Abstract
Objectives
There are few data regarding the accuracy of short frailty tools as predictors of mortality and other clinical outcomes of older patients admitted to a geriatric ward. We therefore analyzed the accuracy of Rockwood et al's Clinical Frailty Scale and an easy and quick to perform operationalization of Fried et al's frailty phenotype, as predictors of mortality and other clinical outcomes in our cohort of patients.
Design
Prospective analysis with a follow-up period of 6 months.
Setting and Participants
307 patients who were 65 years of age or older were included in the study. The patients were assessed in terms of the two frailty measures during their stay in a geriatric ward.
Results
The Clinical Frailty Scale and the frailty phenotype were both suitable for differentiating between patients who died due to any cause from those who survived during followup (primary outcome) (area under the ROC curves (AUC) values 0.867 (95% CI 0.807–0.926), p<0.001 and 0.754 (95% CI 0.688–0.821), p<0.001, respectively). Regarding the secondary outcomes: 1. unplanned admission to hospital and 2. a fall during follow-up, the Clinical Frailty Scale discriminated or tended to discriminate between patients to whom these criteria applied and those to whom they did not (AUC=0.569 (95% CI 0.502–0.636), p=0.046 and AUC=0.574 (95% CI 0.501–0.647), p=0.071, respectively). The frailty phenotype did not show such a differentiation when applied to secondary outcomes (AUC=0.500 (95% CI 0.432–0.568), p=0.994 and AUC=0.518 (95% CI 0.439–0.598), p=0.658, respectively).
Conclusion
Both short frailty instruments are suitable predictors of mortality in older patients who were admitted to a geriatric ward. The Clinical Frailty Scale, but not the frailty phenotype, predicted at least some of the secondary outcomes, i.e., the outcome unplanned admission to hospital during follow-up.
Key words: Clinical Frailty Scale, frailty phenotype, mortality, unplanned admission to hospital, fall
Introduction
Frailty is a major health burden in older people (1, 2). It is commonly referred to as a condition of reduced resistance to stressors, due to a decline in physiological reserves and which is associated with adverse health outcomes, such as mortality, hospitalization, and falls, amongst others (3., 4., 5., 6.). Frailty is associated with the ageing process per se (3). However, particular diseases or adverse conditions, such as malnutrition (7), sarcopenia or polypharmacy, can also result in impairment of multiple organ function capacity and consequently frailty (3, 4, 8). Of clinical interest, frailty is a dynamic process that, in general, deteriorates over time, but may also be sensitive to treatment and intervention (3, 9).
Recently, consensus was achieved (9) in distinguishing between the wider construct of frailty (10., 11., 12., 13.), which takes several aspects, such as multi-morbidity and central nervous system disorders of the patients, into account, and physical frailty (14., 15., 16.). Several tools that allow the evaluation of patients in regard to frailty have been developed and validated (3, 9). Bearing in mind the busy daily clinical practice in hospitalized settings, short tools that allow rapid recognition of patients at risk in relation to frailty and the identification of frail patients are desirable (9, 17). Such short tools include, in terms of the wider construct of frailty, for example, the Clinical Frailty Scale (18), and, in terms of physical frailty, for example, the frailty phenotype (14).
However, to the best of our knowledge, no study has so far analyzed the significance of two different short frailty tools that relate to the two different, major frailty constructs together as predictors of mortality and other clinical outcomes in a cohort of old patients who were admitted to geriatric wards. Against this background, using a prospective, longitudinal analysis, we aimed to evaluate the accuracy of the Clinical Frailty Scale and the frailty phenotype as predictors of mortality and other clinical outcomes based on a cohort of older patients in geriatric wards.
Methods
Study population and study design
The study population used in the analysis consisted of hospitalized patients staying in geriatric wards at the Geriatrics
Centre Erlangen. The Geriatrics Centre Erlangen is a geriatric department of the Hospital of the Congregation of St. Francis Sisters of Vierzehnheiligen in the city of Erlangen which is located in southern Germany. Patients admitted due to acute disease or exacerbation of a chronic disease were enrolled. Inclusion criteria were 1) 65 years of age or older, 2) the ability to provide informed consent or the availability of a legal guardian to provide informed consent for the study patient.
