Fax: (206) 221-5414
Clinical presentation of delirium in patients undergoing hematopoietic stem cell transplantation
Delirium and distress symptoms and time course
Article first published online: 10 JAN 2005
Copyright © 2005 American Cancer Society
Volume 103, Issue 4, pages 810–820, 15 February 2005
How to Cite
Fann, J. R., Alfano, C. M., Burington, B. E., Roth-Roemer, S., Katon, W. J. and Syrjala, K. L. (2005), Clinical presentation of delirium in patients undergoing hematopoietic stem cell transplantation. Cancer, 103: 810–820. doi: 10.1002/cncr.20845
- Issue published online: 3 FEB 2005
- Article first published online: 10 JAN 2005
- Manuscript Accepted: 28 OCT 2004
- Manuscript Revised: 15 OCT 2004
- Manuscript Received: 2 AUG 2004
- American Cancer Society. Grant Number: RPG-97-035-01-PBR
- University of Washington Royalty Research Fund
- National Cancer Institute. Grant Numbers: CA63030, CA78990, CA92408
- bone marrow transplantation;
Delirium is common in patients undergoing hematopoietic stem cell transplantation (HSCT) and is associated with considerable morbidity and excess mortality in diverse patient samples. Although delirium can be treated successfully, it is largely undiagnosed. Understanding the clinical presentation of delirium may help improve the recognition of delirium in these patients. In the current study, the authors investigated the clinical presentation of delirium in HSCT patients, including the time course of these symptoms and comorbid affective distress, fatigue, and pain.
Ninety patients ages 22–62 years were recruited prior to undergoing their first allogeneic or autologous HSCT. Delirium, distress, and pain symptom assessments were conducted prospectively 3 times per week from pretransplantation through Day 30 posttransplantation.
Delirium episodes occurred in 50% of patients and lasted approximately 10 days, with peak severity at the end of the second week posttransplantation. Factor analysis revealed three groups of delirium symptoms representing psychosis-behavior, cognition, and mood-consciousness. Delirium episodes were characterized by rapid onset of psychomotor and sleep-wake cycle disturbance that persisted and cognitive symptoms that continued to worsen throughout much of the episode. Rises in psychosis-behavior and cognitive symptoms predated the start of delirium episodes by ≈ 4 days. Affective distress and fatigue were common and appeared to be associated most with psychosis-behavioral delirium symptoms.
The results describe in detail the clinical presentation of delirium in patients undergoing HSCT. Affective distress and fatigue commonly were associated with delirium. These findings may aid clinicians in improving the recognition and treatment of delirium in this population and avoiding further morbidity and potential mortality. Cancer 2005. © 2005 American Cancer Society.
Delirium is common in patients who undergo hematopoietic stem cell transplantation (HSCT), occurring in 50% of patients during the 4 weeks posttransplantation.1 Delirium is associated with considerable negative outcome, including increased morbidity, functional decline, and utilization of health care resources.2–6 Delirium interferes with the patient's ability to inform physicians and nurses of changes in symptoms and can affect participation in treatment-related decisions. Thus, it can result in delayed recognition and treatment of serious medical conditions and may lead to significant risk of self-injurious behavior, such as falling and pulling out intravenous lines, inadequate self-care, and adherence with medical treatment, and secondary medical problems, such as decubiti and aspiration pneumonia.7 In patients with malignancies, delirium is associated with increased mortality, with excess mortality ratios of 6.2 during hospitalization and 14.1 during 1–5 years of follow-up.8 Moreover, the experience of delirium also is highly distressing for the majority of patients, spouses/caregivers, and nurses caring for patients with delirium.9
There is growing evidence that, once identified, delirium can be treated successfully.10–15 Because delirium is typically a reversible syndrome signaling an underlying medical condition, it is important to make a diagnosis early in its course.3 However, delirium is undiagnosed in 32–67% of patients and remains largely undertreated, despite its association with numerous negative outcomes.3, 16 One potential reason for the lack of diagnosis and treatment of delirium is that delirium can present clinically in many forms. Like in other medical populations, early symptoms of delirium in patients with malignant disease often are overlooked or misdiagnosed as primary anxiety, depression, anger, or psychosis.17 Early and accurate recognition of delirium allows both for treatment of symptoms and for investigation and treatment of the underlying etiologies.
