Abstract
Aims
Delirium is associated with adverse outcomes following hip fracture, but the prevalence and significance of delirium for the prognosis and ongoing rehabilitation needs of patients admitted from home is less well studied. Here, we analyzed relationships between delirium in patients admitted from home with 1) mortality; 2) total length of hospital stay; 3) need for post-acute inpatient rehabilitation; and 4) hospital readmission within 180 days.
Methods
This observational study used routine clinical data in a consecutive sample of hip fracture patients aged ≥ 50 years admitted to a single large trauma centre during the COVID-19 pandemic between 1 March 2020 and 30 November 2021. Delirium was prospectively assessed as part of routine care by the 4 A’s Test (4AT), with most assessments performed in the emergency department. Associations were determined using logistic regression adjusted for age, sex, Scottish Index of Multiple Deprivation quintile, COVID-19 infection within 30 days, and American Society of Anesthesiologists grade.
Results
A total of 1,821 patients were admitted, with 1,383 (mean age 79.5 years; 72.1% female) directly from home. Overall, 87 patients (4.8%) were excluded due to missing 4AT scores. Delirium prevalence in the whole cohort was 26.5% (460/1,734): 14.1% (189/1,340) in the subgroup of patients admitted from home, and 68.8% (271/394) in the remaining patients (comprising care home residents and inpatients when fracture occurred). In patients admitted from home, delirium was associated with a 20-day longer total length of stay (p < 0.001). In multivariable analyses, delirium was associated with higher mortality at 180 days (odds ratio (OR) 1.69 (95% confidence interval (CI) 1.13 to 2.54); p = 0.013), requirement for post-acute inpatient rehabilitation (OR 2.80 (95% CI 1.97 to 3.96); p < 0.001), and readmission to hospital within 180 days (OR 1.79 (95% CI 1.02 to 3.15); p = 0.041).
Conclusion
Delirium affects one in seven patients with a hip fracture admitted directly from home, and is associated with adverse outcomes in these patients. Delirium assessment and effective management should be a mandatory part of standard hip fracture care.
Cite this article: Bone Jt Open 2023;4(6):447–456.
Take home message
Delirium can be ascertained in a large, near-complete hip fracture population in routine clinical practice using a well-validated, easy-to-use tool (4AT).
One in seven patients admitted from their own home had delirium, and this was associated with adverse outcomes.
Delirium assessment on admission using a tool such as the 4AT should be part of standard care for all acute hip fracture patients.
Early delirium assessment could guide prognostication, facilitate proactive onward care planning, and be used to identify modifiable clinical risk factors.
Introduction
Delirium is an acute-onset, fluctuating neuropsychiatric syndrome characterized by disturbed consciousness, attention, and cognition. It is common in patients with an acute hip fracture, and is associated with adverse outcomes, including reduced mobility, prolonged hospital stay, in-hospital mortality, and higher care requirements following discharge.1 There is increasing recognition of the importance of delirium in the management and prognosis following hip fracture. Local and national hip fracture guidelines now urge effective delirium detection and care.2-4
The prevalence of delirium in hip fracture patients has mainly been investigated in research studies, with delirium assessment performed by research or specialist clinical teams. Most studies have assessed postoperative delirium (POD) rates: in a 2015 systematic review of ten studies, the prevalence of POD ranged from 13% to 56%.5 Three recent studies reported POD prevalences of 19%, 22%, and 25%.1,6,7 A smaller group of studies have reported preoperative delirium prevalences, of 19% (only home-dwelling patients included), 58%, and 18%.8-10
Another source of data on delirium in hip fracture is large clinical registries. However, according to the most recent published reports, delirium is prospectively assessed as part of routine care in a minority: the National Hip Fracture Database (NHFD; England, Wales and Northern Ireland), the Scottish Hip Fracture Audit (SHFA), the Irish Hip Fracture Database, and the Australian and New Zealand Hip Fracture Registry.11-15 Of these, only the NHFD has reported POD prevalence rates in three subsets of data: 25% in a single-centre study of 1,224 patients (2016 to 2018), 26% in a single centre of 175 patients (2018 to 2019), and 29% in 107,028 patients from the whole NHFD (2018 to 2019).16-18 In a USA National Surgical Quality Improvement Program (NSQIP) study, in which delirium was ascertained retrospectively using International Classification of Diseases (ICD)-9 codes, the POD rate in hip fracture patients was 27%.19
Several studies have assessed outcomes following delirium in hip fracture patients.1,19,20 These have analyzed whole cohorts, comprising patients admitted from home, all care settings, and those sustaining fractures as inpatients. One study using local NHFD data reported that POD was associated with increased mortality at 30 days and one year.17 Less is known about delirium in patients admitted from home, including the prevalence and significance of delirium for prognosis and ongoing rehabilitation needs in these patients. Returning to live at home on discharge is important, with days alive and at home following a hip fracture validated as a patient-centred outcome for patients aged > 50 years.21 In a study of 207 patients admitted directly from private homes, Krogseth et al8 reported an inpatient delirium incidence of 39%, and associations with higher mortality and risk of new institutionalization. Determining whether similar associations exist in large, population-representative datasets with routine delirium assessment is important for clinical practice.
In Scotland, hip fracture services are delivered in accordance with the Scottish Standards of Care for Hip Fracture Patients (SSCHFP). Delirium assessment is mandated for all patients on admission using the 4 A’s Test (4AT).22 This study aim was to determine any associations of delirium in patients admitted from home with: 1) mortality at 30 and 180 days; 2) total length of stay in acute and subacute care settings; 3) need for post-acute inpatient rehabilitation; and 4) readmission to hospital from home within 180 days.
Methods
This study is reported in line with STROBE guidelines for observational studies.
