Journals of Gerontology: Medical Sciences

cite as: J Gerontol A Biol Sci Med Sci, 2019, Vol. XX, No. XX, 1–6
doi:10.1093/gerona/glz109

Downloaded from https://academic.oup.com/biomedgerontology/advance-article-abstract/doi/10.1093/gerona/glz109/5485916 by Universidad Nacional de Colombia user on 28 May 2019

Advance Access publication May 6, 2019

Research Practice

10-Minute Targeted Geriatric Assessment Predicts

Disability and Hospitalization in Fast-Paced Acute Care
Settings
Márlon J. R. Aliberti, MD,1,2,3,*, Kenneth E. Covinsky, MD, MPH,2,3
Daniel Apolinario, MD, PhD,1 Alexander K. Smith, MD, MS, MPH,2,3 Sei J. Lee, MD,
MAS,2,3 Sileno Q. Fortes-Filho, MD,1 Juliana A. Melo, MD,1 Natalia P. S. Souza, BS,4
Thiago J. Avelino-Silva, MD, PhD,1 and Wilson Jacob-Filho MD, PhD1
1
Division of Geriatrics, Department of Internal Medicine, University of Sao Paulo Medical School, Brazil. 2Division of Geriatrics, Department
of Medicine, University of California, San Francisco (UCSF). 3Veterans Affairs Medical Center, San Francisco, CA. 4Gerontology, School of
Arts, Sciences and Humanities, University of Sao Paulo, Brazil.

*Address correspondence to: Márlon J. R. Aliberti, MD, Hospital das Clinicas da Faculdade de Medicina da Universidade de Sao Paulo, Av. Dr.
Eneas de Carvalho Aguiar 255, 8º Andar, Bloco 8—Nucleo de Apoio a Pesquisa e Ensino em Geriatria e Gerontologia / Sao Paulo (SP), 05403-000,
Brazil. E-mail: maliberti@usp.br

Received: October 8, 2018; Editorial Decision Date: April 17, 2019

Decision Editor: Anne Newman, MD, MPH

Abstract
Background: Limited time and resources hinder the use of comprehensive geriatric assessment in acute contexts. We investigated the predictive
value of a 10-minute targeted geriatric assessment (10-TaGA) for adverse outcomes over 6 months among acutely ill older outpatients.
Methods: Prospective study comprising 819 acutely ill outpatients (79.2 ± 8.4 years; 63% women) in need of intensive management (eg,
intravenous therapy, laboratory test, radiology) to avoid hospitalization. The 10-TaGA provided a validated measure of cumulative deficits.
Previously established 10-TaGA cutoffs defined low (0–0.29), medium (0.30–0.39), and high (0.40–1) risks. To estimate whether 10-TaGA
predicts new dependence in activities of daily living and hospitalization over the next 6 months, we used hazard models (considering death
as competing risk) adjusted for standard risk factors (sociodemographic factors, Charlson comorbidity index, and physician estimates of
risk). Differences among areas under receiver operating characteristic curves (AUROC) examined whether 10-TaGA improves outcome
discrimination when added to standard risk factors.
Results: Medium- and high-risk patients, according to 10-TaGA, presented a higher incidence of new activities of daily living dependence
(21% vs 7%, adjusted subhazard ratio [aHR] = 2.4, 95% CI = 1.3–4.5; 40% vs 7%, aHR = 5.0, 95% CI = 2.8–8.7, respectively) and
hospitalization (27% vs 13%, aHR = 2.0, 95% CI = 1.2–3.3; 37% vs 13%, aHR = 2.9, 95% CI = 1.8–4.6, respectively) than low-risk patients.
The 10-TaGA remarkably improved the discrimination of models that incorporated standard risk factors to predict new activities of daily
living dependence (AUROC = 0.76 vs 0.71, p < .001) and hospitalization (AUROC = 0.71 vs 0.68, p < .001).
Conclusions: The 10-TaGA is a practical and efficient comprehensive geriatric assessment tool that improves the prediction of adverse
outcomes among acutely ill older outpatients.
Keywords: Geriatric assessment, Risk factors, Functional performance, Frailty, Prognosis.

