Insight and Personal Narratives of Illness in Schizophrenia

Psychiatry 65(3) Fall 2002 197

PAUL H. LYSAKER,CATHERINE A. CLEMENTS,CYNTHIA D. PLASCAK-HALLBERG, STACY J. KNIPSCHEER, AND DUSTIN E. WRIGHT

Insight in schizophrenia tends to be assessed as the degree to which one possesses specific knowledge. It therefore often fails to account for the fact that awareness of illness is an inextricable part of a personal narrative and may be incoherent or incomplete for many different narrative reasons. Accordingly, we have developed a means of eliciting narratives of illness: the Indiana Psychiatric Illness Interview, and a method for rating the coherence of those narratives: the Narrative Coherence Rating Scale. In this article we describe these methods and present data on their reliability and validity in a study of the illness narratives of 33 outpatients with schizophrenia or schizoaffective disorder. Results suggest our measures possess sufficient internal consistency and good to excellent interrater reliability. Addition- ally, as predicted, our measures of narrative coherence were significantly correlated with traditional measures of insight and with measures of cognitive impairment and hopelessness gathered earlier.

Relative to persons with other psychiat- to as “poor insight,” are commonly seen as ric disorders, persons with schizophrenia spec- clinically significant because of their links with trum disorders are often unaware of what poorer treatment compliance (Bartko, Herc- others readily perceive as their symptoms, psy- zeg, and Zador 1988; Cuffel, Alford, Fischer, chosocial deficits and/or need for rehabilita- and Owen 1996; Smith et al. 1999), clinical tion or treatment (Amador, Strauss, Yale, and outcome (Schwartz 1998), social function (Ly- Gorman 1991; Amador 1995; David 1990). saker, Bell, Bryson, and Kaplan 1998a), and These deficits in awareness, generally referred response to vocational rehabilitation (Lysaker, Bell, Milstein, Bryson, & Goulet 1994, Ly- saker, Bryson, & Bell 2002). Additionally, Paul H. Lysaker, PhD, is a clinical psycholo- some degree of unawareness of one’s own gist at the Roudebush VA Medical Center and the mental state is also thought to be a significant Indiana University School of Medicine. Catherine barrier to the development of a working rela- A. Clements, MS, is a research assistant, Cynthia D. Plascak-Hallberg, MS, is a research assistant, and tionship with mental health professionals. Stacy J. Knipscheer, MSW is a social worker at the To date, research on the nature of poor Roudebush VA Medical Center. Dustin E. Wright, insight in schizophrenia has tended to evolve MS, is a research assistant at Indiana University— out of one of two theoretical perspectives. Purdue University, Indianapolis. First, one arm of the literature has emphasized Research for this article was funded by the Institute for Psychiatric Research, Indiana Univer- how unawareness of illness may result from sity School of Medicine, Indianapolis, Indiana, cognitive impairments. Various authors have USA. suggested that, paralleling observations about Address correspondence to Paul Lysaker, unawareness of deficits in neurological disor- PhD, Day Hospital 116H, 1481 West 10th St., ders, persons with schizophrenia may fail to Roudebush VA Medical Center, Indianapolis, IN 46202; E-mail: Lysaker.Paul_H_PHD@Indian grasp the breadth or presence of their illness apolis.va.gov because of generalized impairments in the 198 INSIGHT AND PERSONAL NARRATIVES OF ILLNESS IN SCHIZOPHRENIA ability to think in an abstract and flexible man- treatment y”) or if he or she endorses certain ner (Amador, Strauss, Yale, and Gorman concrete beliefs (e.g., “I accept that I have z,” 1991). Supporting research includes findings or “I am different from others because I suffer that poor insight concurrently and prospec- from q”). For example, the “Self-appraisal of tively predicts poorer performance on neuro- Illness Questionnaire,” a questionnaire used psychological tests of executive function to assess insight via self-report, measures in- (Lysaker & Bell 1994; 1998b; Marks, Fas- sight by asking participants to rate the extent tenau, Lysaker, and Bond 2000; Mohamed, to which they agree or disagree with state- Fleming, Penn, and Spaulding 1999; Young, ments which endorse views that they are ill Davila, and Scher 1993; Young, Zakzanis, and (Marks, Fastenau, Lysaker, and Bond 2000). Baily 1998). Conversely, impairments in exec- Similarly, the Scale to Assess Unawareness of utive function have been found to predict the Mental Disorders (SUMD; Amador et al. intransigence of deficits in insight (Lysaker 1995), a widely used interviewer rating scale and Bell 1995). for insight, calls for an interviewer to deter- From a second perspective, however, it mine what symptoms are present and then has been suggested that poor insight may re- decide whether the participants’ verbaliza- flect an elective denial of painful realities tions contain a sufficient understanding of the (Frese 1993). It has been proposed that un- phenomena and their consequences. awareness of illness may be an adaptive way The problem with these approaches is of avoiding a role that is stigmatizing (Corri- that awareness of any type of illness is not gan and Penn 1999; Wahl and Harman 1989; merely an isolated cognition, but is also an Warner, Taylor, Powers, and Hyman 1989). element of a larger personal and narrative un- Evidence