CHAPTER 14

Defining and Assessing Competency to Stand Trial

by Ronald Roesch and Stephen L. Golding

Competency to stand trial is an important issue for many psychologists. Because competency to stand trial is raised significantly more often than the insanity defense, psychologists involved in forensic assessment and consultation are likely to have frequent experience with it. In this chapter, we will present an overview of the competency laws, research, and methods of assessment with the aim of providing forensic psychologists with the basic information they need to conduct competency evaluations. We do not believe, however, that this chapter will sufficiently prepare a novice forensic psychologist to carry out an evaluation. As we will make clear, the issues surrounding a competency determination are highly complex. An evaluator needs not only a high level of clinical knowledge and skills but also considerable knowledge of the legal system.

We urge the reader interested in pursuing work in the competency area to supplement this chapter with other materials (e.g., Golding, Roesch, & Schreiber, 1984; Petrella & Poythress, 1983; Roesch & Golding, 1985; Williams & Miller, 1981; Winick, 1983) as well as workshops and other forms of continuing education.

DEFINING COMPETENCY

Provisions allowing for a delay of trial because a defendant was not capable of proceeding have long been a part of the legal process. English common law allowed for an arraignment, trial, judgment, or execution of an alleged capital offender to be stayed if he or she "becomes absolutely mad" (Hale, 1736, cited in Silten & Tullis, 1977, p. 1053). Over time, statutes have been created in the United States and Canada that have further defined and extended the common law practice (see Verdun-Jones, 1981, and Webster, Menzies, & Jackson, 1982, for reviews of the Canadian competency law and practice). The modern standard in U.S. law was established in Dusky v. United States (1960). Although the exact wording varies, all states use the Dusky standard to define competency (Favole, 1983). In Dusky, the Supreme Court held that

It is not enough for the district judge to find that "the defendant is oriented to time and place and has

______________________

*This chapter was prepared while the first author was a visiting professor in the Department of Psychology, Arizona State University. Preparation of this chapter was supported in part by a fellowship to the first author from the Social Sciences and Humanities Research Council.

some recollections of events," but that the test must

be whether he has sufficient present ability to consult with his lawyer with a reasonable degree of rational understanding -- and whether he has a rational as well as factual understanding of the proceedings against him. (p.402)

Though the concept of competency to stand trial has been established in law, the legal definition, as exemplified by Dusky, has never been explicit. What is meant by "sufficient present ability?" How does one determine whether a defendant "has a rational as well as factual understanding?" To be sure, the courts have provided some direction to evaluators. For example, one might plausibly interpret the Dusky standard as implying that a defendant with amnesia would be incompetent to stand trial. But a number of legal decisions have held that this is not necessarily the case (e.g., Ritchie v. Indiana, 1984; Wilson v. United States, 1968). The reasons for this are varied. In Missouri v. Davis (1983), the defendant had memory problems due to brain damage. Nevertheless, the Missouri Supreme Court held that amnesia by itself is not a sufficient reason to bar the trial of an otherwise competent defendant. In Montana v. Austed (1982), the court held that the bulk of the evidence against the defendant was physical and not affected by amnesia. Finally, in a Maryland decision (Morrow v. Maryland, 1982), the court held that, because of the potential for fraud, amnesia does not justify a finding of incompetence. The court also stated that everyone has amnesia to some degree since the passage of time erodes memory. These decisions are of interest because they support the view that evaluators cannot reach a finding of incompetency independent of the facts of the legal case -- an issue we will return to later.

The problems in defining and assessing competency should be clear from these few examples. As might be expected, one finds a broad range of interpretations of the Dusky standard. Since the courts have given mental health professionals a large share of the responsibility for defining and evaluating competency, it should not be surprising to find that mental status issues such as presence or absence of psychosis have played a dominant role in the findings of evaluators. In fact, evaluators initially involved in assessing competency seemed to be equating psychosis with incompetency (Cooke, 1969; McGarry, 1965; Roesch & Golding, 1980). Furthermore, evaluators in the past rarely took into account the specific demands of a defendant's case.

