Acute Care of the Elderly Unit

	ACE Home \| Contact Us \| Center for Aging and Health School of Medicine \| UNC Home \|
Mission Goals Meet the Team CQI Assessment Tools Educational Resources Unit Orientation and Expectations Admitting Clinical Practice
Critically Appraised Topic: Are geriatric patients with depression at increased risk of developing delirium when admitted to an acute care hospital? Appraised by: Tiffany L. Reed, D.O., PGY-2, Internal Medicine, The Reading Hospital & Medical Center Date: August 24, 2007 Why did I choose to examine this question? During inpatient geriatrics, we had seen two patients with a known diagnosis of depression develop delirium early in the hospital stay. I wondered if depression alone could increase a patient’s risk of experiencing delirium in the acute care setting. This is cause for concern, of course, because delirium is associated with negative outcomes including prolonged hospital stays, increased comorbid illness, increased rate of long-term care placement, and increased mortality. Upon reviewing the literature, many studies were conflicting as to the relationship between depression and delirium. As pointed out on the first page of the selected article, 8 studies found an association, and 8 did not. Reference: McAvay GJ, VanNess PH, Bogardus ST, et al. Depressive symptoms and the risk of incident delirium in older hospitalized adults. J Am Geriatr Soc 2007; 55: 684-691. Patients: There were 416 patients aged 70 and older who were at intermediate to high risk of developing delirium and were not taking antidepressants at the time of admission to an acute care teaching hospital. They were members of the control group from the Delirium Prevention Trial. Baseline patients’ characteristics are shown in Table 1, p. 687. The patients’ comorbid illnesses, medication use, laboratory results, and surgeries are displayed in Table 2, p. 688. Type of study: Secondary analysis of a prospective cohort study of patients from the Delirium Prevention Trial. Clinical Bottom Lines: 1. Patients over 70 years of age who expressed depressive symptoms including dysphoric mood and hopelessness (more psychological symptoms) as assessed by the Geriatric Depression Scale (GDS) were at greater risk of developing incident delirium (as identified with the Confusion Assessment Method) during acute hospitalization. 2. Those patients with withdrawal-apathy-vigor symptoms of depression (more somatic symptoms) were not at increased risk for delirium during hospitalization. My patients: Patients 70+ with a known diagnosis of depression admitted to UNC acute care teaching hospital. One patient’s antidepressants had been held due to sedation. The other patient had been continued on her antidepressant at admission. In this study, the patients did not have a known diagnosis of depression. They were screened using the Geriatric Depression Scale to assess for dysphoric mood symptoms and hopelessness symptoms (psychological aspects of depression that had previously been correlated with postoperative delirium following vascular surgery in a study that used the Hamilton Scale). Cohort study: The cohort study takes a large population and follows patients with a specific condition over time and compares them with another group that has not been affected by that condition. Disadvantages: Observational and not always as reliable as randomized, controlled studies because of potential differences between the 2 groups. However, when assessing prognosis, cohort studies are said to be better than case control or case series studies. Were patients accounted for at the conclusion of the study? Yes. There were 426 matched patients in the initial control group from the Delirium Prevention Trial. 43 additional unmatched patients were also included for a total of 526 eligible patients. 469 (82.7%) agreed to participate. 8 did not survive the hospitalization (excluded), leaving 461 patients. 43 of them were taking antidepressants at the time of admission (excluded). 2 had missed 3 or more questions on the Geriatric Depression Scale (excluded). 416 patients were included in the final study sample. Also excluded from the group were those with profound dementia, aphasia, or intubation unable to participate in interviews. Were patients, clinicians, and study personnel blinded? Interviewers performing the Geriatric Depression Scale, Mini-Mental Status Examination, and Confusion Assessment Method were blinded to the study aims and hypothesis. A computerized statistical analysis was performed using SAS. It was not clear whether the patients were blinded; however, informed consent was performed according to the Yale Institutional Review Board standards. The purpose of the study was to determine whether the psychological symptoms of depression from the Geriatric Depression Scale, assessed at hospital admission, were associated with the incidence of delirium. Some of the previous studies that did not find such a correlation, looked at the somatic symptoms such as withdrawal, apathy, and vigor (fatigue, insomnia) that can be seen in older adults, particularly the acutely ill, for other reasons (comorbid illness). Patients were assessed at admission and daily for delirium with the Confusion Assessment Method. They were evaluated for psychological symptoms of depression including dysphoric mood and hopelessness using the GDS-15 (on the last page) at the screening interview. The MMSE was administered at admission and during daily interviews, and the medical record was reviewed for a diagnosis of dementia. Functional status was evaluated with ADL and IADL scales. Visual and hearing impairment were tested. The Charlson Comorbidity Index (burden of chronic comorbid conditions) and the APACHE II score (risk index for mortality in patients in ICU to predict LOS) were determined using the medical record. Meds taken before admission and BUN:creatinine ratio were determined. Stats in Tables 1–3 . Analyzed delirium only in first 5 days to decrease effect of hospitalization events. Cox proportional hazards regression models were used. Medical comorbidity was taken into account. Results: Delirium: 36 (8.6%)/416 patients developed delirium within 5 days of hospitalization. They were older, less were Caucasian, lower education level, poorer cognitive functioning, more had dementia, more had functional impairments, had more severe illness, higher BUN/cr ratio, same amt of surgery. Dementia: Of those with delirium, they had 5.7 depressive symptoms at baseline vs 4.2 symptoms without delirium. Delirious patients had higher total depression scores (table 3), more dysphoric mood symptoms, and more hopelessness symptoms. Withdrawal-apathy-vigor symptoms did not differ. Number of mood symptoms was higher in patients taking anticonvulsants, anxiolytics, sedatives, antipsychotics, hypnotics (specifically hopelessness). Covariates were controlled for (last paragraph p 687). Minimal degree of confounding as more covariates were added. Meds also included without prediction of delirium. Strengths: Medical patients, p. 689 second-to-last paragraph. Weaknesses: Psych dx of depression not used. Conclusions: See first page of this. Determine correlation b/w depression and delirium, screen for depression to keep us more vigilant (looking for delirium) when these patients are admitted to the hospital.

