Climbing the Walls
A Survey of Psychotropic Ignorance
The defining architectural features of the prison are the walls—the real physical boundaries that separate the inside from the outside. Obviously, this closure captures the people who live and work in the prison; people do not flow freely across those physical boundaries. The doors remain locked, the walls are nearly impenetrable. Not so obviously, this closure also walls off information about what takes place on the inside. The prison is closed off not only to people but also to data, which have a hard time making their way into and out of the prison. The first problem I encountered when I began my research into the use of psychotropics in prison was finding out exactly what government-collected information was available on the topic. Cursory searches of both the scientific literature and government publications pointed me to an exceptionally narrow body of knowledge that reveals some of what we can know and, more important, what we cannot know about the use of psychotropics in prisons.
Administrative surveys of prisoners and institutional censuses are the central means by which we might know things about psychotropic use during incarceration. What they can tell us about these practices is limited, however—so limited that it might be said to qualify as a kind of nonknowledge, willful ignorance, or a front for malicious state secrecy. It is not just that we don’t know much about psychotropic drug use in prisons, it is that few procedures are in place to enable the collection of this information at all. Our knowledge is hampered by built-in limits to what can be known. I have examined all of the major population-based surveys of prescription drug use in prisons, and in this chapter I outline the scope of what prison pharmacoepidemiology can and cannot tell us about psychotropic use among imprisoned populations. Official government knowledge flows directly from the mechanisms of data collection that the government deploys. What the government knows is directly linked to the questions it asks. If the government does not ask certain questions, that leaves us in the dark.
The specific science that remains in the dark is pharmacoepidemiology, the study of the use and effects of drugs in populations. This combination of pharmacology (the science of drugs and their effects on the body) and epidemiology (the science of population health) applies the theories and methods of epidemiology to the study of the distribution and effects of drugs in populations after the drugs’ approval by the U.S. Food and Drug Administration (FDA). This science has its origins in the early twentieth century, when the nascent pharmaceutical industry experienced systemic problems with the safety of its products, leading to passage of the Pure Food and Drug Act of 1903. In 1961, pharmacoepidemiology experienced a seminal moment and began to play an important role in the U.S. drug regulatory process after the widely prescribed drug thalidomide, a sleeping aid marketed as safe for pregnant women, led to the births of thousands of babies worldwide with malformed limbs. This tragic episode spurred the development of the FDA’s four-phase clinical trial process, in which drugs are subjected to scientific scrutiny before they can be sold to the general public. (Relatedly, in chapter 3, I discuss the pivotal role that prisons and prisoners played in the unfolding of clinical drug trials in the United States.)
In addition to clinical trials, which are supposed to catch unsafe or ineffective drugs before they can go on the market, a parallel system has been established to monitor the safety of drugs after they have been approved for use in the general population. One key mechanism of this system is the FDA Adverse Event Reporting System (FAERS), which collects voluntary reports on “adverse events” and “medication errors” from health professionals, consumers, and drug companies. Using the interactive Public Dashboard tool available on the FDA’s website, members of the public can enter the names of drugs and see detailed and historical information about adverse events related to those drugs. This system, the very best source of public information about adverse drug events that the U.S. government has to offer, is not perfect; as the FDA website notes, the Public Dashboard tool “cannot be used to calculate the incidence of an adverse event or medication error in the U.S. population.” In 2007, the FDA added another system called Postmarket Drug and Biologic Safety Evaluations, which monitors the safety of drugs eighteen months after their FDA approval or after they have been used by ten thousand people.
Beyond these formal federal monitoring systems, numerous population health surveys collect data on psychotropic drug use in the general U.S. population. A growing body of research has analyzed sociological patterns in access to mental health services and psychotropic use in the noninstitutionalized U.S. population. Ryne Paulose-Ram and colleagues used nationally representative data from the National Health and Nutrition Examination Survey for the years 1988–91 to analyze psychotropic drug use and found that women used psychotropics at nearly twice the rate of men (4.6 percent versus 7.5 percent); they also found that African Americans and Mexican Americans self-reported lower use than did non-Hispanic whites (5.1 percent, 4.5 percent, and 6.6 percent, respectively). Studies of prescribing patterns for psychotropics suggest that members of racial and ethnic minority groups are less likely than whites to receive these drugs for a range of psychiatric diagnoses. Using 1992–2000 data from more than 5,000 patient visits recorded in the National Ambulatory Medical Care Survey and the National Hospital Ambulatory Care Medical Survey, Gail Daumit and her colleagues found that African Americans and Hispanics were less likely than whites to receive atypical antipsychotics, drugs that produce more side effects and are generally considered less effective than other drugs in managing psychotic disorders. This finding was supported by the results of a smaller, local study of 1,245 psychiatric patients. Other studies have found that African Americans and Hispanics with depression are less likely than whites with depression to receive pharmacotherapy, and the same racialized pattern holds true for anxiety disorders and among veterans with bipolar disorder.
