Abstract
Background
frailty screening facilitates the stratification of older adults at most risk of adverse events for urgent assessment and subsequent intervention. We assessed the validity of the Identification of Seniors at Risk (ISAR), Clinical Frailty Scale (CFS), Programme on Research for Integrating Services for the Maintenance of Autonomy seven item questionnaire (PRISMA-7) and InterRAI-ED at predicting adverse outcomes at 30 days and 6 months amongst older adults presenting to the Emergency Department (ED).
Methods
a prospective cohort study of adults ≥65 years who presented to the ED was conducted. The ISAR, CFS, PRISMA-7 and InterRAI-ED were assessed. Blinded follow-up telephone interviews were completed at 30 days and 6 months to assess the incidence of mortality, ED re-attendance, hospital readmission, functional decline and nursing home admission. The sensitivity, specificity, negative predictive value and positive predictive value of the screening tools were calculated using 2 × 2 tables.
Results
a total of 419 patients were recruited; 47% female with a mean age of 76.9 (Standard deviation = 7.2). The prevalence of frailty varied across the tools (CFS 57% versus InterRAI-ED 70%). At 30 days, the mortality rate was 5.1%, ED re-attendance 18.1%, hospital readmission 14%, functional decline 47.6% and nursing home admission 7.1%. All tools had a high sensitivity and positive predictive value for predicting adverse outcomes.
Conclusion
older adults who screened positive for frailty were at significantly increased risk of experiencing an adverse outcome at 30 days with the ISAR being the most sensitive tool. We would recommend the implementation of the ISAR in the ED setting to support clinicians in identifying older adults most likely to benefit from specialised geriatric assessment and intervention.
Keywords: screenings, instruments, adverse, outcomes, frailty, older people
Key Points
High prevalence of frailty in older adults presenting to the Emergency Department.
Screening tools predict adverse outcomes of mortality, Emergency Department re-attendance, nursing home admission and functional decline.
Screening tools may be a method to risk stratify older adults for comprehensive geriatric assessment.
Background
Older adults presenting to the Emergency Department (ED) comprise of an increasingly complex cohort of patients [1]. Given their higher burden of comorbidities, it is important to provide adequate specialist support. There is an ever-growing population, it is necessary to screen those most at risk of adverse outcomes, who are most likely to benefit from Comprehensive Geriatric Assessment (CGA) that is resource intensive [2, 3].
Screening tools are frequently used in the ED setting to risk stratify patient cohorts. There is great variability in the screening tools used, particularly in relation to frailty screening. In a scoping review by Theou et al., 89 measures were used to indicate frailty, these included 13 established frailty tools and 35 non-frail tools that were used as surrogate markers [4]. The multitude of tools assessed in research makes choice of a single tool for the ED setting difficult. O’Caoimh et al. assessed the diagnostic accuracy of three such tools in the ED environment when compared to the reference standard of CGA, namely the Clinical Frailty Scale (CFS), Identification of Seniors at Risk Tool (ISAR) and the Programme on Research for Integrating Services for the Maintenance of Autonomy seven item questionnaire (PRISMA-7) [5]. This study found that all scales were accurate and reliable at identifying those who were frail. However, the study did not assess their predictive accuracy to identify adverse outcomes [5]. A systematic review by Jorgenson et al. assessed the predictive accuracy of frailty tools in the ED setting. These frailty screening tools predicted those who were at risk of hospitalisation, nursing home admission, mortality and prolonged length of stay over a 6-month period after an initial ED visit but not 30-day readmission [6].
The British Geriatric Society (BGS) recommends frailty assessment for all older adults in primary care and community settings if they are in contact with medical services [7]. Several tools are discussed in this report but no one tool is proposed for widespread use. The ‘Same-Day Acute Frailty Services’ report published by National Health Service (NHS) Improvement, NHS England, the Ambulatory Emergency Care Network and the Acute Frailty Network advocates for the use of the CFS within 30 min of an individual over 65 years presenting to acute services [8]. In the Irish setting, the National Clinical Programme for Integrated Care of Older People highlights the importance of identifying individuals at risk of frailty and completing Comprehensive Geriatric Assessment [9]. This is further expanded on in the Enhanced Community Care Implementation Guidance which mandates frailty screening at the front door ED setting [10]. Similar to the BGS, no one frailty screening tool is recommended for implementation.
We aimed to assess the validity of commonly used screening tools including the ISAR, CSF, PRISMA-7 and InterRAI-ED at predicting the adverse outcomes of mortality, ED re-attendance, hospital readmission, functional decline and nursing home placement at 30 days and 6 months amongst older adults who presented to the ED at a University Teaching Hospital in Ireland.
Methods
Study design and setting
This was a prospective cohort study. The STrengthening the Reporting of OBservational studies in Epidemiology standardised reporting guidelines were used to ensure a standardised approach to reporting our findings [11]. The study took place in the ED of a University Teaching Hospital catering for medical and surgical patients in a catchment area of 465,000 people. The protocol for this study is published elsewhere [12].