The escalating number of older people demanding care for acute con- focused model of acute care health services leads clinicians not to
ditions has challenged the health care systems (1). In these patients, recognize this broad spectrum of vulnerability among older persons
social, functional, and psychological issues often interact with epi- (5). By neglecting these multiple dimensions of risk, clinicians may
sodes of disease exacerbation (2–5). Nevertheless, the typical disease-

© The Author(s) 2019. Published by Oxford University Press on behalf of The Gerontological Society of America.
1
All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
2 Journals of Gerontology: MEDICAL SCIENCES, 2019, Vol. XX, No. XX
underestimate older adults’ risk for adverse outcomes such as repeated
hospitalizations, functional disability, and death (1–5).
Previous work has demonstrated the importance of compre-
hensive geriatric assessment (CGA) in estimating prognosis and
guiding the care of acutely ill older adults (6). However, these
studies were mostly conducted in the hospital setting and examined
time-consuming CGA tools, which are unfeasible in the fast-paced
acute care context (6,7). Other less time-consuming screening tools,
including the Identification of Seniors at Risk (ISAR) (8,9), have
been primarily used for prognosis assessment. Since such instru-
ments do not often provide a comprehensive perspective of geriatric
syndromes, more practical and efficient CGA tools, that are suitable
for busy clinical settings, are needed (10).
The 10-minute targeted geriatric assessment (10-TaGA) is a
quick and easy-to-administer CGA tool recently developed to screen
ten domains (social support, recent hospitalizations, falls, number
of medications, activities of daily living [ADL], cognition, self-rated
health, depressive symptoms, nutritional status, and gait speed) (10).
This instrument provides an overall measure of cumulative deficits
from multiple domains. In previous research, the 10-TaGA score
showed excellent accuracy in identifying frail individuals in the
context of a time-constrained acute care setting (10). In the present
study, we investigated the predictive value of the 10-TaGA for new
dependence in ADL and unplanned hospitalization in 6 months
among acutely ill older outpatients.
Methods
Design, Setting, and Participants
We conducted a prospective study comprising individuals aged
60 years old or more with acute medical conditions or exacerba-
tion of chronic illnesses demanding intensive management to avoid
full-time hospitalization. Patients were referred by their physicians
to a geriatric day hospital (GDH) from the emergency department,
home care, ambulatory services, and primary health care unit of the
University of Sao Paulo Medical School, which is the largest public
medical center in Latin America, attending 1.5 million people (28%
older adults) from the metropolitan region of Sao Paulo, Brazil. The
GDH operates 12 hours a day and offers short-term treatment (ie,
rapid access to diagnostic testing facilities and intravenous therapy)
as an alternative to emergency department use or hospitalization for
older adults (11). The major reasons for referral to this service in-
volved infections, acute anemia, refractory pain, symptomatic heart
failure, decompensated diabetes, and uncontrolled hypertension.
Further details about the GDH can be found elsewhere (11).
All individuals (n = 957) consecutively referred to the GDH from
May 2014 to April 2017 were assessed for eligibility on admission.
Exclusion criteria comprised patients who were under exclusive pal-
liative care (n = 29), were totally dependent in ADL (n = 41), required
immediate hospitalization (n = 28), had an elective surgical procedure
within 6 months (n = 22), and/or refused to participate (n = 18). After
screening, the final sample consisted of 819 participants.
The University of Sao Paulo Medical School Institutional Review
Board approved the study. Signed informed consent was obtained
from participants or their proxies. Staff members responsible for pa-