It is important to note that institutionalized populations are regularly excluded from large-scale population health surveys, and they are definitely not included in large-scale studies of the generalized effects of prescription drugs. There is so much we do not know, and cannot know, about psychotropic drug use in prisons because of the particular ways in which the government has designed health surveys of prisoners and prison health care practices, both historically and today. Despite the large numbers of prisoners with mental health problems, it is impossible to know whether the patterns of psychotropic use in U.S. prisons are consistent with those in noninstitutionalized contexts (see chapter 4 for discussion of other institutionalized populations). The ignorance that permeates prison pharmacoepidemiology contrasts with the proliferation of knowledge about psychotropics in so-called free society.
Quantifying Psychotropics in Prisons
The U.S. prison system is not a unified, coherent thing. Rather, it is made up of many individual systems at different jurisdictional levels: city/county, state, federal, military, extralegal. This institutional heterogeneity presents a challenge to researchers and advocates who are trying to make sense of patterns of practice across and within systems. For the purposes of this chapter, I have set aside the array of small-scale studies conducted ad hoc by researchers in particular jurisdictions, examining an individual jail, for instance, or one state-level department of corrections. These local studies, which are relatively small in number, do provide important information about localized practices in particular places at particular moments in time. The Texas system is a good example. In the late 1990s, Jacques Baillargeon and colleagues conducted a pair of studies on prescribing patterns within the Texas prison system, one focused on antidepressants and the other on antipsychotics. They found that among Texas prisoners with formal diagnoses of depressive disorders, black and Latino prisoners were less likely than whites to be placed on selective serotonin reuptake inhibitors and more likely not to be prescribed any antidepressant treatment. Black prisoners were more likely than either Hispanics or whites to be prescribed tricyclic antidepressants. I imagine that this analysis was possible because the UT Medical System, for which Baillargeon and his colleagues worked, was the primary provider of medical care in the Texas prison system. The researchers were able to collect data on Texas prisoners’ psychotropic use through the medical records system shared by the UT Medical System and the prisons.
Such local studies may be both interesting and important, but they do not contribute much to the goal of constructing an empirical bird’s-eye view of the U.S. prison system as a whole that will allow us to evaluate claims about the use of psychotropics at the levels of all U.S. prisoners and across all U.S. prison institutions. A recent international meta-analysis found that issues of polypharmacy (the use of more than one drug at the same time), high and long-term dosing, and lack of documentation and monitoring are important in prison contexts; unfortunately, the authors identify only two articles reporting on research conducted within U.S. systems, neither of which actually documents system-wide use of psychotropics. One of the articles reports on individual cases of quetiapine abuse in Ohio prison facilities, and the other discusses the metabolic syndrome monitoring system used by the New Jersey Department of Corrections and focuses on prisoners taking second-generation antipsychotics.
If we had comprehensive and consistent information, going back decades, on the numbers of prisoners who have been administered psychotropics and for what reasons, that would be incredibly valuable for the science of pharmacoepidemiology. We do not have such information, however. First of all, the federal government has been collecting any form of nationally representative data about prisoners’ health only since 1974; before then, no systematic effort was made to collect any health information on prisoners. It would be hard to prove, but my guess is that the collection of prisoner health data began in response to the prisoner revolt and subsequent massacre at Attica Correctional Facility in 1971, the same year that the National Prisoner Statistics Program was transferred to the Bureau of Justice Statistics from the Federal Bureau of Prisons. The program started collecting statistical data each year on federal and state prisons back in 1926, although episodic raw counts of U.S. prisoners had been made as early as the 1850 census. Figure 1 lists the major administrative surveys of prisoners and institutional prison censuses that ask any questions about psychotropic use during incarceration. In the next section, I explore the questions these instruments ask in order to bring to light what they can tell us about psychotropic use in U.S. prisons and jails.
Administrative Surveys of Prisoners
Prisoners are seldom asked whether they have been prescribed or have taken psychotropic drugs during their period of incarceration, and because such queries were not included in multiple survey years over time, cross-sectional analysis is difficult, and historical analysis is impossible. Surveys have inconsistently assessed prisoners’ mental health status over time, complicating any effort to pose questions about the appropriateness of reported psychotropic use as a therapy for diagnosed psychiatric illness or reported psychological distress. Relatedly, surveys have conflated prisoners who report taking psychotropics during incarceration with prisoners who have psychiatric diagnoses, as if everyone who takes such drugs in prison is, by definition, mentally ill.
