Abstract

Objective. To assess the effects of delirium on the institutionalization rate, taking into account geriatric syndromes and nutritional status.

Methods. This population-based study took place in an acute care unit and included participants older than 75 years, arriving from home and later discharged. Confusion Assessment Method (CAM) symptoms were recorded by the nurses within 24 hours after admission and every 3 days. Delirium was defined using the CAM algorithm, and subsyndromal delirium responded to symptoms not fulfilling the CAM algorithm. These delirium categories were either present at admission (prevalent) or occurred during the hospital stay (incident). Participants were classified as having a low dietary intake when energy intake was at any time lower than 600 kcal/d. Age, sex, known cognitive impairment, weight, functional dependency, and laboratory testing as well as diagnoses were also recorded. Step-by-step backward logistic regression was used to identify predictors of institutionalization.

Results. Among 427 patients, 310 (72.6%) were discharged and were compared with 117 (27.4%) participants admitted to an institution. Female sex (odds ratio [OR]: OR 2.15, 95% confidence interval [CI]: CI 1.22–3.78), prevalent delirium (OR 3.19, 95% CI 1.33–7.64), subsyndromal delirium (OR 2.72, 95% CI 1.48–5.01), incident subsyndromal delirium (OR 4.27, 95% CI 2.17–8.39), low dietary intake (OR 2.50, 95% CI 1.35–4.63), and a fall (OR 2.16, 95% CI 1.22–3.84) or a diagnosis of stroke (OR 2.03, 95% CI 1.04–3.94) were independent predictors of institutionalization.

Conclusions. Symptoms of delirium and severe nutritional impairment led patients to geriatric institutions. Therefore, these institutions need to implement policies that address both of these issues.

DELIRIUM, defined as “an etiologically non-specific organic cerebral syndrome, characterized by concurrent disturbances of consciousness and attention, perception, thinking, memory, psychomotor behavior, emotion and the sleep-wake cycle” in the classification of mental disorders (1) is a frequent feature in older acutely ill hospitalized patients. The rate of delirium at admission in acute care units seems to range between 10.4% to 41% for patients aged older than 65 years (2–7). This rate is higher in surgery patients than in medical ones (6). Furthermore, incident delirium is even more frequent, ranging from 24% to 31.3% (4,5). Both this high frequency and the documented adverse outcomes associated with this disorder have led to well-designed preventive studies. For example, Inouye and colleagues showed that it was possible to decrease the rate of delirium from 15.0% to 9.9% by using interventions targeting the predisposing factors for delirium (8). However, this rate is still clinically significant in terms of adverse outcomes.

To date, the adverse outcomes documented are higher in-hospital mortality and subsequent mortality (9), increased incidence of dementia (10), decrease in functional autonomy (11), and higher rate of institutionalization.

Frailty seems to be an important risk factor for delirium as well as previous cognitive impairment, severity of the disease, polypharmacy, and particularly psychoactive drugs and opiates (12). Undernutrition may be related to institutionalization, and frailty is understandably a risk factor for institutionalization. Therefore, the risk of institutionalization of patients suffering from prevalent or incident delirium associated with these risk factors needs to be further investigated.

Routine assessment of delirium requires a screening tool that is quick and has good sensitivity and specificity. Inouye and colleagues (13) developed such a test, the Confusion Assessment Method (CAM), with an external rating time of 5 minutes. With this tool, comparable sensitivity and specificity can be obtained by both nurses and clinicians (14). Another advantage of the CAM is that it is an external test that does not require the direct participation of patients and can be applied to all patients including those unable or unwilling to participate. Furthermore, the constituent symptoms of the CAM can be analyzed one by one when the diagnostic algorithm does not make it possible to establish the diagnosis of delirium. In fact, patients suffering from 1 or more symptoms of delirium seem to be twice as numerous as those with the CAM-defined syndrome (15). Such patients might share some of the risks of poor outcome with those who have the full syndrome. Thus, routine use of the CAM by nursing staff makes it possible to monitor a large cohort of older patients during their hospital stay.