Population
All adults aged ≥65 years who presented to the ED between September 2019 and April 2020 were considered eligible for enrolment if they met the following eligibility criteria: a Manchester Triage Category of 2–5 [13], resident in the catchment area and English speaking. Older adults were informed of the study by the research nurse (RN). Capacity was presumed in all patients, however, in the context of a clinical concern regarding a patient’s capacity to consent, where the patient agreed in principle to participate, the study was discussed with the patient’s next of kin. Written informed consent was obtained from all participants.
Patients were not eligible for inclusion if the patients or their caregivers were unable to speak English and therefore unable to consent or provide baseline demographic information. Patients who were acutely unwell were excluded. Recruitment by the RN was 8 am to 5 pm Monday to Friday, therefore patients presenting outside these hours were excluded.
Initial assessment
Baseline demographic information including age, gender and existing comorbidities (Charlson Comorbidity Score) [14] was collected by the RN. Four frailty screening tools were then administered during this index visit namely the CFS, PRISMA-7, ISAR and InterRAI-ED. The CFS is a pictorial scale rated from 1 to 9 that is based on the patient’s functional status 2 weeks prior to assessment [15]. The PRISMA-7 is a seven-item questionnaire to identify those at risk of frailty [16]. The ISAR is a six-item questionnaire with yes/no answers that is validated to predict mortality, ED re-attendance, hospital readmission, functional decline and nursing home admission [17]. The InterRAI-ED is an app-based screening tool that forms part of the InterRAI Management system [18]. Functional assessment at baseline was documented using the Barthel Index [19].
Outcome assessment
Outcomes were assessed by a blinded RN by telephone interview at 30 days and 6 months. Data collected included mortality, ED re-attendance, hospital readmission, subjective change in functional ability noted by the patient, nursing home admission and healthcare utilisation (GP visit, Public Health Nurse visit or health and social care professional review). Hospital management systems (PAS system and MAXIMS system) were used to determine the outcomes of mortality, ED revisit and hospital readmission for those who did not answer the follow-up phone call. The RN tried to contact patients five times at differing times of the day to maximise data capture.
Statistical analysis
A large sample was required to ensure sufficient observations across the adverse outcome categories. The sample size was calculated using logistic regression estimating a sample size of at least 400 was required for an analysis with six potential predictors as detailed in the study protocol [20]. Descriptive statistics were analysed, using aggregate anonymised participant data linked across baseline, 30 days and 6 months follow-up. Categorical data were described by counts and percentages. Continuous data that approximated a normal distribution were described using means and standard deviations, otherwise the median and interquartile range were presented. The unadjusted relative risk (RR) and corresponding 95% confidence interval (CI) for adverse outcomes (mortality, ED re-admission, hospital readmission, functional decline and nursing home admission) at 30 days and 6 months, were presented at pre-specified cut-offs scores for the four frailty tools. Sensitivity, specificity, negative predictive and positive predictive values and corresponding 95% CIs were calculated to determine the use of the frailty tools to predict adverse outcomes in older adults in the ED. A 5% level of significance was used for all statistical tests. All statistical analysis were undertaken using SPSS Version 24.
Patient and public involvement
An older adult Patient and Public Involvement (PPI) representative was consulted in relation to the aims of the study and the outcome measures most relevant to older adults. On completion of the first draft of the paper, the results of the study were discussed with the PPI representative and their perspective on meaningful outcome measures in particular were included in further drafts. They reviewed the final draft of the paper and are named as an author of the submitted manuscript. All meetings were conducted over telephone due to COVID restrictions.
Results
A total of 419 patients were recruited. Figure 1 describes the flow of patients in the study. Table 1 illustrates baseline patient characteristics and measures including frailty scores (ISAR, PRISMA-7, CFS and InterRAI), Barthel Index, and the Charlson co-morbidity score. Females represented 46.7% of the total population and the mean age was 76.9 years [standard deviation (SD) = 7.2 years]. The majority of patients were living rurally (74.2%), with family (65.0%) and were white Irish (95.9%). One third (34%) of patients had visited their GP prior to review in the ED and half of the patients presented by ambulance. The median Barthel score was 18. Overall, the prevalence of frailty in the cohort identified by ISAR, PRISMA-7, CFS and InterRAI-ED was 63.6%, 59%, 57% and 70%, respectively.
Figure 1.
PRISMA diagram.
Table 1.