Study population
This cohort study was based on the International Multicentre Project Auditing COVID-19 in Trauma & Orthopaedics (IMPACT) Hip Fracture Audit,23,24 using prospectively collected, retrospectively validated clinical audit data from a single high-volume orthopaedic major trauma centre in Scotland. IMPACT is a global multicentre project, originally purposed to coordinate urgent collaborative research, report near real-time data, and guide the clinical response to COVID-19. All patients ≥ 50 years admitted with an acute hip fracture between 1 March 2020 and 30 November 2021 were included. Direct admission from home was determined based on contemporaneous clinical documentation, coded by specialist clinical auditors, and validated using live electronic health records (EHRs) by a study author (AJH). Patients were included if they had an intracapsular or extracapsular fracture of the proximal femur, up to and including the subtrochanteric region. Fractures around an existing implant and isolated fractures of the public ramus, acetabulum, or greater trochanter were excluded, since these patients were excluded from data collection for national audit purposes at the time of the study.
Data collection
Data were collected from EHRs (TrakCare; InterSystems Corporation, USA) and clinical documentation by trained specialist audit coordinators as part of routine audit activity. Data were collated by specialist auditors, and verified by a senior analyst independent of the study. Variables included: patient demographics (age, sex, type of residence on admission); Scottish Index of Multiple Deprivation (SIMD); patient factors (American Society of Anesthesiologists (ASA) grade); COVID-19 within 30 days; injury factors (fracture type); inpatient care factors (type of operation; length of stay in acute and subacute inpatient settings); delirium assessed using the 4AT; discharge destination; number and date of readmissions; and mortality status up to 180 days. SIMD is an area-based relative measure of deprivation.25 COVID-19 infection was considered within 30 days, as this was previously demonstrated to be independently associated with increased mortality.23 Delirium was defined as a 4AT score ≥ 4. The 4AT is a well-validated, brief clinical tool for delirium recommended in numerous guidelines and hip fracture registries.12,13,26,27 The first available 4AT score from the emergency department (ED) was used; if none recorded, then the first 4AT score from the admission orthopaedic ward was used instead. The vast majority of 4AT assessments were performed within 24 hours of admission, as per admission protocol. Data underwent verification through manual cross-referencing against live EHRs by AJH, and additional variables were collected manually including mortality status following admission, and date of death up to 180 days. The EHRs for all patients were reviewed for a minimum of 180 days after date of admission, or until date of death.
Patients with no 4AT score from the ED or admission orthopaedic ward were excluded. Patients from outside the service catchment area were included in all analyses except readmission, where they were excluded due to lack of reliable follow-up. Data were compiled using the bespoke IMPACT Audits data collection tool using data-validated fields to ensure coding accuracy and consistency.28 This dataset has been used in previously published studies.29,30 Data were collected and handled as part of service evaluation in accordance with UK Caldicott principles.
Statistical analysis
Statistical analyses were performed using RStudio version 4.2.1 and packages: finalfit; dplyr; survival; survminer; TableOne, epiDisplay, and tidyverse (Integrated Development for R; Rstudio, PBC, USA). Delirium was coded as a binary variable: 4AT score ≥ 4 suggestive of delirium.22,26 Continuous variables were assessed for differences between groups using an independent-samples t-test (parametric) or Mann-Whitney U test (non-parametric), and categorical variables using a chi-squared test. Multivariable logistic regression was used to analyse associations of delirium with: 1) mortality at 30 and 180 days; 2) need for post-acute inpatient rehabilitation; and 3) for patients discharged from the acute stay to a home within the catchment area, likelihood of readmission to an acute hospital within 180 days. All multivariable models included covariates: age, sex, SIMD quintile, COVID-19 within 30 days, and ASA grade. A p-value of < 0.05 was deemed significant. Assumptions of logistic regression were tested and satisfied.
Supplementary analyses
Supplementary analyses were performed to assess baseline characteristics and outcomes in patients admitted from higher care settings, including care homes, rehabilitation facilities, and transfer from other acute care settings. A further multivariable logistic regression analysis was performed to analyse associations of delirium with mortality at 30 and 180 days in patients admitted from all care settings. This included the same covariates as the above regression models (age, sex, SIMD quintile, COVID-19 within 30 days, ASA grade), with an additional binary variable for type of residence on admission (home/higher care setting). A p-value of < 0.05 was deemed significant.
Sensitivity analyses
Sensitivity analyses were performed including only patients with a 4AT score from the ED. In these patients it is known that any delirium was present on admission, and that assessment was preoperative.
Results
Delirium prevalence
The study flowchart is shown in Figure 1. A total of 1,822 patients (mean age 80.7 years; 71.7% female) were admitted during the study period. One died within hours of admission and was excluded. A total of 87 (4.8%) had missing 4AT scores, and were excluded. Of the 1,734 included patients, 460 (26.5%) had delirium.
Fig. 1
Flow chart demonstrating inclusion and exclusion of patients in the study. ED, emergency department.
There were 1,383 (mean age 79.5 years; 72.1% female) patients admitted directly from home. Of these, 43 had missing 4AT scores and were excluded. Of the 1,340 included patients (mean age 79.5 years; 72.2% female), 189 (14.1%) had delirium. In the remaining 394 patients admitted from care homes (283 (71.8%)), rehabilitation facilities (31 (7.9%)), and transferred from inpatient settings (80 (20.3%)), delirium prevalence was 68.8%.
Baseline characteristics
Demographics of the 1,340 patients admitted from home are shown in Table I, stratified by delirium status. Patients with delirium were older and had a higher ASA grade, indicating greater severity of comorbid disease. There was no difference in the prevalence of COVID-19 within 30 days between the groups.
Table I.