supporting these views includes derstanding of one’s life (Kleinman 1988). findings that acceptance of the label “mentally Awareness or denial of illness is a story embed- ill” predicts more recalcitrant psychosocial ded in a larger life story which is, by necessity, deficits (Thompson 1988) and greater despair inextricably linked to a variety of past suc- (Schwartz 2001). Other studies have found cesses and failures as well as to future dreams that gains in insight often have a sobering and expectations (Davidson and Strauss 1995; effect on persons with schizophrenia, leading Kirmayer and Corin 1998; Lysaker and France to increases in levels of despair and/or decre- 1999; Williams and Collins 1999). ments in mood elevation (Carroll et al. 1999). But does it really make a difference if Thus, perhaps unawareness of illness may one sees insight as a singular cognition as serve a protective function in the same way opposed to an element of narrative? By con- that some positive symptoms are thought to ceptualizing insight as an isolated piece of shield ill persons from low self-esteem (Black- knowledge and not as a story bound to a life’s wood, Howard, and Bentel 2001). Lastly, a trajectory, we reason that it is possible to ne- recent study has suggested that unawareness glect the fact that an understanding of illness of symptoms may be linked to a greater prefer- may be incomplete for any combination of ence for positive reappraisal, which is a coping different reasons. As an illustration, consider strategy in which stressors are recast as posi- person A., who presents a minimalist story tive events (Lysaker, Bryson, Lancaster, Ev- of his/her illness. Perhaps (s)he says, “I have ans, & Bell in press). nerves” and cannot give any details and is not While the work of both of these per- sure of what (s)he should be doing to “calm spectives has been instructive, we would sug- his/her nerves.” Certainly A. might be said to gest that both views have tended to share a have limited insight. For one, his/her story of similar methodological and conceptual limita- illness lacks sufficient details about symptoms, tion. In particular, both arms have tended to treatment needs and the consequences of the view a person as “having” insight if he or she disorder. Alternatively, consider person B., can demonstrate acceptance of specific facts whose story of schizophrenia contains many (e.g., “I know I have symptom x,” or “need historical details of illness, but does not logi- LYSAKER ET AL. 199 cally connect those details. (S)he might also ered researcher ratings of cognitive symptoms be said to have limited insight because, with of illness using the Positive and Negative Syn- a lack of logical connections, no clear picture drome Scale (Kay, Fizszbein, and Opler 1987), of what is wrong or should be done emerges. and self-reports of despair or hopelessness. Lastly, consider person C., whose story per- We predicted that in general, lesser ratings of haps contains a sufficient number of logically coherence would be related to greater levels connected details but lacks plausibility overall. of cognitive symptoms and greater reports of For instance, maybe (s)he says that (s)he hears hopelessness. Toinsure that the NCRS scores voices because “the reincarnation process has were not merely a reflection of verbal ability gone awry.” Thus, (s)he might be said to have or gross symptomatology, we lastly correlated poor insight. However, in this instance it is those scores with the Vocabulary subtest of because his/her story strikes others as implau- the WAIS-III (Wechsler 1997), a measure of sible and is unlikely to receive consensual vali- verbal intelligence, and the Positive and Neg- dation from his/her community. Our point is ative Component scores of the PANSS. that traditional ratings of insight gathered from self-report or standardized interviews for persons A., B. and C. might give identical METHODS insight scores despite the fact that each person’s limited understanding of his/her illness Participants arises for very different narrative reasons. Thus, we propose that current assess- Thirty-three males with DSM-IV diag- ments of insight may be limited because they noses of schizophrenia (n = 23) or schizoaffec- fail to account for a) that lack of insight repre- tive disorder (n = 10) were recruited from an sents an incomplete or alternative understand- outpatient psychiatry clinic at a Midwestern ing of illness as contained within a story of VA Medical Center. The mean age was 47.2 that illness, and b) how a story of illness may (SD = 9.01) and mean education was 13.8 (SD = be incomplete or not understandable to others 4.29) years. Participants had a mean of 7.8 for any combination of reasons. Accordingly, (SD = 7.90) lifetime hospitalizations with the we have set out to develop both a means of first occurring on average at age 27 (SD = eliciting narratives of illness and a reliable and 6.34). Twenty-four participants were White, valid way of measuring how understandable eight were African American and one was or acceptable those stories are to others. The Latino. All participants were in a post-acute aims of this article are, specifically, twofold. phase of illness as defined by having no hos- We first present the Indiana Psychiatric Illness pitalizations or changes in medication or Interview (IPII), our method for eliciting housing in the month prior to entering the rating narratives of illness; and the