Commentary: Carceral Suicides—Some Exceptions and Paradoxes

Dr. Anasseril Daniel¹ has put forth a timely, concise summary of selected aspects of prison suicide, a concern of inestimable importance to the criminal justice system. Even allowing for the “inflation” of suicide rates resulting from incarcerated population figures that are not adjusted to account for variable admission rates,² and even allowing for the approximate equivalent rates in prison and extramural communities, suicide remains a serious problem in facilities with barred windows. Regardless of the setting and circumstances, some suicides may simply not be preventable. Nonetheless, prisons and jails are institutions that can aggravate or alleviate the potential for suicide in vulnerable inmates and that have an obligation to provide and properly deploy preventive, therapeutic, and remedial resources.

Prison administrators and clinicians will find Daniel's article most useful and relevant to their efforts to prevent suicide. Several of the findings concerning prison suicides correspond to features of jail suicides, and preventive measures are remarkably similar, as one might expect. To further the discussion, then, this response will broaden the scope to include jail suicides. When jail and prison suicides, or carceral suicides, are considered together, remarkable exceptions and paradoxes to general findings and recommendations present themselves—exceptions and paradoxes that are also important in developing preventive strategies.

Demographic and other actuarial correlates are typically regarded as more accurate and reliable, if highly limited, predictors of future adverse events than are clinical findings. Demographic trends must certainly be considered in risk assessment and management, but placing undue emphasis on them would be a mistake.³ Well recognized is the disproportionately high rate of carceral suicides by whites in comparison with African Americans. Citing Haycock,⁴ Daniel suggests that African-American suicides may be underreported by prison staff. Perhaps. Even so, the observation that whites are overrepresented compared with African Americans is supported by numerous studies. Though imprecise, racial/ethnic grouping should be more easily ascertainable than variables such as subjective stressors and individual motivations. Moreover, the predominance of suicides among whites outside of correctional systems⁵ is consistent with the finding behind bars.

Exceptions to this general finding should be recognized in fathoming the significance of race and ethnicity. In several studies of jail suicides, the discrepancy based on race was reversed, with African Americans the more frequent victims of suicide.^6,7 If African-American status were given diminished predictive weight in facilities wherein African-American suicides predominate, this consideration of race would misguide attempts at suicide risk assessment. Not to be overlooked, however, is the frequent omission in reports of the racial distribution in the facility or in the corresponding extramural population for comparison.

This is not to discount the significance of race and ethnicity altogether. Native American males have an extraordinarily high rate of suicide⁸ that should not be overlooked when they are incarcerated. Although comprising only two percent of federal prisoners, Native Americans account for five percent of suicides from 1993 to 1997.⁸ The suicidal risk of unilingual inmates who speak an uncommon language remains problematically unknown until they can be evaluated in their own language. Recognition of a pattern of racial/ethnic suicide in a facility can be instructive. New York jails once had a higher rate of suicide among Hispanic inmates,⁹ who were grouped together and able to converse in Spanish. The rate of suicide among Hispanics diminished substantially¹⁰ after they were integrated into the rest of the jail population. Although causal connections are elusive in a single naturalistic observation, perhaps in some cases integration results in more uniform access to medical and social services.

A troublesome paradox is that the least is known about the most critically important aspect of suicidal risk assessment in carceral settings: stressors and motivators. Studies and surveys more often report demographic findings than these two elusive but vital factors. In earlier work,¹¹ we categorized all prison suicidal stressors as acute and chronic and as institutional, interpersonal, legal, or medical in origin. In this scheme, interpersonal stressors are related to intimates, friends, and family outside of prison, whereas institutional stressors included those arising from other inmates. A particularly difficult institutional factor to study is the stressor of sexual threat, coercion, or rape. For a variety of reasons, this is likely to be underreported as a stressor, and may therefore be underappreciated as a stressor except by clinicians with sufficient professional experience in carceral facilities. Attention given by Daniel¹ to this potential suicidal stressor is therefore of particular interest.