Patient baseline characteristics according to delirium status ascertained on admission for patients admitted from home. Univariable analyses were used to assess the difference between those with and without delirium for each variable. Delirium = 4 AT ≥ 4.
| Variable | Whole cohort (n = 1,340) | Delirium (n = 189) | No delirium (n = 1,151) | p-value |
|---|---|---|---|---|
| Mean age, yrs (SD) | 79.5 (10.3) | 82.9 (8.9) | 79.0 (10.4) | < 0.001* |
| Age (yrs), n | < 0.001† | |||
| 50 to 59 | 69 | 7 | 62 | |
| 60 to 69 | 154 | 7 | 147 | |
| 70 to 79 | 356 | 36 | 320 | |
| 80 to 89 | 553 | 98 | 455 | |
| 90+ | 208 | 41 | 167 | |
| Female sex, n (%) | 968 (72.2) | 129 (68.3) | 839 (72.9) | 0.218† |
| SIMD quintile, n (%) | 0.078† | |||
| 1 (most deprived) | 168 (12.6) | 15 (7.9) | 153 (13.4) | |
| 2 | 297 (22.3) | 52 (27.5) | 245 (21.4) | |
| 3 | 221 (16.6) | 31 (16.4) | 190 (16.6) | |
| 4 | 236 (17.7) | 39 (20.6) | 197 (17.2) | |
| 5 (least deprived) | 412 (30.9) | 52 (27.5) | 360 (31.4) | |
| ASA grade, n (%) | < 0.001† | |||
| 1 (healthy) | 38 (2.8) | 1 (0.5) | 37 (3.2) | |
| 2 | 413 (30.8) | 22 (11.6) | 391 (34.0) | |
| 3 | 789 (58.9) | 136 (72.0) | 653 (56.7) | |
| 4 | 88 (6.6) | 23 (12.2) | 65 (5.6) | |
| 5 (moribund) | 12 (0.9) | 7 (3.7) | 5 (0.4) | |
| COVID-19 positive (< 30 days), n (%) | 53 (4.0) | 10 (5.3) | 43 (3.7) | 0.415† |
-
*
Independent-samples t-test
-
†
Chi-squared test
-
ASA, American Society of Anesthesiologists; SD, standard deviation; SIMD, Scottish Index of Multiple Deprivation.
Univariable analyses
Outcomes of the 1,340 patients admitted from home are shown in Table II, stratified by delirium status. Patients with delirium had a higher mortality rate at 30 and 180 days. Median total length of stay was 20 days longer for patients with delirium, versus those without (p < 0.001, Mann-Whitney U test). Overall, 444 patients (33.1%) went to post-acute inpatient rehabilitation facilities following the acute hospital stay (37 to care homes, 363 to rehabilitation facilities, and 44 to other acute care settings). Only 62/189 (32.8%) patients with delirium directly returned to private homes following the acute stay, versus 767/1151 (66.6%) of those without delirium (p < 0.001, chi-squared test).
Table II.
Patient outcomes according to delirium status, as assessed on admission for patients admitted from home. Univariable analyses assess the difference between those with and without delirium for each variable. Delirium = 4AT ≥ 4.
| Variable | Whole cohort (n = 1,340) | Delirium (n = 189) | No delirium (n = 1,151) | p-value |
|---|---|---|---|---|
| Mortality, n (%) | ||||
| Within 30 days | 63 (4.7) | 21 (11.1) | 42 (3.6) | < 0.001* |
| Within 180 days | 197 (14.7) | 54 (28.6) | 143 (12.4) | < 0.001* |
| Discharge destination following acute hospital stay, n (%) | < 0.001* | |||
| Own home | 829 (61.9) | 62 (32.8) | 767 (66.6) | |
| Care home | 37 (2.8) | 10 (5.3) | 28 (2.3) | |
| Rehabilitation facility | 363 (27.1) | 82 (43.4) | 281 (24.4) | |
| Other acute care | 44 (3.3) | 9 (4.8) | 35 (3.0) | |
| Died in hospital | 67 (5.0) | 26 (13.8) | 41 (3.6) | |
| Discharge to post-acute rehabilitation, n (%) | 444 (33.1) | 101 (53.4) | 343 (29.8) | < 0.001* |
| Median total length of stay in inpatient facilities, days (IQR) | 16 (9 to 36) | 34 (16 to 66) | 14 (8 to 31) | < 0.001† |
| Patients discharged home in catchment area (n = 813) | Delirium (n = 61) | No delirium (n = 752) | ||
| Readmission, n (%) | ||||
| Within 30 days | 153 (18.8) | 22 (36.1) | 131 (17.4) | 0.001* |
| Within 180 days | 307 (37.8) | 36 (59.0) | 271 (36.0) | 0.001* |
-
*
Chi-squared test.
-
†
Mann-Whitney U test.
-
IQR, interquartile range.
Multivariable analyses
The following results relate to those patients admitted directly from home. In multivariable analyses adjusting for age, sex, SIMD, ASA grade, and COVID-19 within 30 days, delirium was independently associated with a higher 180-day mortality rate (odds ratio (OR) 1.69 (95% confidence interval (CI) 1.13 to 2.54); p = 0.011) (Table III). Delirium was independently associated with increased likelihood of requiring post-acute inpatient rehabilitation following the acute stay (OR 2.80 (95% CI 1.97 to 3.96); p < 0.001) (Table IV). In total, 16 patients were discharged to residences outside the catchment area and were excluded from readmission analysis. In the 813 patients who went home within the catchment area, delirium was independently associated with an increased risk of readmission to an acute hospital within 180 days (OR 1.79 (95% CI 1.02 to 3.15); p = 0.041) (Table V).
Table III.