The Survey of Inmates in State Correctional Facilities (SISCF), a nationally representative survey of inmates housed in state prisons within all U.S. states, began collecting data on state prisoners in 1974 and continued every five years. In 2004, it was combined with the Survey of Inmates in Federal Correctional Facilities (SIFCF) to form the Survey of Inmates in State and Federal Correctional Facilities (SISFCF), which was later renamed the Survey of Prison Inmates. Each wave of the SISCF was structured to link the probability for selection of any inmates to the sizes of the facilities where they were incarcerated. In the first stage of the study design, statisticians created independent sampling frames for male and female prisons and then stratified prisons by census region and facility type (i.e., confinement versus community corrections) within each frame. In 1991, an additional sampling frame was added for security level of the facility (maximum, medium, minimum). In the second stage, prison officials provided rosters to census officials who then selected inmates randomly according to predetermined targets established by the Bureau of Justice Statistics. Interviewers from the U.S. Census Bureau conducted face-to-face interviews with consenting inmates in their facilities. Inmate participation was voluntary, and prisoners’ individual responses to survey questions were confidential and anonymous.
With each successive wave of the survey, questions about psychotropic use changed in important ways, making cross-wave comparisons nearly impossible. First, the key question of whether the inmate was currently taking a psychotropic drug was asked of the entire sample of inmates only in 1979, 1986, and 1991, and the wording and clarifying follow-up questions changed over time. In 1979, prisoners were asked a series of questions. First, they were asked if they were currently taking any medications. If they said yes, they were then asked to identify the names of up to three medications they were currently taking. In 1986, prisoners were asked if they were currently taking medications for “mental problems,” and in 1991, they were asked if they had taken medications for emotional or mental problems since they had been admitted to prison. In 1986 and 1991, inmates were not asked the follow-up question about their current medication use. To my knowledge, official government reports have not referenced these data about mental health care in state prisons, in part because of these differences in the questionnaires over time.
The 1997 and 2004 waves of the SISCF did ask whether inmates were taking psychotropic drugs during their incarceration; rather, only inmates who had ever taken psychotropic drugs prior to their imprisonment were asked this question. In 1999, the Bureau of Justice Statistics published a report titled Mental Health and Treatment of Inmates and Probationers based on data from the 1997 SISCF, the 1996 Survey of Inmates in Local Jails, and the 1995 Survey of Adults on Probation. For the purposes of the analysis, state inmates were defined as having a mental illness if they had “a current mental or emotional condition” or had experienced “an overnight stay in a mental hospital or treatment program.” Given these criteria, 16.2 percent of state prison inmates reported mental illness. From this survey, we know that in 1997, 60 percent of mentally ill prisoners received some form of mental health treatment; 50 percent of those who received any treatment received psychotropics. By contrast, 44 percent had received counseling or therapy and 24 percent had been hospitalized overnight in a mental hospital. The way in which these questions are nested, and the fact that they are not asked of every state inmate, really limits what we know about drugging practices in state prisons. In 1996, according to data from the Survey of Inmates in Local Jails, 41 percent of mentally ill inmates received some form of mental health treatment—only 34 percent of them were given psychotropics. This survey has been conducted seven times, in 1972, 1978, 1983, 1989, 1992, 1996, and 2002. But 1996 was the only survey year in which a question about psychotropic use during incarceration was asked.
In 2004, as noted above, the SISCF and SIFCF were conducted at the same time, together becoming the Survey of Inmates in State and Federal Correctional Facilities. The 2004 wave asked a series of questions about prisoner mental health and included, for the first time, a modified structural clinical interview focused on psychiatric symptoms for major depression, mania, and psychosis as specified in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders. The survey also asked prisoners if they were taking a prescription psychotropic medication in the year prior to their incarceration; 18 percent of state prisoners, 14.4 percent of jail inmates, and 10.3 percent of federal prisoners answered yes. A secondary analysis of data from the 2004 SISFCF and the 2002 Survey of Inmates in Local Jails found that 69.1 percent of federal prisoners, 68.6 percent of state prisoners, and 45.5 percent of jail inmates with diagnosed disorders had taken psychotropics during their incarceration. Taken together, these data also show that at each level of duration of incarceration, from jail detention to long-term imprisonment, prisoners are prescribed psychotropics more often than they receive other mental health treatments.
In 2016, the SISFCF was renamed the Survey of Prison Inmates. In that survey year, the following sequence of questions about psychotropic drug use was introduced:
- At the time of [your offense], were you taking prescription medicine for any problem you were having with your emotions, nerves, or mental health?