There is some evidence that this situation is changing. Early evaluators did not usually have any training in legal issues. Most were employed by state mental hospitals -- the site of the majority of evaluations of competency. Typically, they had no training in either the assessment of competency or matters of law. As a consequence, the evaluations were based on the same standard mental status examinations that had been used with other patients in the hospital. If psychological tests were used at all, they were used as a diagnostic tool to determine presence or absence of psychosis.

In the past 10 years or so, these entrenched practices have been challenged. Research has provided evidence that the presence of psychosis is not sufficient by itself for a finding of incompetency (Roesch & Golding, 1980). It is also the case that better training programs for professionals in forensic psychology and psychiatry have been developed. Many graduate psychology programs and law schools cooperate to provide instruction in psychology as well as law, and a number of departments of psychology include forensic psychology as an area of expertise (Roesch, Grisso, & Poythress, 1985). Another reason for the changes in existing practices has been the work of McGarry and his colleagues on the assessment of competency (Lipsitt, Lelos, & McGarry, 1971; McGarry, 1965; McGarry et al., 1973). Their work was the starting point for a more sophisticated and systematic approach to the assessment of competency. Before turning to a review of assessment methods, we will provide a brief overview of the legal procedures involved in competency questions.

OVERVIEW OF PROCEDURES

Laws regarding competency vary from state to state, although most jurisdictions follow procedures similar to the overview we will describe in this section. Clinicians should consult their own state statute for the specific law and procedure applicable in each state.

The issue of competency may be raised at any point in the criminal process, and if the court determines that a bona fide doubt exists as to a defendant's competency, it must consider the issue formally and usually will order an evaluation (Drope v. Missouri, 1975; Pate v. Robinson, 1966). Though many states will allow these evaluations to be conducted on an outpatient basis, most of them are conducted in institutional settings, primarily forensic units within state mental hospitals. We have argued (Roesch & Golding, 1980) that inpatient evaluation is unnecessary in all but perhaps a small percentage of cases. As we will discuss later in this chapter, this is because most determinations of competency are easily made on the basis of a screening interview. This being the case, inpatient evaluations are unnecessarily lengthy and costly.

Efforts in recent years have been aimed at creating community-based evaluation procedures (e.g., Beran & Toomey, 1979; Fitzgerald, Peszke, & Goodwin, 1978; Melton, Weithorn, & Slobogin, 1985). Laben, Kashgarian, Nessa, and Spencer (1977) estimated the cost of the community-based evaluations they conducted in Tennessee was one-third the cost of the typical mental hospital evaluation (see also Fitzgerald, et al., 1978). We expect a major change in the competency procedures in the next 10 years. By then a majority of competency evaluations will take place in outpatient settings or in local jails. The assessment procedures we describe later in this chapter are explicitly designed for use in such settings.

One legal issue that may concern evaluators is whether information obtained in a competency evaluation can be used against a defendant during a trial or other legal proceeding. While some concerns have been raised about possible self-incrimination (Berry, 1973; Pizzi, 1977), it is generally agreed that information obtained in an evaluation will be limited to legal hearings on the issue of competency unless the defendant is informed prior to the evaluation of its potential uses and competently consents (Estelle v. Smith, 1981). The defendant may object but still be required to attend an evaluation but may, of course, refuse to cooperate with evaluators. In such a case, any information that was obtained would be limited to a hearing on the competency issue only.

Once a competency evaluation has been completed and the written report submitted (see Petrella & Poythress, 1983, for a discussion of the content of these reports), the court may schedule a hearing. If, however, both the defense and the prosecution accept the findings and recommendations in the report, a hearing does not have to take place. It is likely that in the majority of states, a formal hearing is not held for most cases. If a hearing is held, the evaluators may be asked to testify, but most hearings are quite brief and usually only the written report of an evaluator is used. In fact, the majority of hearings last only a few minutes and are held simply to confirm the findings of evaluators (Steadman, 1979). The ultimate decision about competency rests with the court, which is not bound by the evaluators' recommendations (e.g., North Dakota v. Heger, 1982). In most cases, however, the court accepts the recommendations of the evaluators (Steadman, 1979; Williams & Miller, 1981).