Patient demographics and other characteristics
| Demographics | N (%) | |
|---|---|---|
| Gender | Male | 223 (53.3) |
| Female | 195 (46.7) | |
| Age (years) | 76.9 (7.2 SD) | |
| Location | Rural | 305 (74.2) |
| Urban | 106 (25.8) | |
| Ethnicity | White Irish | 395 (95.9) |
| Other | 17 (4.1) | |
| Marital Status | Married/Partner | 203 (49.4) |
| Single | 51 (12.4) | |
| Separated/Divorced | 29 (7.1) | |
| Widowed | 128 (31.1) | |
| Living Status | Family | 267 (65.0) |
| Alone | 125 (30.4) | |
| Nursing home | 15 (3.6) | |
| Other | 4 (1.0) | |
| Education | Primary | 178 (43.8) |
| Secondary | 162 (39.9) | |
| Tertiary | 66 (16.3) | |
| ED entry mode | Ambulance | 216 (51.5) |
| Car | 191 (45.6) | |
| Other | 12 (0.03) | |
| Referral Source | Self/Family | 208 (50.7) |
| GP | 142 (34.6) | |
| Other Healthcare | 60 (14.6) | |
| Patient Measures | Median (IQR) | |
| ISAR (0–6) | 2 (2) | |
| Inter Rai (0–12) | 3 (5) | |
| PRISMA-7 (0–7) | 2 (3) | |
| CFS (0–9) | 5 (2) | |
| Barthel index (0–20) | 18 (7) | |
| Charlson (0–24) | 2 (3) | |
Follow-up at 30 days and 6 months
The incidence of adverse outcomes at 30 days and 6 months follow-up are presented in Table 2.
Table 2.
Patient follow-up
| Outcome | Follow-Up | |
|---|---|---|
| 30-days n (%) | 6-months n (%) | |
| Mortality | 21/415 (5.1%) | 37/416 (8.9%) |
| ED re-attendance | 75/414 (18.1%) | 127/387 (32.8%) |
| Unplanned Hospital Visit | 58/414 (14.0%) | 108/385 (28.1%) |
| Nursing Home admission | 28/392 (7.1%) | 22/379 (5.8%) |
| Functional decline | 186/391 (47.6%) | 193/379 (50.9%) |
Note different denominators were due to data on outcomes obtained from variable sources, i.e. telephone interview and hospital databases.
At 30 days, the mortality rate was 5.1%. ED re-attendance was 18.1%. A total of 14% had an unplanned hospital readmission, 47.6% of the cohort reported functional decline and 7.1% had a nursing home admission.
The rates of adverse outcomes at 6 months, were as follows: mortality rate was 8.9%, 32.8% had an ED re-attendance, 28.1% had an unplanned hospital readmission, 50.9% self-reported a decline in functional ability and 5.8% had a nursing home admission. There were low levels of missing data for the outcomes of mortality (<1%) and ED re-attendance at 30 days (<1%). Due to mortality, follow-up data were not available on the other adverse outcomes for some patients (<10%). No statistical adjustment was required given the low levels of missing data. Outcome data were compiled from both telephone interviews and hospital database interrogation explaining the discrepancy in numbers of patients in each group for outcome and at different timepoints.
Risk of adverse outcomes at 30 days and 6 months
The risk of adverse outcomes at 30 days for each cut-off of the frailty tools is presented in Table 3. Older adults who screen positive for frailty are more likely to experience an adverse outcome. A score of ≥2 on the ISAR and ≥5 on the CFS indicates that an older person is ˃14 times more likely to be dead at 30 days in comparison to their non-frail counterparts.
Table 3.
Risk of adverse outcomes at 30 days and 6 months
| Frailty Tool & Cut-off Score | Follow-up Period | ||||
|---|---|---|---|---|---|
| 30-Days | 6 Months | ||||
| N (%) | RR (95% CI) | N (%) | RR (95% CI) | ||
| Mortality | |||||
| ISAR | <2 | 0/151 (0%) | 1/152 (0.7%) | ||
| ≥2 | 21/264 (8%) | NC (1.47, 395.36) | 36/264 (13.6%) | 20.73 (2.87, 149.67) | |
| PRISMA | <3 | 1/170 (0.6%) | 2/171 (1.2%) | ||
| ≥3 | 20/245 (8.2%) | 13.87 (1.88, 102.42) | 35/245 (14.3%) | 12.21 (2.98, 50.10) | |