Recognizing delirium in patients who undergo HSCT is complex. The phenomenology of delirium traditionally has been discussed in terms of motoric subtypes: hypoactive, hyperactive, or mixed.18, 19 However, this distinction does not describe the complex symptom constellation of delirium adequately in patients who undergo HSCT. The clinical presentation of delirium symptoms also may vary by suspected etiology (e.g., drug related, infectious/electrolyte, hormonal).20, 21 The results of the few studies that classified delirium symptoms using factor analytic techniques suggest that the clinical presentation of delirium may be variable in different medical populations (e.g., geriatric patients, general medical/surgical unit patients, oncology patients).18, 22–24 Further complicating the diagnostic picture is the comorbidity between delirium symptoms, psychological distress,9 and pain.25 It is possible that heightened pain and distress may serve as a flag, alerting the clinician to the possibility of delirium in these patients.
To aid oncologists in accurately identifying delirium in their patients, it is essential to understand the clinical picture of delirium symptoms among patients who undergo HSCT. To our knowledge, there have been no studies describing the phenomenology of delirium specifically in patients undergoing HSCT. Patients who undergo HSCT represent a population with unique pathophysiologic, treatment, psychosocial, and environmental factors that may influence their neuropsychiatric condition, and they may present with unique delirium symptoms. The objectives of the current study were 1) to use descriptive and factor analytic techniques to describe the domains and time course of delirium symptoms presenting in patients who experience delirium; 2) to describe how these symptoms differ in patients with no delirium, brief delirium, and prolonged delirium; and 3) to investigate the comorbidity of delirium symptoms with affective distress, fatigue, and pain over time.
MATERIALS AND METHODS
Ninety patients ages 22–62 years at the Fred Hutchinson Cancer Research Center (FHCRC) were screened and approached for entry into the study prior to their first myeloablative allogeneic or autologous bone marrow or peripheral blood HSCT from 1997 to 1999. A broad range of malignant diagnoses were represented, including: chronic myeloid leukemia (42%), acute lymphoblastic leukemia/acute myeloid leukemia (28%), breast carcinoma/ovarian carcinoma (12%), myelodysplasia/multiple myeloma (11%), and non-Hodgkin lymphoma (7%), with a range of conditioning regimens (Table 1). More details about the patient characteristics and study methods can be found elsewhere.1
|Characteristic||No. of patients (%)a|
|No delirium (n = 24)b||Delirium event (n = 21)b||Delirium episode (n = 45)b|
|22–35 yrs||11 (46)||7 (33)||10 (22)|
|36–45 yrs||8 (33)||7 (33)||13 (29)|
|46–55 yrs||4 (17)||5 (24)||16 (36)|
|56–62 yrs||1 (4)||2 (10)||6 (13)|
|Male||15 (62)||15 (71)||24 (53)|
|Female||9 (38)||6 (29)||21 (47)|
|White||23 (96)||20 (95)||42 (93)|
|Hispanic||0 (0)||1 (5)||2 (5)|
|Native American||0 (0)||0 (0)||1 (2)|
|Other||1 (4)||0 (0)||0 (0)|
|CML||11 (46)||10 (48)||17 (38)|
|ALL or AML||8 (33)||7 (33)||10 (22)|
|BR or OV||1 (4)||1 (5)||9 (20)|
|MDS or MM||2 (8)||3 (14)||5 (11)|
|NHL||2 (8)||0 (0)||4 (9)|
|Bone marrow||20 (83)||16 (76)||30 (67)|
|Stem cell||4 (17)||5 (24)||14 (33)|
|Allogeneic||21 (88)||19 (90)||33 (73)|
|Autologous||3 (12)||2 (10)||12 (27)|
|CSP/methotrexate||19 (79)||14 (67)||28 (62)|
|Methotrexate||3 (13)||2 (9)||13 (9)|
|RFT5.dgA||0 (0)||4 (19)||4 (9)|
|BC-3||2 (8)||1 (5)||0 (0)|
|Total body irradiation|
|No||11 (46)||8 (38)||18 (40)|
|Yes||13 (54)||13 (62)||17 (60)|
The protocol and all study procedures were approved by the FHCRC Institutional Review Board. All study patients signed written informed consent to participate. Assessment was conducted prospectively from pretransplantation through Day 30 posttransplantation, with study consent obtained prior to the start of transplantation conditioning. Patients completed a comprehensive pretransplantation battery assessing functional status, affective and cognitive function, and delirium at the time of consent. Additional demographic and medical information was obtained from medical charts and computerized demographic records. At 7 days pretransplantation, during conditioning, and through Day 30 posttransplantation, trained research nurses and investigators (J.R.F. and S.R.-R.) assessed patients with a brief delirium, distress, and pain assessment battery on a schedule of three times per week targeted to the same time of day (Monday, Wednesday, and Friday). During the course of the study, if a psychiatric consultation was requested, then the patient received treatment according to standard practice at FHCRC. This treatment was recorded but did not affect participation in the study.