At this point defendants found competent proceed with their case. For defendants found incompetent, either trials are postponed until competency is regained or the charges are dismissed. The disposition of incompetent defendants is perhaps the most problematic area of the competency procedures. Until the case of Jackson v. Indiana (1972) virtually all states allowed the automatic and indefinite commitment of incompetent defendants. In Jackson, the Supreme Court held that defendants committed solely on the basis of incompetency "cannot be held more than the reasonable period of time necessary to determine whether there is a substantial probability that he will attain that capacity in the foreseeable future" (p.738). The Supreme Court did not specify how long a period of time would be reasonable nor did it indicate how progress toward the goal of regaining competency could be assessed.

The Jackson decision did have a substantial effect on commitment laws. Many states revised their statutes to provide for alternatives to commitment as well as limits on the length of commitment (Roesch & Golding, 1979). Twenty-six of the 50 states had, as of 1979, limits ranging from 6 to 18 months. A few states based length of treatment on the length of the sentence which would have been given if the defendant was convicted. It is interesting to note that 24 states continued to allow indefinite commitment, seemingly in contradiction to Jackson. It is possible that the laws in some states have changed since our review.

Once defendants are found incompetent, they may have only limited rights to refuse treatment (see Winick, 1983, for a review). Medication is the most common form of treatment, although some jurisdictions have established treatment programs designed to increase understanding of the legal process (e.g., Pendleton, 1980; Webster, Jenson, Stermac, Gardner, & Slomen, 1985) or that confront problems that hinder a defendant's ability to participate in the defense (Davis, 1985).

This brief overview of the competency procedures is intended to provide a basic understanding of the process. For a more complete discussion of the legal issues as well as a review of empirical research on the various aspects of the competency procedures, the reader is referred to Roesch & Golding (1980), Steadman and Hartstone (1983), and Winick (1983).

ASSESSING COMPETENCY

Though there has been some confusion over the definition of competency, there nevertheless appears to be good agreement among evaluators about whether a defendant is competent or not. The few studies of reliability report that pairs of evaluators agree in 80 percent or more of the cases (Golding, et al., 1984; Goldstein & Stone, 1977; Poythress & Stock, 1980; Roesch & Golding, 1980). When base rates of findings of competency are considered, however, these high levels of agreement are less impressive, and they do not suggest that evaluators are necessarily in agreement about the criteria for a determination of competency. A psychologist, without even directly assessing a group of defendants, could achieve high levels of agreement with an examining clinician, simply by calling all defendants competent. Since in most jurisdictions, approximately 80 percent of all referred defendants are competent (for reasons discussed later in this chapter), the psychologist and the examiner would have an excellent percentage of agreement. Though the problem of base rates can be corrected through the use of certain statistics (e.g., kappa), the studies reporting reliability usually have small samples overall and consequently few incompetent defendants. It is the more difficult decisions, involving cases where competency is truly a serious question, that are of concern. How reliable are decisions about these cases? To date, no study has accumulated enough of these cases to answer this question.

High levels of reliability do not, of course, ensure that valid decisions are being made. Two evaluators could agree that the presence of psychosis automatically leads to a finding of incompetency. As long as the evaluators are in agreement about their criteria for determining psychosis, the reliability of their final judgments about competency will be high. As we suggest throughout this chapter, it is possible that the criteria used by too many evaluators inappropriately rely on traditional mental status issues without considering the functional aspects of a particular defendant's case. We know of no studies that have addressed the validity issue except, to a limited degree, the research we will report later in this chapter (Golding, et al., 1984).

Validity is, of course, difficult to assess because of the criterion problem. Criterion-related validity is usually assessed by examining concurrent validity and predictive validity (Messick, 1980). Predictive validity is impossible to assess fully because only defendants who are considered competent are allowed to proceed. It is feasible to look at the predictive validity of decisions about competent defendants, but not possible, of course, to assess the decisions about incompetent defendants, since they are referred for treatment, and judicial proceedings are suspended. Concurrent validity is also difficult to determine because it does not make sense to look simply at correlations with other measures (e.g., diagnosis, intelligence) if one adopts a functional, case-by-case assessment of a defendant's competency. For these reasons, then, there is no "correct" decision against which to compare judgments.