| CFS | <5 | 1/174 (0.6%) | 1/174 (0.6%) | ||
| ≥5 | 20/240 (8.3%) | 14.5 (1.96, 107.02) | 36/204 (15%) | 26.25 (3.63, 189.63) | |
| Inter RAI | <3 | 1/124 (0.8%) | 1/124 (0.8%) | ||
| ≥3 | 20/291 (6.9%) | 8.52 (.16, 62.80) | 36/255 (12.4%) | 15.46 (2.14, 111.54) | |
| ED Re-attendance | |||||
| ISAR | <2 | 18/154 (11.7%) | 35/151(23.2%) | ||
| ≥2 | 57/260(21.9%) | 1.88 (1.15, 3.06) | 92/236 (39%) | 1.68 (1.21, 2.34) | |
| PRISMA | <3 | 24/172(14%) | 45/170 (26.5%) | ||
| ≥3 | 51/242 (21.1%) | 1.51 (0.97, 2.35) | 82/217 (37.8%) | 1.43 (1.05, 1.93) | |
| CFS | <5 | 25/176 (14.2%) | 42/175(24%) | ||
| ≥5 | 50/237 (21.1%) | 1.49 (0.96, 2.30) | 85/212(40.1%) | 1.67 (1.22, 2.28) | |
| Inter RAI | <3 | 17/126(13.5%) | 33/125(26.4%) | ||
| ≥3 | 58/288 (20.1%) | 1.49 (0.91, 2.46) | 94/262 (35.9%) | 1.36 (0.97, 1.90) | |
| Hospital Readmission | |||||
| ISAR | <2 | 10/153 (6.5%) | 27/150 (18%) | ||
| ≥2 | 48/261 (18.4%) | 2.81 (1.47, 5.40) | 81/235 (34.5%) | 1.91 (1.30, 2.81) | |
| PRISMA | <3 | 13/172 (7.6%) | 38/169 (22.5%) | ||
| ≥3 | 45/242 (18.6%) | 2.46 (1.37, 4.42) | 70/216 (32.4%) | 1.44 (1.03, 2.02) | |
| CFS | <5 | 14/176(8%) | 34/174 (19.5%) | ||
| ≥5 | 44/237 (18.6%) | 2.33 (1.32, 4.12) | 74/211 (35.1%) | 1.79 (1.26, 2.55) | |
| Inter RAI | <3 | 8/126 (6.3%) | 25/124 (20.2%) | ||
| ≥3 | 50/288 (17.4%) | 2.73 (1.34, 5.60) | 83/261 (31.8%) | 1.58 (1.07, 2.34) | |
| Nursing Home Admission | |||||
| ISAR | <2 | 4/149 (2.7%) | 5/148 (3.4%) | ||
| ≥2 | 24/243 (9.9%) | 3.68 (1.30, 10.39) | 17/231 (7.4%) | 2.18 (0.82, 5.78) | |
| PRISMA | <3 | 4/168 (2.4%) | 6/166 (3.6%) | ||
| ≥3 | 24/224(10.7%) | 4.5 (1.59, 12.72) | 16/213 (7.5%) | 2.08 (0.83, 5.19) | |
| CFS | <5 | 5/173 (2.9%) | 9/172 (5.2%) | ||
| ≥5 | 23/219 (10.5%) | 3.63 (1.41, 9.36) | 13/207 (6.3%) | 1.20 (0.53 2.74) | |
| Inter RAI | <3 | 3/123 (2.4%) | 5/123 (4.1%) | ||
| ≥3 | 25/269 (9.3%) | 3.81 (1.172, 12.38) | 17/257 (6.6%) | 1.61 (0.61 4.27) | |
| Functional Decline | |||||
| ISAR | <2 | 58/148 (39.2%) | 56/148 (37.8) | ||
| ≥2 | 128/243 (52.7%) | 1.34 (1.06, 1.70) | 137/231 (59.3%) | 1.57 (1.24, 1.98) | |
| PRISMA | <3 | 64/167 (38.3%) | 63/166 (38%) | ||
| ≥3 | 122/224 (54.5%) | 1.42 (1.13, 1.78) | 130/213 (61%) | 1.61 (1.29, 2.01) | |
| CFS | <5 | 73/172 (42.4%) | 68/172(39.5%) | ||
| ≥5 | 113/219 (51.6%) | 1.22 (0.98, 1.51) | 125/207(60.4%) | 1.53 (1.23, 1.89) | |
| Inter RAI | <3 | 41/122 (33.6%) | 42/122 (34.4%) | ||
| ≥3 | 145/269 (53.9%) | 1.60 (1.22, 2.11) | 151/257 (58.8%) | 1.71 (1.31, 2.23) | |
Individuals who were frail according to any of the screening tools were more likely to have adverse outcomes at both time points with a similar performance amongst tools. The PRISMA-7 performed best at predicting nursing home admission at 30 days. Those who screened positive for frailty on the ISAR had the highest risk of nursing home admission at 6 months with a relative risk of 2.18 (0.82, 5.78). The risk of functional decline was greater amongst frail older adults.
Sensitivity and specificity analysis
Table 4 presents the sensitivity and specificity for each frailty tool for the adverse outcomes at 30 days and 6 months. All four frailty screening tools assessed had at least 95% sensitivity for mortality at both 30 days and 6 months. Specificity was much lower. In addition, all screening tools had 82% sensitivity or more for nursing home admission at 30 days. There was high negative predictive value for all tools with the ISAR having the highest negative predictive value for mortality and ED re-attendance.
Table 4.