tients’ clinical management had no access to the study protocol.
Baseline Assessment
A research team (nurse, social worker, and pharmacist) conducted
a multidimensional assessment on admission. Sociodemographic
factors (age, sex, race, and household income) and the Charlson
Downloaded from https://academic.oup.com/biomedgerontology/advance-article-abstract/doi/10.1093/gerona/glz109/5485916 by Universidad Nacional de Colombia user on 28 May 2019
comorbidity index (12), a measurement of overall disease burden,
were assessed using participant reports and medical chart reviews.
Referring physicians provided their best estimate of the likelihood of
new dependence in ADL and hospitalization of their patients within
the next 6 months. They were asked to indicate one of the following
likelihood categories when reporting their opinion regarding each
estimated risk (dependence in ADL and hospitalization): less than
10%, 25%, 50%, 75%, and greater than 90%.
A team member was responsible for the ISAR (0–6) (8). This
screening tool is composed of six questions that concisely address
functional status, previous hospital admission, presence of visual
and cognitive impairments, and polypharmacy. Another team
member was responsible for the 10-TaGA (10). This multi-domain
screener assesses social support (living arrangement and availability
of help), emergency department visits and hospitalizations in the
prior 6 months, the number of falls in the last year, the number of
medications, dependence in ADL (Katz index), a 10-point cognitive
screener, self-rated health, the 4-item Geriatric Depression Scale, nu-
tritional status (weight loss in the last year and body mass index),
and walking speed. Each domain is classified into three categories:
normal (0 points), mild impairment (0.5 points), and severe im-
pairment (1 point). A single numerical score from 0 (no deficit is
present) to 1 (all deficits are present) is calculated by dividing the
total sum of points by the number of evaluated domains. In a pre-
vious comparison with the physical frailty phenotype classification
(10), 10-TaGA cutoffs of at least 0.30 and at least 0.40 showed ex-
cellent accuracy to identify pre-frail and frail status, respectively.
The operational and scoring instructions for 10-TaGA are shown in
Supplementary Figure 1.
Clinical and performance measures of a standard CGA were
also evaluated to construct a Frailty Index (FI-CGA) by counting
the number of deficits in ten health domains (7,13), including cog-
nition (Mini-Mental State Examination), mood (15-item Geriatric
Depression Scale), communication (ie, hearing, vision, and speech),
mobility (Timed-Up and Go test), balance test and history of
falls, bladder and bowel functions, nutrition (Mini Nutritional
Assessment), ADL and instrumental activities of daily living, and so-
cial resources. Following the method used in previous work (7,13),
each domain was scored as no problem (0), minor problem (0.5),
major problem (1). Supplementary Table 1 presents the criteria and
scoring rules for the FI-CGA.
Outcomes
Primary outcomes were time to new dependence in ADL and time
to unplanned hospitalization within 6 months of GDH admis-
sion. Investigators who were blinded to the baseline assessments
performed monthly phone calls using standardized follow-up
interviews. Unplanned hospitalization was defined as a hospital
stay for unexpected medical treatment for 24 hours or longer.
Participants were classified as having developed new ADL de-
pendence when they reported requiring help in a previously pre-
served ADL (including bathing, toileting, dressing, transferring,
and eating), as compared to baseline. Investigators made up to 10
additional telephone calls if patients or their proxies could not
be reached in the first monthly attempts. By using this method,
all participants were successfully contacted for their follow-up
interviews. Participants who did not develop any of the outcomes,
and were alive at the end of the 6-month follow-up period, were
censored.
Journals of Gerontology: MEDICAL SCIENCES, 2019, Vol. XX, No. XX 3