Uni- and multivariable logistic regression analysis of factors associated with 30- and 180-day mortality for patients admitted from home. Multivariable analysis was adjusted for age, sex, Scottish Index of Multiple Deprivation, American Society of Anesthesiologists grade, and COVID-19 within 30 days.
| Variable | Within 30 days, OR (95% CI) | p-value | Within 180 days, OR (95% CI) | p-value | ||
|---|---|---|---|---|---|---|
| Univariable analysis | Multivariable analysis | Univariable analysis | Multivariable analysis | |||
| Age | ||||||
| Per additional year | 1.03 (1.00 to 1.06) | 1.04 (1.00 to 1.07) | 0.037 | 1.05 (1.03 to 1.07) | 1.05 (1.03 to 1.07) | < 0.001 |
| Sex | ||||||
| Male | Reference | Reference | Reference | Reference | ||
| Female | 0.70 (0.41 to 1.19) | 0.63 (0.34 to 1.14) | 0.126 | 0.55 (0.40 to 0.76) | 0.49 (0.34 to 0.69) | < 0.001 |
| SIMD quintile | ||||||
| 1 (most deprived) | Reference | Reference | Reference | Reference | ||
| 2 | 0.72 (0.31 to 1.69) | 0.61 (0.24 to 1.53) | 0.293 | 1.08 (0.63 to 1.82) | 0.95 (0.54 to 1.67) | 0.854 |
| 3 | 0.75 (0.30 to 1.84) | 0.47 (0.16 to 1.39) | 0.173 | 0.54 (0.29 to 1.02) | 0.42 (0.21 to 0.84) | 0.014 |
| 4 | 0.63 (0.25 to 1.58) | 0.50 (0.18 to 1.39) | 0.187 | 0.93 (0.53 to 1.63) | 0.74 (0.40 to 1.36) | 0.333 |
| 5 (least) | 0.85 (0.39 to 1.84) | 0.65 (0.27 to 1.55) | 0.331 | 1.23 (0.75 to 2.01) | 1.04 (0.61 to 1.79) | 0.874 |
| ASA grade * | ||||||
| 1 (healthy) | - | - | - | - | - | - |
| 2 | 0.27 (0.10 to 0.68) | 0.32 (0.12 to 0.84) | 0.021 | 0.32 (0.20 to 0.50) | 0.41 (0.25 to 0.65) | < 0.001 |
| 3 | Reference | Reference | Reference | Reference | ||
| 4 | 3.07 (1.50 to 6.29) | 2.80 (1.33 to 5.88) | 0.007 | 3.09 (1.93 to 4.95) | 3.23 (1.97 to 5.29) | < 0.001 |
| 5 (moribund) | - | - | - | - | - | - |
| Delirium status: 4AT ≥ 4 | 3.28 (1.90 to 5.68) | 1.72 (0.89 to 3.34) | 0.109 | 2.80 (1.95 to 4.02) | 1.69 (1.13 to 2.54) | 0.011 |
| COVID-19-positive (< 30 days) | 2.74 (1.13 to 6.68) | 3.02 (1.17 to 7.78) | 0.022 | 1.54 (0.78 to 3.05) | 1.41 (0.67 to 2.96) | 0.359 |
-
*
No patients ASA grade 1 died within 180 days; all patients ASA grade 5 died within 30 days.
-
ASA, American Society of Anesthesiologists; CI, confidence interval; OR, odds ratio; SIMD, Scottish Index of Multiple Deprivation.
Table IV.
Uni- and multivariable logistic regression analysis of factors associated with requirement for post-acute inpatient rehabilitation following the acute inpatient stay (including care home, rehabilitation facility, or other acute care setting) for patients admitted from home. Multivariable analysis adjusted for age, sex, Scottish Index of Multiple Deprivation, American Society of Anesthesiologists grade, and COVID-19 within 30 days.
| Variable | OR (95% CI) | p-value | |
|---|---|---|---|
| Univariable analysis | Multivariable analysis | ||
| Age | |||
| Per year | 1.06 (1.05 to 1.07) | 1.05 (1.04 to 1.06) | < 0.001 |
| Sex | |||
| Male | Reference | Reference | |
| Female | 0.97 (0.75 to 1.24) | 0.92 (0.70 to 1.21) | 0.537 |
| SIMD quintile | |||
| 1 (most deprived) | Reference | Reference | |
| 2 | 1.24 (0.84 to 1.83) | 1.00 (0.65 to 1.55) | 0.971 |
| 3 | 0.95 (0.62 to 1.45) | 0.79 (0.50 to 1.25) | 0.312 |
| 4 | 1.01 (0.67 to 1.53) | 0.80 (0.50 to 1.26) | 0.329 |
| 5 (least) | 1.22 (0.84 to 1.77) | 1.01 (0.67 to 1.53) | 0.960 |
| ASA grade | |||
| 1 (healthy) | 0.07 (0.02 to 0.28) | 0.13 (0.03 to 0.57) | 0.007 |
| 2 | 0.30 (0.23 to 0.39) | 0.37 (0.28 to 0.50) | < 0.001 |
| 3 | Reference | Reference | |
| 4 | 1.62 (1.04 to 2.53) | 1.60 (1.20 to 2.58) | 0.052 |
| 5 (moribund) | 12.91 (1.66 to 100.51) | 11.38 (1.42 to 91.19) | 0.022 |
| Delirium status: 4AT ≥ 4 | 4.09 (2.95 to 5.68) | 2.80 (1.97 to 3.96) | < 0.001 |
| COVID-19-positive (< 30days) | 2.79 (1.58 to 4.92) | 2.85 (1.54 to 5.28) | < 0.001 |
-
ASA, American Society of Anesthesiologists; CI, confidence interval; OR, odds ratio; SIMD, Scottish Index of Multiple Deprivation.
Table V.
Uni- and multivariable logistic regression analysis of factors associated with readmission to an acute hospital from home within 180 days. Multivariable analysis adjusted for age, sex, Scottish Index of Multiple Deprivation, American Society of Anesthesiologists grade, and COVID-19 within 30 days.