As we have indicated, the courts usually accept mental health judgments about competency. Does this mean that the judgments are valid? Not necessarily, because the court is simply accepting the evaluators' definition of competency. We have argued (Roesch & Golding, 1980) that the only direct way of assessing the validity of decisions about incompetency is to allow defendants who are believed to be incompetent to proceed with a trial anyway. This would be a provisional trial in which assessment of a defendant's performance could continue. If a defendant were unable to participate, then the trial could be stopped. If a verdict a verdict had already been reached and the defendant convicted, the verdict could be set aside.

We suspect that in a significant percentage of trials alleged incompetent defendants will be able to participate. In addition to the obvious advantages to both defendants, in terms of briefer periods of confinement, the use of a provisional trial could provide valuable information about what should be expected of a defendant in certain judicial proceedings. Short of a provisional trial, it may be possible to address the validity issue by having independent experts evaluate the information provided by evaluators and other sources. We have used this technique in our research and will discuss this later in the chapter. In the next section, we will review various methods for assessing competency.

We believe the most reasonable approach to the assessment of competency is based on a functional evaluation of a defendant's ability. While an assessment of the mental status of a defendant is important, it is not sufficient as a method of evaluating competency. Rather, the mental status information must be related to the specific demands of the legal case, as has been suggested by legal decisions such as the ones involving amnesia discussed earlier in this chapter. A defendant may be hallucinating and still be found competent to stand trial if those hallucinations do not impair the defendant's ability to consult with his or her attorney and otherwise participate in the legal process.

Some cases are more complex than others and may, as a result, require a higher level of competency. Thus, it may be that the same defendant is competent for one type of legal proceeding but not for others. In certain cases, a defendant may be required to testify. In this instance, a defendant who is likely to withdraw in a catatoniclike state may be incompetent. But the same defendant may be able to proceed if the attorney intends to plea bargain (the way in which the vast majority of all criminal cases are handled).

Florida has passed a statute that requires evaluators to relate a defendant's mental condition to each of 11 legal factors. Each of the following factors must be addressed in the evaluator's report to the court:

1. Defendant's appreciation of the charge

2. Defendant's appreciation of range and nature of possible penalties

3. Defendant's understanding of the adversary nature of the legal process

4. Defendant's capacity to disclose to attorney pertinent facts surrounding the alleged offense

5. Defendant's ability to relate to attorney

6. Defendant's ability to assist attorney in planning defense

7. Defendant's capacity to realistically challenge prosecution witness

8. Defendant's ability to manifest appropriate courtroom behavior

9. Defendant's capacity to testify relevantly

10. Defendant's motivation to help[ himself in the legal process

11. Defendant's capacity to cope with the stress of incarceration prior to trial (Florida Rule of Criminal Procedure 3.2(a)(1) (Winick, 1983, p. 7)

A functional approach was adopted in the case of Wilson v. United States (1968). In that decision, the Court of Appeals held that six factors should be considered in determining whether a defendant's amnesia impaired the ability to stand trial:

1. The extent to which the amnesia affected the defendant's ability to consult with and assist his lawyer

2. The extent to which the amnesia affected the defendant's ability to testify in his own behalf

3. The extent to which the evidence in suit could be extrinsically reconstructed in view of the defendant's amnesia. Such evidence would include evidence relating to the crime itself as well as any reasonably possible alibi

4. The extent to which the Government assisted the defendant and his counsel in the reconstruction

5. The strength of the prosecution's case. Most important here will be whether the Government's case is such as to negate all reasonable hypotheses of innocence. If there is any substantial possibility that the accused could, but for his amnesia, establish an alibi or other defense, it should be presumed that he would have been able to do so.

6. Any other facts and circumstances which would indicate whether or not the defendant had a fair trial (Wilson v. United States, 1968, pp. 463-464).

One could substitute any symptom for amnesia in the receding quote. If this were done, the evaluation of competency would certainly be based on a determination of the manner in which a defendant's incapacity may have an effect on the legal proceedings.

Our view of competency has led us to develop a method for assessing competency which would allow consideration of both the mental status questions and the legal issues. Before we review this measure, we will present an overview of the history of competency assessment methods.