Statistical Analysis Table 1. Participant Characteristics on Admission at the Geriatric

Downloaded from https://academic.oup.com/biomedgerontology/advance-article-abstract/doi/10.1093/gerona/glz109/5485916 by Universidad Nacional de Colombia user on 28 May 2019

The observed outcome incidences were reported with their 95% Day Hospital (n = 819)
confidence interval (CI) and according to three levels of physician Variable Mean (SD) or n (%)
estimate of risk (≤25%, 50%, and ≥75%) at baseline. Areas under
the receiver operating characteristic curves (AUROC) were used to Sociodemographic factors
estimate the accuracy of 10-TaGA to predict the risk of new ADL Age (y) 79.2 (8.4)
dependence and hospitalization. Sensitivity, specificity, predictive Female 517 (63)
values, and likelihood ratios were calculated for every 0.1-point Ethnicity
  White 500 (61)
increase in 10-TaGA score. We computed differences between the
  Black 88 (11)
AUROC of 10-TaGA and ISAR and FI-CGA for both primary out-
  Mixed 189 (23)
comes using nonparametric methods (14).   Asian 42 (5)
The cumulative incidence function was used to measure the like- Annual household income per capita
lihood of outcomes according to previously determined categories    < 4000 USD 252 (31)
of 10-TaGA scores (low-risk = 0–0.29, medium-risk = 0.3–0.39,   4000–8000 USD 434 (53)
high-risk = 0.4–1) (10). To examine whether 10-TaGA was a pre-    > 8000 USD 133 (16)
dictor of the outcomes, we used competing risk hazards models, Charlson comorbidity index
considering death as a competing risk (15). Four models were fitted 0 points 120 (15)
for each outcome: (a) unadjusted, (b) adjusted for sociodemographic 1–2 points 349 (42)
≥ 3 points 350 (43)
factors and Charlson comorbidity index, (c) adjusted for physician
Physician estimate of risk
estimate of the likelihood of each studied outcome (new ADL de-
Likelihood of new ADL dependence in 6 months
pendence and hospitalization) in 6 months, and (d) fully adjusted.   ≤ 25% 231 (28)
Moreover, by assessing differences between AUROCs, we estimated   50% 229 (28)
the impact on outcome discrimination obtained by adding 10-TaGA   ≥ 75% 126 (15)
to these models. Finally, we analyzed the participants’ risk of adverse Likelihood of hospitalization in 6 months
outcomes in 6 months by stratifying their incidences according to the   ≤ 25% 235 (29)
physician estimate of risk and 10-TaGA scores (≥0.4 or <0.4).   50% 272 (33)
All analyses were performed using Stata, version 15 (Stata Corp.,   ≥ 75% 103 (13)
College Station, TX). 10-min targeted geriatric assessment
Living alone 163 (19.9)
Hospitalization in the previous 6 months 250 (30.5)
Results History of fall(s) in the last year 383 (46.8)
Table 1 shows participant baseline characteristics. The primary causes Polypharmacy (≥5 medications) 699 (85.3)
Need help in performing any ADL (Katz index) 457 (55.8)
of referral to the GDH were decompensated diabetes (17%), acute
Cognitive impairment (10-point cognitive 280 (34.2)
anemia (16%), congestive heart failure (13%), and infections (12%)
screener ≤ 5points)
(Supplementary Table 2). The mean 10-TaGA score was 0.41 ± 0.16, Self-rated health poor or very poor* 136 (18.1)
with 176 (21%) patients being classified as low-risk, 168 (21%) as Four-item Geriatric Depression Scale ≥3 141 (18.8)
medium-risk, and 475 (58%) as high-risk. During the follow-up points (0–4)*
period, 236 (29%) older patients developed new ADL dependence, and At risk of malnutrition (weight loss >10 pounds 390 (47.6)
243 (30%) were hospitalized. Physician estimate of risk was associ- in the last year and/or BMI <22)
ated with the incidence of both outcomes. However, clinicians overesti- Mobility disability (gait speed <0.6 m/s 394 (48.1)
mated the rate of new ADL dependence and unplanned hospitalization or unable to complete the walking test)
in patients at higher risk at baseline (see Supplementary Table 3).
Notes: ADL = activities of daily living; BMI = body mass index (kg/m2);
The 10-TaGA showed good accuracy to identify risk of new ADL
SD = standard deviation.
dependence (AUROC = 0.72; 95% CI = 0.68–0.75) and moderate ac-
*Excluded 67 patients unable to answer because of severe dementia (Mini-
curacy to detect risk of unplanned hospitalization (AUROC = 0.64; Mental State Examination < 10 points).
95% CI = 0.60–0.68). The cutoffs of 10-TaGA scores that classified
participants as being at medium (≥0.3) and high risk (≥0.4) presented
high sensitivity (>90% and >70%, respectively) but low specificity physician estimate of risk, older patients at medium and high-risk,
(≤50%) for both outcomes. Higher specificity values (>70%) re- according to 10-TaGA scores, presented a higher incidence of new
quired a cutoff ≥0.5 (see Supplementary Table 4). ADL dependence and unplanned hospitalization, compared to those
Compared to ISAR, the 10-TaGA showed a significantly higher who were classified as low-risk patients (Table 2). Table 3 shows
accuracy to predict both new ADL dependence (AUROC 0.72 vs that the 10-TaGA introduction remarkably improved the discrim-
0.66; p = .002) and hospitalization (AUROC 0.64 vs 0.60; p = .04); ination of all models containing standard risk factors for both new
compared to FI-CGA, the 10-TaGA showed a significantly higher ADL dependence and hospitalization. Finally, Figure 2 illustrates
accuracy to predict hospitalization (AUROC 0.64 vs 0.60; p = .02) that 10-TaGA scores improved the prediction of adverse outcomes,
and a slightly inferior accuracy to predict new ADL dependence regardless of the physician estimate of risk.
(AUROC 0.72 vs 0.74; p = .07) (see Supplementary Figure 2).
Figure 1 shows the cumulative incidence of new ADL depend-
ence and unplanned hospitalization according to different 10-TaGA
Discussion
categories, highlighting the increased rate of outcomes among parti- In this study, we showed that 10-TaGA was a strong predictor of
cipants with higher cumulative deficits at baseline. After adjustment new ADL dependence and unplanned hospitalizations in 6 months
for sociodemographic factors, Charlson comorbidity index, and among acutely ill older patients in need of intensive management to
4 Journals of Gerontology: MEDICAL SCIENCES, 2019, Vol. XX, No. XX