| Variable | OR (95% CI) | p-value | |
|---|---|---|---|
| Univariable analysis | Multivariable analysis | ||
| Age | |||
| Per year | 1.04 (1.03 to 1.06) | 1.03 (1.02 to 1.05) | < 0.001 |
| Sex | |||
| Male | Ref | Ref | |
| Female | 0.76 (0.55 to 1.04) | 0.74 (0.53 to 1.04) | 0.081 |
| SIMD quintile | |||
| 1 (most deprived) | Ref | Ref | |
| 2 | 0.74 (0.45 to 1.21) | 0.69 (0.41 to 1.17) | 0.166 |
| 3 | 0.66 (0.39 to 1.10) | 0.59 (0.34 to 1.03) | 0.063 |
| 4 | 0.86 (0.52 to 1.43) | 0.79 (0.46 to 1.37) | 0.408 |
| 5 (least) | 0.83 (0.52 to 1.32) | 0.81 (0.49 to 1.34) | 0.409 |
| ASA grade * | |||
| 1 (healthy) | 0.07 (0.02 to 0.30) | 0.11 (0.03 to 0.47) | 0.003 |
| 2 | 0.41 (0.30 to 0.57) | 0.48 (0.35 to 0.67) | < 0.001 |
| 3 | Ref | Ref | |
| 4 | 1.88 (0.94 to 3.75) | 1.96 (0.95 to 4.02) | 0.067 |
| 5 (moribund) | - | - | - |
| Delirium status: 4AT ≥ 4 | 2.56 (1.50 to 4.35) | 1.79 (1.02 to 3.15) | 0.041 |
| COVID-19-positive (< 30days) | 2.9 (1.13 to 7.45) | 2.72 (1.02 to 7.25) | 0.040 |
-
*
Patients discharged out of catchment area were excluded from readmission analysis. No patients with ASA grade 5 went home.
-
ASA, American Society of Anesthesiologists; CI, confidence interval; OR, odds ratio; SIMD, Scottish Index of Multiple Deprivation.
Supplementary analyses
Demographics of the 394 patients admitted from higher care settings are provided in Supplementary Table i, and outcomes for these patients in Supplementary Table ii. The mortality rate in these patients was 37.8% within 180 days, and 65% were discharged to a care home. In multivariable analysis adjusting for age, sex, SIMD, ASA grade, COVID-19 within 30 days, and type of residence, delirium was independently associated with an increased mortality rate within 180 days (Supplementary Table iii). Admission from a higher care setting was associated with a two-fold increased mortality rate within 30 days, and nearly two-fold greater mortality rate within 180 days.
Sensitivity analyses
Of the 1,734 patients with a 4AT score admitted during the study period, 1,185 (68.3%) had an admission 4AT score recorded in the ED. Of these, 272 patients (23.0%) had delirium. Of 1,340 patients admitted directly from home, 975 (72.8%) had an initial 4AT score recorded in the ED. Of these, 119 (12.2%) had delirium. In multivariable analyses adjusting for age, sex, SIMD, ASA grade, and COVID-19 within 30 days, associations of delirium on admission with a higher mortality rate at 180 days, increased likelihood of requiring post-acute inpatient rehabilitation, and readmission within 180 days remained significant (Supplementary Tables iv to vi).
Discussion
This study reported the prevalence of delirium and associated outcomes in patients admitted directly from home with an acute hip fracture, using robust interrogation of standardized, validated routine data. Only 4.8% of patients did not have a 4AT score recorded in their EHR. More than one in seven patients admitted from home had delirium. After adjusting for age, sex, level of social deprivation, ASA grade, and COVID-19 infection within 30 days, delirium was independently associated with an increased mortality rate at 180 days, nearly three-fold increased risk of requiring post-acute rehabilitation following the acute stay, and nearly two-fold increased readmission risk within 180 days. Observed associations persisted in sensitivity analyses, including only patients with a 4AT score recorded in the ED. This study adds to the existing literature, by demonstrating that delirium can feasibly be ascertained at scale in routine practice in a large, near-complete acute hip fracture population, and that clinically ascertained delirium is associated with unfavourable outcomes in these patients.
Few studies have assessed prevalence and outcomes of delirium in patients admitted directly from home, who likely have lower levels of frailty, dementia, and comorbidities associated both with delirium and adverse outcomes, and in whom days at home following admission with a hip fracture has been demonstrated to be a valid and important patient-centred outcome.21,31 The reported prevalence of delirium (14.1%) in this study is slightly lower than reported by Krogseth et al8 (19%). This may reflect the much smaller, selected sample with high levels of missing data in the latter study (85/207 did not have a preoperative delirium assessment), and different assessment methods (4AT embedded in routine practice, versus Confusion Assessment Method-based specialist assessment). In this study, prevalence of delirium in the whole cohort (26.5%; 23.0% in those with 4AT scores in ED) is similar to reported prevalences of POD in other hip fracture cohorts, including studies using NHFD data.1,16-18
Our observed associations of delirium with adverse outcomes are aligned with those reported in a USA-based study using data from the NSQIP, in which preoperative delirium was ascertained retrospectively using ICD-9 codes and outcomes were assessed up to 30 days.32 Our findings highlight the importance of systematic delirium assessment on admission in this vulnerable population, suggesting this could guide early prognostication, facilitate communication with patients and relatives, improve resource allocation, and ensure delivery of specialized multidisciplinary services associated with improved outcomes.33,34 Tools such as the 4AT are quick to use and can be widely implemented by non-specialist clinicians.
There was no significant difference in the prevalence of delirium in patients with COVID-19 infection versus those without, despite delirium being a well-documented symptom of COVID-19 during the time period of this study.35,36 This may reflect the many other precipitating factors for delirium in patients with an acute hip fracture, and/or the relatively small number of patients admitted from home with COVID-19 within 30 days (n = 53). COVID-19 was independently associated with a higher 30-day mortality and need for post-acute inpatient rehabilitation, consistent with the findings from a previous study using the IMPACT dataset.23 Higher rates of delirium were observed in older patients and those with higher ASA grade, consistent with previous studies.37,38 The effect size of delirium was reduced by inclusion of ASA grade in multivariable models, suggesting that observed outcomes of delirium may partly reflect its associations with comorbidity. However, effects of delirium remained significant, suggesting either an independent effect of delirium, or the presence of unmeasured factors. Notably, frailty status and dementia were not measured or controlled for, both of which are associated with increased risk of delirium, and with adverse outcomes.31 ASA grade may partly control for frailty; in a study of postoperative outcomes in patients undergoing unscheduled surgery, associations between the Clinical Frailty Scale (CFS) score and unfavourable outcomes were not independent of ASA grade.39 Our study highlights the unaddressed need for robust prospective frailty assessment in routine hip fracture care, to better stratify and communicate risks. Cognitive impairment has been described as the primary risk factor for institutionalization following hip fracture.40 Linkage of EHR data with other routine datasets could facilitate identification of these and other confounding variables.
