Background

Delirium occurs frequently in older hospitalized patients with a prevalence at admission of 18–35% and an incidence during hospitalization of 11–56% [1].

Delirium is a disorder of brain function characterized by acute onset of mental state fluctuations. A main clinical feature is reduced ability to maintain attention. Moreover, patients become incoherent in thinking and speech and have impaired memory and altered level of consciousness, moving from a wakeful state to drowsiness, or becoming unresponsive [2]. Many old patients are predisposed to delirium due to cognitive impairment [1].

Previous studies have demonstrated an association between the development of delirium and an increased risk of morbidity and mortality, poor rehabilitation outcomes, prolonged hospitalization and increased institutionalization [3,4,5,6,7,8,9,10].

The influence of hospital ward design on the risk of delirium in older patients has received little research attention. One study in an intensive care unit found that patients (mean age 59 years) staying in single-bed rooms had a lower risk of developing delirium than patients staying in multiple-bed rooms, 6.8% vs. 15.1%; p < 0.01 [11]. To our knowledge, no study has investigated the influence of single-bed rooms as opposed to multiple-bed rooms on delirium in geriatric patients.

The aim of the present study was to investigate the risk of delirium among patients 75 years or older admitted to a geriatric department in relation to the architecture of the ward, especially single-bed rooms vs. multiple-bed rooms. We had the opportunity to perform the study because a geriatric department was moved from an old hospital building with multiple-bed rooms to a new hospital with single-bed rooms, without any changes in uptake area, staff or admission criteria.

Patients and methods

Study design and patients

The project was conducted as an observational, prospective cohort study as part of a quality development project by the Geriatric Department at Aarhus University Hospital, Denmark. The study population was patients 75 years or older, admitted to the old hospital with multiple-bed rooms (old wards) in the period from 15 September 2016 to 19 March 2017, and to the new hospital with single-bed rooms (new wards) from 20 March to 19 December 2017. Patients were excluded if they (1) had already been included once after a prior admission to the same hospital; (2) were dying or somnolent at the time of admission, as assessed by a specialist in geriatrics; (3) were unable to communicate for different reasons, e.g., aphasia, dementia or deaf-mute; (4) were unable to understand or speak Danish; (5) were hospitalized during Easter 2017 (1 week) and summer holiday 2017 (4 weeks) when the new wards had holiday staffing; (6) were admitted to the new wards less than 30 days after discharge from the old wards.

Differences between the old and new wards

The old wards consisted of two geriatric wards located at different addresses. There were 5 three-bed rooms, 11 two-bed rooms and two single-bed rooms with 13 shared bathrooms. The rooms had one or two windows at the end of the room next to one of the beds. The geriatric wards constituted a total area of approximately 1550 m2. Visiting time was unrestricted within the limits of consideration for other patients in the room.

The two geriatric wards in the new hospital are located on two different floors in the same building. In total, there are 32 single-bed rooms with large windows and with own bathrooms and a total area of roughly 2350 m2. Visiting time is unrestricted, and a relative can stay overnight in the room, which was very inconvenient in the old wards.

Measurements of delirium

At admission, all enrolled patients were examined for delirium by the geriatric staff using a Danish translation of the Confusion Assessment Method (CAM), originally developed by Inouye [12]. Every day at 8 a.m. (± 2 h) and at 8 p.m. (± 3 h) until discharge, the patients were assessed with CAM or in few cases by a physician diagnosing delirium (ICD-10 code F05). All CAM measurements were registered in the electronic patient record and subsequently copied to the research database by one of the authors (SB). If CAM was missing in the electronic patient record, the relevant staff member was immediately contacted for completion.

Delirium present at admission was defined as a positive CAM score or by a diagnosis of delirium at the first scheduled screening, which typically took place within the first 10 h after admission. Cases of delirium diagnosed after the first scheduled screening were defined as new cases.

The British National Institute for Health and Care Excellence (NICE) recommends CAM for diagnosing delirium. The method is an easy-to-use screening tool for patients undergoing surgical, medical, geriatric, and palliative-care treatment. It includes four criteria: (1) an acute beginning and a fluctuating course, (2) inattention, (3) disorganized thinking, and (4) an altered level of consciousness. A delirium diagnosis requires the presence of criteria 1 and 2, and at least one of criteria 3 and 4 [13, 14].

A German study of frail elderly patients showed a sensitivity of 0.77, a specificity of 0.96–1.00 and a Cohen’s kappa coefficient of 0.95 [15]. Similarly, an American systematic review found a sensitivity of 0.94 and a specificity of 0.89 [16]. A Danish clinical guideline recommends using the Danish version of the training manual and coding guide for CAM [12, 17, 18].

