Comparison of diagnostic classification systems for delirium with new research criteria that incorporate the three core domains

https://doi.org/10.1016/j.jpsychores.2016.03.011Get rights and content

Highlights

  • New TMF research diagnostic criteria uniquely capture delirium's three core domains.

  • An agnostic reference standard compared DSM, ICD, and new TMF criteria performances.

  • TMF criteria have high/balanced specificity and sensitivity compared to DSM & ICD.

  • TMF criteria may improve accuracy of delirium research.

Abstract

Objective

Diagnostic classification systems do not incorporate phenomenological research findings about the three core symptom domains of delirium (Attentional/Cognitive, Circadian, Higher Level Thinking). We evaluated classification performances of novel Trzepacz, Meagher, and Franco research diagnostic criteria (TMF) that incorporate those domains and ICD-10, DSM-III-R, DSM-IV, and DSM-5.

Methods

Primary data analysis of 641 patients with mixed neuropsychiatric profiles. Delirium (n = 429) and nondelirium (n = 212) reference standard groups were identified using cluster analysis of symptoms assessed using the Delirium Rating Scale-Revised-98. Accuracy, sensitivity, specificity, positive and negative predictive values (PPV, NPV), and likelihood ratios (LR +, LR −) are reported.

Results

TMF criteria had high sensitivity and specificity (87.4% and 89.2%), more balanced than DSM-III-R (100% and 31.6%), DSM-IV (97.7% and 74.1%), DSM-5 (97.7% and 72.6%), and ICD-10 (66.2% and 100%). PPV of DSM-III-R, DSM-IV, and DSM-5 were < 90.0%, while PPV for ICD-10 and TMF were > 90%. ICD-10 had the lowest NPV (59.4%). TMF had the highest LR + (8.06) and DSM-III-R the lowest LR − (0.0). Overall, values for DSM-IV and DSM-5 were similar, whereas for ICD-10 and DSM-III-R were inverse of each other. In the pre-existing cognitive impairment/dementia subsample (n = 128), TMF retained its highest LR + though specificity (58.3%) became less well balanced with sensitivity (87.9%), which still exceeded that of DSM.

Conclusions

TMF research diagnostic criteria performed well, with more balanced sensitivity and specificity and the highest likelihood ratio for delirium identification. Reflecting the three core domains of delirium, TMF criteria may have advantages in biological research where delineation of this syndrome is important.

Introduction

Research using symptom rating instruments has advanced the phenomenological understanding of delirium and found that delirium has three core symptom domains [1]. These domains are Cognitive (attention with other cognitive abilities), Higher Level Thinking (thought process, semantic language and executive function) and Circadian (sleep–wake cycle and motor activity patterns). These were delineated from studies using descriptive, regression, and exploratory and confirmatory factor analyses [2], [3], [4], [5], [6], [7], [8]. These domains are also consistent with findings from delirium research on sleep–wake cycle, motor activity and attention [9], [10], [11], [12], [13], [14]. Further, these core domain symptoms are likely generated from associated underlying neural disturbances implicated in delirium. As a state of impaired consciousness, delirium alters functioning of highly distributed neural networks for information processing across all higher cerebral cortical regions, as well as gating and circadian rhythm in diencephalic regions (thalamus and hypothalamus) [15], [16], [17].

Different versions of the Diagnostic and Statistical Manual of Mental Disorders (DSM) and International Classification of Diseases (ICD) have been employed over recent decades for delirium diagnosis, but comparison studies indicate that these systems vary in their identification of delirium [18], [19], [20], [21], [22], [23], [24], [25]. Though their cardinal criterion involves inattention, inclusion of other symptoms varies. Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) [26] and Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) [27] require the fewest symptoms of any classification system to obtain a diagnosis of delirium, therefore capturing cases that could be termed subsyndromal [18], [19], [20], [21], [22], [23], [24], [25]. Further, none requires nor recommends that at least one symptom be present from each of delirium's three core domains.

When diagnostic criteria are loose in their requirements they may increase detection rates by nonspecialists, where false positives are better tolerated than false negatives given the prognostic implications of delirium. However, research benefits from a more accurate classification system that increases certainty regarding delirium presence, especially critical for translational and treatment research.

More specific delirium diagnostic criteria are needed because of the inconsistencies among current classification systems that are principally derived through expert consensus. Our proposed Trzepacz, Meagher, and Franco (TMF) Research Diagnostic Criteria for delirium (see Box 1 below, and Supplemental Box 1 in e-component for Spanish version) capture and require elements from all three core domains with the intent that researchers seeking to understand phenomenology, pathophysiology, treatment and translational relationships have criteria that would be more enduring and rigorous (see Table 1 for comparisons across classification systems). Requiring presence of core domains should also enhance clinical diagnosis of delirium. In fact, Kean et al. (2010) developed a 3-item delirium diagnostic tool for nonspecialists that only assessed symptoms representing each core domain (sleep–wake cycle, vigilance and comprehension) in acute traumatic brain injury patients and found a very strong relationship using receiver operating characteristic analysis with an independent DSM-IV diagnosis, performing similarly to the Delirium Rating Scale-Revised-98 (DRS-R98) [6].