Delirium Rating Scale
The Delirium Rating Scale (DRS) is a 10-item, clinician-rated scale that rates the severity of delirium symptoms over a 24-hour period using all available information from the patient interview, mental status examination, medical history and tests, nursing observations, and family reports.26 The maximum possible DRS score is 32. It has been shown that a cut-off score >12 distinguishes patients with delirium from patients with other neuropsychiatric disorders.26, 27
Memorial Delirium Assessment Scale
The Memorial Delirium Assessment Scale (MDAS)16 is a 10-item, clinician-rated scale that assesses the severity of delirium and has been validated against a number of commonly used delirium assessment measures.28 It has been validated in cancer populations.22, 23 Items on the MDAS range from 0 to 3, depending on severity and frequency. On the few assessments for which items on the DRS or MDAS assessments could not be scored (4.1% of DRS assessments and 4.5% of MDAS assessments), mostly due to severe medical illness, the total score was computed as the prorated score of the completed items. Excellent interrater reliability was established in this study.1
Definition of delirium
To distinguish between delirium that was transient and potentially of less clinical significance and delirium that was prolonged and potentially of greater clinical significance, we defined a delirium event as a single DRS score > 12 and a delirium episode as a DRS score > 12 for at least 2 of 3 consecutive assessments. The resolution of the delirium episode occurred when there were at least 2 consecutive assessments with a DRS score ≤ 12. Delirium rating severity was determined using the MDAS.
Profile of Mood States
The Profile of Mood States (POMS) is a 30-item, abbreviated form of the original 65-item POMS,29, 30 which is a valid and reliable measure of distress and mood disturbance. The POMS has six subscales. We also calculated composite scores of affective distress (the sum of the Tension-Anxiety, Depression-Dejection, and Anger-Hostility scores) and fatigue less vigor (the Fatigue-Inertia score minus the Vigor-Activity score). The POMS was administered at each delirium assessment.
The Pain Score is a general verbal measure of degree of reported pain using a 0–10 numeric rating scale (0 = no pain; 10 = worst possible pain) and was administered at each delirium assessment.
Charlson Comorbidity Index
The Charlson Comorbidity Index31 rates the severity of comorbid medical illness based on weighted scores assigned to the presence of common chronic medical illnesses obtained from chart extraction pretransplantation and is associated with mortality in the subsequent year.
To describe and determine group differences on delirium, distress, and pain symptoms among delirium categories, the average item score means and percentile distributions were determined, and a log-linear regression on the delirium category (i.e., comparable to an analysis of variance) was used to obtain the multiple r2 and the P values. For the POMS and pain items, the regression provides a traditional test for association between these items and the delirium category. Because the DRS and MDAS are measures of delirium, they have a direct association with the delirium category; therefore, we did not report P values. For these item scores, the multiple r2 (ranging from 0 to 1) may be interpreted as a comparison of their relative influence on delirium determination. The r2 is analogous to a squared correlation coefficient that describes the strength of association between each average item score and the delirium categories.
We conducted a factor analysis of delirium symptoms combining the DRS items and the MDAS items to describe the major domains of delirium symptoms presented by patients who experienced delirium. The factor analysis was performed using maximum likelihood with a varimax rotation. Other rotations produced interpretively similar results. Scores for each delirium symptom (combined DRS and MDAS items) were averaged over each patient's delirium events or episodes (DRS > 12), so that the data for the factor analysis comprised a single average score per symptom for each patient. Of the 66 patients who had a DRS score > 12 (delirium episodes and events), 2 patients had missing data for all assessments for at least 1 item. Missing data were not imputed in any analyses presented here; thus, those two patients were excluded from the factor analysis.
To describe the time course of symptoms, delirium (DRS, MDAS), distress (POMS), and pain symptom scores were presented graphically for patients without evidence of delirium, patients with a delirium event only, and patients with a delirium episode. We used the LOESS smoothing function in the SPLUS software package (version 6.1, release 1; Insightful Corp., Seattle, WA) for ease of interpretation.