Narrative study. Excluded from the study were partici- Coherence Rating Scale (NCRS), our method pants with a history of mental retardation or for rating the coherence of those narratives. active substance abuse. Second, we present data from a study of the interrater reliability, internal consistency, con- Instruments current and predictive validity of the NCRS. To assess the concurrent validity of the Indiana Psychiatric Illness Interview (IPII) NCRS we correlated it with previously is a semi-structured interview that we have gathered researcher ratings of insight using developed to assess illness narrative. It is an the Scale to Assess Unawareness of Mental individual interview conducted by a research Disorders (Amador et al. 1995). As a test of the assistant that generally lasts between 30 and predictive validity of the NCRS, we examined 90 minutes that can either be typed verbatim whether it was correlated with other variables during the course of the interview or taped often found to be related to traditional assess- and later transcribed. The interview is divided ments of insight, including previously gath- conceptually into four sections. In the first 200 INSIGHT AND PERSONAL NARRATIVES OF ILLNESS IN SCHIZOPHRENIA section, rapport is established and the partici- tone of the interview is intended to be conver- pant is asked “What kind of person do you sational rather than interrogatory or judg- see yourself as? Is there a story about your mental. The interviewer’s task is to elicit life? Are there other sides to you?” In the enough information to understand the story second section the participant is asked “Do the participant is telling about mental illness, you think you have a mental illness and if so not to confirm or disagree with that story. what do you think it is?” If the participant Narrative Coherence Rating Scale (NCRS) suggests through her or his answers that she is a six item, 18-point rating scale created to or he in any manner thinks she or he has a assess narrative coherence of illness narratives. mental illness, the interviewer proceeds to ask It is completed by a trained rater following a four questions about that illness. In this sec- review of an IPII transcript. The six items are tion it is crucial for the interviewer to proceed as follows: regardless of whether the participant’s understanding of illness concurs with the opinion 1. “Logical connections” among the of the interviewer. If an illness is acknowl- narrative of past psychiatric illnesses; edged, the participant is asked to “say more 2. “Logical connections” among the about your experience of mental illness in the narrative of psychiatric illnesses in the past, about what caused these problems, how present; you feel about having this mental illness and 3. “Richness of historical detail” among about what is going to happen in the future?” the narrative of past psychiatric ill- Next the participant is asked “Since you be- nesses; came mentally ill (or condition as identified 4. “Richness of historical detail” among by the participant), what about you or your life the narrative of psychiatric illnesses has changed and what has stayed the same?” If in the present; the participant does not mention vocational, 5. “Plausibility” among the narrative of community, family, and/or cognitive emo- psychiatric illnesses in the past; tional function, these are queried about, both 6. “Plausibility” among the narrative of in terms of what has changed and what has psychiatric illnesses in the present. stayed the same. In the third section the participant is asked “To what extent and in what These six scores are used to generate ways does your mental illness (or condition three general scores: “logical connections,” as identified by the participant) control your “richness of historical detail” and “plausibil- life?” Lastly the participant is asked “what is ity,” which are merely the sums of the past expected to be different and what will be the and present ratings for each category. Anchors same in the future?” Again if the participant for each subscale are presented in Table 1. does not mention vocational, community, Vocabulary subtest of the WAIS-III family, and/or cognitive emotional function, (Wechsler 1997) assesses participants’ knowl- these are queried about, both in terms of what edge of vocabulary. This subtest is the best is and what is not expected to change. The single correlate of verbal intelligence and has interviewer is allowed to ask for clarification been widely used as a brief assessment of pre- at any point when confused or when the par- morbid intellectual function. ticipant offers only a single sentence in re- Scale to Assess Unawareness of Mental sponse to a question. Queries are made as Disorders (SUMD; Amador et al. 1995) is a non-directively as possible and are consistent rating scale completed by clinically trained with the participant’s own words. Thus, if the staff following a semi-structured interview and participant indicates that she or he has a con- chart review. For the purposes of this study dition that is “a gift from God,” all queries we used the total score, which is the sum of at this point will be about the “gift.” If the the three central items of the SUMD: a) participant later refers to that “gift” as an “ill- awareness of mental disorder; b) awareness of ness,” only then is “illness” queried about. The the consequences of mental disorder; and c) LYSAKER ET AL. 201