Sensitivity, specificity, positive predictive value and negative predictive value analysis
| 3 month follow-up | 6 month follow-up | |||||||
|---|---|---|---|---|---|---|---|---|
| Sensitivity (95% CI) | Specificity (95% CI) | PPV | NPV | Sensitivity (95% CI) | Specificity (95% CI) | PPV | NPV | |
| Mortality | ||||||||
| ISAR ≥2 PRISMA ≥3 CFS ≥5 Inter RAI ≥3 |
1.00 (0.84, 1.00) |
0.38 (0.34, 0.43) |
0.08 | 1.00 | 0.97 (0.86, 1.00) |
0.40 (0.35, 0.45) |
0.14 | 0.99 |
| 0.95 (0.76, 1.00) |
0.43 (0.38, 0.48) |
0.08 | 0.99 | 0.95 (0.82, 0.99) |
0.45 (0.40, 0.50) |
0.14 | 0.99 | |
| 0.95 (0.76, 1.00) |
0.44 (0.39, 0.49) |
0.08 | 0.99 | 0.97 (0.86, 1.00) |
0.46 (0.41, 0.51) |
0.15 | 0.99 | |
| 0.95 (0.76, 1.00) |
0.31 (0.27, 0.36) |
0.07 | 0.99 | 0.97 (0.86, 1.00) |
0.33 (0.28, 0.38) |
0.12 | 0.99 | |
| ED Re-attendance | ||||||||
| ISAR ≥2 PRISMA ≥3 CFS ≥5 Inter RAI ≥3 |
0.76 (0.65, 0.85) |
0.40 (0.35, 0.46) |
0.22 | 0.88 | 0.72 (0.64, 0.80) |
0.45 (0.38, 0.51) |
0.39 | 0.77 |
| 0.68 (0.56, 0.78) |
0.44 (0.38, 0.49) |
0.21 | 0.86 | 0.65 (0.56, 0.73) |
0.48 (0.42, 0.54) |
0.38 | 0.74 | |
| 0.67 (0.55, 0.77) |
0.45 (0.39, 0.50) |
0.21 | 0.86 | 0.67 (0.58, 0.75) |
0.51 (0.45, 0.57) |
0.40 | 0.76 | |
| 0.77 (0.66, 0.86) |
0.32 (0.27, 0.37) |
0.20 | 0.87 | 0.74 (0.65, 0.81) |
0.35 (0.30, 0.42) |
0.36 | 0.74 | |
| Hospital Readmission | ||||||||
| ISAR ≥2 PRISMA ≥3 CFS ≥5 Inter RAI ≥3 |
0.83 (0.71, 0.91) |
0.40 (0.35, 0.45) |
0.18 | 0.93 | 0.75 (0.66, 0.83) |
0.44 (0.38, 0.50) |
0.34 | 0.82 |
| 0.78 (0.65, 0.87) |
0.45 (0.39, 0.50) |
0.19 | 0.92 | 0.65 (0.55, 0.74) |
0.47 (0.41, 0.53) |
0.32 | 0.78 | |
| 0.76 (0.63, 0.86) |
0.46 (0.40, 0.51) |
0.19 | 0.92 | 0.69 (0.59, 0.77) |
0.51 (0.44, 0.57) |
0.35 | 0.80 | |
| 0.86 (0.75, 0.94) |
0.33 (0.28, 0.38) |
0.17 | 0.94 | 0.77 (0.68, 0.84) |
0.36 (0.30, 0.42) |
0.32 | 0.80 | |
| Nursing Home Admission | ||||||||
| ISAR ≥2 PRISMA ≥3 CFS ≥5 Inter RAI ≥3 |
0.86 (0.67, 0.96) |
0.40 (0.35, 0.45) |
0.10 | 0.97 | 0.77 (0.55, 0.92) |
0.40 (0.35, 0.45) |
0.07 | 0.97 |
| 0.86 (0.67, 0.96) |
0.45 (0.40, 0.50) |
0.11 | 0.98 | 0.73 (0.50, 0.89) |
0.45 (0.40, 0.50) |
0.08 | 0.96 | |
| 0.82 (0.63, 0.94) |
0.46 (0.41, 0.51) |
0.11 | 0.97 | 0.59 (0.36, 0.79) |
0.46 (0.40, 0.51) |
0.06 | 0.95 | |
| 0.89 (0.72, 0.98) |
0.33 (0.28, 0.38) |
0.09 | 0.98 | 0.77 (0.55, 0.92) |
0.33 (0.28, 0.38) |
0.07 | 0.96 | |
| Functional Decline | ||||||||
| ISAR ≥2 PRISMA ≥3 CFS ≥5 Inter RAI ≥3 |
0.69 (0.62, 0.75) |
0.44 (0.37, 0.51) |
0.53 | 0.61 | 0.71 (0.64, 0.77) |
0.49 (0.42, 0.57) |
0.59 | 0.62 |
| 0.66 (0.58, 0.72) |
0.50 (0.43, 0.57) |
0.54 | 0.62 | 0.67 (0.60, 0.74) |
0.55 (0.48, 0.63) |
0.61 | 0.62 | |
| 0.61 (0.53, 0.68) |
0.48 (0.41, 0.55) |
0.52 | 0.58 | 0.65 (0.58, 0.71) |
0.56 (0.48, 0.63) |
0.60 | 0.60 | |
| 0.78 (0.71, 0.84) |
0.40 (0.33, 0.47) |
0.54 | 0.66 | 0.78 (0.72, 0.84) |
0.43 (0.36, 0.50) |
0.59 | 0.66 | |
Discussion
In this prospective cohort study, the prevalence of frailty amongst a cohort of older adults who presented to the ED, during 8 am to 5 pm from Monday to Friday, ranged from 57% to 70%. Despite this, the median Barthel score was 18 indicating relative pre-morbid independence. In keeping with the high prevalence of frailty, one in five of the overall cohort represented to ED within 30 days and nearly one in 10 patients died at 6 months. Older adults who screened positive for frailty on any of these tools were more likely to experience all of these adverse outcomes at 30 days and 6 months. The incidence of both mortality and ED re-attendance was higher in the frail group regardless of screening tool administered. The ISAR tool identified a higher risk of mortality, ED re-attendance and hospital readmission at 30 days.