Critically Appraised Topic:

Are geriatric patients with depression at increased risk of developing delirium when admitted to an acute care hospital?

Appraised by: Tiffany L. Reed, D.O., PGY-2, Internal Medicine, The Reading Hospital & Medical Center

Date: August 24, 2007

Why did I choose to examine this question?

During inpatient geriatrics, we had seen two patients with a known diagnosis of depression develop delirium early in the hospital stay. I wondered if depression alone could increase a patient’s risk of experiencing delirium in the acute care setting. This is cause for concern, of course, because delirium is associated with negative outcomes including prolonged hospital stays, increased comorbid illness, increased rate of long-term care placement, and increased mortality. Upon reviewing the literature, many studies were conflicting as to the relationship between depression and delirium. As pointed out on the first page of the selected article, 8 studies found an association, and 8 did not.

Reference: McAvay GJ, VanNess PH, Bogardus ST, et al. Depressive symptoms and the risk of incident delirium in older hospitalized adults. J Am Geriatr Soc 2007; 55: 684-691.

Patients: There were 416 patients aged 70 and older who were at intermediate to high risk of developing delirium and were not taking antidepressants at the time of admission to an acute care teaching hospital. They were members of the control group from the Delirium Prevention Trial. Baseline patients’ characteristics are shown in Table 1, p. 687. The patients’ comorbid illnesses, medication use, laboratory results, and surgeries are displayed in Table 2, p. 688.

Type of study: Secondary analysis of a prospective cohort study of patients from the Delirium Prevention Trial.

Clinical Bottom Lines:

1. Patients over 70 years of age who expressed depressive symptoms including dysphoric mood and hopelessness (more psychological symptoms) as assessed by the Geriatric Depression Scale (GDS) were at greater risk of developing incident delirium (as identified with the Confusion Assessment Method) during acute hospitalization.

2. Those patients with withdrawal-apathy-vigor symptoms of depression (more somatic symptoms) were not at increased risk for delirium during hospitalization.