TABLE 1 Rating Criteria for the Narrative Coherence Rating Scale

NCRS Scale 0 1 2 3

Logical Many major instances Some major or many Some minor Not missing connections of no logical minor instances of instances of no logi- logical connections no logical connec- cal connections connections tions Historical Missing many major Missing some major or Missing some minor No details details details many minor details details missing Plausibility Many major moments Some major or many Some minor moments Not lacking lacking realism minor moments lack- lacking realism realism ing realism

awareness of the effects of medication. Each Greig, and Kaplan 1999). Of note, the original of these items is rated on a 5-point scale which Cognitive component includes the PANSS in- ranges from “1” or “complete awareness” to sight and judgment item. Given, however, that “5” or “severe unawareness.” Information re- we thought that this item might bias us to find garding the interrater reliability and validity positive results because of its high correlation of the SUMD have been presented elsewhere with the SUMD total (Lysaker, Bell, Bryson, (Amador et al. 1995; Lysaker, Bell, Bryson, and Kaplan 1998b), we calculated the Cogni- and Kaplan 1998b). tive Component without the PANSS insight Attitude Questionnaire (AQ; Magaletta item. and Oliver 1999) contains the 28 items pub- lished by Magaletta and Oliver (1999) that Procedures were used in their factor analysis of multiple measures assessing hope and related con- Following informed consent, partici- structs. Scores for hopelessness for this study pants were given the IPII. The interview was were calculated by obtaining an average of generally conducted by two interviewers— responses to the 14 items contained in the one whose primary task was to conduct the hopelessness factor with higher scores reflect- interview and the other who typed the partici- ing higher levels of hopelessness. Items in- pants’ responses verbatim. Ratings using the clude “I’m always optimistic about my future” NCRS were generally made more than a and “I always look on the bright side of month following the interview. For the pur- things.” poses of interrater reliability, the first 16 tran- Positive and Negative Syndrome Scale scripts were also rated blindly by two separate (PANSS; Kay, Fizszbein, and Opler 1987) is raters, neither of whom was present during a 30 item rating scale completed by clinically the IPII. trained research staff at the conclusion of chart Of the 33 participants who completed review and a semi-structured interview. For the IPII, researcher ratings of insight and the purposes of this study the Positive, Nega- symptoms and self-reports of hopelessness tive and Cognitive PANSS factor analytically were available for 30 from the database of an derived components were utilized (Bell, Ly- earlier study that examined the correlates of saker, Gouet, Milstein, and Lindenmayer hopelessness and outcome in schizophrenia 1994). The factor structure of the PANSS has (c.f. Lysaker and Lysaker 2001). Information been widely replicated (e.g., Frederickson et from this database was gathered between 6 al. 1997) and information about its predictive months and 1 year prior to obtaining the IPII validity presented elsewhere (Bryson, Bell, transcripts. Researcher ratings of insight 202 INSIGHT AND PERSONAL NARRATIVES OF ILLNESS IN SCHIZOPHRENIA and psychopathology obtained in the data- dictive validity of the NCRS, the criterion for base were made blind to participants’ self re- accepting an individual correlation as signifi- port. cant was set at .025 instead of the traditional .05. Additionally, two-tailed tests were used Results despite the fact that unidirectional hypotheses were made a priori. Means and standard deviations for the As presented in Table 3, all NCRS NCRS scores were as follows: Logical consis- scores were significantly related to the SUMD tency = 5.3 (2.0); Details = 5.5 (2.2); Plausibil- total score, with poorer insight predicting ity = 4.7 (1.6); and the Total = 15.6 (4.7). lower levels of narrative coherence of illness NCRS scores were unrelated to age, education on all scales. Previous ratings of cognitive and lifetime number of hospitalizations. The symptoms using the PANSS predicted less NCRS scores of participants with schizophre- logical consistency, less plausibility and a nia did not differ from those of participants lower overall coherence total. Previous ratings with schizoaffective disorder. of hopelessness predicted greater levels of log- To assess internal consistency, a coeffi- ical consistency and greater plausibility. cient alpha was calculated using the six indi- PANSS positive and negative components and vidual ratings (i.e., past and present ratings for the WAIS-III Vocabulary subtest scores were all three subscales) for the 33 cases rated by unrelated to NCRS scores. the primary rater. An overall coefficient alpha of .88 was found. Calculation of coefficient alpha with items deleted found no instances DISCUSSION in which overall consistency was improved with the deletion of items. As presented in The data from this study suggest that Table2, good to excellent interrater reliability the IPII and NCRS may offer a meaningful was found using three raters for the same 16 way of assessing awareness of illness as embed- cases with intra-class correlations ranging ded within a narrative of illness among persons from .81 to .95. Also as presented in Table 2, with schizophrenia. An acceptable level of in- individual scales were significantly related to ternal consistency was found among the indi- one another, generally accounting for between vidual items of the NCRS and an equally 25% and 36% of the variance amongst one acceptable degree of agreement was found be- another individually. tween blind raters for the NCRS, suggesting Of note, since multiple correlations that the NCRS measures its constructs reli- were planned to assess the concurrent and pre- ably. Evidence that NCRS ratings validly as-