We aimed to assess the impact of delirium and symptoms of delirium occurring in older patients hospitalized in an acute care geriatric unit having a risk of institutionalization, taking into account other components of frailty. For this purpose, we set up an observational prospective study using each 3-day CAM assessment by nurses and prospective chart data abstraction including geriatric syndromes and energy intake.

Methods

Study Population

All patients older than 75 years admitted between July 2000 and June 2001 to our 80-bed acute care geriatric unit were prospectively included in this observational study. Patients whose stay was shorter than 3 days were excluded from the analyses. For each individual, the first hospital stay occurring during the survey period was considered. Patients usually living in an institution and those deceased before discharge were also excluded from the analysis.

The hospital ethical committee approved the protocol.

Delirium Assessment Procedure

Nurses were instructed to record CAM symptoms within 24 hours following admission and then every 3 days during the hospital stay. The CAM algorithm was applied to define delirium. Diagnosis of delirium resulting from this procedure within the first 4 days of the stay was termed “prevalent delirium.” Among patients free of delirium at admission, subsequent delirium was termed “incident delirium.” Patients having 1 or more CAM symptoms but not fulfilling the CAM algorithm were considered as having “subsyndromal delirium” (15). They were classified as having either prevalent subsyndromal delirium or incident subsyndromal delirium, defined in the same manner as delirium. When subsyndromal delirium was followed by delirium, the worst state (delirium) for each participant was considered for analysis.

Description of Patients

Descriptive data were obtained after prospective chart abstraction. Age and sex were recorded for all patients, and any previously known cognitive impairment was systematically sought with the help of the family practitioner and the family. The following questions were asked about the patient's usual state: “Is the patient diagnosed as having dementia,” “Does the patient display memory impairment?” “Does the patient exhibit difficulties in executing tasks such as administering the household budget, taking public transportation, using the phone independently, or taking care of his or her home?” and “Does the patient have any difficulty in recognizing those close to him?”. The authors then stated if a cognitive impairment existed with respect to DSM-IV criteria (1).

Diagnoses were recorded and were used to calculate the Charlson index reflecting illness severity (16). For the purpose of describing frailty, the diagnoses of depression, diabetes mellitus, stroke, Parkinson's disease, and falls (17) were considered as well as known visual impairment. Medications taken prior to admission and at discharge were also considered, including psychoactive drugs, opiates, and the total number of medications. The Katz Activities of Daily Living (ADL) scale (18) was used to assess the level of functional dependency of the participants. This scale scores dependency for 6 activities from 0 to 2: bathing, eating, dressing, continence, transfer to toilets, and locomotion. A score of 2 reflected the maximal level of dependency for each. ADL was scored by nurses at the moment of admission but not at discharge.

Nutritional assessment in this unit included weight measurement at admission and continuous visual monitoring of dietary intake (19). At the end of each meal, nurses estimated the volume ingested for each food category in units of quarters (from 0 to 4 quarters). The exact amount of calories was not calculated, but assuming that the daily calorie content of the tray given to the patients was 1800 kcal per day plus nutritional supplement, the participants were classified according to their nutritional behavior in 3 categories: 1) acceptable intake: more than 1200 kcal per day; 2) impaired intake: between 600 and 1200 kcal per day, and 3) low intake: less than 600 kcal/day or artificial feeding. When patients displayed changes in nutritional categories during their stay, the worst state was considered for analysis.

Routine laboratory tests included hemoglobin count (g/100 mL), creatinemia (μmol/L), natremia (mmol/L), serum albumin (normal range 35–45 g/L), and C-reactive protein (normal range <5 mg/L). The latter two were determined by latex immunonephelometry (BNA, Behring, Rueil-Malmaison, France). C-reactive protein was used to assess the intensity of the inflammatory process.