This study has several strengths. Delirium was prospectively ascertained, with assessment performed during routine care rather than for research purposes. This supports generalizability. Assessment was performed on admission, using a well-established delirium assessment tool with high sensitivity and specificity for delirium.26 This study used an established mechanism for data collection and verification by trained specialist auditors, which has been used in peer-reviewed studies published by the IMPACT Collaborative and SHFA, as well as in annual reporting by Public Health Scotland endorsed by the Scottish Government.29,30 Hip fracture services at this centre are standardized and delivered according to the Scottish Standards of Care for Hip Fracture Patients (SSCHFP).3 This has led to collection of granular data with high levels of delirium ascertainment. The current study employed an additional data validation process conducted by a study author (AJH), independent of the original auditors. A 180-day follow-up period facilitated assessment of outcomes in a group of patients from home with lower risk of adverse short-term outcomes.
A limitation was that dementia and frailty were not measured and adjusted for. Although a recent study demonstrated that EHRs can be used by non-orthogeriatricians to assign an accurate retrospective frailty score,41 this was not performed due to large patient numbers. Similarly, estimation of dementia prevalence for the whole cohort, for example through accessing GP records, was not performed due to resource constraints. Nevertheless, delirium, as ascertained by a robust assessment tool embedded in routine clinical practice and EHR, provided valuable prognostic information. Although the vast majority of 4ATs were performed at or near admission as per local clinical and national SSCHFP protocols, precise timing of the first 4AT score in relation to the time of admission was not recorded. However, sensitivity analyses were performed including only 4AT scores from the ED (all within a few hours of admission), and key associations remained. This study was conducted in a single centre, which may introduce bias with regard to the study population and clinical management factors. However, the sample is consecutive, relatively large, and the centre delivers care in accordance with the SSCHFP, with performance consistent with other Scottish centres. Data on periprosthetic fractures were not collected for national audit purposes at the time of the study; this may limit generalisability to these patients. Cohort studies involving retrospective data analysis are vulnerable to bias during data aggregation; risk was minimized by two-stage independent validation of the clinical audit data.
In summary, one in seven patients admitted from their own home had delirium on admission assessment. Adjusting for age, sex, ASA grade, level of social deprivation, and COVID-19 infection, delirium in these patients was independently associated with an increased mortality rate at 180 days, increased risk of requiring post-acute rehabilitation, and increased likelihood of being readmitted to acute services within 180 days of discharge home. Assessment for delirium on admission with a tool such as the 4AT should be part of routine care for hip fracture patients. This could guide early prognostication, facilitate proactive onward care planning, and identify modifiable clinical risk factors. Future studies should seek to identify mediators of adverse outcomes for patients admitted with delirium, and link data from the acute admission with other datasets to better characterize risk factors for delirium.
References
1. Lisk R , Yeong K , Enwere P , et al. Associations of 4AT with mobility, length of stay and mortality in hospital and discharge destination among patients admitted with hip fractures . Age Ageing . 2020 ; 49 ( 3 ): 411 – 417 . Crossref PubMed Google Scholar
2. No authors listed . Hip fracture: management . National Institute for Health and Care Excellence . 2011 . https://www.nice.org.uk/guidance/cg124 ( date last accessed 25 May 2023 ). Google Scholar
3. No authors listed . Scottish Standards of Care for Hip Fracture Patients 2019 . Scottish Government . https://www.shfa.scot.nhs.uk/_docs/2019/Scottish-standards-of-care-for-hip-fracture-patients-2019.pdf ( date last accessed 23 May 2023 ). Google Scholar
4. Griffiths R , Babu S , Dixon P , et al. Guideline for the management of hip fractures 2020 . Anaesthesia . 2021 ; 76 ( 2 ): 225 – 237 . Crossref PubMed Google Scholar
5. Oh ES , Li M , Fafowora TM , et al. Preoperative risk factors for postoperative delirium following hip fracture repair: a systematic review . Int J Geriatr Psychiatry . 2015 ; 30 ( 9 ): 900 – 910 . Crossref PubMed Google Scholar
6. Wang C-G , Qin Y-F , Wan X , Song L-C , Li Z-J , Li H . Incidence and risk factors of postoperative delirium in the elderly patients with hip fracture . J Orthop Surg Res . 2018 ; 13 ( 1 ): 186 . Crossref PubMed Google Scholar
7. Wang N-Y , Hirao A , Sieber F . Association between intraoperative blood pressure and postoperative delirium in elderly hip fracture patients . PLoS One . 2015 ; 10 ( 4 ): e0123892 . Crossref PubMed Google Scholar
8. Krogseth M , Wyller TB , Engedal K , Juliebø V . Delirium is a risk factor for institutionalization and functional decline in older hip fracture patients . J Psychosom Res . 2014 ; 76 ( 1 ): 68 – 74 . Crossref PubMed Google Scholar
9. Freter S , Dunbar M , Koller K , MacKnight C , Rockwood K . Prevalence and characteristics of pre-operative delirium in hip fracture patients . Gerontology . 2016 ; 62 ( 4 ): 396 – 400 . Crossref PubMed Google Scholar