TABLE 2 Intercorrelations and Interrater Reliability of the Narrative Coherence Rating Scale

Logical Interrater Connec- Historical Total1 reliabil- tions1 Details1 Plausibility1 n = ity2 NCRS Scale n = 33 n = 33 n = 33 33 n = 16

Logical Connections 1.00 ns ns ns .84** Historical Details .46* 1.00 ns ns .95** Plausibility .59** .44* 1.00 ns .81** Total .84** .80** .78** 1.00 .90**

1Pearson correlation coefficient. 2Intraclass correlation for 3 raters. *p < .01; **p < .0001 LYSAKER ET AL. 203

TABLE 3 Intercorrelations of Ratings of Insight Symptoms and Self-report of Trauma and Hopelessness with the Narrative Coherence Rating Scale (n = 30)

PANSS PANSS PANSS Self- SUMD Positive Negative Cognitive Reported WAIS-III NCRS Scale Total Component Component Component Hopelessness Vocabulary

Logical Connections −.53** .24 .09 −.59*** .47* .30 Historical Details −.41* .18 −.21 −.35 .20 .28 Plausibility −.47* .20 .08 −.48* .44* .07 Total −.59*** .23 −.02 −.54** .35 .28

*p < .025; **p < .01; ***p < .001 sess insight were found in several analyses. to ask whether impairments in executive func- First, NCRS scores were significantly related tion versus verbal memory are related to dif- to ratings of insight gathered using standard- ferent narrative problems with a story of ill- ized interview methods 6 months to 1 year ness. Also, are different forms of disruptions earlier. Second, paralleling other insight re- of an illness narrative more or less effective search, the NCRS total score and the logical as shields against some of the causes of despair connections and plausibility subscales were re- in schizophrenia, such as stigma and social lated to cognitive symptoms, while the logical isolation? In all, by conceptualizing insight as connections and plausibility subscales were re- more than an isolated piece of knowledge as lated to hopelessness. There was no evidence well as a story bound to a life’s trajectory, we that NCRS scores merely reflect global im- reason that the methods detailed here might pairment or psychopathology. NCRS scores uniquely provide a window into the manner were not significantly correlated with positive in which an understanding of illness may be symptoms, negative symptoms or verbal intel- incomplete and thus illuminate a person’s ligence. Of note, only modest evidence was individual needs. In other words, whereas tra- found with regard to discriminant validity of ditional measures may measure whether the individual subscales. The logical connec- someone knows or doesn’t know that he or tions and plausibility subscales were both she is ill, the measures presented here have equally related to cognitive symptoms and the potential to portray in greater detail the hopelessness while the historical details scale unique ways in which his or her story may was not. capture or fail to capture his or her condition. With replication across broader sam- Additionally, more comprehensive means ples and settings, we would suggest that the of measuring the coherence of an illness narra- IIPI and NCRS, relative to traditional mea- tive could have important implications for un- sures which assess the accuracy of appraisals derstanding a person’s individual needs and of illness, might provide a unique method for may therefore be important for both clinical examining the correlates and etiology of in- practice and outcome research. It is widely sight as well as a venue for studying the ways noted that one of the unique outcomes that in which persons with mental illness narrate psychotherapy seeks to effect in schizophrenia their lived experience of that illness. By assess- is increased narrative coherence (Fenton ing the different ways in which a narrative of 2000; Haugsjerd 1994; Lysaker and Lysaker illness can fail to achieve coherence, perhaps 2001). Perhaps the methods presented in this the complex interrelationship between aware- article could serve as a means of assessing ness, denial, neurocognition and pain can be whether or not participation in psychotherapy better understood. For instance, future studies effects narrative in schizophrenia and if so, with the IPII and NCRS have the potential in what ways. One might similarly use these 204 INSIGHT AND PERSONAL NARRATIVES OF ILLNESS IN SCHIZOPHRENIA methods to ask whether pharmacological in- treatment. Thus, replication is needed with terventions, which appear to increase cogni- larger populations, including women, persons tive capabilities, are related to increases in nar- in an earlier phase of illness and/or those re- rative coherence in severe mental illness. For fusing treatment. A wider variety of predictor clinicians these methods might provide in- variables also needs to be studied to determine sights into the narrative struggles of persons whether each subscale is in fact differentially they are endeavoring to treat. It may prove related to outcome. Second, the narratives useful for both treatment planning and assess- elicited here, as is true of narratives in ment of change to know more about the ways general (cf. Hermans 1996), came into being in which stories of “what is wrong” do and do fundamentally through dialogue with another not cohere. One might imagine, for instance, in a particular context. Replication with inter- that different psychotherapeutic interventions viewers in other sites and nonclinical settings might be useful for the person whose story are also essential, as well as is research on the lacks historical detail versus. the person whose influence of the interviewer. Third, the NCRS story lacks plausibility. It may also be particu- assesses three facets of narrative we believed to larly useful to track changes in the disparity be intricately involved in narrative coherence. between hopelessness and coherence of narra- There may be more aspects of narrative yet tive of disorder. As found here and elsewhere, to be articulated that could be incorporated greater coherence may have a sobering effect in future versions of the scale. Finally, since such as that mood elevation may be eroded multiple correlations were performed in the and/or realistic despair ensues. This may be validity analyses, the chances of spurious find- especially important to monitor and to address ings are inflated, even though more conserva- clinically given literature that this process may tive alphas and two-tailed tests were em- be accompanied by resistance and/or client’s ployed. wishes to abandon their gains (Duckworth, As a final note, it is the intent of this Nair, and Patel 1997). study to supplement current insight research Of note, this was the first study of the by providing a means of assessing various as- reliability and validity of the IPII and NCRS pects of awareness of illness as it is embedded and there are several limitations. First, our within a personal narrative. We would suggest sample size was modest, we utilized few pre- that both traditional and narrative approaches dictor variables to determine discriminant va- to assessing insight complement one another lidity and all participants were male. Addition- and enrich our understanding of the experi- ally, many years had passed since the onset of ences and dialogues contained within a life- their illness and all were in some form of active time of living with mental illness.