avoid full-time hospital admission. This quick and easy-to-administer

Downloaded from https://academic.oup.com/biomedgerontology/advance-article-abstract/doi/10.1093/gerona/glz109/5485916 by Universidad Nacional de Colombia user on 28 May 2019

Figure 1. Cumulative incidence of (A) new ADL dependence and (B)
unplanned hospitalization over 6 months according to the 10-minute targeted
geriatric assessment (10-TaGA) on admission at the geriatric day hospital
(n = 819). 10-TaGA categories were validated in a previous study against the
physical frailty phenotype; low-risk (0–0.29), medium-risk (0.3–0.39), and
high-risk (0.4–1). ADL = activities of daily living.
multidimensional tool provided valuable prognostic information by
identifying factors that are not traditionally considered in medical
assessments and are not captured by physicians’ clinical impression.
Most methods assessing vulnerability in acutely ill patients were
tested in the emergency department and relied on self-report question-
naires that capture only part of their complex needs (5,9). Considering
that the context of patients who seek medical attention, but do not re-
quire inpatient hospitalization, or whose providers hope to avoid hos-
pital admission, is becoming very common, we need information on
tools for identifying the prevalent number of frailty patients in innovative
settings such as day hospital and home care, where the risk to develop
poor outcomes is substantial. In these settings, similar to the emergency
department, health care professionals deal with high numbers of patients
and limited resources, eventually discharging older adults directly in the
community, without further investigation or support (16).
Among fast-screening tools available in acute settings, ISAR is
the most validated and widely used instrument (8,9). Our results
corroborate previous reports showing that ISAR is a useful predictor
of adverse outcomes in acutely ill patients. Nevertheless, our findings
also indicate that 10-TaGA presented a higher performance than
ISAR to stratify the risk of new dependence in ADL and unplanned
hospitalization after acute medical illness, with the additional benefit
of screening for geriatric syndromes such as social support, depres-
sive symptoms, cognitive and nutritional status, falls, and mobility.
Previous studies have highlighted the value of CGA tools to esti-
mate prognosis and guide the care of acutely ill older adults (6). The
Multidimensional Prognostic Index and FI-CGA are well-validated
measures that combine information about physical, psychological,
and social domains to identify high-risk older adults and promote
patient-centered care (6,7). Most studies using these extended CGA
tools involved patients admitted in the hospital setting where such
time-consuming tools are feasible (6,7). In our work, compared to
the FI-CGA, the 10-TaGA showed similar prognostic power to iden-
tify older patients at risk to develop new disability and significantly
higher accuracy to detect those at risk of hospitalization, indicating