Outcome

Admission to a geriatric institution (nursing home or residential care home) at discharge was considered here for analysis. The group of patients transferred to a rehabilitation unit was also depicted. Although the rehabilitation staff is used to selecting such patients capable of returning home, the final outcome for these patients was not recorded.

Analysis

Statistical analysis was performed by using SAS system software (version 8.2, SAS Institute, Inc., Cary, NC). Values were expressed as means (± standard deviation [SD]) or as percentages. Patients institutionalized at discharge were compared with others for descriptive variables such as age, sex, previous cognitive impairment, delirium categories, diagnoses, medications, Charlson's index score, ADL score at admission, weight, dietary intake categories, and biochemical measures, using the chi-square test or the exact Fisher test in case of expected frequencies below 5. The Student t test was used for numerical variables or the Kruskal-Wallis test for nonnormally distributed parameters. This was completed by a step-by-step backward logistic regression analysis including age, sex, previously known cognitive impairment, and significant risk factors for institutionalization revealed by univariate analysis (p <.1). This allowed estimation of adjusted odds ratio (OR) and 95% confidence intervals (95% CI) associated with the risk of institutionalization.

Results

During the survey period, 928 hospital stays occurred for 847 patients. In this population, 222 came from geriatric institutions, 65 from rehabilitation units or were hospitalized in another unit, and the origin was unknown in 1 participant. Thus, 559 participants came from the community. Among them, 55 patients died, 52 were discharged to different post-acute care units, and 25 were transferred to other specialty departments. These 77 transferred patients were not different from the other surviving patients in terms of age, sex ratio, and delirium diagnosis. Thus, the study population consisted of 427 patients, of which 310 (72.6%) were discharged and 117 (27.4%) were admitted to a geriatric institution. The mean hospital stay of the former was shorter (15.8 ± 9.4 days) than the latter (25.9 ± 13.5 days, p <.0001) due to the usually long duration of the institutionalization procedure. Admission characteristics of the study population are shown in Table 1. Women were more likely to be institutionalized than men (p <.01). Age was unrelated to the rate of institutional admission. After univariate analysis, previously known cognitive impairment (p =.02), higher ADL functional dependency score (p =.02), lower weight (p <.001), and lower serum albumin concentration at admission (p =.03) were associated with institutionalization. When considering the Katz scale items one by one, patients discharged to institutions were more often dependent for bathing (p =.02) and dressing (p =.02). There was no significant relationship between institutionalization risk and treatment at admission (Table 1). The number of medications taken at discharge was 5.3 ± 2.3 in patients discharged to the community compared with 5.4 ± 2.5 in those admitted to geriatric institutions (p =.60). The rate of patients under opiates at discharge was similar in both groups (respectively, 13.7% compared with 13.9%, p =.95). A trend for more patients under psychoactive drugs at discharge was observed in participants discharged to geriatric institutions (69.6% in these patients compared with 59.1% in those discharged to the community, p =.06).

Among this sample, only 230 participants (53.9%) remained free of delirium symptoms during their hospital stay. Full-syndromal delirium occurred in 49 participants (11.5%), 34 had prevalent delirium, and 15 had incident delirium. In contrast, subsyndromal delirium was more frequent, be it prevalent or incident (148 patients, 34.7%). The distribution of delirium categories was different according to the mode of discharge, with higher rates of delirium or subsyndromal delirium in patients discharged to institutions (Table 2; p <.0001). Patients discharged to geriatric institutions had been in lower dietary intake groups during their hospital stay than those returning home (Table 3; p <.01).

Table 4 shows the results of the final model of the step-by-step backward logistic regression predicting the likelihood of admission to a geriatric institute for these patients. Female sex, prevalent delirium and subsyndromal delirium, incident subsyndromal delirium, low dietary intake, and the diagnosis of falls or stroke were independent predictors of institutionalization. In contrast, age and previously known cognitive impairment were not independent predictors of being discharged to an institution in this population.