All tools studied had a high sensitivity, informing healthcare providers on the low risk of adverse outcomes after discharge in a non-frail group of older adults. The sensitivities of the individual tools differed depending on the outcomes assessed. ISAR was most sensitive for mortality, and second to the InterRAI ED tool for all other outcomes. The CFS and PRISMA-7 had higher specificities for all outcomes (43–55% for PRISMA-7 and 45–56% for CFS). All of the tools had low specificity and positive predictive values for predicting adverse outcomes in frail older adults. However, as a screening tool, it is important to ensure that those who screen negatively are less likely to have an adverse outcome so the sensitivity of the tool is more important than the specificity. All tools have good sensitivity at predicting older adults who are most likely to have adverse outcomes, with the ISAR and the InterRAI ED having the best sensitivity across all outcomes. The negative predictive value was excellent with ISAR performing best at correctly identifying those who did not have a negative outcome. Older patients who present to the ED can commonly be undifferentiated in nature and it is necessary to correctly identify those who are at risk of adverse outcomes. These tools albeit with low specificity can do this with high sensitivity and high negative predictive values.
The ISAR, CFS and PRISMA-7 are commonly used in the Irish ED clinical setting. A previous systematic review determining the predictive accuracy of the ISAR showed that it has moderate predictive accuracy for adverse outcomes [20]. The PRISMA-7 has been reviewed in individual cohorts and a systematic review determining its predictive accuracy is currently ongoing [21]. The CFS has been studied extensively in multiple different populations and is predictive of adverse outcomes in hospitalised patients [22], surgical patients [23] and those with COVID-19 [24, 25]. These studies showed that those who screen frail on the CFS had more adverse outcomes. Conversely, InterRAI-ED, an InterRAI management systems domain, is not in routine clinical practice in this setting and there is conflicting evidence of its predictive value. Gretarsdottir et al. reported 100% sensitivity for mortality in the Icelandic population for the Inter-RAI-ED, implementation made feasible with the use of an electronic handheld device [26]. Conversely, Michalski-Monnerat et al., demonstrated that the InterRAI-ED Screener in a Swiss cohort had poor prediction of adverse outcomes including hospital admission (28.8% sensitivity), length of hospital stay (26.3% sensitivity) and 30-day hospital readmission (26.1% sensitivity) [27]. Furthermore, the predictive value for ED re-attendance in an Australian cohort was similarly poor (AUROC 0.55, P = 0.09) [28]. The lack of digitalisation in the majority of Irish ED settings would make the InterRAI ED difficult to implement in clinical practice where handheld devices are not readily available. Furthermore, the lack of integration of IT systems across healthcare settings is also barrier to implementation.
Clinical implications
Within the ED setting, frailty screening has the potential to identify those at highest risk of future adverse events for whom dedicated intervention such as CGA may be considered. This can be achieved by all tools with high sensitivity. CGA has demonstrated improved outcomes in patients admitted to an acute geriatric ward setting [3]. Whilst the beneficial value of CGA has not been consistently demonstrated in the ED setting, shortly after ED discharge [29] or in community dwelling older adults [30] this may in fact primarily relate to the lack of assessment of CGA effectiveness in a high-risk frail cohort. Methodological issues may have also diluted the benefits such as inadequate implementation of study interventions particularly challenges in timely implementation extending beyond ED discharge. This study identifies a cohort of older adults who are at high risk of adverse outcomes including mortality, functional decline, nursing home admission, ED re-attendance and rehospitalisation. This provides a rational for resource allocation of effective interventions.
Due to the competing interests in the ED department, it can be difficult to implement screening strategies. An upcoming qualitative evidence synthesis will explore the barriers and facilitators to screening in the ED. Barry et al. describe the perceived benefits and difficulties regarding screening in a busy ED setting [31]. It is essential to develop strategies to encourage implementation and to support staff to engage with screening programmes.
Overall, frailty screening tools provide a robust method to select a high-risk cohort for assessment and intervention, with the ISAR and InterRAI ED most sensitive in this study. In particular, the ISAR tool is quick and easy to employ in the ED triage environment.
Strengths and limitations
This study has numerous strengths including: the prospective study design; blinded outcome assessment; objective measurement of outcomes measures such as mortality and admission from hospital databases and medical records to avoid recall bias; clinically relevant adverse events; along with robust PPI involvement informing the study from outset to manuscript preparation. There were some limitations noted: recruitment extended across the daytime working week, so this may not be representative of those patients who present outside of normal working hours; acutely unwell patients were excluded, who may be at even greater risk of future adverse outcomes; self-reported functional ability was assessed at 30 days and 6 months, which is subjective and may incur recall bias. The majority of the population lived with their family or a partner, which may have impacted their healthcare utilisation and incidence of functional decline. Furthermore, the population mainly comprised of white Irish and this may affect generalisability of the findings outside of this cohort. Of note, the diagnostic accuracy of the screening tools was not assessed as part of the study but this has recently been investigated by O’Caoimh et al. in a similar cohort [5].