To investigate the comorbidity of delirium symptoms with distress and pain, we computed correlations between delirium factors and total scores, POMS subscale scores and composite scores, and pain scores. Patient averages across all assessments were used in the correlation calculations, yielding population-level correlations based on independent summaries.
Log-linear, multivariate, mixed models were employed to describe the magnitude (i.e., coefficients) and the strength (i.e., P values) of the longitudinal associations between the POMS affective distress and delirium factors and the POMS fatigue less vigor and delirium factors. Within-factor correlations were modeled with an autoregressive correlation structure of order 2. The P values may be interpreted as describing the relative predictive strength of the factor scores. All analyses were performed using SPLUS software.
Ninety patients who were admitted to FHCRC for HSCT consented to participate in the study over a 2-year period and completed between 5 and 13 posttransplantation assessments (mean ± standard deviation, 12.2 ± 2.0 assessments). Of 8 patients who dropped out due to severe illness between Assessments 5 and 8, 4 patients had not experienced a delirium episode. Patient characteristics stratified by delirium status are shown in Table 1. Details regarding delirium prevalence, incidence, duration, severity, and pretransplantation risk factors are reported elsewhere.1
Delirium, Distress, and Pain Symptoms
Table 2 presents the mean scores and the 25th, 50th, and 75th percentile distributions of delirium symptoms, as rated by the DRS and MDAS for those patients with no delirium, a delirium event, or a delirium episode. Mean scores and distributions of POMS subscales and pain scores are presented for comparison with delirium symptoms. Patients who experienced delirium episodes, as expected (and by definition), experienced more severe delirium symptoms compared with patients who experienced isolated delirium events. The most severe symptoms other than definitional items, such as temporal onset (abruptness of symptom onset), variability in symptoms (waxing and waning), and physical disorder, were sleep-wake cycle, psychomotor behavior, and cognitive disturbance. The occurrence of hallucinations and delusions was low. The clinical presentation of delirium was similar whether measured by the MDAS or the DRS. The relative uniformity of the percentile scores listed for most symptoms indicated narrow and skewed distributions of most variables, suggesting that the majority of patients experienced delirium symptoms similarly.
|Characteristic||Mean score (25th/50th/75th percentile)a||r2d|
|No delirium (n = 24)b||Delirium event (n = 21)bc||Delirium episode (n = 45)bc|
|Temporal onset||1.0 (0.7/1.2/1.4)||3.0 (3.0/3.0/3.0)||3.0 (3.0/3.0/3.0)||0.81|
|Perceptual disturbance||0.1 (0.0/0.0/0.0)||0.5 (0.0/0.0/1.0)||0.6 (0.0/0.7/1.0)||0.23|
|Hallucinations||0.0 (0.0/0.0/0.0)||0.1 (0.0/0.0/0.0)||0.2 (0.0/0.0/0.3)||0.10|
|Delusions||0.0 (0.0/0.0/0.0)||0.0 (0.0/0.0/0.0)||0.2 (0.0/0.0/0.0)||0.10|
|Psychomotor behavior||0.6 (0.4/0.6/0.8)||1.1 (1.0/1.0/1.0)||1.6 (1.3/1.8/2.0)||0.57|
|Cognitive status||0.4 (0.1/0.2/0.5)||1.2 (0.8/1.0/2.0)||1.4 (0.8/1.3/2.0)||0.22|
|Physical disorder||1.8 (1.8/2.0/2.0)||2.0 (2.0/2.0/2.0)||2.0 (2.0/2.0/2.0)||0.16|