Discussion

In this study, we show that delirium or symptoms of delirium either prevalent or incident were strong independent predictors of institutionalization in this cohort of geriatric hospitalized patients. Furthermore, severe nutritional impairment reflected by very low dietary intake, female sex, and diagnoses associated with frailty such as falls or stroke were also independent predictors of institutionalization.

There was a very high frequency of symptoms of delirium in this cohort and, as a result, only half of the participants remained symptom free during their hospital stay. Similar findings were made in a post-acute care facility at admission (15). Our data highlight the importance of delirium symptoms, and not only full CAM-defined delirium, on the occurrence of negative outcome such as institutionalization. Marcantonio and colleagues have already demonstrated the clinical importance of subsyndromal delirium, evidencing its negative effect on functional recovery (20).

Our observational study has some limitations. Long-term outcome was not explored, and it cannot be excluded that patients discharged to the community were subsequently admitted to geriatric institutions. In studies concerning smaller cohorts of hospitalized participants, delirium seems to remain a predictor for subsequent admissions to geriatric institutions (9,21). Thus, we may have underestimated the impact of delirium on the rate of institutionalization. One consequence of this finding (i.e., delirium symptoms being associated with a high risk of institutionalization at discharge from hospital) is that there is probably a high prevalence of these symptoms at admission to geriatric institutions, and that the latter should establish managing these symptoms as a goal. This opens up a broad area of clinical research, since it seems difficult to reduce the rate of these symptoms (and probably their outcome) by using behavioral interventions (22).

The most surprising result was that previously known cognitive impairment was not an independent risk factor for being discharged to an institution. However, it was particularly frequent in this sample of elderly patients, concerning half of them. Cognitive impairment is known to be a major risk factor for delirium and may go unrecognized prior to hospitalization. During the hospital stay, precise diagnosis of dementia is also difficult, mostly owing to the high frequency of delirious symptoms, as seen in this study in approximately 50% of the survivors. On the other hand, delirium seems to be associated with a higher risk of subsequent dementia in intact participants (10) and could be an early marker of brain dysfunction (9). Thus, the effect of dementia on risk of admission to geriatric institutions could have been underestimated or obscured by the delirium symptoms. Delirium symptoms could simply be the origin of caregiver burnout, in the same way that dementia-associated behavior disturbances are, as has been shown previously (23). Furthermore, delirium symptoms generally occur suddenly, and care is usually not organized to address this new problem. Indeed, in this study, a significant percentage of the cognitively impaired patients (about two thirds) returned home. Therefore, cognitive impairment itself did not seem to play a predominant role in the choice of place of residence.

In the present sample, loss of daily autonomous activity was not an independent predictor for not returning home. One limitation of our study is the fact that the discharge Katz score was not available. However, worsening of functional dependency during hospitalization has been shown to be a predictor of institutionalization (24), and delirium has been shown to predict worsening of functional status during hospitalization (5). Furthermore, delirium seems to be a predictor of loss of autonomy independent of loss of function 3 months after hospital discharge (11). According to Marcantonio and colleagues (6), delirium does not seem to be a short-duration feature and could thus durably impair functional autonomy in the participants concerned. In the final model of predictors of institutionalization, we also introduced diagnosis and features strongly associated with loss of autonomy, such as falls and stroke or severe malnutrition. Stroke has previously been shown to be a risk factor for not returning home in all age groups (25), clearly due to loss of autonomy.

Diagnosis of falls associated with frailty (17) was also an independent predictor for not returning home. Falls are strongly associated with the fear of being alone at home, such as may be understood on the basis of patient and family interviews. Another limitation of our study was that we did not record social data requiring the consent of the participants such as marital status, financial income, or the existence of a social or family caregiver. The fact that female sex was associated with the risk of institutionalization could be due to the lack of a caregiver or to the relatively low income of elderly women. Women in this age group are more likely than men to be widowed and to live alone at home. According to the last French population census performed in 1999 (INSEE, National Institute for Statistics and Epidemiology), 80% of French women aged 85 years or older were widowed and 49% of them lived alone. Furthermore, since the income of elderly women is relatively low, it is rare to have a permanently employed caregiver at home. Therefore, the fear of falling and not receiving help immediately might be a major determinant in the choice of living in an institution.