REFERENCES

AMADOR, X. F., FLAUM, M., ANDREASEN, J. B., MILSTEIN, R. M., and LINDENMAYER,J.P. N., STRAUSS, D. H., YALE, S. A., CLARK, S. C., Five component model of schizophrenia: Assessing and GORMAN, J. M. Awareness of illness in schizo- the factorial invariance of the PANSS. Psychiatry phrenia and mood disorders. Archives of General Research (1994) 52:295–303. Psychiatry (1995) 51:826–36. BLACKWOOD, N. J., HOWARD, R. J., and AMADOR, X. F., STRAUSS, D. H., YALE, BENTEL, R. B. Cognitive neuropsychiatric models S. A., and GORMAN, J. M. Awareness of illness of persecutory delusions. American Journal of Psychi- in schizophrenia. Schizophrenia Bulletin (1991) 17: atry (2001) 158:527–39 113–32. BRYSON, G. J., BELL, M. D., GREIG,T., BARTKO, G., HERCZEG, I., and ZADOR,G. and KAPLAN, E. Internal consistence and temporal Clinical symptomatology and drug compliance in stability and neuropsychological correlates of three schizophrenia. Acta Psychiatrica Scandinavica (1988) cognitive components of the PANSS. Schizophrenia 77:74–6. Research (1999). 38:27–35. BELL, M. D., LYSAKER, P. H., GOULET,CARROLL, A., FATTAH, S., CLYDE, Z., LYSAKER ET AL. 205