As preparation for the present study, the geriatric staff members, comprising nurses, nurse assistants, occupational therapists and physiotherapists completed a course on delirium and treatment in geriatric patients, including the Danish version of the training manual and coding guide for CAM followed by a test [17]. This preparation took place before inclusion of the first patients, and new staff tended the same course before performing CAM screenings. During the initial course period, the reliability of the Danish version of CAM was evaluated on 52 patients. The inter-rater agreement as measured by Cohen’s kappa was 0.85.

Data collection

From the patient records, the following baseline characteristics were extracted: age, gender, housing conditions (nursing homes, sheltered home or own home), prior diagnosis of dementia, main hospitalization diagnosis, body mass index (BMI), physical functional ability (Barthel 100), body temperature, pulse and respiratory rates, and blood samples: hemoglobin, sodium, creatinine, albumin, white cell count and C-reactive protein.

Dementia was identified using the International Classification of Diseases 10th Revision (ICD-10) codes Dementia in Alzheimer’s disease (F00 and G30.1), Vascular dementia (F01), Dementia in other diseases (F02), Unspecified dementia (F03), and Delirium superimposed on dementia (F05.1) [19].

Charlson Comorbidity Index was calculated from all diagnoses recorded in the patient journal for the current admission. The Systemic Inflammatory Response Syndrome (SIRS) was defined as temperature > 38 °C or < 36 °C, pulse > 90 bpm, respiratory rate > 20 breaths per minute and white cell count > 12 × 109/L or < 4 × 109/L with at least two positive criteria for the definition of sepsis [20].

Statistics

Patients’ baseline characteristics were compared by Student’s t test or Wilcoxon’s rank sum test for continuous variables, and Pearson’s chi-squared test or Fisher’s exact test for categorical variables.

The incidence of delirium during hospital stay was compared between the old and the new ward using a Cox regression model adjusted for potential confounders: age, comorbidity, housing conditions, prior diagnosis of dementia, SIRS and main diagnosis (infection, fracture, other). Time at risk began at admission and ended at the first delirium episode or with censoring at the date of discharge or death. Patients with delirium present on admission were excluded from the incidence analysis. Some patients were admitted both once to the old and once to the new ward. To adjust for this clustering, we used robust variance estimates. A test of the proportional hazards assumption was supplemented with inspection of “log–log” plots.

A comparison of time to recovery from a patient’s first delirium episode was performed using Cox regression with censoring at discharge and adjusted for Charlson’s Comorbidity Index. It included patients with delirium present at admission, but in a secondary analysis these patients were excluded. For these analyses, the duration of a delirium period was defined by consecutive, positive results of the twice-daily CAM assessments after these modifications: a missing assessment was substituted by the result of a non-missing neighbor assessment with positive neighbour assessments taking precedence over negative assessments. If one or two negative assessments were surrounded by positive assessments, they were substituted by positive assessments.

Statistical analyses were performed with Stata software, version 15.1 (StataCorp LLC, College Station, Texas). p values less than 0.05 were considered statistically significant.

Ethics

As a quality development project with no intervention, the study was exempted from notification to the Central Denmark Region Ethical Committee (Inquiry number 200/2017). The study protocol was approved by the Danish Data Protection Agency, case no. 1-16-02-254-16. Data were stored according to good research practice in Research Electronic Data Capture (REDCap) hosted at Aarhus University, Denmark [21]. The study was registered at ClinicalTrials.gov (Identifier NCT03199768).

Results

A total of 1,014 consecutive patients were enrolled in the study (Fig. 1). Baseline patient characteristics are shown in Table 1; they were well balanced between the old and the new wards, except for age which was slightly lower for the new wards; mean difference 0.75 years (95% CI 0.04–1.47). A test for interaction showed no effect modification by age. For 105 patients (10%), delirium was present at admission, with no difference between the old and new wards. Among patients with delirium at admission, 21% had a prior diagnosis of dementia; for patients without delirium at admission, this was the case for 7%.

Fig. 1
figure 1

Flowchart for inclusion

Full size image
Table 1 Baseline characteristics of 1014 geriatric patients
Full size table

Length of stay (LOS) ranged from 1 to 24 days in the old wards, and 15 (4%) patients had a LOS of 14 days or more. In the new wards, LOS ranged from 1 to 36 days, and 5 (1%) patients had a LOS of 14 days or more.

In total, 14,246 CAM examinations were planned, and 13,923 (98%) of these were completed. At admission, 105 patients had delirium, and among the remaining 909 patients, 140 developed delirium during the hospital stay. During the first 14 days of hospital stay, the cumulative incidence of delirium was 29% in patients admitted to the old wards and 16% in patients admitted to the new wards. The patients in the new wards had a significantly reduced risk of delirium compared with the patients in the old wards; HR = 0.66 (95% CI 0.48–0.93, p = 0.02), adjusted for age, comorbidity, housing conditions, prior diagnosis of dementia, SIRS and main diagnosis. Figure 2 shows the cumulative incidence of delirium among patients in the old and the new wards. Development of delirium during hospital stay was associated with a prior diagnosis of dementia (HR: 1.74, 95% CI 1.02–2.98, p = 0.04) and with age (HR per 10 years: 1.58, 95% CI 1.18–2.11, p = 0.002).