Our data also show that severe malnutrition as assessed by very low dietary intake was an independent predictor for not returning home. This has not been previously shown. We did not explore the different causes associated with this low dietary intake, so the strength of our findings is attenuated. In this sample, patients with the worst prognosis were excluded since the sample did not include patients deceased at hospital. However, like delirium symptoms, low dietary intake seems to reflect a great need for care. Efficient management of severe malnutrition at home could be very difficult and could cause much worry for family or caregivers. The ability of geriatric institutions to manage nutritional impairment compared with the home environment has not yet been analyzed and could vary between types of institutions and from types of home care systems. However, our data point to an implicit advantage for institutions. Geriatric institutions should therefore implement specific policies to address nutritional issues (26), particularly the nutritional risk at the moment of admission.

Conclusion

Both full CAM-defined delirium and subsyndromal delirium were strong predictors for being discharged to geriatric institutions in this study, as were other geriatric syndromes such as severe malnutrition, falls, and stroke. We therefore suggest that geriatric institutions should implement policies for care of delirium symptoms, risk of falling, and nutritional impairment.

Table 1.

Admission Characteristics of Study Population (427 Patients) According to Their Mode of Discharge.

 Discharged to Community Discharged to Geriatric Institutions  
Admission Characteristics 310 Patients 117 Patients p Value 
Age* (y) 84.6 ± 6.2 85.6 ± 6.8 .14 
Sex ratio (M/F) 0.52 0.26 .006 
Previously known cognitive impairment, n (%) 149 (48.4) 71 (60.7) .02 
ADL score* 6.0 ± 3.4 7.0 ± 3.0 .02 
Visual impairment, n (%) 15 (4.9) 5 (4.4) .81 
Weight* (kg) 60.5 ± 14.1 54.9 ± 11.7 <.001 
Blood Na* (mmol/L) 138.0 ± 4.3 137.6 ± 5.9 .49 
Hemoglobin* (g/100 mL) 12.5 ± 1.8 12.1 ± 1.8 .06 
Creatinemia* (μmol/L) 98.7 ± 34.9 100.3 ± 38.0 .67 
Serum albumin (g/L) 33.7 ± 5.0 32.5 ± 5.1 .03 
C-reactive protein (mg/L) 35.0 ± 54.9 37.0 ± 55.5 .93 
Charlson index score* 1.89 ± 1.71 1.87 ± 1.73 .91 
Diabetes mellitus, n (%) 40 (12,9) 13 (11.1) .74 
Depression, n (%) 86 (27.7) 35 (29.9) .72 
Stroke, n (%) 36 (11.6) 22 (18.8) .06 
Falls, n (%) 50 (16.1) 32 (27.4) .01 
Parkinson's palsy, n (%) 17 (5.5) 5 (4.3) .81 
Admission treatment    
    Number of medications* 5.8 ± 3.0 5.6 ± 2.9 .75 
    Psychoactive drugs, n (%) 168 (54.7) 73 (62.9) .15 
    Opiates, n (%) 50 (16.1) 11 (9.4) .08 
 Discharged to Community Discharged to Geriatric Institutions  
Admission Characteristics 310 Patients 117 Patients p Value 
Age* (y) 84.6 ± 6.2 85.6 ± 6.8 .14 
Sex ratio (M/F) 0.52 0.26 .006 
Previously known cognitive impairment, n (%) 149 (48.4) 71 (60.7) .02 
ADL score* 6.0 ± 3.4 7.0 ± 3.0 .02 
Visual impairment, n (%) 15 (4.9) 5 (4.4) .81 
Weight* (kg) 60.5 ± 14.1 54.9 ± 11.7 <.001 
Blood Na* (mmol/L) 138.0 ± 4.3 137.6 ± 5.9 .49 
Hemoglobin* (g/100 mL) 12.5 ± 1.8 12.1 ± 1.8 .06 
Creatinemia* (μmol/L) 98.7 ± 34.9 100.3 ± 38.0 .67 
Serum albumin (g/L) 33.7 ± 5.0 32.5 ± 5.1 .03 
C-reactive protein (mg/L) 35.0 ± 54.9 37.0 ± 55.5 .93 
Charlson index score* 1.89 ± 1.71 1.87 ± 1.73 .91 
Diabetes mellitus, n (%) 40 (12,9) 13 (11.1) .74 
Depression, n (%) 86 (27.7) 35 (29.9) .72 
Stroke, n (%) 36 (11.6) 22 (18.8) .06 
Falls, n (%) 50 (16.1) 32 (27.4) .01 
Parkinson's palsy, n (%) 17 (5.5) 5 (4.3) .81 
Admission treatment    
    Number of medications* 5.8 ± 3.0 5.6 ± 2.9 .75 
    Psychoactive drugs, n (%) 168 (54.7) 73 (62.9) .15 
    Opiates, n (%) 50 (16.1) 11 (9.4) .08 

