Clinical Interview Research Paper

1. Advantages Of A Clinical Interview

Many of the advantages of a clinical interview are somewhat obvious, but worth noting for the record nevertheless (Groth-Marnat 1997). First, clinical interviews are substantially more flexible than alternative methods. Interviewers can alter the focus, depth, or even the style of an interview to be optimally responsive to the particular demands, interests, or needs of the respondent or the assessment. Response sets (intentional, habitual, or unconscious tendencies to provide false or misleading responses) can affect the validity of a clinical interview (Rogers 1995), but interviewers can themselves be sensitive and responsive to symptom exaggeration, distortion, or denial during the course of an interview. Interviewers may notice if a respondent is being excessively acquiescent or defensive, if the mood state of a respondent is contributing to excessive self-denigration, or if the responses are inconsistent across the interview. The interviewer can then alter immediately the format, style, or scoring of the interview to make adjustments for the response sets, can conduct follow-up queries to assess for the presence of problematic response sets, or even discuss a response set directly with the respondent in order to decrease its effects.

A disadvantage of other methods of clinical assessment is that they will routinely cover domains of functioning that will not be particularly relevant or necessary for an issue or patient at hand and yet, at the same time, fail to cover in adequate depth the domain of functioning of most interest or relevance to the patient and clinician. For example, most omnibus self-report inventories attempt to cover virtually all domains of psychopathology but must then provide an inadequate assessment for any one of them. An interviewer has the unique advantage of being able virtually to abandon a focus of inquiry during the course of an assessment to spend more time and effort on a particular line of investigation.

Finally, the clinician conducting the interview is usually the person who will ultimately provide the clinical report, and seeing or hearing for oneself is usually much more compelling than being told by something or someone else. The clinician is able to see and experience firsthand the person’s behaviors, feelings, statements, and manner of relatedness. The presentation of the patient’s psychopathology in his or her speech, affect, and behavior within the clinician’s office can provide a powerfully vivid and compelling portrayal.

2. Limitations Of Unstructured Clinical Interviews

Many of the benefits and advantages of a clinical interview, however, fail to be realized in routine clinical practice, as the freedom and authority pro-vided by a clinical interview does not come without substantial responsibilities, costs, and limitations. The reliability and validity of a clinical interview depend substantially upon the conscientiousness, skills, and talents of the clinician. Seeing for oneself can be very compelling, but it can be equally illusory and deceiving. Many clinicians may place excessive faith, or at least have excessive confidence, in their own perceptions and judgments, despite the fact that studies have shown repeatedly that unstructured clinical assessments often obtain poor agreement across different interviewers (Dawes 1995, Garb 1998). Two clinicians relying upon their own skills, talents, and abilities will often provide different conclusions regarding the same patient. At least one of them will be wrong, but both will believe it is the other clinician. The instrument of the unstructured clinical interview is for the most part the clinician, and there are perhaps few clinicians who truly recognize or adequately appreciate their own limitations, deficits, and flaws.

Many studies have documented that unstructured clinical interviews tend to be unreliable and are highly susceptible to primacy effects, halo effects, false expectations, misleading assumptions, and confirmatory biases (Dawes 1994, Garb 1998, Widiger and Saylor 1998), and this research appears to have had only a marginal effect upon the beliefs or behavior of most individual practitioners (e.g., Westen 1997). The lack of an adequate impact of this research is perhaps due in part to the ability of persons to believe that the research is primarily relevant to persons other than themselves. An advantage of other methods of assessment is that the research indicating, for example, systematic errors within a laboratory instrument, will clearly be relevant to almost any administration of that instrument. Research indicating the systematic errors of a sample of clinicians might not be applicable to a clinician who did not actually participate in that particular study. Clinicians can then argue and believe that the decision-making research is not really applicable to them because they are in fact adequately sensitive and responsive to the issues, errors, biases, or concerns identified in this research.

A major source for the failure of unstructured clinical interviews to provide reliable or valid assessments is the failure to conduct systematic or comprehensive assessments. An innovation of the American Psychiatric Association’s (APA) Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; APA 1994) is the provision of relatively specific and explicit diagnostic criteria for each mental disorder. Reliable and valid clinical diagnoses are now readily obtained as long as the interviewer does indeed comprehensively assess every diagnostic criterion in a systematic manner (Nathan and Langenbucher 1999). Many studies, however, have indicated that clinicians will often reach a conclusion after determining the presence of only a small subset of the diagnostic criteria set, and will fail to assess for the presence of additional symptomatology of other possible disorders (Garb 1998, Widiger and Saylor 1998). Zimmerman and Mattia (1999), for example, compared the clinical diagnoses provided for 500 patients who were assessed with unstructured clinical inter-views with the diagnoses provided by a semistructured interview that systematically assessed for the presence of the diagnostic criteria for most of the commonly occurring Axis I mental disorders (i.e., mental disorders other than personality disorders or mental retardation). More than 90 percent of the patients receiving the unstructured clinical interview were provided with only one diagnosis, whereas more than a third of the patients assessed with the semi-structured interview were discovered to have met the diagnostic criteria for at least three different mental disorders. Comorbidity among mental disorders has substantial significance and importance to clinical treatment, yet it appears to be grossly under-recognized in general clinical practice (Zimmerman and Mattia 1999).