Our study aim is to compare the discriminant capacity of existing delirium diagnostic systems with our proposed TMF research diagnostic criteria. In order to evaluate the TMF criteria in an unbiased fashion we could not rely on an existing system as the reference standard. Therefore, we first developed an agnostic reference standard using cluster analysis of DRS-R98 items from a prospectively collected pooled research database of 641 neuropsychiatric cases to determine delirium and nondelirium groups. These delirium and nondelirium clusters (groups) provided the independent reference standard against which we compared the discriminant performances of DSM, ICD and our proposed TMF criteria for delirium status.

Section snippets

Study population and design

This report includes cross-sectional data prospectively collected during research assessments of delirium in 8 patient cohorts from 5 inter-related studies of delirium phenomenology conducted in Ireland and India as part of a collaborative consortium, the Cognitive Impairment Research Group at the University of Limerick, in Ireland. Data collection and rater training were standardized and consistent across all studies where all collaborators were experienced delirium researchers (DM, ML, FJ,

Comparability of DRS-R98 scores across studies included in the analysis

Cronbach's alpha showed a very good internal consistency of our merged dataset. It was 0.90 (range = 0.88 to 0.92 after removing each of the eight cohorts that comprised the sample) for the DRS-R98 Total scale, and 0.89 for DRS-R98 Severity scale (range = 0.87 to 0.91).

Reference standard groups delineated by DRS-R98 cluster analysis

Boxplot distributions of DRS-R98 Total scores for DSM-IV clinically diagnosed delirium and nondelirium cases are shown in Fig. 1, Part A. Reference standard groups obtained by cluster analysis, 212 nondelirium and 429 delirium cases,

Discussion

Our aim was to evaluate a newly proposed classification system for delirium (TMF) that more fully represented the phenomenological research data for the uniqueness of delirium by requiring presence of symptoms from all three of its core domains. Researchers, in particular, want to ensure their subjects are truly representative of a condition's core characteristic features in order to delve beyond phenotype into endophenotype and genotype levels of understanding. Therefore, we developed and

Competing interest statement

All authors have completed the Unified Competing Interest form at www.icmje.org/coi_disclosure.pdf and declare that (1) None of the authors received support for the submitted work; (2) Dr. Paula T. Trzepacz is retired employee of Eli Lilly and Company, though they have no product for delirium; (3) None of the spouses, partners, or children of the authors have financial relationships that may be relevant to the submitted work; and (4) Dr. Paula T. Trzepacz holds the copyright of the DRS-R98 but

Role of funding source

There was no formal funding for this study.

Special acknowledgment

We dedicate this work to honor the late Dr. Maeve Leonard, our wonderful and esteemed collaborator in delirium research whose efforts for over a decade, even right up to her hospice admission, enabled us to develop these new research diagnostic criteria.

Other acknowledgments

Suzanne Timmons and Niamh O′Regan from the Centre for Gerontology and Rehabilitation, School of Medicine, University College Cork, Cork, Ireland; Sandeep Grover from the Department of Psychiatry, Postgraduate Institute of Medical Education & Research, Chandigarh 160012, India; and Faiza Jabbar from the Psychiatry for Later Life Service, University College Hospital, Galway, Ireland, for their efforts in collecting the clinical data used in this pooled analysis.

References (39)

  • S.K. Mattoo et al.

    Symptom profile and etiology of delirium in a referral population in northern India: factor analysis of the DRS-R98

    J. Neuropsychiatr. Clin. Neurosci.

    (2012)
  • J. Kean et al.

    Initial validation of a brief provisional diagnostic scale for delirium

    Brain Inj.

    (2010)
  • S. Thurber et al.

    Confirmatory factor analysis of the Delirium Rating Scale Revised-98

    J. Neuropsychiatr. Clin. Neurosci.

    (2015)
  • D.P. Lowery et al.

    Quantifying the association between computerised measures of attention and confusion assessment method defined delirium: A prospective study of older orthopaedic surgical patients, free of dementia

    Int. J. Geriatr. Psychiatry

    (2008)
  • M. Van Uitert et al.

    Rest-activity patterns in patients with delirium

    Rejuvenation Res.

    (2011)
  • K. Hatta et al.

    Preventive effects of ramelteon on delirium: a randomized placebo-controlled trial

    JAMA. Psychiatry.

    (2014)
  • S. Balan et al.

    The relation between the clinical subtypes of delirium and the urinary level of 6-SMT

    J. Neuropsychiatr. Clin. Neurosci.

    (2003)
  • P.T. Trzepacz

    Update on the neuropathogenesis of delirium

    Dement. Geriatr. Cogn. Disord.

    (1999)
  • P.T. Trzepacz et al.

    Delirium: a subcortical phenomenon?

    J. Neuropsychiatr. Clin. Neurosci.

    (1989)
  • Cited by (0)

    View full text