Elsevier

Psychiatry Research

Volume 101, Issue 1, 14 February 2001, Pages 47-54
Psychiatry Research

Diagnosis of manic episodes in adolescent inpatients: structured diagnostic procedures compared to clinical chart diagnoses

https://doi.org/10.1016/S0165-1781(00)00248-1Get rights and content

Abstract

This study examined the accuracy of clinical chart diagnoses of manic episodes in adolescent psychiatric patients, as well as treatment selection and patient outcome. A consecutive sample of 120 consenting adolescent patients was assessed at admission, discharge, and 30 and 120 days post-discharge. Clinical chart diagnoses were compared to research-quality diagnoses involving structured interview, chart review, and consensus. Agreement statistics were computed, and the symptom and treatment differences were compared between patients for whom there was and was not diagnostic agreement. Clinical diagnoses of manic episodes were more common than research diagnoses, and the rate of agreement between diagnoses was low (kappa=0.15). Patients diagnosed as experiencing a manic episode by the clinical chart, but not via the research procedure, had reduced severity scores on elation and activity, and higher scores on depression. These patients also had more severe scores on depressive symptoms at follow-up. Manic episodes were diagnosed more frequently by clinicians relative to research-quality procedures. Patients who were diagnosed as experiencing manic episodes by the clinician, but not the research procedure, appeared to have depression and hostility, but not elation. The depression in these patients may not be adequately treated, and there are potential clinical implications of over-diagnosis of manic episodes in adolescents.

Introduction

The occurrence of bipolar disorder in children and adolescents was reported as early as 1921 by Kraepelin, and recent epidemiological data (Carlson and Kashani, 1988, Lewinsohn et al., 1995) and adult retrospective reports (Joyce, 1984, Lish et al., 1994) continue to document the presence of bipolar disorder in adolescents. Prevalence estimates between 6 (Lewinsohn et al., 1993) and 7% have been reported in adolescents (Kashani et al., 1987), and some researchers have suggested a possible recent increase in the overall incidence of adolescent-onset bipolar disorder (Geller and Luby, 1997). Other researchers have reported peak risk rates for the onset of bipolar disorder between 15 and 19 years of age and a median onset of 19 years of age, indicating that roughly half the cases of bipolar disorder have onset during the teenage years or even earlier (Burke et al., 1990). This research, together with significant advances in diagnostic nomenclature, has contributed to an increased recent awareness of the prevalence and characteristics of bipolar disorder in young people.

Several authors have previously written about the under-recognition and under-diagnosis of bipolar disorder in children (e.g. Weller et al., 1986, Weller et al., 1994) and adolescents (Carlson and Kashani, 1988). For instance, many adolescents and children who met formal criteria for bipolar disorder were found to be receiving clinical treatment for other disorders, such as attention-deficit disorder. In addition, the use of structured diagnostic interviews, such as the K-SADS (Ambrosini, 2000), has resulted in the identification of cases of adolescent bipolar disorder missed by clinical diagnosticians (Gammon et al., 1983).

A specific focus of the present study was on the identification of manic episodes. Bipolar disorder is intrinsically variable in its clinical presentation and a manic episode is required for the diagnosis of bipolar I disorder according to DSM-IV criteria. An individual with a history of a manic episode who is later found to be depressed will be given a diagnosis of bipolar disorder at subsequent episodes of depression. Consequently, the identification of a history of a manic episode will have a marked impact on treatment planning and future expectations regarding prognosis and outcome. Little direct information on the accuracy of identification of current manic episodes in adolescents is available and this study presents preliminary data regarding this issue.

In the present study, we examined the clinician versus research-derived diagnoses of bipolar disorder, manic type, in a consecutive sample of 120 adolescent psychiatric inpatients using a highly standardized data-collection procedure. All patients were examined with a structured diagnostic interview, comprehensive chart review, and informant interview while in the hospital. This information was presented at a consensus meeting with a senior clinician who reviewed all of the information and discussed each case in detail with the diagnostic rater. After this procedure, an ‘expert consensus’ diagnosis was generated and this diagnosis was compared to clinician diagnoses recorded in the chart. The treatments received by the patients in this study were examined and their 120-day outcome after discharge was also examined. Among the outcome measures collected were levels of severity of affective and non-affective psychiatric symptoms, readmission to inpatient psychiatric care, and compliance with medication and psychotherapy recommendations that were made at the time of discharge from the hospital.