MEASURES OF COMPETENCY

Prior to the 1960s, there were no standard methods for assessing competency. One of the first was a checklist developed by Robey (1965), which focuses on court process issues such as understanding of the legal process. Another early procedure used a checklist and a set of interview questions devised by Bukatman, Foy, and DeGrazia (1971). Neither of these early measures was used often (Schreiber, 1978). By far, the greatest impact on competency assessment came first from the work of A. Louis McGarry and his colleagues at the Harvard Medical School's Laboratory of Community Psychiatry. McGarry, a psychiatrist, was involved in the development of two measures: the Competency Screening Test and the Competency Assessment Instrument.

The Competency Screening Test (CST) was created by Lipsitt, et al. (1971) as a screening measure to identify clearly competent defendants and thus minimize the need for lengthy inpatient evaluations. Such a screening process was considered important because the vast majority of defendants referred for evaluations are competent. The reason is that many other factors influence referrals, including the use of the evaluation commitment as a method for denying bail, as a tactical maneuver to delay a trial, a way of providing a basis for a reduction in charges or sentences, and as a means of getting defendants who are seen to be in need of mental health treatment out of the jails and into the hospitals (Chernoff & Schaffer, 1972; Dickey, 1980; Golten, 1972; Kaufman, 1972; Lewin, 1969; Menzies, Webster, Butler, & Turner, 1980; Roesch & Golding, 1985; Teplin, 1984).

The CST, however, has not often been used as a screening device. Many evaluators have not chosen to use the CST because of the underlying values reflected in its scoring procedures. The CST is a 22-item measure in sentence completion format. Typical items are "Jack felt that the judge _______," or "If the jury finds me guilty _____________." Defendants are asked to fill in the blanks to complete the sentence. Each item is given a score of 2 (competent), 1 (questionable), or 0 (incompetent). The CST is designed so that a low total score (Lipsitt et al. used a cutoff score of 20) would identify possibly incompetent defendants.

The scoring method has been criticized (Brakel, 1974; Roesch & Golding, 1980) because of its idealized perception of the criminal justice system. On one item, "Jack felt that the judge _____," responses like "was right" or "was fair" would receive a score of 2, while responses such as "was unjust," "was too harsh," or "was wrong" would get a score of 0. On another item, "When Bob disagreed with his lawyer on his defense, he ______," a score of 0 would be given to "figured there was no sense arguing." We have suggested (Roesch & Golding, 1980) that such responses may actually reflect a sense of powerlessness in controlling one's outcome in the legal system. This may be based on past experiences with the legal system and may well be an accurate interpretation.

The CST has been examined in a number of studies. In their initial study on the CST, Lipsitt et al. (1971) reported interjudge reliability was .93. Unfortunately, only one other study using the CST has assessed reliability. Randolph, Hicks, and Mason (1981) reported reliability coefficients quite similar to Lipsitt and others in the .90 range. Keeping in mind that these are the only two studies, it would appear that the CST can be scored in a reliable manner. There have been at least four validity studies comparing classification based on CST cutoff scores and hospital evaluation decisions. In general, the CST appears to function reasonably well as a screening device. The errors appear to be in the desired direction--that is, the CST appears to label defendants as incompetent when later evaluations find them competent (false negatives). In the Lipsitt and colleagues study (1971), the false-negative rate was 16 percent (7 of 43 cases) while only three cases were misclassified as competent. Nottingham and Mattson (1981) found that 9 of 50 cases were misclassified as incompetent by the CST, but none was misclassified as competent. In the Randolph et al. study (1981) of the CST, they report an exceedingly high false-positive rate of 47 percent but a false-negative rate of zero. The overall accuracy of the CST, in terms of agreement with a psychiatric evaluation, was 72 percent. Shatin (1979) found that the CST in his study of 21 female defendants misclassified only four cases. Shatin also reported his results on the use of a five-item short form of the CST and found good concordance with the full 22-item CST.

The results of these studies lead one to give a mixed review of the CST. While it appears that the CST is a reliable instrument, serious questions can be raised about its usefulness as a screening device because of the potential for misclassifying possibly incompetent defendants. At this point, it is not possible to recommend that it be used as a sole method of screening defendants.