- Since you were admitted to prison, have you taken prescription medicine for any problem you were having with your emotions, nerves, or mental health?
If the respondent answered yes to the second question, he or she was then asked:
- Are you currently taking prescription medicine for any problem with your emotions, nerves, or mental health?
All respondents, regardless of their prior mental health diagnoses, were asked this sequence of questions, an improvement over previous surveys, because it enables the evaluation of psychotropic use within prisons across nationally representative populations of state and federal inmates. However, because the questions are new, the data gathered using them can be analyzed only for the 2016 survey year so far, and these data are not yet publicly available.
The National Inmate Survey is a collaborative survey of a 10 percent sample of correctional facilities that includes a minimum of one prison and one jail in each U.S. state. Conducted by the Research Triangle Institute under contract with the Bureau of Justice Statistics, the National Inmate Survey is the survey mechanism for ensuring correctional institutions’ compliance with the Prison Rape Elimination Act of 2003. The survey was first conducted in 2007, with additional waves following in 2008–9 and 2011–12. The 2011–12 wave was conducted in 233 state and federal prisons, 358 jails, and 15 special facilities (e.g., military, ICE, and Native American tribal facilities) between February 2011 and May 2012. In this wave, adult prisoners in federal and state prisons (but not the special facilities) were asked about their mental health histories and prescription psychotropic use since their incarceration. More than 100,000 prisoners and detainees participated in the survey. The 2007 and 2008–9 surveys did not ask about drug use. A report on the 2011–12 survey’s findings was published in June 2017.
In the survey, only prisoners who reported having diagnosed mental health problems prior to their incarceration were asked the following questions about psychotropic drugs and alternative mental health treatments:
- Have you ever taken any prescription medicine for any problem you were having with your emotions, nerves, or mental health?
- At the time of the offense for which you are currently serving time, were you taking prescription medicine for any problem you were having with your emotions, nerves, or mental health?
- Since you were admitted to any facility to serve time on your current sentence, have you taken prescription medicine for any problem you were having with your emotions, nerves, or mental health?
- Are you currently taking prescription medicine for any problem with your emotions, nerves, or mental health?
The survey then asked a detailed battery of questions about five categories of mental health problems: manic depression, bipolar disorder, or mania; depressive disorder; schizophrenia or psychotic disorder; posttraumatic stress disorder (PTSD) or anxiety disorder; and personality disorder. For each category, the following sequence of questions was asked:
- In the 30 days before you were admitted to this facility, were you taking any prescription medicine for your [mental health problem]?
- In the 30 days before you were admitted to any facility to serve time on your current sentence, were you taking any prescription medicine for your [mental health problem]?
- Since you were admitted to this facility, have you taken any prescription medicine for your [mental health problem]?
- Since you were admitted to any facility to serve time on your current sentence, have you taken any prescription medicine for your [mental health problem]?
- Think about when you were first told that you had [this mental health problem] after you were admitted to any facility to serve time on your current sentence. How soon after you were told did you start taking prescription medicine for your [mental health problem]?
- Are you currently taking prescription medicine for your [mental health problem]?
If the respondent answered no to the preceding question, this follow-up question was asked:
- Why aren’t you currently taking prescription medicine for your [mental health problem]?
So, while the National Inmate Survey provides the most comprehensive data on psychotropic use among prisoners with mental health diagnoses to date, it is structured in such a way that it does not capture respondents who have not been diagnosed with mental health problems but are nevertheless prescribed psychotropics or forced to take them in prison. It also provides no information about inmates who did not take such drugs prior to incarceration but may be taking them now. On one hand, this narrow approach makes sense given the survey’s focus on mental health care delivery—prison administrators want to know how well they are meeting the mental health needs of their inmates with acknowledged mental health problems. But on the other hand, the questions that are not being asked are the ones that might make it possible to tease out information about potential forced medication, overdosing, misdiagnoses, and prescribing errors.
Institutional Censuses of Prisons
While prisoner surveys provide useful but limited information about psychotropic use at the level of the prisoner, institutional censuses also shed light on the practice. In these censuses, prison officials themselves are queried about the policies and practices of their institutions. In 2000, the Census of State and Federal Adult Correctional Facilities included all 84 federal facilities, 1,320 state facilities, and 264 private facilities in operation on June 30, 2000. Across the years this census has been conducted—1974, 1979, 1984, 1990, 1995, 2000, and 2005—only the 1979, 1984, and 2000 iterations have asked about psychotropics, and the question was different each time:
- In 1979: “How many prisoners are receiving prescription medication for mental or emotional stress, anxiety, or depression?”