Notes: *Mean ± SD (standard deviation).

Table 2.

Delirium Category Distribution of Study Population (427 Patients) According to Mode of Discharge, p <.0001.

 Discharged to Community Discharged to Geriatric Institutions Total 
Delirium Category 310 Patients 117 Patients 427 (100) 
Prevalent delirium, n (%) 21 (6.8) 13 (11.1) 34 (8.0) 
Incident delirium, n (%) 9 (2.9) 6 (5.1) 15 (3.5) 
Prevalent subsyndromal delirium, n (%) 57 (18.4) 31 (26.5) 88 (20.6) 
Incident subsyndromal delirium, n (%) 33 (10.6) 27 (23.1) 60 (14.0) 
Symptom free, n (%) 190 (61.3) 40 (34.2) 230 (53.9) 
 Discharged to Community Discharged to Geriatric Institutions Total 
Delirium Category 310 Patients 117 Patients 427 (100) 
Prevalent delirium, n (%) 21 (6.8) 13 (11.1) 34 (8.0) 
Incident delirium, n (%) 9 (2.9) 6 (5.1) 15 (3.5) 
Prevalent subsyndromal delirium, n (%) 57 (18.4) 31 (26.5) 88 (20.6) 
Incident subsyndromal delirium, n (%) 33 (10.6) 27 (23.1) 60 (14.0) 
Symptom free, n (%) 190 (61.3) 40 (34.2) 230 (53.9) 
Table 3.

In-Hospital Dietary Intake Distribution of Study Population (417 Patients) According to Their Mode of Discharge, p <.01.

 Discharged to the Community Discharged to Geriatric Institutions Total 
Dietary Intake Group 304 Patients 113 Patients 417 (100) 
Acceptable (always >1200 kcal/d), n (%) 151 (49.7) 42 (37.2) 193 (46.3) 
Insufficient (<1200 kcal/d and >600 kcal/d), n (%) 108 (35.5) 38 (33.6) 146 (35.0) 
Low intake (600 kcal/d), n (%) 45 (14.8) 33 (29.2) 78 (18.7) 
 Discharged to the Community Discharged to Geriatric Institutions Total 
Dietary Intake Group 304 Patients 113 Patients 417 (100) 
Acceptable (always >1200 kcal/d), n (%) 151 (49.7) 42 (37.2) 193 (46.3) 
Insufficient (<1200 kcal/d and >600 kcal/d), n (%) 108 (35.5) 38 (33.6) 146 (35.0) 
Low intake (600 kcal/d), n (%) 45 (14.8) 33 (29.2) 78 (18.7) 
Table 4.