10. Costa-Martins I , Carreteiro J , Santos A , et al. Post-operative delirium in older hip fracture patients: a new onset or was it already there? Eur Geriatr Med . 2021 ; 12 ( 4 ): 777 – 785 . Crossref PubMed Google Scholar
11. Werner M , Macke C , Gogol M , Krettek C , Liodakis E . Differences in hip fracture care in Europe: a systematic review of recent annual reports of hip fracture registries . Eur J Trauma Emerg Surg . 2022 ; 48 ( 3 ): 1625 – 1638 . Crossref PubMed Google Scholar
12. No authors listed . Scottish Hip Fracture Audit annual report 2022 (2021 data) . NHS National Services Scotland . August 16 , 2022 . https://www.shfa.scot.nhs.uk/Reports/index.html ( date last accessed 23 May 2023 ). Google Scholar
13. No authors listed . NHFD annual report 2021 . Royal College of Physicians . October 5 , 2021 . https://www.rcplondon.ac.uk/projects/outputs/nhfd-annual-report-2021 ( date last accessed 23 May 2023 ). Google Scholar
14. No authors listed . Annual Report 2022 . Australian & New Zealand Hip Fracture Registry . https://anzhfr.org/wp-content/uploads/sites/1164/2022/08/ANZHFR-2022-Annual-Report-Print-version-v2.pdf ( date last accessed May 2023 ). Google Scholar
15. No authors listed . Irish Hip Fracture Database National Report 2020 . National Office of Clinical Audit . https://s3-eu-west-1.amazonaws.com/noca-uploads/general/NOCA_IHFD_REPORT_2017-2021.pdf ( date last accessed 9 June 2023 ). Google Scholar
16. Uzoigwe CE , O’Leary L , Nduka J , et al. Factors associated with delirium and cognitive decline following hip fracture surgery . Bone Joint J . 2020 ; 102-B ( 12 ): 1675 – 1681 . Crossref PubMed Google Scholar
17. Rajeev A , Railton C , Devalia K . The crucial factors influencing the development and outcomes of postoperative delirium in proximal femur fractures . Aging Med (Milton) . 2022 ; 5 ( 2 ): 94 – 100 . Crossref PubMed Google Scholar
18. Matharu GS , Shah A , Hawley S , et al. The influence of mode of anaesthesia on perioperative outcomes in people with hip fracture: a prospective cohort study from the National Hip Fracture Database for England, Wales and Northern Ireland . BMC Med . 2022 ; 20 ( 1 ): 319 . Crossref PubMed Google Scholar
19. Haynes MS , Alder KD , Toombs C , Amakiri IC , Rubin LE , Grauer JN . Predictors and sequelae of postoperative delirium in a geriatric patient population with hip fracture . J Am Acad Orthop Surg Glob Res Rev . 2021 ; 5 ( 5 ): e20.00221 . Crossref PubMed Google Scholar
20. Albanese AM , Ramazani N , Greene N , Bruse L . Review of postoperative delirium in geriatric patients after hip fracture treatment . Geriatr Orthop Surg Rehabil . 2022 ; 13 : 21514593211058948 . Crossref PubMed Google Scholar
21. McIsaac DI , Talarico R , Jerath A , Wijeysundera DN . Days alive and at home after hip fracture: a cross-sectional validation of a patient-centred outcome measure using routinely collected data . BMJ Qual Saf . 2021 ; bmjqs-2021-013150 . Crossref PubMed Google Scholar
22. Bellelli G , Morandi A , Davis DHJ , et al. Validation of the 4AT, a new instrument for rapid delirium screening: a study in 234 hospitalised older people . Age Ageing . 2014 ; 43 ( 4 ): 496 – 502 . Crossref PubMed Google Scholar
23. Hall AJ , Clement ND , Ojeda-Thies C , et al. IMPACT-Global Hip Fracture Audit: Nosocomial infection, risk prediction and prognostication, minimum reporting standards and global collaborative audit: Lessons from an international multicentre study of 7,090 patients conducted in 14 nations during the COVID-19 pandemic . Surgeon . 2022 ; 20 ( 6 ): e429 – e446 . Crossref PubMed Google Scholar
24. Hall AJ , Clement ND , MacLullich AMJ , Simpson AHRW , White TO , Duckworth AD . The IMPACT of COVID-19 on trauma & orthopaedic surgery provides lessons for future communicable disease outbreaks: minimum reporting standards, risk scores, fragility trauma services, and global collaboration . Bone Joint Res . 2022 ; 11 ( 6 ): 342 – 345 . Crossref PubMed Google Scholar
25. No authors listed . Scottish Index of Multiple Deprivation 2020 . Scottish Government . 2020 . https://www.gov.scot/collections/scottish-index-of-multiple-deprivation-2020/ ( date last accessed 6 June 2023 ). Google Scholar
26. Tieges Z , Maclullich AMJ , Anand A , et al. Diagnostic accuracy of the 4AT for delirium detection in older adults: systematic review and meta-analysis . Age Ageing . 2021 ; 50 ( 3 ): 733 – 743 . Crossref PubMed Google Scholar
27. No authors listed . Delirium: prevention, diagnosis and management in hospital and long-term care . National Institute for Health and Care Excellence . July 28 , 2010 . https://www.nice.org.uk/guidance/cg103 ( date last accessed 31 May 2023 ). Google Scholar
28. Hall AJ , Clement ND , MacLullich AMJ , et al. The delivery of an emergency audit response to a communicable disease outbreak can inform future orthopaedic investigations and clinical practice: lessons from IMPACT Hip Fracture Global Audits . Bone Joint Res . 2022 ; 11 ( 6 ): 346 – 348 . Crossref PubMed Google Scholar
29. Hall AJ , Clement ND , MacLullich AMJ , White TO , Duckworth AD , IMPACT Revisited Group . COVID-19 during the index hospital admission confers a “double-hit” effect on hip fracture patients and is associated with a two-fold increase in 1-year mortality risk . Musculoskeletal Care . 2022 ; 20 ( 3 ): 705 – 717 . Crossref PubMed Google Scholar
30. Kay RS , Hall AJ , Duckworth AD , Clement ND . Socioeconomically-deprived patients suffer hip fractures at a younger age and require more hospital admissions, but early mortality risk is unchanged: The IMPACT Deprivation Study . Musculoskeletal Care . 2022 . Crossref PubMed Google Scholar
31. Wilson JE , Mart MF , Cunningham C , et al. Delirium . Nat Rev Dis Primers . 2020 ; 6 ( 1 ): 90 . Crossref PubMed Google Scholar