Fig. 2
figure 2

Cumulative incidence of delirium (105 patients with delirium at admission not included)

Full size image

Figure 3 shows the time from diagnosis to recovery of the first period of delirium. There was no difference between the old and the new wards in the duration of first delirium episode; HR: 1.06 (95% CI 0.86–1.32, p = 0.57). Restricting the analysis to patients without delirium at admission gave a very similar result; HR: 1.04.

Fig. 3
figure 3

Time from diagnosis to recovery of the first period of delirium

Full size image

In the old ward, 35 (8%) patients stayed in single rooms throughout the hospital stay; of these, 13 patients had delirium at admission, and 4 patients developed delirium during hospital stay.

Discussion

Incidence of delirium

We found a reduced incidence of delirium among geriatric patients staying in single-bed rooms compared with patients staying in multiple-bed rooms, and this finding remained robust after adjustment for baseline risk factors. This supports the hypothesis that a change from multiple-bed rooms to single-bed rooms in geriatric wards can prevent delirium. To our knowledge, there are no similar studies with a geriatric population.

Our observations are in accordance with a study of younger patients conducted by Caruso and colleagues, showing that single-bed rooms could prevent delirium in an intensive care unit (odds ratio: 4.03, 95% CI 2.13–7.62) [11]. Patients were younger, but the methods were similar to those used in the present study, except for the fact that CAM-ICU was used. Like CAM, CAM-ICU is a recognized psychometric instrument validated for intensive-care patients [22]. Caruso et al. excluded patients with prior diagnoses of dementia, while we included patients previously diagnosed with dementia. This may explain the high incidence (16% vs. 29%) in our study compared with the study by Caruso et al. (6.8% vs. 15.1%). It is well known that the risk of delirium increases with age and that delirium is associated with dementia [1, 23]. This was also the case in the present study.

Time to recovery from delirium

When delirium occurred, we found no difference between the old and the new wards in the duration of the first delirium episode. This is in accordance with the study conducted by Caruso et al., who found no difference in the duration of delirium between single-bed rooms and multiple-bed rooms [11]. It seems that single-bed rooms do not reduce the duration of delirium episodes.

Effects of the hospital environment

Problems of noise, sleep disturbance, stress and lack of privacy are frequent in multiple-bed rooms. In a study by Leung and colleagues, preoperative sleep disruption was associated with postoperative delirium [24]. In another study by Weng and colleagues, sleep disturbance was reduced in single-bed rooms compared with multiple-bed rooms in a geriatric department [25].

Our study shows that multiple-bed rooms are a risk factor for developing delirium during hospital stay. Obviously, this is a modifiable risk factor related to the hospital environment. Decision-makers should incorporate this knowledge when building new wards or renovating old ones. Single-bed room design is associated with higher building cost, but the costs of delirium are a prolonged length of hospital stay, an increased risk of complications during admission and lower functional levels afterwards [26]. Leslie and colleagues estimated the cost of treatment of patients with delirium to be substantially increased compared with costs for other patients [27].

Strengths and limitations of the study

A major strength of the present study is the inclusion of a large population of geriatric patients. A so-called natural experiment allowed us to compare the risk of delirium in two different wards. In the study period, there were no changes in uptake area, staff composition or admission criteria. We, therefore, assume that the patient populations in the old and the new wards were rather similar. Moreover, delirium was identified prospectively by a trained geriatric staff twice a day during hospitalization. The staff completed the same course before using CAM, and the clinical guidelines for delirium did not change. The staff turnover did not change in relation to the relocation of the department. Before the study was performed, a test sample found a satisfactory inter-rater agreement of the diagnostic instrument. Finally, delirium fluctuates, and screening the patients twice a day during the hospital stay reduced the risk of missing episodes of delirium.

Another strength of the present study is the almost complete measurements. Only 2% of all planned CAM examinations were missing. One percent of the patients were out of the department for examinations or were somnolent, and for 0.9%, the CAM score was missing for unknown reasons.

It is well known that transfer of older patients is associated with an increased incidence of delirium [25]. Transfer of patients may have influenced the incidence, but we have no reason to believe that the frequency of transfer was different in the two types of wards.

It is a limitation of the present study that moving the geriatric department to modern hospital premises involved several changes, and the study does not allow to decide with certainty which changes were responsible for the reduced incidence of delirium. We believe that the change to single-bed rooms had a major impact, but other changes such as more space, quiet rooms, better access to daylight, better working conditions for the staff, or improved indoor climate might also play a role.

Conclusion

Compared with multiple-bed rooms, single-bed rooms seem to prevent delirium during hospitalization of geriatric patients, but other differences between older and newer hospital premises may also be of importance.