|Sleep-wake cycle||1.2 (1.0/1.2/1.4)||1.7 (1.4/2.0/2.0)||1.9 (1.8/2.0/2.2)||0.37|
|Lability of mood||0.2 (0.1/0.2/0.3)||0.9 (0.9/1.0/1.0)||0.8 (0.5/0.8/1.0)||0.36|
|Variability of symptoms||0.2 (0.0/0.0/0.3)||3.8 (4.0/4.0/4.0)||3.7 (3.7/4.0/4.0)||0.93|
|Prorated total||5.5 (5.0/5.4/6.3)||14.2 (13/14/15)||15.4 (15/15/16)||0.83|
|Awareness||0.0 (0.0/0.0/0.0)||0.1 (0.0/0.0/0.0)||0.2 (0.0/0.0/0.3)||0.16|
|Disorientation||0.0 (0.0/0.0/0.0)||0.1 (0.0/0.0/0.0)||0.2 (0.0/0.0/0.4)||0.13|
|Short-term memory||0.2 (0.0/0.1/0.2)||0.5 (0.0/0.0/1.0)||0.6 (0.0/0.5/1.0)||0.10|
|Impaired digit span||0.3 (0.0/0.0/0.3)||0.5 (0.0/0.0/1.0)||0.8 (0.1/0.7/1.2)||0.13|
|Reduced attention||0.0 (0.0/0.0/0.0)||0.3 (0.0/0.0/0.0)||0.5 (0.2/0.5/0.8)||0.26|
|Disorganized thinking||0.0 (0.0/0.0/0.0)||0.0 (0.0/0.0/0.0)||0.3 (0.0/0.2/0.4)||0.24|
|Perceptual disturbance||0.1 (0.0/0.0/0.0)||0.4 (0.0/0.0/0.6)||0.6 (0.2/0.5/0.9)||0.24|
|Delusions||0.0 (0.0/0.0/0.0)||0.0 (0.0/0.0/0.0)||0.1 (0.0/0.0/0.0)||0.09|
|Psychomotor activity||0.6 (0.4/0.6/0.9)||1.2 (1.0/1.0/1.5)||1.7 (1.5/1.8/2.0)||0.56|
|Sleep-wake cycle||1.2 (1.0/1.2/1.4)||1.7 (1.4/2.0/2.0)||1.9 (1.8/2.0/2.0)||0.10|
|Prorated total||2.3 (1.8/2.0/2.9)||4.7 (3.4/4.0/5.4)||6.9 (4.9/6.8/8.4)||0.39|
|Profile of Mood States|
|Tension-Anxiety||0.4 (0.0/0.2/0.6)||0.5 (0.2/0.4/0.6)||0.9 (0.4/1.0/1.3)||0.17e|
|Depression-Dejection||0.1 (0.0/0.0/0.1)||0.2 (0.0/0.2/0.4)||0.6 (0.1/0.5/0.8)||0.21e|
|Anger-Hostility||0.1 (0.0/0.1/0.2)||0.2 (0.0/0.0/0.2)||0.5 (0.1/0.4/0.8)||0.20e|
|Vigor-Activity||0.9 (0.5/0.9/1.2)||0.7 (0.2/0.6/1.0)||0.6 (0.2/0.5/0.9)||0.04|
|Fatigue-Inertia||0.9 (0.5/0.7/1.2)||1.1 (0.6/0.9/1.3)||1.7 (1.0/1.7/2.4)||0.15e|
|Confusion-Bewilderment||0.3 (0.1/0.3/0.4)||0.3 (0.0/0.2/0.4)||0.7 (0.3/0.6/1.0)||0.21e|
|Affective distressf||0.6 (0.1/0.5/0.9)||0.9 (0.6/0.8/1.3)||1.9 (0.7/1.8/2.6)||0.22e|
|Fatigue less Vigorg||0.1 (−0.5/−0.2/0.7)||0.4 (−0.3/0.5/0.9)||1.1 (0.3/1.3/1.9)||0.10h|
|Pain score||2.2 (0.9/2.0/3.1)||2.7 (1.3/2.8/3.8)||5.1 (3.9/5.3/6.8)||0.26e|
Among those who experienced delirium events shown in Table 2, 18 patients (86%) with psychomotor disturbance were scored as hypoactive, and 3 patients (14%) were scored as mixed hypoactive/hyperactive. No patients were scored hyperactive, and only two patients had no psychomotor disturbance. Among the patients with delirium episodes, 162 assessments (86%) in patients with a psychomotor disturbance were scored as hypoactive, 5 assessments (3%) were scored as hyperactive, and 22 assessments (12%) were scored as mixed. Only four assessments showed no psychomotor disturbance. Within the delirium episode group, 59% of patients scored as hypoactive at all assessments with DRS scores > 12.
All POMS subscale and pain scores differed significantly between the delirium groups, with the exception of the POMS Vigor-Activity score. All POMS subscale scores were higher (lower, in the case of Vigor) in patients who experienced a delirium episode compared with patients who experienced delirium events or who had no delirium. Similarly, affective distress and fatigue less vigor composite scores were higher in patients who had delirium episodes. Scores on most POMS subscales generally were similar for the delirium event group and the no delirium group. Pain scores were much higher for patients who had delirium episodes compared with patients who had no delirium or who had delirium events only.