Conclusion
The ISAR, PRISMA-7, CFS and InterRAI-ED are frailty screening tools that identify older adults at a high risk of experiencing adverse outcomes at 30 days and 6 months. In particular, the ISAR is a brief, practical frailty screening tool which has demonstrated a high sensitivity for prediction of key adverse outcomes in older adults in the ED setting. This tool is suitable for intervention within the busy ED context and would distinguish a vulnerable cohort who may benefit from targeted resource intensive specialist geriatric assessment and intervention.
Contributor Information
Aoife Leahy, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; Department of Ageing and Therapeutics, University Hospital Limerick, Limerick, Ireland.
Gillian Corey, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; ALERT, Department of Emergency Medicine, University Hospital Limerick, Limerick, Ireland.
Helen Purtill, Department of Mathematics & Statistics, University of Limerick, Limerick, Ireland.
Aoife O’Neill, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; Central Statistics Office, Cork, Ireland.
Collette Devlin, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland.
Louise Barry, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; School of Nursing and Midwifery, Faculty of Education and Health Sciences, University of Limerick, Limerick, Ireland.
Niamh Cummins, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; School of Medicine, University of Limerick, Limerick, Ireland.
Ahmed Gabr, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; Department of Ageing and Therapeutics, University Hospital Limerick, Limerick, Ireland.
Abdirahman Mohamed, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; Department of Ageing and Therapeutics, University Hospital Limerick, Limerick, Ireland.
Elaine Shanahan, Department of Ageing and Therapeutics, University Hospital Limerick, Limerick, Ireland.
Denys Shchetkovsky, ALERT, Department of Emergency Medicine, University Hospital Limerick, Limerick, Ireland.
Damien Ryan, ALERT, Department of Emergency Medicine, University Hospital Limerick, Limerick, Ireland; School of Medicine, University of Limerick, Limerick, Ireland.
Monica O’Loughlin, Ageing Research Centre, University of Limerick, Limerick, Ireland.
Margaret O'Connor, Department of Ageing and Therapeutics, University Hospital Limerick, Limerick, Ireland.
Rose Galvin, School of Allied Health, Faculty of Education and Health Sciences, Ageing Research Centre, Health Research Institute, University of Limerick, Limerick, Ireland; Ageing Research Centre, University of Limerick, Limerick, Ireland.
Declaration of Conflicts of Interest
None.
Declaration of Sources of Funding
Health Research Board (ILP-HSR-2017-014). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
Data Availability Statement
Anomymised data is stored at Open Science Framework, available at doi 10.17605/OSF.IO/4NG6H.
References
- 1. Thompson D, Rumley-Buss M, Conroy S. Transforming emergency services for frail older people in hospital. Nurs Manag (Harrow) 2015; 22: 18–9. [DOI] [PubMed] [Google Scholar]
- 2. Ellis G, Marshall T, Ritchie C. Comprehensive geriatric assessment in the emergency department. Clin Interv Aging 2014; 9: 2033. 10.2147/CIA.S29662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ellis G, Gardner M, Tsiachristas A et al. Comprehensive geriatric assessment for older adults admitted to hospital. Cochrane Database Syst Rev 2017; 2017: 9. 10.1002/14651858.CD006211.pub3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Theou O, Campbell S, Malone ML, Rockwood K. Older adults in the emergency department with frailty. Clin Geriatr Med 2018; 34: 369–86. [DOI] [PubMed] [Google Scholar]
- 5. O'Caoimh R, Costello M, Small C et al. Comparison of frailty screening instruments in the emergency department. Int J Environ Res Public Health 2019; 16: 3626. 10.3390/ijerph16193626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Jørgensen R, Brabrand M. Screening of the frail patient in the emergency department: a systematic review. Eur J Intern Med 2017; 45: 71–3. [DOI] [PubMed] [Google Scholar]
- 7. Turner G, Clegg A, Youde J. Fit for Frailty-Consensus Best Practice Guidance for the Care of Older People Living in Community and Outpatient Settings-a Report from the British Geriatrics Society 2014. London: British Geriatrics Society, 2014. [Google Scholar]
- 8. Same Day Frailty Services . Available at https://www.england.nhs.uk/wp-content/uploads/2021/02/SDEC_guide_frailty_May_2019_update.pdf
- 9. Making a start in Integrated Care for Older Persons . A Practical Guide to the Local Implementation of Integrated Care Programmes for Older Persons. Health Service Executive. Available at https://www.icpop.org/_files/ugd/29601c_d8cfeac3dc5343c9945517a4995cd06b.pdf.
- 10. Enhanced Community Care Implementation Guidance. Health Service Executive. Available at https://www.icpop.org/_files/ugd/29601c_9cb8a3787eef4e4ebccefbdea164353a.pdf.