Independent Predictors of Institutionalization of Hospitalized Geriatric Patients (Step-by-Step Backward Logistic Regression, 415 Participants).

Independent Predictors OR 95% CI p Value 
Age (for 1-year increase) 1.02 0.98–1.06 .37 
Sex (female versus male) 2.15 1.22–3.78 .008 
Previously known cognitive impairment 1.21 0.73–1.99 .46 
Delirium categories*    
    Prevalent delirium 3.19 1.33–7.64 .009 
    Incident delirium 2.64 0.83–8.45 .10 
    Prevalent subsyndromal delirium 2.72 1.48–5.01 .001 
    Incident subsyndromal delirium 4.27 2.17–8.39 <.0001 
Dietary intake group    
    Insufficient (<1200 kcal/d and >600 kcal/d) 1.11 0.65–1.92 .70 
    Low intake (600 kcal/d) 2.50 1.35–4.63 .003 
Diagnosis    
    Falls 2.16 1.22–3.84 .008 
    Stroke 2.03 1.04–3.94 .04 
Admission biological data    
    Hemoglobin (for 1-g/100 mL decrease) 1.10 0.96–1.27 .18 
Independent Predictors OR 95% CI p Value 
Age (for 1-year increase) 1.02 0.98–1.06 .37 
Sex (female versus male) 2.15 1.22–3.78 .008 
Previously known cognitive impairment 1.21 0.73–1.99 .46 
Delirium categories*    
    Prevalent delirium 3.19 1.33–7.64 .009 
    Incident delirium 2.64 0.83–8.45 .10 
    Prevalent subsyndromal delirium 2.72 1.48–5.01 .001 
    Incident subsyndromal delirium 4.27 2.17–8.39 <.0001 
Dietary intake group    
    Insufficient (<1200 kcal/d and >600 kcal/d) 1.11 0.65–1.92 .70 
    Low intake (600 kcal/d) 2.50 1.35–4.63 .003 
Diagnosis    
    Falls 2.16 1.22–3.84 .008 
    Stroke 2.03 1.04–3.94 .04 
Admission biological data    
    Hemoglobin (for 1-g/100 mL decrease) 1.10 0.96–1.27 .18 

Notes: *These delirium categories are compared to symptom-free category.

These dietary intake groups are compared to acceptable intake group.

OR = odds ratio; 95% CI = 95% confidence interval.

This work was funded by the Clinical Research section of the Centre Hospitalier Universitaire de Bordeaux, France.