The other measure developed by McGarry, the Competency Assessment Instrument (CAI), is significantly more useful than the CST. The CAI contains 13 items related to legal issues. The items include "appraisal of available legal defenses," "quality of relating to attorney," and "capacity to disclose pertinent fact." Each item is scored on a 1 to 5 scale, ranging from "total incapacity" to "no incapacity." The CAI manual contains clinical examples of levels of incapacity as well as suggested interview questions.

The CAI has been used in a number of jurisdictions, although perhaps more as an interview structuring device than in the two-stage screening manner (with the CST) as originally intended by McGarry (see Laben et al., 1977; Schreiber, 1978). A revised and extended version has been adopted for use in Canada (Roesch, Webster, & Eaves, 1984). Unfortunately, there are few studies reporting either reliability or validity data. We used the CAI in a North Carolina study (Roesch & Golding, 1980). Thirty interviews conducted by pairs of interviewers yielded item-percent agreement ranging from 68.8 percent to 96.7 percent, with a median of 81.2 percent. The interviewers were in agreement on the competency status of 29 of the 30 defendants (26 competent, 3 incompetent). The interviewers' decisions were in concordance with the more lengthy hospital evaluation decisions in 27 of 30 cases, or 90 percent. Obviously, more studies are needed, but the CAI appears to hold promise as a brief screening device.

The most recent work on competency assessment is our research on the development of the Interdisciplinary Fitness Interview (IFI). The IFI is designed to assess both the legal and psychopathological aspects of competency. The IFI consists of three major sections: (1) legal issues (5 items); (2) psychopathological issues (11 items); and (3) overall evaluation (4 items). The three items in the counsensual judgment section reflect post-assessment resolution of differences between judges. Table 14.1 contains a list of the items.

Each of the general items represents an organizing scheme for more specific subareas that have been seen to influence competency decisions. For example, six subareas are subsumed under the broad "capacity to appreciate" which forms the core of item 1. These are (1) appreciating the nature of the state's criminal allegation; (2) ability to provide a reasonable account of one's behavior prior to, during, and subsequent to the alleged crime; (3) ability to provide an account of relevant others during the same time period; (4) ability to provide relevant information about one's own state of mind at the time of the alleged crime, including intentions, feelings, and cognitions; (5) ability to provide information about the behavior of the police during apprehension arrest, and interrogation; and (6) projected ability to provide feedback to an attorney about the veracity of witness testimony during trial, if a trial is likely to be involved. Note, however, in line with the open-textured nature of the competency construct, that a complete enumeration is not possible; rather, an attempt is made to summarize the general "lay of the land," allowing for specifics to be a matter of personal judgment.

The IFI was designed so that evaluators would have to consider both legal and mental status issues, but neither in isolation. The format of the IFI requires evaluators to relate their observations to the specific demands of the legal situations. For each item, evaluators are asked to rate the degree of incapacity of the defendant, as well as to give the item a score to indicate the influence that the incapacity might have on the overall decision about competency. Thus, a defendant may receive a score indicating the presence of hallucinations (item 10) but receive a low weight score because the evaluator has determined that the presence of hallucinations would not have much effect on the conduct of the legal case. Another defendant with the same symptom may receive a high weight score because the hallucinations are considered to be more of a potential problem during the legal proceedings.

A lengthy training manual for use of the IFI has been developed as a guide for evaluators. For each item, the manual provides a set of suggested questions and follow-up probes and also gives clinical guidance for the handling of typical interview problems.

The IFI can be used by a single evaluator, and we expect that this would be the manner in which it is most frequently employed. However, in our research on the IFI, we have used two evaluators so that reliability could be assessed. Moreover, we were interested in what the IFI, as a research instrument, could tell us about how competency is being defined. Because we believe that competency is as much a legal construct as it is a mental health one, we were also interested in determining if there were any differences in the way in which legal and mental health professionals would reach decisions about competency. Thus, we departed from the tradition of having only mental health professionals assess competency by inviting lawyers to participate also. In our research, a lawyer and a mental health professional have worked together as co-interviewers of defendants whose competency had been questioned. We believe that this collaboration is an important method that might result in a broader, and it is hoped, more valid assessment of competency. The presence of lawyers forces a different consideration of the legal issues specific to a case, and it also provides a method of training mental health professionals in the law. One of the problems in the past has been that mental health evaluators lacked knowledge about the legal system. While the creation of graduate training programs in law and psychology as well as joint Ph.D./J.D. degree programs may change this situation in the future (see Melton, Chapter 26 in this volume; Roesch, Grisso, & Poythress, 1985), collaborating on forensic assessment could serve as an excellent means of educating existing evaluators about legal issues.