Several specific questions were addressed:

  • 1.

    What was the rate of clinician diagnosis of a manic episode in this sample of 120 consecutive admissions?

  • 2.

    What was the rate of agreement between research-quality diagnostic procedures and clinician chart diagnoses?

  • 3.

    What were the research-based diagnoses of subjects on whom there was diagnostic disagreement?

  • 4.

    What treatments did the patients diagnosed as experiencing a manic episode by their clinical staff receive?

  • 5.

    What was the clinical outcome as a function of diagnostic agreement?

Section snippets

Subjects

Subjects in this study were 120 adolescent psychiatric inpatients who participated in a longitudinal study of treatment outcome at a private psychiatric hospital in Westchester, New York. The sample was drawn from consecutive admissions to adolescent inpatient services, stratified across inpatient treatment units as a function size of the unit. Each patient received a full research-quality diagnostic assessment, as well as a separate comprehensive assessment of symptoms and treatments received

What was the rate of clinical diagnosis of bipolar disorder, manic type, in this sample of 120 consecutive admissions?

Of the 114 adolescent inpatients who received a discharge assessment, 29 (25%) had a chart diagnosis of bipolar disorder, currently manic, at discharge, whereas 14 patients (12%) had a research diagnosis of bipolar disorder, currently manic, at the time of their discharge from inpatient care.

What was the rate of agreement between research-quality diagnostic procedures and clinical chart diagnoses?

Of the 29 adolescents who received a diagnosis of bipolar disorder, currently manic, in their hospital chart, 11 also received this research diagnosis. This is a raw level of agreement of 40%. When

Discussion

The results of this study indicated that clinical diagnosis of manic episodes was more common than research-based diagnosis in adolescent psychiatric inpatients. This finding occurred despite the fact that the research diagnosticians had access to every chart note written by the clinicians that generated the diagnoses. Similar to previous studies on inpatient units, approximately 15% of the adolescent inpatients received a research diagnosis of bipolar disorder, currently manic, which was

References (17)

There are more references available in the full text version of this article.

Cited by (54)

  • Cognitive behavior therapy for COVID-19 related distress

    2021, Mental Health Effects of COVID-19
  • The Use of Evidence-Based Assessment for Anxiety Disorders in an Australian Sample

    2020, Journal of Anxiety Disorders
    Citation Excerpt :

    Structured or semi-structured diagnostic interview schedules provide the clinician an opportunity to assess symptoms using scientifically tested questions directly derived from diagnostic criteria (Silverman, 1994). Research has demonstrated that such structured interviews, when compared to the clinical interview, are more accurate (Basco et al., 2000; Endicott, 2001; Jensen-Doss & Weisz, 2008; Miller, 2001; Pogge et al., 2001; Silverman & Ollendick, 2005) and comprehensive (Pinninti, Madison, Musser, & Rissmiller, 2003; Shear et al., 2000; Zimmerman & Mattia, 1999). Structured interviews have also been shown to predict better therapy engagement (such as lower rates of appointment ‘no-shows’, cancellations, and reduced rates of therapy drop-out) and enhance treatment outcomes (Jensen-Doss & Weisz, 2008).

  • Training teachers' diagnostic competence with problem-based learning: A pilot and replication study

    2019, Teaching and Teacher Education
    Citation Excerpt :

    A construct that at least builds on professional interactions is the diagnostic agreement. Studies in medicine showed that diagnostic agreement not only has a positive effect on patients' compliance and motivation, but also on treatment success (Jensen-Doss & Weisz, 2008; Pogge et al., 2001). As a first step in schools, different settings of collaborative and interactive settings could be identified (e.g., case conferences, documentations), as well as their effects on individual diagnostic competence and students' academic achievement, learning behavior, and achievement emotions.

View all citing articles on Scopus
View full text