A variety of studies have also indicated that clinicians relying upon unstructured clinical interviews routinely fail to assess for the presence of the specified diagnostic criteria (Widiger and Sanderson 1995). One of the more compelling demonstrations of this failure was provided by Morey and Ochua (1989). Morey and Ochua provided 291 clinicians with the 166 DSM-III (APA 1994) personality disorder diagnostic criteria and asked them to indicate which DSM-III personality disorder(s) were present in one of their patients and to indicate which of the 166 diagnostic criteria were present. Kappa for the agreement between their diagnoses and the diagnoses that would be given based upon the diagnostic criteria they indicated to be present was very poor, ranging from 0.11 (schizoid) to only 0.58 (borderline). In other words, their clinical diagnoses agreed poorly with their own assessments of the diagnostic criteria for each of the personality disorders. Comparable results have since been re-ported in many subsequent studies (Widiger and Saylor 1998).

Among the more consistently documented errors in clinical practice are gender and racial biases in the application of diagnostic criteria (Garb 1998). DSM-IV diagnostic criteria sets will contain a degree of gender and racial bias (Hartung and Widiger 1998). However, racial and gender biases that have been documented empirically have been due in large part to a failure of clinicians to adhere to the specified diagnostic criteria set for a respective mental disorder (Whaley 1997, Widiger 1998). When clinicians are compelled to follow closely the criteria set for a mood, psychotic, or personality disorder, gender and racial biases are significantly less likely to occur.

3. Advantages Of Semistructured Clinical Interviews

Limitations of unstructured clinical interviews can be addressed in part through the administration of more structured interview schedules in which a set of specified questions must be administered, the interpretation and scoring of the responses to which are guided by an accompanying manual. These interview schedules can vary substantially in the extent to which the questions are open-ended, observations of the respondent are included, and the interviewer is allowed to conduct follow-up queries. A fully structured interview would be essentially equivalent to a verbally administered self-report inventory (Widiger and Saylor 1998). Most interview schedules, however, are more accurately described as being semistructured, as they will include subtle and indirect questioning, will require follow-up queries, and will include open-ended questions, the responses to which will require professional judgment and expertise for interpretation and scoring. The only difference between some semistructured interviews and a skilled clinician is that the inclusion of a semistructured interview documents explicitly the obtainment of a reliable, replicable, systematic, objective, and comprehensive assessment of all of the relevant symptomatology.

Semistructured interviews are the preferred method for obtaining clinical assessments in research, but are perhaps rarely used in general clinical practice. Clinicians perceive semistructured interviews as being constraining, impractical, or superficial (Westen 1997). Semistructured interviews are indeed constraining, as they are a means by which to ensure that the findings are reliable, replicable, systematic, comprehensive, and objective by constraining the clinician from failing to assess all of the necessary criteria in a minimally adequate manner (Segal 1997). Most semi-structured interviews, however, allow and do in fact encourage clinicians to have a significant impact through the administration of follow-up queries and the reliance upon their professional judgment for the scoring of the responses.

One of the major impediments to the implementation of a semistructured interview in general clinical practice is the amount of time that is required for their complete administration. Researchers will often pay both interviewers and patients for two or three hours of interviewing; no such funding luxury is available in general clinical practice. However, the amount of time required for the administration of a semistructured interview can be reduced substantially by first ad-ministering a screening questionnaire to narrow the line of inquiry. Screening questionnaires with a high false positive rate (i.e., err in the direction of identifying too much rather than too little psychopathology) are also useful in alerting the clinician to domains of functioning that might have been otherwise neglected.

Many clinicians will also perceive some of the required questioning to be simplistic or superficial. However, it is important to appreciate that a substantial amount of research has informed the development of a particular line of questioning. Semi-structured interviews can in fact be an excellent source for discovering new and effective methods of inquiry. Semistructured interviews, however, will not be as effective as an unstructured interview in establishing rapport. Most clients will appreciate the comprehensive and thorough nature of a semistructured interview, but if the establishment of rapport is a major clinical issue, then a lengthy semistructured interview will at times be problematic.

4. Recommendations For Future Research

A clinical and scientific limitation of many semi-structured interviews is the absence of data normally obtained through the course of the development and validation of a psychometric instrument. For example, semistructured interview reliability data are often simply confined to an agreement with respect to the scoring of a previous or concurrent administration of the interview. The poor reliability obtained in general clinical practice is due to inconsistent, incomplete, or idiosyncratic interviewing. It is unclear if some of the semistructured interviews have actually resolved this problem given the absence of studies on the interrater (or test–retest) reliability of independent administrations of the interview (Rogers 1995, Segal 1997, Widiger and Saylor 1998).