My patients: Patients 70+ with a known diagnosis of depression admitted to UNC acute care teaching hospital. One patient’s antidepressants had been held due to sedation. The other patient had been continued on her antidepressant at admission.

In this study, the patients did not have a known diagnosis of depression. They were screened using the Geriatric Depression Scale to assess for dysphoric mood symptoms and hopelessness symptoms (psychological aspects of depression that had previously been correlated with postoperative delirium following vascular surgery in a study that used the Hamilton Scale).

Cohort study: The cohort study takes a large population and follows patients with a specific condition over time and compares them with another group that has not been affected by that condition.

Disadvantages: Observational and not always as reliable as randomized, controlled studies because of potential differences between the 2 groups. However, when assessing prognosis, cohort studies are said to be better than case control or case series studies.

Were patients accounted for at the conclusion of the study? Yes.

There were 426 matched patients in the initial control group from the Delirium Prevention Trial.

43 additional unmatched patients were also included for a total of 526 eligible patients.

469 (82.7%) agreed to participate.

8 did not survive the hospitalization (excluded), leaving 461 patients.

43 of them were taking antidepressants at the time of admission (excluded).

2 had missed 3 or more questions on the Geriatric Depression Scale (excluded).

416 patients were included in the final study sample.

Also excluded from the group were those with profound dementia, aphasia, or intubation unable to participate in interviews.

Were patients, clinicians, and study personnel blinded?

Interviewers performing the Geriatric Depression Scale, Mini-Mental Status Examination, and Confusion Assessment Method were blinded to the study aims and hypothesis. A computerized statistical analysis was performed using SAS. It was not clear whether the patients were blinded; however, informed consent was performed according to the Yale Institutional Review Board standards.

The purpose of the study was to determine whether the psychological symptoms of depression from the Geriatric Depression Scale, assessed at hospital admission, were associated with the incidence of delirium. Some of the previous studies that did not find such a correlation, looked at the somatic symptoms such as withdrawal, apathy, and vigor (fatigue, insomnia) that can be seen in older adults, particularly the acutely ill, for other reasons (comorbid illness).

Patients were assessed at admission and daily for delirium with the Confusion Assessment Method. They were evaluated for psychological symptoms of depression including dysphoric mood and hopelessness using the GDS-15 (on the last page) at the screening interview. The MMSE was administered at admission and during daily interviews, and the medical record was reviewed for a diagnosis of dementia. Functional status was evaluated with ADL and IADL scales. Visual and hearing impairment were tested. The Charlson Comorbidity Index (burden of chronic comorbid conditions) and the APACHE II score (risk index for mortality in patients in ICU to predict LOS) were determined using the medical record. Meds taken before admission and BUN:creatinine ratio were determined.

Stats in Tables 1–3 . Analyzed delirium only in first 5 days to decrease effect of hospitalization events. Cox proportional hazards regression models were used. Medical comorbidity was taken into account.

Results:

Delirium: 36 (8.6%)/416 patients developed delirium within 5 days of hospitalization. They were older, less were Caucasian, lower education level, poorer cognitive functioning, more had dementia, more had functional impairments, had more severe illness, higher BUN/cr ratio, same amt of surgery.

Dementia: Of those with delirium, they had 5.7 depressive symptoms at baseline vs 4.2 symptoms without delirium.

Delirious patients had higher total depression scores (table 3), more dysphoric mood symptoms, and more hopelessness symptoms. Withdrawal-apathy-vigor symptoms did not differ.

Number of mood symptoms was higher in patients taking anticonvulsants, anxiolytics, sedatives, antipsychotics, hypnotics (specifically hopelessness).

Covariates were controlled for (last paragraph p 687). Minimal degree of confounding as more covariates were added. Meds also included without prediction of delirium.

Strengths: Medical patients, p. 689 second-to-last paragraph.

Weaknesses: Psych dx of depression not used.

Conclusions: See first page of this. Determine correlation b/w depression and delirium, screen for depression to keep us more vigilant (looking for delirium) when these patients are admitted to the hospital.