Delirium and Catatonia in Critically Ill Patients: The Delirium and Catatonia Prospective Cohort Investigation

Abstract

Objectives: Catatonia, a condition characterized by motor, behavioral, and emotional changes, can occur during critical illness and appear as clinically similar to delirium, yet its management differs from delirium. Traditional criteria for medical catatonia preclude its diagnosis in delirium. Our objective in this investigation was to understand the overlap and relationship between delirium and catatonia in ICU patients and determine diagnostic thresholds for catatonia.

Design: Convenience cohort, nested within two ongoing randomized trials.

Setting: Single academic medical center in Nashville, TN.

Patients: We enrolled 136 critically ill patients on mechanical ventilation and/or vasopressors, randomized to two usual care sedation regimens.

Measurements and main results: Patients were assessed for delirium and catatonia by independent and masked personnel using Confusion Assessment Method for the ICU and the Bush Francis Catatonia Rating Scale mapped to Diagnostic Statistical Manual 5 criterion A for catatonia. Of 136 patients, 58 patients (43%) had only delirium, four (3%) had only catatonia, 42 (31%) had both, and 32 (24%) had neither. In a logistic regression model, more catatonia signs were associated with greater odds of having delirium. For example, patient assessments with greater than or equal to three Diagnostic Statistical Manual 5 symptoms (75th percentile) had, on average, 27.8 times the odds (interquartile range, 12.7-60.6) of having delirium compared with patient assessments with zero Diagnostic Statistical Manual 5 criteria (25th percentile) present (p < 0.001). A cut-off of greater than or equal to 4 Bush Francis Catatonia Screening Instrument items was both sensitive (91%; 95% CI, 82.9-95.3) and specific (91%; 95% CI, 87.6-92.9) for Diagnostic Statistical Manual 5 catatonia.

Conclusions: Given that about one in three patients had both catatonia and delirium, these data prompt reconsideration of Diagnostic Statistical Manual 5 criteria for "Catatonic Disorder Due to Another Medical Condition" that preclude diagnosing catatonia in the presence of delirium.

Figure 1. Distribution of Delirium and/or Catatonia

Bar Graph showing distribution of delirium and/or catatonia over the entire ICU stay by individual patients (N=136 patients). This figure demonstrates that 31% of patients met criteria for both delirium and catatonia during their ICU or hospital stay, which goes against current DSM catatonia nosology precluding diagnosis of catatonia due to another medical condition when delirium is present. †Delirium diagnosed using CAM-ICU: Confusion Assessment Method for the ICU. Catatonia diagnosed using ≥ 3 DSM-5 Criterion A items. † Delirium was diagnosed with the CAM-ICU: Confusion Assessment Method for the ICU. * Catatonia was defined according to DSM 5 Criterion A.

Figure 2. Frequency and Distribution of Catatonic Signs According to Delirium Status

Percent of catatonic signs that were present in delirious (CAM-ICU +) versus non-delirious (CAM-ICU −) assessments (N=452, all paired assessments rather than by patient). Note that for each catatonic sign shown on the Y axis, the frequency is higher in delirious than in non-delirious patients, demonstrating overlap of individual catatonic signs and delirium in critically ill patients. †Delirium diagnosed using CAM-ICU: Confusion Assessment Method for the ICU. ‡ Catatonic signs were measured using the Bush Francis Catatonia Rating Scale. † Delirium was diagnosed with the CAM-ICU: Confusion Assessment Method for the ICU.

Figure 3. Probability of Delirium at Different Catatonia Thresholds

Probability of being delirious (CAM-ICU+) at various DSM-5 Criterion A cut-off points. * Clinical Interpretation: Again using the interquartile ranges of the cohort distributions to paint a clinically interpretable picture of the meaning of these data, a patient with 3 DSM-5 catatonia signs present had 27.8 times the odds (95% CI: 12.7, 60.6) of being delirious in comparison to a patient with 0 DSM-5 catatonia signs present; p= <0.001. NOTE: This figure uses the reference standard DSM-5 criteria for catatonia. This threshold of ≥3 DSM-5 items for catatonia is distinct from the Bush Francis Catatonia Screening Instrument (BFCSI) thresholds shown in Table 2, which are presenting validation data of a bedside instrument that could be used by non-psychiatrists in routine ICU care. See Figure Legend for clinical examples that explain the application of these data. ‡ Delirium was diagnosed using CAM-ICU: Confusion Assessment Method for the ICU. † The gray shaded area represents the 95% Confidence Interval (CI).