Delirium Factor Structure
Table 3 presents the results of a factor analysis of the combined DRS and MDAS items for the 64 patients who experienced a delirium event or episode and had complete data for analysis, comprising 216 assessments. The factor analysis revealed a 3-factor structure that accounted for 44% of the variance in symptoms. The P value for the test of the hypothesis that the 3-factor solution was sufficient was 0.01: The addition of a fourth factor merely refined the third factor reported here, contrasting lability (negative loading) with perceptual disturbances (positive loading). This structure indicated the presence of three main clusters of delirium symptoms in these patients. The first symptom cluster was psychosis-behavior (perceptual disturbance, hallucinations, delusions, psychomotor disturbance, reduced attention, and disorganized thinking) and accounted for 18% of the variance in the data. The second symptom cluster represented cognitive symptoms (cognitive status, disorientation, short-term memory, and impaired digit span) and accounted for an additional 16% of the variance in the data. The third symptom cluster represented mood-consciousness symptoms (sleep-wake cycle, lability of mood, variability of symptoms, and awareness) and accounted for a final 10% of the variance. Overall, these three symptom clusters generally were differentiated well (i.e., items within a factor loaded highly on that factor and did not load highly on other factors).
|Factor description||Itemb||Factor 1||Factor 2||Factor 3|
|Perceptual disturbance||DRS2 + MDAS7||0.63||0.25|
|Delusions||DRS4 + MDAS8||0.60||0.19|
|Psychomotor disturbance||DRS5 + MDAS9||0.52||0.11|
|Impaired digit span||MDAS4||0.71||0.13|
|Sleep-wake cycle||DRS8 + MDAS10||0.53|
|Lability of mood||DRS9||0.12||−0.21||−0.29|
|Variability of symptoms||DRS10||−0.48|
Time Course of Delirium Symptoms
Figure 1 shows the time course of the three symptom cluster factor scores as a function of days since stem cell transplantation for patients who had no delirium, patients who had isolated delirium events, and patients who had a delirium episode. The symptom clusters showed the same patterns in all three patient groups, in that psychosis-behavior and cognitive symptoms tracked together with similar intensity, and mood-consciousness symptoms tracked with the others but with much lower scores. The degree of elevation in this pattern distinguished the delirium episode group, the delirium event group, and the no delirium group. The three symptom clusters were relatively stable in the no delirium group. There was a small peak of all 3 symptom clusters in the delirium events group (peak, around 9–12 days posttransplantation) with a rapid remission of symptoms. There was a more dramatic rise and a gradual fall from a higher peak of symptoms in the delirium episode group, and symptoms generally returned to baseline by Day 30.
Figure 2 shows the time course of symptoms in each symptom cluster for patients who had a delirium episode, aligned to the start of their delirium episode. Symptoms of psychosis-behavior (Factor 1) began several days prior to the start of a delirium episode, were constant from the start of the episode through 7–9 days, then generally resolved over time. An exception was psychomotor behavior, which remained elevated throughout. Cognitive symptoms (Factor 2) began several days prior to a delirium episode and continued to climb in severity for 7–10 days before resolving. Variability in symptoms began to rise prior to the start of the delirium episode and peaked, by definition, at the start of the episode. With this exception, the symptoms of mood-consciousness (Factor 3) tended to be relatively stable throughout a delirium episode. Disturbance of the sleep-wake cycle remained elevated throughout.
Comorbidity of Delirium with Distress and Pain
Figure 3 shows the time course of delirium severity (MDAS total) with POMS affective distress and fatigue less vigor composite and pain scores for patients in the no delirium group, the delirium event group, and the delirium episode group. Both POMS affective distress scores and fatigue less vigor scores were elevated and tracked with delirium scores in patients who had delirium episodes relative to patients who had delirium events or no delirium. POMS affective distress scores and fatigue less vigor scores were constant over time in patients who had no delirium or who had delirium events and were slightly elevated during peak delirium in patients who had delirium episodes. These scores also remained elevated over time in the delirium episode group versus the no delirium group and the delirium event group.
Pain severity tracked with delirium severity for all 3 patients groups, but the pattern was shifted leftward by about 3 days. Pain scores were similar for patients who had no delirium and patients who had delirium events and were characterized by a distinct peak (Day 9 postinfusion) followed by a resolution of pain to below the baseline level, then a slight rise. In the delirium episode group, the peaks of delirium and pain were much higher compared with the peaks in the other two groups.
