent in that gender. It is recommended that male inmates who answer five or more questions "yes" and female inmates who answer four or more questions "yes" be referred for further evaluation.

BJMHS. The BJMHS is an eight-item yes/no questionnaire that takes about two to three minutes and requires minimal training to administer; it asks six questions about current mental disorders and two about any history of hospitalization or medication for mental or emotional problems. Inmates who answer "yes" to two or more questions about current mental disorders or acknowledge having been hospitalized or taking medication for mental or emotional problems are referred for further evaluation. Instructions for administering the screen appear on the back of the form. Correctional classification officers, intake staff or nursing staff can administer the screen without specialized mental health training, but may receive brief informal training before administration.

The Correctional Mental Health Screen

First phase. The researchers combined into one composite interview questions from five screening modules for a range of mental disorders. (5) The resulting Composite Mental Health Screen consisted of 53 items and took about 25 minutes to administer.

Researchers then administered the composite screen to randomly selected adult detainees in Connecticut's five jails (four for men and one for women) within 24 to 76 hours after admission. Inmates younger than 18, those considered high-risk or in restricted housing, and those already under medical or mental health care were excluded. Twenty percent of those screened underwent a comprehensive (45-minute to three-hour) clinical assessment using a combination of instruments. The results of the composite screen were compared with the clinical assessments and information about the inmates from correctional records, including mental health scores and overall risk scores.

Second phase. After comparing the test results, the researchers eliminated the questions with the fewest variations in answers. They then separately analyzed two samples, each consisting of one-half of the composite screen interviews. They used the results of these analyses to determine which questions best predicted nine diagnoses of mental illness associated with emotional and behavioral instability (including risk of harm to self or others). Problems adhering to the facility's activity schedule and disciplinary standards were also considered.

Based on their analyses, the researchers selected 12 questions for male inmates and eight questions for female inmates, which they tested on 206 inmates. Follow-up clinical assessments showed that the screens identified both male and female inmates with serious mental disorders in all nine categories. The screens proved highly valid in identifying depression, anxiety, full and partial posttraumatic stress disorders (PTSD), selected personality disorders, and the presence of any current mental disorder. Using a cutoff score of five or more "yes" answers, the CHMS-M was 75.5 percent accurate in identifying male inmates with a previously undetected mental illness. Using a cutoff score of four or more "yes" answers, the CHMS-F was 75 percent accurate in identifying female inmates with a previously undetected mental illness.

The clinical assessments found the incidence of serious mental illness among the inmates to be far higher than in the general population and comparable to that in psychiatric settings. This finding is especially significant given that inmates who had already been referred for mental health care because of obvious behavioral signs of mental illness or a history of mental health hospitalization were excluded from the screening.

Recommendations. Based on interviews with officers who administered the screens, the researchers suggest the following improvements for the administration procedure:

* Provide additional informal training to clarify the purpose of the screen and improve interviewing and probing techniques;

* Have nurses, if available, administer the tool to uncooperative inmates or those who feel uncomfortable answering questions about symptoms of mental illness; and

* Offer a computer-assisted version of the tool, which may increase responses to questions.

The Brief Jail Mental Health Screen

The BJMHS is adapted from the Referral Decision Scale, a 14-item questionnaire designed to identify inmates with schizophrenia, bipolar disorder and serious depression. Although the Referral Decision Scale is effective in identifying inmates with mental illness, it is less so in diagnosing specific disorders and questions about lifetime rather than current symptoms may overestimate the need for current mental health treatment. (6)

Development/administration. The developers of the BJMHS shortened the Referral Decision Scale to eight questions and revised several questions to ask about current symptoms. Informally trained jail intake staff administered the screen during booking to both male and female detainees in four county jails, two in New York and two in Maryland, from May 2002 to January 2003. Nurses administered the screen to some inmates as part of a health screen or to inmates who were too intoxicated to answer the questions during booking. Twelve percent of the inmates screened using the BJMHS were referred for further assessment.