- 11. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–7. [DOI] [PubMed] [Google Scholar]
- 12. Leahy A, Corey G, O'Neill A et al. Screening instruments to predict adverse outcomes for undifferentiated older adults attending the emergency department: protocol for a prospective cohort study. HRB Open Res 2021; 4: 2. 10.12688/hrbopenres.13131.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Mackway-Jones K, ed. Emergency Triage. London: BMJ Publishing, 1997. [Google Scholar]
- 14. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994; 47: 1245–51. [DOI] [PubMed] [Google Scholar]
- 15. Rockwood K, Song X, MacKnight C et al. A global clinical measure of fitness and frailty in elderly people. Cmaj 2005; 173: 489–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Raîche M, Hébert R, Dubois MF. PRISMA-7: a case-finding tool to identify older adults with moderate to severe disabilities. Arch Gerontol Geriatr 2008; 47: 9–18. [DOI] [PubMed] [Google Scholar]
- 17. McCusker J, Bellavance F, Cardin S, Trepanier S, Verdon J, Ardman O. Detection of older people at increased risk of adverse health outcomes after an emergency visit: the ISAR screening tool. J Am Geriatr Soc 1999; 47: 1229–37. [DOI] [PubMed] [Google Scholar]
- 18. Gray LC, Peel NM, Costa AP et al. Profiles of older patients in the emergency department: findings from the interRAI multinational emergency department study. Ann Emerg Med 2013; 62: 467–74. [DOI] [PubMed] [Google Scholar]
- 19. Mahoney FI, Barthel DW. Functional evaluation: the Barthel index: a simple index of independence useful in scoring improvement in the rehabilitation of the chronically ill. Md State Med J 1965; 14: 61–5. [PubMed] [Google Scholar]
- 20. Galvin R, Gilleit Y, Wallace E et al. Adverse outcomes in older adults attending emergency departments: a systematic review and meta-analysis of the identification of seniors at risk (ISAR) screening tool. Age Ageing 2017; 46: 179–86. [DOI] [PubMed] [Google Scholar]
- 21. Higginbotham O, O'Neill A, Barry L et al. The diagnostic and predictive accuracy of the PRISMA-7 screening tool for frailty in older adults: a systematic review protocol. HRB Open Res 2020; 3: 26. 10.12688/hrbopenres.13042.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Erhag HF, Guðnadóttir G, Alfredsson J et al. The association between the clinical frailty scale and adverse health outcomes in older adults in acute clinical settings–a systematic review of the literature. Clin Interv Aging 2023; 18: 249–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Aucoin SD, Hao M, Sohi R et al. Accuracy and feasibility of clinically applied frailty instruments before surgery: a systematic review and meta-analysis. Anesthesiology 2020; 133: 78–95. [DOI] [PubMed] [Google Scholar]
- 24. Zhang X-M, Jiao J, Cao J et al. Frailty as a predictor of mortality among patients with COVID-19: a systematic review and meta-analysis. BMC Geriatr 2021; 21: 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Zou Y, Han M, Wang J, Zhao J, Gan H, Yang Y. Predictive value of frailty in the mortality of hospitalized patients with COVID-19: a systematic review and meta-analysis. Ann Transl Med 2022; 10: 166. 10.21037/atm-22-274. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Gretarsdottir E, Jonsdottir AB, Sigurthorsdottir I et al. Patients in need of comprehensive geriatric assessment: the utility of the InterRAI emergency department screener. Int Emerg Nurs 2021; 54: 100943. 10.1016/j.ienj.2020.100943. [DOI] [PubMed] [Google Scholar]
- 27. Michalski-Monnerat C, Carron PN, Nguyen S, Büla C, Mabire C. Assessing older Patientsʼ vulnerability in the emergency department: a study ofInterRAI EDScreener accuracy. J Am Geriatr Soc 2020; 68: 2914–20. [DOI] [PubMed] [Google Scholar]
- 28. Taylor A, Broadbent M, Wallis M, Marsden E. The predictive validity of the interRAI ED screener for predicting re-presentation within 28 days for older adults at a regional hospital emergency department. Australasian Emerg Care 2019; 22: 149–55. [DOI] [PubMed] [Google Scholar]
- 29. Jay S, Whittaker P, Mcintosh J, Hadden N. Can consultant geriatrician led comprehensive geriatric assessment in the emergency department reduce hospital admission rates? A systematic review. Age Ageing 2017; 46: 366–72. [DOI] [PubMed] [Google Scholar]
- 30. Briggs R, McDonough A, Ellis G, Bennett K, O'Neill D, Robinson D. Comprehensive geriatric assessment for community-dwelling, high-risk, frail, older people. Cochrane Database Syst Rev 2022; 2022: CD012705. 10.1002/14651858.CD012705.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Barry L, Galvin R, Murphy Tighe S, O'Connor M, Ryan D, Meskell P. The barriers and facilitators to implementing screening in emergency departments: a qualitative evidence synthesis (QES) protocol exploring the experiences of healthcare workers. HRB Open Res 2020; 3: 50. 10.12688/hrbopenres.13073.2. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Anomymised data is stored at Open Science Framework, available at doi 10.17605/OSF.IO/4NG6H.