32. Agrawal S , Turk R , Burton BN , Ingrande J , Gabriel RA . The association of preoperative delirium with postoperative outcomes following hip surgery in the elderly . J Clin Anesth . 2020 ; 60 : 28 – 33 . Crossref PubMed Google Scholar
33. Farrow L , Hall AJ , Ablett AD , Johansen A , Myint PK . The influence of hospital-level variables on hip fracture outcomes . Bone Joint J . 2021 ; 103-B ( 10 ): 1627 – 1632 . Crossref PubMed Google Scholar
34. Farrow L , Hall A , Aucott L , Holt G , Myint PK . Does quality of care in hip fracture vary by day of admission? Arch Osteoporos . 2020 ; 15 ( 1 ): 52 . Crossref PubMed Google Scholar
35. Zazzara MB , Penfold RS , Roberts AL , et al. Probable delirium is a presenting symptom of COVID-19 in frail, older adults: a cohort study of 322 hospitalised and 535 community-based older adults . Age Ageing . 2021 ; 50 ( 1 ): 40 – 48 . Crossref PubMed Google Scholar
36. Harapan BN , Yoo HJ . Neurological symptoms, manifestations, and complications associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and coronavirus disease 19 (COVID-19) . J Neurol . 2021 ; 268 ( 9 ): 3059 – 3071 . Crossref PubMed Google Scholar
37. Raats JW , van Eijsden WA , Crolla R , Steyerberg EW , van der Laan L . Risk factors and outcomes for postoperative delirium after major surgery in elderly patients . PLoS One . 2015 ; 10 ( 8 ): e0136071 . Crossref PubMed Google Scholar
38. Venkatakrishnaiah NK , Anandkumar UM , Wooly S , Rajkamal G , Gadiyar HB , Janakiraman P . Identification of factors contributing to the development of postoperative delirium in geriatric patients with hip fractures - A prospective study . J Family Med Prim Care . 2022 ; 11 ( 8 ): 4785 – 4790 . Crossref PubMed Google Scholar
39. McGuckin DG , Mufti S , Turner DJ , Bond C , Moonesinghe SR . The association of peri-operative scores, including frailty, with outcomes after unscheduled surgery . Anaesthesia . 2018 ; 73 ( 7 ): 819 – 824 . Crossref PubMed Google Scholar
40. Hawley S , Inman D , Gregson CL , Whitehouse M , Johansen A , Judge A . Predictors of returning home after hip fracture: a prospective cohort study using the UK National Hip Fracture Database (NHFD) . Age Ageing . 2022 ; 51 ( 8 ): afac131 . Crossref PubMed Google Scholar
41. Kay RS , Hughes M , Williamson TR , Hall AJ , Duckworth AD , Clement ND . The Clinical Frailty Scale can be used retrospectively to assess the frailty of patients with hip fracture: a validation study . Eur Geriatr Med . 2022 ; 13 ( 5 ): 1101 – 1107 . Crossref PubMed Google Scholar
Author contributions
R. S. Penfold: Conceptualization, Formal analysis, Visualization, Writing – original draft, Writing – review & editing.
A. J. Hall: Conceptualization, Investigation, Project administration, Writing – original draft, Writing – review & editing.
A. Anand: Writing – review & editing, Supervision.
N. D. Clement: Conceptualization, Writing – review & editing, Supervision.
A. D. Duckworth: Conceptualization, Writing – review & editing, Supervision.
A. M. J. MacLullich: Conceptualization, Writing – review & editing, Supervision.
Funding statement
This research was funded in whole, or in part, by the Wellcome Trust [223499/Z/21/Z]. For the purpose of open access, the author has applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission.
ICMJE COI statement
R. S. Penfold is a fellow on the Multimorbidity Doctoral Training Programme for Health Professionals, which is supported by the Wellcome Trust [223499/Z/21/Z]. R. S. Penfold is also an Editorial Fellow and Editorial Board member for Age and Ageing (unpaid). A. M. J. MacLullich is the main author of the 4AT delirium assessment tool, and has no current or future financial conflicts of interest. A. J. Hall is Deputy Chair of the Scottish Hip Fracture Audit Research Group (unpaid). A. D. Duckworth reports research grants from NIHR, OTA, SORT-IT, Stryker, Smith & Nephew, and Acumed, book royalties from Taylor & Francis and Elsevier, a lecture payment from AgNovos Healthware, payment for an educational event from Smith & Nephew, all of which are unrelated to this study. A. D. Duckworth is also a member of the OTA and OTS Research Committees, an Editorial Board member of The Bone & Joint Journal, Bone & Joint Research, and Bone & Joint 360, and an Associate Editor for Trials, OTAI, and JBJS Case Connector. N. D. Clement is an Editorial Board member of The Bone & Joint Journal, Bone & Joint Research, and Arthroplasty (BMC). A. Anand is a co-investigator into multimorbidity and delirium, funded through institutional payments from NIHR and Dunhill Medical Trust, unrelated to this study.
Data sharing
The datasets generated and analyzed in the current study are not publicly available due to data protection regulations. Access to data is limited to the researchers who have obtained permission for data processing. Further inquiries can be made to the corresponding author.
Ethical review statement
Data were collected and handled as part of service evaluation in accordance with UK Caldicott principles.
Open access funding
The open access fee for this article was funded by the Wellcome Trust.
Follow R. S. Penfold @rosespenfold
Follow A. J. Hall @andrewhallortho and @IMPACTaudits
Follow A. Anand @atula_tweets
Follow A. D. Duckworth @DuckworthOrthEd and @EdinOrthopaedic
Follow A. M. J. MacLullich @A_MacLullich
Supplementary material
Supplementary analyses for patients admitted from higher care settings (including care homes, rehabilitation facilities, and inpatient transfer) and sensitivity analyses including only patients with a 4AT recorded in the emergency department.
© 2023 Author(s) et al. This article is distributed under the terms of the Creative Commons Attributions (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original author and source are credited.