- In 1984: “How many residents/inmates are receiving psychotropic medications (such as Thorazine and Stelazine)?”
- In 2000: “Of all inmates confined in your facility on June 30, 2000, how many were receiving psychotropic medications (drugs having a mind-altering effect, e.g., antidepressants, stimulants, sedatives, tranquilizers, and other anti-psychotic drugs?)”
From the 2000 census, we know that, as a matter of policy, 73 percent of state prisons distributed psychotropics to prisoners and 71 percent provided therapy or counseling, although only 65 percent conducted psychiatric assessments. As a matter of practice, 10 percent of all state inmates were receiving psychotropics; 13 percent were in therapy or counseling. That amounts to 114,400 state prisoners receiving psychotropics in 2000. In five states at least 20 percent of prisoners were receiving psychotropics: Hawaii, Maine, Montana, Nebraska, and Oregon. In Alabama, Arkansas, and Michigan, in contrast, less than 5 percent of prisoners were taking such drugs. The California Medical Facility at Vacaville stands out among all state prisons in that year: 42 percent of the 3,070 inmates were taking psychotropics. Perhaps coincidentally, this facility was also the site of an extensive biomedical research program in the 1960s (see chapter 3).
From this census, we know that 95 percent of maximum-security prisons had policies concerning psychotropic distribution, compared to 88 percent of medium-security prisons and only 62 percent of minimum-security facilities. Additionally, the proportion of prisoners receiving psychotropics was higher in maximum-security (11 percent) than in medium- (10 percent) and minimum-security facilities (6 percent). More prisoners in public prisons than in private prisons took psychotropics (10 percent versus 7.7 percent). As far as I have been able to discern, this is the only time that an official report has mentioned the issue of psychotropic use in privately operated prisons. Prison size was also related to psychotropic policies: lower proportions of prisoners from large prisons (750 prisoners or more) took psychotropics compared with prisoners in smaller prisons (9.6 percent versus 10.4 percent).
The National Survey of Prison Health Care, conducted in 2011, was a onetime collaboration between the National Center for Health Statistics and the Bureau of Justice Statistics that asked prison medical officials about the provision and delivery of health care to prisoners. The survey targeted all fifty state prison systems and all correctional institutions overseen by the Federal Bureau of Prisons. It asked respondents if the prisons where they worked conducted mental health assessments and whether they had contract agreements for pharmaceutical services, but it did not ask about the provision of psychotropics to prisoners. This is a remarkable omission for a survey that was supposed to focus on health care services, and a strange omission given the history of questioning that preceded it.
Equitable access to high-quality mental health treatment should be an important goal in U.S. prisons and jails, but more prison pharmacoepidemiology is necessary to document patterns of practice and inequity in the system. The fact that vast public resources are allocated to mental health treatment for incarcerated persons further justifies efforts toward closer scientific scrutiny of psychotropic drug distribution in U.S. prisons. This research should cover many aspects of psychotropic prescribing practices in prisons. No survey has ever asked prisoners if they were forcibly administered psychotropics or if they were prescribed psychotropics without first receiving psychiatric evaluation. We cannot know about psychotropic use among prisoners who have not been diagnosed with psychiatric disorders. We cannot know whether psychotropics have been used forcibly in cases of crisis intervention or in emergency psychiatric units. We cannot know about drug-specific uses within and across prison systems. We cannot document historical use trends that account for changes in prison population growth in relationship to patterns of mental illness. We cannot know anything about undocumented uses of psychotropics in any prison system.
To make this problem even more difficult, we cannot know any and all of these things within and across racialized and gendered groups of prisoners. It has long been known that men and people of color are disproportionately incarcerated in the United States; what is needed now is a better understanding of how race and gender structure the provision of health treatment among these vulnerable institutionalized populations. Unfortunately, the research conducted on psychotropic drug use thus far has focused either on racial and ethnic minority groups or on specific gender groups, with no attention to the intersection of racial and gender structures. No study has documented or analyzed patterns of psychotropic drug use among state prisoners within specific race and gender groups. This narrow vision has led to a partial understanding of the mental health crisis among prisoners. A number of small-scale case studies suggest that women prisoners are more likely than their male counterparts to receive psychotropic drugs, but we have no idea how this difference looks when race and ethnicity are also considered.
Surveys and censuses foreclose many pressing questions concerning the use of psychotropics in prisons. Why do these government instruments do such a relatively poor job of documenting psychotropic drug use in the prison system? Just as the prison itself is designed to keep people locked in, these instruments are designed to keep prying eyes out. The next chapter describes an effort to get up and over the walls of the prison to learn more about how prisons deliver prescription drugs.