In preventing carceral suicides, especially in jails, perhaps the single most important measure is initial screening at booking or reception, followed by effective follow-through. A variety of screening instruments have been used.³ Further research should help to establish which is the best instrument for this purpose. Any screening must include questions about current suicidal thoughts and past attempts as well as whether the individual presently appears depressed. Booking officers are taught to administer a multi-question screening instrument; add up the points; and, if a predesignated threshold score is achieved, notify a clinician for further assessment. Of importance, screeners who lack clinical backgrounds must be instructed to report specific signs of suicidality, such as acknowledgment of present thoughts of suicide, regardless of whether the inmate's score achieves the indicated threshold.

In prison settings, Daniel recommends administration of a suicide screening instrument to all inmates at reception. Those who show signs of suicidal thoughts or behavior should be administered a multi-dimensional suicide risk assessment. In addition, we suggest that clinical assessment for suicidality should not rely solely on a standardized questionnaire but should include as well a clinical interview that allows the inmate to express his own subjective experience. Whether a suicide risk assessment should be conducted at “every clinical encounter” is an interesting proposition, but certainly it should be performed with regularity and with more frequency in those identified to be at high risk.

The apparent lethality of single-cell placement is mentioned repeatedly in the literature, and indeed people tend to kill themselves when alone and unobserved. Daniel correctly stresses the recommendation from the NCCHC Standard and Guidelines,¹² that suicidal inmates should not be placed in a single cell unless they have constant supervision. This standard should be applied in prisons uniformly throughout the country. Moreover, the reality that some inmates who are not in single cells also kill themselves suggests that better protection would be one-to-one constant observation, regardless of whether the inmate's placement is in a single cell. Even constant and direct observation cannot be expected to be preventive unless done properly. Observation must either include direct visualization of the inmate’s head and neck or removal of ligatures and other potentially lethal instruments.

More critical than the single cell itself is the lack of appropriate observation, the omission of which exacerbates the risk when the suicidal inmate is isolated with potentially self-destructive items.¹³ In jail, the availability of personnel resources for observation is much more variable than in prisons. Even in comparatively better staffed hospital settings, clinicians can encounter objections to constant observation from administrators who are concerned with cost containment. In some jails, staffing is inadequate for general security needs, and extra assignments such as constant observation are not negotiable. The consulting mental health professional is faced with agonizing questions: Is the effort to improve the quality of services a worthy enough goal even if the desired standard is not soon achievable? Or should the professional wash his or her hands of any involvement, thereby avoiding compromise of the sound ethics and clinical standards so well established in hospitals and increasingly in prisons?

Beyond the need to establish and consistently apply full suicidal precautions for inmates who are presently acutely suicidal, intermediate levels should also be established for those at risk of non-lethal self-injury, whose risk is substantial but not immediate. Extremes between maximum precautions and no precautions, without intermediate levels, are inadequate in any hospital, jail, or prison.

The significant risks of “cheeking,” hoarding, swapping, and overdosing on medications in jails and prisons require extra precautionary measures. Daniel's recommendation that medications with low lethality be prescribed¹ preferentially should be followed. At the same time, inmates should not be denied the quality of psychiatric treatment that has become the standard of psychopharmacotherapy in the general population. Institutional formularies and other systems for accessing psychotropic medications should be sufficiently inclusive to allow for algorithmic medical decisions. Managed-care companies have limited antidepressants in jail formularies to tricyclic drugs because of the higher costs of newer selective reuptake inhibitors. The former should be avoided because of their potential lethality, but not completely. Individuals vary widely in sensitivity and allergy to various medicines, and mental disorders respond to different classes of medicines. Likewise, first- and second-generation antipsychotic drugs should be on hand. In or out of carceral settings, suicide can result when a mental disorder is inadequately treated with pharmacotherapy. Because of the cost of medicine in general, collaboration between administrative and clinical staff is key to ensuring a useful, but not spendthrift, formulary.

• American Academy of Psychiatry and the Law