There are a variety of different interview schedules to assess the same domains of psychopathology. An advantage of this diversity is the availability of different options to choose from. However, current research suggests that these different interview schedules are providing different findings and it is unclear if the failure to replicate findings across studies is due to idiosyncratic administration of interviews, differences in setting, or differences in the interview schedules. One suggestion has been to confine future research to just one interview schedule (Regier et al. 1998). This confinement would contribute to the obtainment of more uniform results, but at the cost of the failure to appreciate the extent to which the results in fact reflect unique aspects of a particular interview schedule. What is needed are studies comparing directly the concurrent and predictive validity of alternative interview schedules within the same patient sample.

Normative data are also lacking for many of the semistructured interviews. The test manuals that accompany the publication of a semistructured interview are often surprisingly weak in their coverage of reliability and validity data. Diagnoses obtained through the administration of a semistructured interview are used as the criterion by which the validity of other instruments is evaluated, but semistructured interview schedules may rely too heavily for their own derivation on simply face validity. In defense of the validity of semistructured interviews, the most compelling published research concerning the etiology, course, pathology, and treatment responsivity of various mental disorders has relied substantially on the administration of a semistructured interview. The results of this extensive research provide considerable support for the construct validity of the respective semistructured interviews that were used in these studies. In addition, detailed summaries of the re-liability and validity of alternative interview schedules are provided in a number of published papers and texts (e.g., Rogers 1995, Segal 1997, Widiger and Sanderson 1995).

5. Conclusions

In sum, the many advantages of semistructured interviews clearly outweigh their limitations and dis-advantages. Many are now being used in general clinical practice when the results of the clinical assessment might be subsequently questioned or re-viewed (e.g., custody, disability, and forensic assessments). A highly talented clinician can provide a more valid assessment than a semistructured interview, but it is risky to assume that one is indeed that talented clinician. It would at least seem desirable for a talented and insightful clinician to be fully informed by a systematic and comprehensive assessment. Semi-structured interviews are used routinely in general clinical research and perhaps will eventually be used routinely in general clinical practice. Individually administered intelligence tests are comparable to a fully structured clinical interview, particularly an assessment of verbal intelligence that involves a series of specified questions, the responses to which are scored according to a test manual. Very few clinicians would attempt to diagnose mental retardation in the absence of the administration of one of these structured interviews. Perhaps in the future no clinician will attempt to diagnose an anxiety, mood, psychotic, dissociative, personality, or other mental disorder without at least considering the results obtained by the administration of a respective semi-structured interview.

Bibliography:

1. American Psychiatric Association 1994 Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association, Washington DC
2. Dawes R M 1994 House of Cards: Psychology and Psychotherapy Built on Myth. Free Press, New York
3. Garb H N 1998 Studying the Clinician. Judgment Research and Psychological Assessment. American Psychological Association, Washington, DC
4. Groth-Marnat G 1997 Handbook of Psychological Assessment, 3rd edn. Wiley, New York
5. Hartung C M, Widiger T A 1998 Gender differences in the diagnosis of mental disorders: Conclusions and controversies of DSM-IV. Psychological Bulletin 123: 260–78
6. Morey L C, Ochua E S 1989 An investigation of adherence to diagnostic criteria: Clinical diagnosis of the DSM-III personality disorders. Journal of Personality Disorders 3: 180–92
7. Nathan P, Langenbucher J W 1999 Psychopathology: Description and classification. Annual Review of Psychology 50: 79–107
8. Regier D A, Kaelber C T, Rae D S, Farmer M E, Knauper B, Kessler R C, Norquist G S 1998 Limitations of diagnostic criteria and assessment instruments for mental disorders. Implications for research and policy. Archives of General Psychiatry 55: 109–15
9. Rogers R 1995 Diagnostic and Structured Interviewing. A Handbook for Psychologists. Psychological Assessment Resources, Odessa, FL
10. Segal D L 1997 Structured interviewing and DSM classification. In: Turner S M, Hersen M (eds.) Adult Psychopathology and Diagnosis. Wiley, New York, pp. 24–57
11. Westen D 1997 Divergences between clinical and research methods for assessing personality disorders: Implications for research and the evolution of Axis II. American Journal of Psychiatry 154: 895–903
12. Whaley A L 1997 Ethnicity race, paranoia, and psychiatric diagnoses: Clinician bias versus sociocultural diff Journal of Psychopathology and Behavioral Assessment 19: 1–20
13. Widiger T A 1998 Sex biases in the diagnosis of personality disorders. Journal of Personality Disorders 12: 95–118
14. Widiger T A, Sanderson C J 1995 Assessing personality disorders. In: Butcher J N (ed.) Clinical Personality Assessment. Practical Approaches. Oxford University Press, New York, pp. 380–94
15. Widiger T A, Saylor K I 1998 Personality assessment. In: Bellack A S, Hersen M (eds.) Comprehensive Clinical Psychology. Pergamon, New York, pp. 145–67
16. Zimmerman M, Mattia J I 1999 Psychiatric diagnosis in clinical practice: Is comorbidity being missed? Comprehensive Psychiatry 40: 182–91