Table 2. Association of 10-minute Targeted Geriatric Assessment (10-TaGA) With 6-Month Adverse Outcomes (n = 819)

Subhazard Ratio (95% confidence interval)

Model 1 = Adjusted for Model

Sociodemographic Factors Model 2 = Adjusted for 3 = Fully
Unadjusted and Comorbidities Physician Estimate of Risk Adjusted

New ADL dependence

10-TaGA categories
  Low-risk (0–0.29) (Reference) (Reference) (Reference) (Reference)
  Medium-risk 2.9 (1.5–5.5) 2.7 (1.4–5.0) 2.4 (1.3–4.5) 2.4 (1.3–4.5)
(0.3–0.39)
  High-risk (0.4–1) 6.7 (3.8–11.7) 5.8 (3.3–10.2) 5.3 (3.0–9.2) 5.0 (2.8–8.7)
Hospitalization
10-TaGA categories
  Low-risk (0–0.29) (Reference) (Reference) (Reference) (Reference)
  Medium-risk 2.2 (1.3–3.6) 2.2 (1.3–3.7) 1.9 (1.2–3.2) 2.0 (1.2–3.3)
(0.3–0.39)
  High-risk (0.4–1) 3.3 (2.1–5.1) 3.4 (2.2–5.4) 2.6 (1.7–4.1) 2.9 (1.8–4.6)