Figure 4 shows the time course of the six POMS subscales throughout a delirium episode for patients who had a delirium episode. Fatigue and vigor, as expected, tracked inversely with one another. Aside from vigor, POMS scores tended to rise prior to the onset of a delirium episode, peaked around Days 3–5, then declined to baseline levels approximately 12 days after onset of the delirium episode. Fatigue was characterized further by a second relative peak toward the end of the episode.
Table 4 is a correlation matrix of delirium factors and DRS and MDAS scores with POMS and pain scores for the entire sample. Overall, the POMS subscales (with the exception of Vigor) and composite scores were correlated moderately with delirium symptoms (DRS and MDAS total scores). Symptoms of psychosis-behavior (Factor 1) were correlated most with the POMS confusion-bewilderment scores and had stronger correlations with mood symptoms, fatigue, and with the POMS affective distress composite score than the other symptom clusters. Pain scores were correlated moderately with symptoms of psychosis-behavior (Factor 1), cognitive symptoms (factor 2), and with the DRS and MDAS total delirium scores.
|Subscale||DRS/MDAS Factor||POMS||DRS total||MDAS total|
|POMS Affective Distress||0.43||0.32||0.08||1.00||0.41||0.48||0.39|
|POMS Fatigue less Vigor||0.15||0.15||0.04||0.41||1.00||0.41||0.33|
Table 5 shows a regression analysis for delirium symptom and symptom change predictors of POMS affective distress. The regression adjusts for age, gender, Charlson comorbidity score, and the previous assessment's POMS affective distress score. The results from unadjusted regressions were similar. Other available disease and treatment adjustment variables (see Table 1) did not affect the coefficients or P values markedly. For example, a 1-unit increase in psychosis-behavior symptoms from the previous delirium assessment was associated significantly with a 17% increase in affective distress scores, whereas a 1-unit increase in mood-consciousness symptoms since the previous delirium assessment was associated with a 6% decrease in affective distress scores. We found a similar significant association between psychosis-behavior symptoms and fatigue less vigor scores in a separate regression analysis (not shown).
|Variable||Fold difference (95%CI)a||P value|
|Previous Factor 1 score||1.17 (1.11–1.23)||< 0.001|
|Change in Factor 1 score||1.17 (1.13–1.21)||< 0.001|
|Previous Factor 2 score||1.02 (0.98–1.06)||0.3|
|Change in Factor 2 score||1.00 (0.98–1.03)||1.0|
|Previous Factor 3 score||0.97 (0.91–1.04)||0.4|
|Change in Factor 3 score||0.94 (0.90–0.99)||0.01|
|Previous POMS Affective Distress score||1.06 (1.02–1.09)||0.001|
|Charlson comorbidity score||1.01 (0.93–1.10)||0.8|
This prospective study is the first to examine in detail the clinical presentation and course of psychological, behavioral, and cognitive symptoms associated with delirium in the setting of HSCT. The most prominent delirium symptoms that distinguish delirious patients from nondelirious patients include sleep-wake cycle, psychomotor behavior, and cognitive disturbance. Hallucinations and delusions were rare. Fatigue, affective distress, and pain were more pronounced among patients with delirium. Generally, there was a “dose-response” relation between delirium severity (i.e., no delirium, delirium event, delirium episode) and the levels of affective distress, fatigue, and pain.
Almost all delirium events and episodes exhibited a psychomotor disturbance; among these, 86% of assessments revealed a hypoactive subtype. We were able to look more specifically at delirium symptoms and found that symptoms were separated into three distinct primary factors—psychosis-behavior, cognition, and mood-consciousness—suggesting three distinct symptom domains present in delirious patients undergoing stem cell transplantation that differ from the domains found in previous studies in more homogeneous cancer populations,22, 23 confirming the unique clinical presentation of delirium in different clinical settings. There is some consistency, however, that cognitive and perceptual symptoms are common symptom domains.
The examination of the time course of specific symptoms of delirium and distress among patients with delirium, compared with patients without delirium, allows clinicians to more readily recognize common symptom constellations that may alert them to potential delirium. In patients who experience a delirium episode, delirium symptoms tend to peak near the end of the second week posttransplantation. However, several prodromal symptoms begin to emerge prior to the start of a delirium episode. In particular, attention, perceptual disturbance, cognition, and evidence of variability of symptoms appear to rise precipitously about 4 days prior to the onset of an episode. Psychomotor disturbances remained high throughout most assessments in the current study. Deficits in cognitive status, particularly digit span (a test of working memory), did not peak until 1 week into a delirium episode.