Our first study of the IFI has been published (Golding, et al., 1984), and the results will be summarized here. The subjects were pretrial defendants in the Boston area who were referred by court clinics to a state mental hospital for competency evaluation. They were interviewed by teams composed of a lawyer and either a psychologist or a social worker. The interviews lasted approximately 45 minutes. While the interviews were conducted jointly, each evaluator independently completed the rating form of the IFI. The results demonstrated that judgments about competency can be made in a reliable manner by lawyers and mental health evaluators. They were in agreement on 97 percent of their final determinations of competency. By type of decision, the interviewers found 58 defendants to be competent, 17 incompetent, and disagreed on the remaining 2 cases. While overall agreement was excellent, there were some differences between the professions at the item level.

Interjudge reliability (kappa) ranged from .40 to .91. The interviewers were less in agreement on the legal items (mean kappa = .48) compared to the psychopathology items (.67). This finding reinforces our view that the assessment of competency will be improved as lawyers and mental health professionals come to understand the reasons for their disagreement. These were also significant differences in the strategies used by each profession to arrive at a decision. Principal component analyses (see Golding et al., 1984) suggested that mental health professionals have a more integrated view of competency because the legal and psychopathology items overlap to a greater degree. For example, the first component for mental health professionals was a combination of legal items (appreciate changes and options) and psychopathology items that might be viewed as clinical correlates of the legal items (insight, absence of thought disturbance, and psychotic belief systems).

Since the defendants in our study were also independently assessed by evaluators at the mental hospital, it was of interest to compare our interviewers' decisions with those of the hospital. The overall agreement was only 76 percent, and there was greater disagreement with respect to decisions that a defendant was incompetent. While the two groups agreed 83 percent of the time on competent decisions, the agreement for incompetency was only 58 percent. It is difficult to interpret these differences because there is no absolute criterion against which to compare them. It is impossible to determine which decision was right and which was wrong. It is conceivable, or course, that both decisions were right because the defendants were seen at different times. A defendant who was incompetent at the first interview could have been competent by the time the second one took place, especially if medication was prescribed.

This discussion illustrates one of the many research problems inherent in studies of competency assessment. Since most defendants are competent (77 percent in our study), it is difficult to obtain a sufficiently large sample of incompetent defendants. Based on our studies in North Carolina and Boston, it has become clear to use that decisions about most defendants referred for competency evaluations are straight-forward--that is, they are competent to stand trial, a finding which is evident no matter what the method of assessment. The potential value of the IFI, we believe, is in assessing defendants whose competency is truly questionable. It is here that we also believe that the collaboration of lawyers and mental health professionals will be of greatest value.

We have proposed elsewhere (Roesch & Golding, 1980) that competency can most effectively be evaluated through the use of a screening evaluation that could be conducted on an outpatient basis. The IFI should be a useful guide for these initial evaluations and also for later evaluations by others if competency cannot be determined at the screening evaluation. The initial research on the IFI appears promising. As more evaluators use it, we could begin to assess how effective it is in evaluating incompetency. It would be of particular interest to see how the IFI operates with the more difficult "gray-area_ cases in which clear decision about competency is difficult. Furthermore, data on specific decisions will be useful to determine the extent to which the particular demands of a case influence these decisions.

GUIDELINES FOR EVALUATORS

We conclude our chapter with a discussion of several issues to which an examiner must pay special attention when conducting an evaluation of competency. Even before seeing a defendant face to face, it is good clinical practice to speak with both the defense and prosecuting attorneys in order to determine as accurately as possible why the fitness issue was raised, what evidence was offered, and what sort of trial and dispositional alternatives are being considered by both sides. This type of information is most useful if the IFI is used in an interdisciplinary fashion but is also quite helpful to a mental health examiner alone.