Notes: Estimates were calculated using the Fine and Gray method, considering death as a competing risk. Model 1 = sociodemographic factors (age, gender,
ethnicity, and household income) + Charlson comorbidity index. Model 2 = physician estimate of the likelihood of new dependence in ADL and hospitalization
in 6 months (<25%, 50%, 75%). Model 3 = sociodemographic factors + Charlson comorbidity index + physician estimate of risk (full model). The 10-TaGA
categories were validated against the physical frailty phenotype in a previous study. ADL = activities of daily living.
Journals of Gerontology: MEDICAL SCIENCES, 2019, Vol. XX, No. XX
Table 3. Impact of 10-minute Targeted Geriatric Assessment (10-TaGA) on 6-month Adverse Outcomes Discrimination (n = 819)
New ADL dependence
Model 1 = sociodemographic factors + Charlson comorbidity index
Model 2 = physician estimate of the likelihood of new ADL dependence in 6 months
 Model 3 = sociodemographic factors + Charlson comorbidity index + physician
estimate (full model)
Hospitalization
Model 1 = sociodemographic factors + Charlson comorbidity index
Model 2 = physician estimate of the likelihood of hospitalization in 6 months
 Model 3 = sociodemographic factors + Charlson comorbidity index + physician
estimate (full model)
Notes: Sociodemographic factors include age, gender, ethnicity, and household income. Physician estimate of the likelihood of new ADL dependence and hos-
pitalization in 6 months (<25%, 50%, >75%). 10-TaGA score (0–1) was classified as low (0–0.29), medium (0.3–0.39), and high (0.4–1) risk. ADL = activities of
daily living; CI = confidence interval; ROC = receiver operating characteristic.
*p-value compares the areas under the ROC curve of the models with and without 10-TaGA.
Figure 2. Incidence of (A) new ADL dependence and (B) unplanned
hospitalization, combining physician estimate of risk and 10-minute targeted
geriatric assessment (10-TaGA) on admission at the Geriatric Day Hospital
(n = 819). 10-TaGA score ≥ 0.4 define individuals at high risk of adverse
outcomes according to a previous study. Physician estimate of the likelihood
of new ADL dependence and hospitalization in 6 months. All comparisons
between 10-TaGA score <0.4 versus ≥0.4 within the same physician estimate
stratum showed a p-value < .05. ADL = activities of daily living.
that this instrument may be a practical and efficient CGA model in
busy clinical settings.
Despite the findings favoring 10-TaGA, our study determines that
the best strategy for assessing prognosis in older adults combines CGA
measures with other predictors commonly available in clinical prac-
tice (16). Although 10-TaGA showed an excellent ability (high nega-
tive predictive values) to identify those who would not be likely to
5
Downloaded from https://academic.oup.com/biomedgerontology/advance-article-abstract/doi/10.1093/gerona/glz109/5485916 by Universidad Nacional de Colombia user on 28 May 2019
Area Under the ROC Curve (95% CI)
Without 10-TaGA With 10-TaGA p-value*
0.68 (0.64–0.72) 0.74 (0.70–0.78) <.001
0.67 (0.63–0.70) 0.73 (0.69–0.76) <.001
0.71 (0.67–0.75) 0.76 (0.72–0.79) <.001
0.63 (0.58–0.67) 0.68 (0.64–0.72) <.001
0.63 (0.59–0.67) 0.68 (0.64–0.72) <.001
0.68 (0.64–0.72) 0.71 (0.67–0.75) <.001
have the outcome, its performance alone was still weak (low positive
predictive values) to detect accurately those who would be most likely
to have the outcome within 6 months. Our findings indicate an ex-
cessive risk for adverse outcomes in those patients at higher scores in
10-TaGA (≥0.4) if they also present other risk factors such as advanced
age (≥85 years old), multimorbidity, a high likelihood of physician
estimate of risk (>50%). Further work is still needed to investigate
longer-term outcomes and to confirm whether a battery of predictors
would achieve better predictive values when combined (16).
The 10-TaGA also has potential advantages for use in fast-paced
settings. Identifying deficits in different domains of functioning may
be essential to measure how severely chronic and acute diseases
impact the patient, as well as his/her overall reserve and ability to
recover from an acute event (5). Moreover, implementing a multi-
dimensional assessment tool could open a window of opportunity
to plan patient-centered care that goes beyond the conventional
disease-specific models, and that can make a change in the trend
toward disability (16).
This study has notable strengths. First, we succeeded in obtaining
complete information from all participants. Second, we were able to
make direct comparisons between 10-TaGA and ISAR and FI-CGA.
Finally, we performed robust analyses that considered the competing
risk of death and standard risk factors for adverse events. Some
limitations should also be noted. First, therapeutic interventions in
the GDH may have affected our results by reducing outcome rates
during the 6-month follow-up. However, the attending staff was
blinded to the study protocol, and the incidence of the outcomes
here reported was sufficiently high to test our hypotheses. Second,
while in developing countries the epidemiology of aging involves
people aged 60 years old or more, we could consider the study popu-