Validation. To validate the BJMHS, trained clinical assessors administered the Structured Clinical Interview for DSM-IV, a longer and more detailed instrument that identifies lifetime and current mental disorders, on a sample of inmates who had undergone screening. The sample included both male and female inmates, and inmates who had and had not been referred for assessment on the basis of the BJMHS. They found that the BJMHS correctly classified 74 percent of male inmates and 62 percent of female inmates.

However, 35 percent of the female inmates who were not identified for referral by the BJMHS were diagnosed as mentally ill based on the clinical interview. The researchers discovered that the instrument did not measure symptoms of anxiety associated with the high incidence of PTSD among female detainees. NIJ is currently funding additional research through spring of 2007 to examine these cases and adapt the BJMHS to include questions that would measure these conditions.

Conclusion

The CMHS, because it uses separate questionnaires for men and women, proved effective in identifying both male and female inmates in need of mental health treatment. Also, excluding obviously mentally ill inmates from the screen highlighted its ability to identify those inmates whose symptoms were less obvious. The high rates of mental disorder found in the follow-up clinical assessment indicate the screen's potential utility in helping provide the correct diagnosis and treatment for those inmates. The BJMHS proved effective in screening male inmates but was less effective for female inmates. Thus, the instrument is being refined by adding items related to stress and anxiety disorders that are more prevalent among female inmates.

Both screens hold promise as powerful tools for standardizing and increasing the accuracy of initial mental health screening in correctional facilities. Their effectiveness in identifying inmates in need of mental health treatment compares favorably with the longer, more cumbersome, and training-intensive tools used in the clinical assessments. Their brevity, use of yes/no questions, simple scoring techniques and availability at no cost make them well-suited for quick mental health screening of large numbers of inmates during booking.

For copies of the instruments and the full grant reports, visit www.ncjrs.org/pdffiles1/nij/grants/210829.pdfandwww.ncjrs.org/pdffiles1/nij/grants/213805.pdf.

ENDNOTES

(1) Beck, A. and L. Maruschak. 2001. Mental health treatment in state prisons, 2000. Special Report. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Available at http://www.ojp.usdoj.gov/bjs/abstract/mhtsp00.htm.

Steadman, H. and B. Veysey. 1997. Providing services for jail inmates with mental disorder. Research in Brief. Washington, DC: U.S. Department of Justice, National Institute of Justice. Available at http://www.ncjrs.gov/pdffiles/162207.pdf.

(2) Parsons, S., L. Walker and D. Grubin. 2001. Prevalence of mental disorder in female remand prisons. Journal of Forensic Psychiatry, 12 (1): 194-202.

Teplin, L. 1990. Detecting disorder: The treatment of mental illness among jail detainees. Journal of Consulting and Clinical Psychology, 58(2):233-236.

(3) Ford, J. and R. Trestman. 2005. Evidence-based enhancement of the detection, prevention, and treatment of mental illness in correctional systems, Final report. Available at www.ncjrs.org/pdffilesl/nij/grants/210829.pdf.

(4) Osher, F., J. Scott, H. Steadman and P. Robbins. 2004. Validating a brief jail mental health screen, Final technical report. Available at www.ncjrs.org/pdffilesl/nij/grants/213805.pdf. For more information on this study, see also: Steadman, H., J. Scott, F. Osher, T. Agnese and P. Robbins. 2005. Validation of the brief jail mental health screen. Psychiatric Services, 56(7): 816-822.

(5) The instruments used were the screening module for the Structured Clinical Interview for DSM-IV for major mental disorders; the Primary Care PTSD Screen for posttraumatic stress disorder; the Iowa Personality Disorder Screen; the Referral Decision Scale; and a substance use screen (not included in the final instruments).

(6) Veysey, B., H. Steadman, J. Morrissey, M. Johnson and J. Beckhead. 1998. Using the Referral Decision Scale to screen mentally ill detainees: Validity and implementation issues. Law and Human Behavior, 22(2): 305-315.

Andrew L. Goldberg, a social science analyst with the Office of Research and Evaluation at NIJ, monitors a number of programs including mental health for corrections. Brian R. Higgins is a senior writer/editor for Lockheed Martin Information Technology. Doris Wells, managing writer/editor for NIJ's update column in Corrections Today, also contributed to this article.