These findings suggest that clinicians should not rely on the presence of “classic symptoms” of delirium, such as agitation, hallucinations, delusions, or disorientation, in identifying HSCT patients who are delirious. In fact, more subtle symptoms of hypoactive psychomotor behavior, altered sleep-wake cycle, and impaired attention and working memory should be early indicators for increased monitoring and identification of possible causes of delirium.
The current study data show that delirium, particularly psychosis-behavioral symptoms (Factor 1), in the HSCT setting is associated with significant affective distress and fatigue symptoms, consistent with the findings of Breitbart et al. of significant distress among a diverse cohort of cancer patients with delirium, including patients with hypoactive delirium.9 Our findings confirm that delirium is associated with increased distress and suffering in the HSCT population. This is particularly important, because the vast majority of delirium in this setting is “hypoactive”; thus, patients exhibit fewer overt signs and symptoms that may alert the clinician. Conversely, the presence of fatigue and affective distress (e.g., depression, anxiety, anger) also should alert the clinician to the possible presence of delirium. The significant comorbidity of these symptoms illustrates why misdiagnosis is common, and screening for delirium in distressed patients is warranted in the acute transplantation setting. Further work is needed to examine specific determinants of these symptom constellations.
The time courses of fatigue, affective distress, and pain posttransplantation were similar in the no delirium group and the delirium event group. However, fatigue and affective distress rose and fell concomitantly with delirium severity. A rise in pain severity seemed to predate delirium severity, suggesting the need to examine pain and its treatment (e.g., opioid medications) as potential longitudinal risk factors for delirium in this population.
Among patients who had a delirium episode, the POMS distress measures peaked around the middle of the first week of a delirium episode. Similar to delirium symptoms, distress symptoms began to rise several days prior to the start of a delirium episode. Fatigue appeared to increase again about 3 weeks after the start of the episode, perhaps due to a discord between a patient's heightened expectation and desire for activity and his continued poor physical functioning.
The psychosis-behavior factor was correlated more strongly than the cognitive or mood-consciousness factors with the POMS affective distress scores. This suggests that many delirious patients may have some insight into their inability to manage their thoughts and behavior, leading to significant distress. Examining this association in greater detail, the regression analysis also showed a significant association between psychosis-behavior delirium symptoms and increased affective distress.
Some of the limitations of the current study are discussed elsewhere1 and include delirium status misclassification due to lack of daily assessments, confounding due to treatment effects, and information bias from missed assessments due to severe medical illness. Furthermore, responses to affective assessments may have been subject to response shifts, whereby delirium status may influence ability to self-report symptoms. Some of the etiologic factors that contribute to delirium, such as medications, anemia, or infection, also may have contributed to the presentation of affective and fatigue symptoms.
These results will assist clinicians working with HSCT patients to identify the presence of delirium and symptom clusters common to delirium in this setting and to educate patients and families about the course of delirium. With a greater ability to identify delirium and differentiate it from other psychiatric disorders (e.g., depressive and anxiety disorders, dementia), clinicians will be equipped better to provide appropriate treatment.
The consistent finding that the longitudinal course of delirium, distress, and pain symptoms is similar among patients with no delirium and patients with delirium events and is elevated markedly in patients with delirium episodes lends validity to the delirium event versus delirium episode distinction, which distinguishes between delirium that is transient and potentially of less clinical significance compared with delirium that is more prolonged and potentially of greater clinical significance. Future research will need to examine specific longitudinal risk factors (including pain and its treatment) for and outcomes of delirium and specific delirium subtypes. Because most of HSCT patients with delirium are hypoactive, studies in this population will be needed to assess the efficacy of different screening and treatment approaches. For example, certain treatment approaches may target specific symptoms of delirium selectively. Epidemiological and treatment studies of delirium should examine the impact of delirium and treatment on distress, in addition to functioning and mortality, as an important outcome variable.
The authors thank Kathy Beach, R.N., and Wendy Brown, R.N., for their invaluable assistance in performing the study.
- 7Acute confusional states in the hospitalized elderly: incidence, risk factors, and complications. Clin Res. 1989; 37: 524A., , .
- 29EdITS manual for the Profile of Mood States, revised 1992. San Diego: Educational and Industrial Testing Service, 1992., , .