lation as relatively young compared to aging research from devel-
oped countries, which uses higher age cutoffs (65 or 75 years old) to
define older patients. Our results from a large medical center located
in a low-middle income country should then be tested and confirmed
in different settings and populations.
In conclusion, the results of this study provide robust, albeit
preliminary evidence supporting the predictive value of a practical
and efficient CGA tool for new dependence in ADL and unplanned
hospitalizations in 6 months among acutely ill older adults. Future
studies should focus on estimating whether 10-TaGA improves pa-
tient care and clinical outcomes.
6 Journals of Gerontology: MEDICAL SCIENCES, 2019, Vol. XX, No. XX
Supplementary Material
Supplementary data are available at The Journals of Gerontology,
Series A: Biological Sciences and Medical Sciences online.
Funding
This work was supported by the Coordenação de Aperfeiçoamento de Pessoal
de Nível Superior—Brazil (CAPES)—Finance Code 001 (88881.131613/2016-
01 to M.J.R.A.). M.J.R.A. conducted part of this work during a visiting
scholar period to the University of California, San Francisco. The sponsor had
no role in the design, methods, subject recruitment, data collections, analysis
and preparation of the manuscript.
Conflict of interest statement
None reported.
References
1. Ellis G, Whitehead MA, Robinson D, O’Neill D, Langhorne P.
Comprehensive geriatric assessment for older adults admitted to hospital:
meta-analysis of randomised controlled trials. BMJ. 2011;343:d6553.
doi:10.1136/bmj.d6553
2. Lohman MC, Scherer EA, Whiteman KL, Greenberg RL, Bruce ML.
Factors associated with accelerated hospitalization and re-hospitalization
among medicare home health patients. J Gerontol A Biol Sci Med Sci.
2017;73(9):1280–1286. doi:10.1093/gerona/glw335
3. Gagliardi C, Corsonello A, Di Rosa M, et al. Preadmission functional de-
cline predicts functional improvement among older patients admitted to
acute care hospital. J Gerontol A Biol Sci Med Sci. 2017;73(10):1363–
1369. doi:10.1093/gerona/glx211
4. Conroy SP, Stevens T, Parker SG, Gladman JR. A systematic review of
comprehensive geriatric assessment to improve outcomes for frail older
people being rapidly discharged from acute hospital: “interface geriatrics”.
Age Ageing. 2011;40:436–443. doi:10.1093/ageing/afr060
5. Theou O, Campbell S, Malone ML, Rockwood K. Older adults in the
emergency department with frailty. Clin Geriatr Med. 2018;34:369–386.
doi:10.1016/j.cger.2018.04.003
6. Pilotto A, Veronese N, Daragjati J, et al. Using the multidimensional prog-
Downloaded from https://academic.oup.com/biomedgerontology/advance-article-abstract/doi/10.1093/gerona/glz109/5485916 by Universidad Nacional de Colombia user on 28 May 2019
nostic index to predict clinical outcomes of hospitalized older persons: a
prospective, multicenter, international study. J Gerontol A Biol Sci Med
Sci. 2018. doi:10.1093/gerona/gly239. [Epub ahead of print].
7. Pilotto A, Rengo F, Marchionni N, et al.; FIRI-SIGG Study Group.
Comparing the prognostic accuracy for all-cause mortality of frailty in-
struments: a multicentre 1-year follow-up in hospitalized older patients.
PLoS One. 2012;7:e29090. doi:10.1371/journal.pone.0029090
8. McCusker J, Bellavance F, Cardin S, Trépanier 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;47:1229–1237. doi:10.1111/j.1532–5415.1999.tb05204.x
9. Di Bari M, Salvi F, Roberts AT, et al. Prognostic stratification of eld-
erly patients in the emergency department: a comparison between the
“Identification of Seniors at Risk” and the “Silver Code”. J Gerontol
A Biol Sci Med Sci. 2012;67:544–550. doi:10.1093/gerona/glr209
10. Aliberti MJR, Apolinario D, Suemoto CK, et al. Targeted geriatric assess-
ment for fast-paced healthcare settings: development, validity, and reli-
ability. J Am Geriatr Soc. 2018;66:748–754. doi:10.1111/jgs.15303
11. Aliberti MJ, Suemoto CK, Fortes-Filho SQ, et al. The geriatric day hos-
pital: preliminary data on an innovative model of care in Brazil for older
adults at risk of hospitalization. J Am Geriatr Soc. 2016;64:2149–2153.
doi:10.1111/jgs.14342
12. Frenkel WJ, Jongerius EJ, Mandjes-van Uitert MJ, van Munster BC,
de Rooij SE. Validation of the Charlson Comorbidity Index in acutely
hospitalized elderly adults: a prospective cohort study. J Am Geriatr Soc.
2014;62:342–346. doi:10.1111/jgs.12635
13. Jones D, Song X, Mitnitski A, Rockwood K. Evaluation of a frailty index based
on a comprehensive geriatric assessment in a population based study of elderly
Canadians. Aging Clin Exp Res. 2005;17:465–471. doi:10.1007/BF03327413
14. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two
or more correlated receiver operating characteristic curves: a nonparametric
approach. Biometrics. 1988;44:837–845. doi:10.2307/2531595
15. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of
a competing risk. J Am Stat Assoc. 1999;94:496–509. doi:10.1080/01621
459.1999.10474144
16. Wou F, Gladman JR, Bradshaw L, Franklin M, Edmans J, Conroy SP. The
predictive properties of frailty-rating scales in the acute medical unit. Age
Ageing. 2013;42:776–781. doi:10.1093/ageing/aft055