COFFEY, I., OWENS, D.G.C., and JOHNSTONE, schizophrenia. Journal of Nervous and Mental Dis- E. C. Correlates of insight and insight change in ease. (1995) 183:103–7. schizophrenia. Schizophrenia Research (1999) 35: LYSAKER, P. H., BELL, M. D., MILSTEIN, 247–53. R. M., BRYSON, G. J., and GOULET, J. Insight and CORRIGAN, P.W., and PENN, D. L. Lessons psychosocial treatment compliance in schizophre- from social psychology on discrediting psychiatric nia. Psychiatry (1994) 57:307–15. stigma. American Psychologist (1999) 54:765–775. LYSAKER, P. H., BELL, M. D., BRYSON, CUFFEL, B. J., ALFOR, J., FISCHER,E.P., G. J., and KAPLAN, E. Psychosocial function and and OWEN, R. R. Awareness of illness in schizo- insight in schizophrenia. Journal of Nervous and phrenia and outpatient treatment compliance. Mental Disease (1998a) 186:432–6. Journal of Nervous and Mental Disease (1996) 184: LYSAKER, P. H., BELL, M. D., BRYSON, 653–59. G. J., and KAPLAN, E. Neurocognitive function and DAVID, A. S. Insight and psychosis. British insight in schizophrenia: Support for an association Journal of Psychiatry (1990) 156:798–805. with impairments in executive function but not with DAVIDSON, L., and STRAUSS, J. S. Beyond impairments in global function. Acta Psychiatrica the biopsychosocial model: Integrating disorder, Scandinavica (1998b) 97:297–301. health and recovery. Psychiatry (1995) 58:44–55. LYSAKER, P. H., and FRANCE, C. M. Psy- DUCKWORTH, K., NAIR, V., and PATEL, chotherapy as an element in support employment J. K. Lost time, found hope and sorrow: The search for persons with severe and persistent mental ill- for self, connection and purpose during awakenings ness. Psychiatry (1999) 62:209–22. on the new antipsychotics. Harvard Review of Psychi- LYSAKER, P.H., and LYSAKER, J. T. Psycho- atry (1997) 5:227–33. sis and the disintegration of dialogical self-struc- FENTON, W. S. Evolving perspectives on ture: Problems posed by schizophrenia for the individual psychotherapy for schizophrenia. Schizo- maintenance of dialogue. British Journal of Medical phrenia Bulletin (2000) 26:47–72. Psychology (2001) 74:23–33. FREDERICKSON, D. H., STEIGER,J.M., LYSAKER, P. H., BRYSON, G. J., and BELL, MACEWAN, G. W., ALTMAN, S., KOPALA, L. C., M. D. Insight and work performance in schizophre- FLYNN,S.W.,LIDDLE, P. F., and HONER,W.G. nia. Journal of Nervous and Mental Disease (2002) PANSS symptom factors in schizophrenia. Schizo- 190: 142–146. phrenia Research (1997) 24:15. LYSAKER, P. H., BRYSON, G. J., LANCAS- FRESE, F. J. Twelve aspects of coping for TER, R., EVANS, J. D., and BELL, M. D. Insight in persons with schizophrenia. Innovations and Re- schizophrenia: Associations with executive function search (1993) 2:39–46. and coping style. Schizophrenia Research (in press). HAUGSJERD, S. Can psychoanalytic theory MAGALETTA, P. R., and OLIVER,J.M.The contribute to the understanding and treatment of hope construct, will and ways: Their relations with schizophrenia? Acta Psychiatrica Scandinavica (1994) self-efficacy, optimism, and general well-being. 