All indications of prior mental health contacts should be pursued before the interview takes place, so that the examiner has as complete a set of mental health records as possible. Similarly, complete police reports of the alleged crime are necessary and a past criminal history record helpful, particularly if the defendant has cycled through the criminal justice and mental health systems several times. Obviously, if the defendant is an inpatient, observational records should be consulted along with all routine psychological test data. Finally, the examiner should maintain an accurate record of all contacts with the defendant, attorneys, and other mental health professionals. These records are invaluable at later stages if legal tactics designed to confuse or mislead a witness are attempted.

Having prepared for an examination in this fashion, one can conduct an efficient and comprehensive interview in a short period of time. Most delays in conducting an evaluation and most time spent in an inpatient status can thus be avoided, and a more relevant examination conducted, if these steps are taken. Prior to the interview, the defendant should be fully informed about any limitations on the interview's confidentiality. The possibility of recording the interview should be discussed, although permission should also be obtained from the defendant's attorney.

The examiner should be aware of any aspects of the interview and the resulting report that are covered by statute or accepted practice within the jurisdiction. As an example of the former, some states require Miranda-like warning that inform the defendant of the limitations of confidentiality that may apply. Similarly, other states dictate the form of the report to the court, and an examiner's report may be excluded if it does not comply with the required format.

In People v. Harris, for example, a psychiatrist's report (that the defendant was competent) was excluded, and the defendant's subsequent conviction was reversed because the opinion was presented in conclusory terms and failed to give the clinical facts and reasons upon which it was based, thus precluding the trier of fact from independently assessing the weight to be given such an opinion. The current competency statute in Illinois is in many ways a model of this developing trend. It requires the examiner to address the facts upon which the conclusion is based, to explain how the conclusion was reached, to describe the defendant's mental and physical disabilities and how these impair the ability to understand the proceedings and assist in the defense, to discuss the likelihood that the defendant will respond to a specified course of treatment, and to explain procedures that would be employed to compensate for the defendant's disabilities, if any. We applaud this sort of specification and urge examiners to adopt the practice, even if it is not mandated in their own jurisdiction.

The conduct of a competency evaluation and the reports prepared for court should therefore be in complete accord with both the spirit and the letter of contemporary legal standards. The examiner must be thoroughly acquainted with the legal literature and in some sense anticipate developments in one's practice. For example, Estelle v. Smith (1981) clearly prohibits the introduction of material obtained under court-ordered competency proceedings at a "critical" (guilt or sentencing) stage of trial. Many states mirror this in their statutes but nevertheless do not regulate the common practice of requesting competency and sanity evaluations at the same time, often resulting in a combined report. We believe this practice is unfortunate and recommend that separate interviews, with distinct reports, be prepared. While a trier of fact is required to separate these issues, it is cognitively almost impossible to do so when the reports are combined. A defendant who is clearly psychotic and "legally insane" at the time of an assault may respond rapidly to treatment upon arrest and be just as nonpsychotic and "legally fit" when actually examined. Caution and fairness dictate keeping reports separate so that the two issues can be considered independently by the courts.

TABLE 1 Interdisciplinary Fitness Interview Items

_______________________________________________________________

Section A: Legal Items

1 Capacity to appreciate the nature of the alleged crime, and to disclose pertinent facts, events, and motives

2 Quality of relationship with one's current attorney

3 Quality of relationship with attorneys in general

4 Anticipated courtroom demeanor and trial conduct

5 Appreciating the consequences of various legal options

Section B: Psychopathological Items

6 Primary disturbance of thought

7 Primary disturbance of communication

8 Secondary disturbance of communication

9 Delusional processes

10 Hallucinations

11 Unmanageable or disturbing behavior

12 Affective disturbances

13 Disturbances of consciousness/orientation

14 Disturbances of memory/amnesia

15 Severe mental retardation

16 General impairment of judgment/insight

Section C: Overall Evaluation

1 Overall fitness judgment

2 Rating of confidence in judgment

3 Comment on basis for decision about defendant

4 Other factors rater might wish to take into account in reaching decision

Section D: Consensual Judgment

1 Fitness judgment after conferring with partner

2 Changes in rating of individual items after conferring

3 Reasons for changes