References

1
American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. DSM-IV. Washington DC: APA Press; 1994.
2
Francis J, Martin D, Kapoor WN. A prospective study of delirium in hospitalized elderly.
JAMA.
 .
1990
;
263
:
1097
-1101.
3
Inouye SK. Delirium in hospitalized older patients: recognition and risk factors.
J Geriatr Psychiatry Neurol.
 .
1998
;
11
:
118
-125 discussion 157–118.
4
Levkoff SE, Evans DA, Liptzin B, et al. Delirium. The occurrence and persistence of symptoms among elderly hospitalized patients.
Arch Intern Med.
 .
1992
;
152
:
334
-340.
5
O'Keeffe S, Lavan J. The prognostic significance of delirium in older hospital patients.
J Am Geriatr Soc.
 .
1997
;
45
:
174
-178.
6
Marcantonio ER, Flacker JM, Michaels M, Resnick NM. Delirium is independently associated with poor functional recovery after hip fracture.
J Am Geriatr Soc.
 .
2000
;
48
:
618
-624.
7
Pompei P, Foreman M, Cassel CK, Alessi C, Cox D. Detecting delirium among hospitalized older patients.
Arch Intern Med.
 .
1995
;
155
:
301
-307.
8
Inouye SK, Bogardus ST, Jr., Charpentier PA, et al. A multicomponent intervention to prevent delirium in hospitalized older patients.
N Engl J Med.
 .
1999
;
340
:
669
-676.
9
Francis J, Kapoor WN. Prognosis after hospital discharge of older medical patients with delirium.
J Am Geriatr Soc.
 .
1992
;
40
:
601
-606.
10
Rockwood K, Cosway S, Carver D, Jarrett P, Stadnyk K, Fisk J. The risk of dementia and death after delirium.
Age Ageing.
 .
1999
;
28
:
551
-556.
11
Murray AM, Levkoff SE, Wetle TT, et al. Acute delirium and functional decline in the hospitalized elderly patient.
J Gerontol A Biol Sci Med Sci.
 .
1993
;
48
:
M181
-M186.
12
Inouye SK. Predisposing and precipitating factors for delirium in hospitalized older patients.
Dement Geriatr Cogn Disord.
 .
1999
;
10
:
393
-400.
13
Inouye SK, van Dyck CH, Alessi CA, Balkin S, Siegal AP, Horwitz RI. Clarifying confusion: the confusion assessment method. A new method for detection of delirium.
Ann Intern Med.
 .
1990
;
113
:
941
-948.
14
Rockwood K, Cosway S, Stolee P, et al. Increasing the recognition of delirium in elderly patients.
J Am Geriatr Soc.
 .
1994
;
42
:
252
-256.
15
Kiely DK, Bergmann MA, Murphy KM, Jones RN, Orav EJ, Marcantonio ER. Delirium among newly admitted postacute facility patients: prevalence, symptoms, and severity.
J Gerontol A Biol Sci Med Sci.
 .
2003
;
58
:
M441
-M445.
16
Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index.
J Clin Epidemiol.
 .
1994
;
47
:
1245
-1251.
17
Rockwood K, Stolee P, McDowell I. Factors associated with institutionalization of older people in Canada: testing a multifactorial definition of frailty.
J Am Geriatr Soc.
 .
1996
;
44
:
578
-582.
18
Katz S, Downs TD, Cash HR, Grotz RC. Progress in development of the index of ADL.
Gerontologist.
 .
1970
;
10
:
20
-30.
19
Bourdel-Marchasson I, Barateau M, Sourgen C, et al. Prospective audits of quality of PEM recognition and nutritional support in critically ill elderly patients.
Clin Nutr.
 .
1999
;
18
:
233
-240.
20
Marcantonio ER, Simon SE, Bergmann MA, Jones RN, Murphy KM, Morris JN. Delirium symptoms in post-acute care: prevalent, persistent, and associated with poor functional recovery.
J Am Geriatr Soc.
 .
2003
;
51
:
4
-9.
21
George J, Bleasdale S, Singleton SJ. Causes and prognosis of delirium in elderly patients admitted to a district general hospital.
Age Ageing.
 .
1997
;
26
:
423
-427.
22
Young LJ, George J. Do guidelines improve the process and outcomes of care in delirium?
Age Ageing.
 .
2003
;
32
:
525
-528.
23
Phillips VL, Diwan S. The incremental effect of dementia-related problem behaviors on the time to nursing home placement in poor, frail, demented older people.
J Am Geriatr Soc.
 .
2003
;
51
:
188
-193.
24
Fortinsky RH, Covinsky KE, Palmer RM, Landefeld CS. Effects of functional status changes before and during hospitalization on nursing home admission of older adults.
J Gerontol A Biol Sci Med Sci.
 .
1999
;
54
:
M521
-M526.
25
Lai SM, Alter M, Friday G, Lai SL, Sobel E. Disposition after acute stroke: who is not sent home from hospital?
Neuroepidemiology.
 .
1998
;
17
:
21
-29.
26
Morley JE, Silver AJ. Nutritional issues in nursing home care.
Ann Intern Med.
 .
1995
;
123
:
850
-859.