90(384):147–52. Journal of Clinical Psychology (1999) 55:539–51. HERMANS, H.J.M. Voicing the self: From MARKS, K. A., FASTENAU, P. S., LYSAKER, information processing to dialogical interchange. P. H., and BOND, G. R. Self-appraisal of illness Psychological Bulletin (1996) 119:31–50. questionnaire: Relationship to researcher-rated in- KAY, S. R., FIZSZBEIN, A., and OPLER, sight and neuropsychological validation. Schizo- L. A. The positive and negative syndrome scale phrenia Research (2000) 45:203–11. for schizophrenia. Schizophrenia Bulletin (1987) 13: MOHAMED, S., FLEMING, S., PENN, P., and 261–76. SPAULDING, W. Insight in schizophrenia: Its rela- KIRMAYER, L. J., and CORIN, E. Inside tionship to measures of executive function. Journal knowledge: Cultural constructions of insight in of Nervous and Mental Disease (1999) 187:525–31. psychosis. In X. F. Amador and A. S. David, eds., SCHWARTZ, R. C. The relationship be- Insight and Psychosis. (pp. 193–220). Oxford Univer- tween insight, illness and treatment outcome in sity Press, 1998. schizophrenia. Psychiatric Quarterly (1998) KLEINMAN,A.The Illness Narratives: Suffer- 69:1–22. ing, Healing and the Human Condition. Basic Books, SCHWARTZ, R. C. Self-awareness in schizo- 1988. phrenia: Its relationship to depressive symptom- LYSAKER, P. H., and BELL, M. D. Insight atology and broad psychiatric impairments. Journal and cognitive impairment in schizophrenia: Perfor- of Nervous and Mental Disease (2001) 189:401–3. mance on repeated administrations of the Wiscon- SMITH, T. E., HULL,J.W.,GOODMAN, M., sin Card Sorting Test. Journal of Nervous and Mental HEDAYAT-HARRIS, A., WILSON, D. F., ISREAL, Disease (1994) 182:656–60. L. M., and MUNICH, R. L. The relative influences LYSAKER, P. H., and BELL,M.D.Work of symptoms, insight and neurocognition on social rehabilitation and improvements in insight in adjustment in schizophrenia and schizoaffective 206 INSIGHT AND PERSONAL NARRATIVES OF ILLNESS IN SCHIZOPHRENIA disorder. Journal of Nervous and Mental Disease WECHSLER,D.Wechsler Adult Intelligence (1999) 187:102–8. Scale—III. Psychology Corporation, 1997. THOMPSON, E. H. Variations in the self- WILLIAMS, C. C., and COLLINS,A.A.De- concept of young adult chronic patients: Chronicity fining new frameworks for psychosocial interven- reconsidered. Hospital and Community Psychiatry tions. Psychiatry (1999) 62:61–78. (1988) 39:771–5. YOUNG, D. A., DAVILA, R., and SCHER, WAHL, O. F., and HARMAN, C. R. Family H. Unawareness of illness and neuropsychological views of stigma. Schizophrenia Bulletin (1989) 15: performance in schizophrenia. Schizophrenia Re- 131–4. search (1993) 10:117–24. WARNER, R., TAYLOR, D., POWERS, M., YOUNG, D. A., ZAKZANIS, K. K., and BAI- and HYMAN, R. Acceptance of the mental illness LEY, C. Further parameters of insight and neuro- label by psychotic patients: Effects on functioning. psychological deficit in schizophrenia and other American Journal of Orthopsychiatry (1989) 59:389– chronic mental diseases. Journal of Nervous and 409. Mental Disease (1998) 186:44–50.