The collection of baseline data was carried out during the stay in hospital of the study participants. Six months after the baseline examination, follow-up data were collected using telephone interviews with patients and documents were received from the study participants, their relatives, legal guardian, family practitioners, or specialists for clinical outcome measures. The primary clinical outcome measure was 1) death due to any cause and the secondary outcome measures were 1) unplanned admission to hospital and 2) a fall.
The present study was conducted in accordance with the Declaration of Helsinki and the guidelines for Good Clinical Practice. The study protocol was approved by the local ethics committee, i.e., the Ethics Committee of the University of Erlangen-Nuremberg. Written informed consent was obtained from each participant or from his or her legal guardian.
The Clinical Frailty Scale
The Clinical Frailty Scale was developed and validated by Rockwood and colleagues (18). It is a tool used to assess the degree of relative fitness and frailty of an older patient, based on clinical judgement, by integrating measures of function, morbidity, and central nervous system impairments. Briefly, the Clinical Frailty Scale allows discrimination of nine categories on a relative fitness to frailty continuum: category 1: very fit, category 2: well, category 3: managing well, category 4: vulnerable, category 5: mildly frail, category 6: moderately frail, category 7: severely frail, category 8: very severely frail, category 9: terminally ill as described elsewhere in detail (18).
The frailty phenotype
The frailty phenotype is based on the concept of five individual phenotypic components: 1. shrinking, i.e. weight loss (unintentional) / loss of muscle mass, 2. poor endurance/exhaustion, 3. slowness, 4. low physical activity and 5. Weakness, as introduced by Fried et al (14). In the current study, we operationalized the individual components as was previously done by Rockwood et al (19): 1. shrinking as identified by an unintentional loss of either ≥ 10 lb or ≥ 5% of body weight in the past year (19), 2. poor endurance/exhaustion as identified by self-report of feeling “tired all the time” (19) 3. slowness as identified by time on timed up and go test (TUG) > 19 seconds (19), 4. low physical activity as identified by the patient needing assistance with walking or he or she being unable to walk (19), and weakness as identified by clearly abnormal strength on physical examination (19). The inability to perform the TUG was considered equivalent to a TUG time > 19 seconds. Patients with none of these components were considered to be robust, those with 1 or 2 to be pre-frail, and those with 3 or more to be frail.
Statistics
All statistical analyses were performed using SPSS software (IBM SPSS Statistics 22). Results are given as mean ± standard deviation or percentage (%) in the tables or text. Comparison of clinical parameters by category of each frailty instrument was performed with the non-parametric Kruskall-Wallis-Test or Chi-Square-Test, where appropriate. Kaplan-Meier estimates were used to determine the percentage of patients who provided an outcome, e.g., aforementioned primary or secondary outcomes, during follow-up. The p value reported for the difference in (event-free) survival curves with reference to the primary and secondary outcomes among the categories of each frailty instrument was based on the logrank test. To test the discriminative ability of each frailty instrument to predict an outcome, e.g., the aforementioned primary or secondary outcome measures, at follow-up, we used receiver-operating characteristic (ROC) curves and evaluated the area under the curve (AUC). A two-tailed p value <0.05 or, in terms of a comparison of the AUCs in relation to the outcome measures of the two different frailty instruments, the lack of an overlap of the 95% confidence intervals of the two different AUCs, were considered statistically significant.
Results
Patient characteristics
307 patients were included into this study. Baseline examinations of all the study participants were performed during their hospital stay in geriatric wards. Complete follow-up data were obtained from all of the patients at 6 months after baseline examination.
The characteristics of the study participants stratified according to the categories of the Clinical Frailty Scale and the frailty phenotype are given in Tables 1 and 2. With increasing categories of each frailty instrument, patients revealed greater morbidity, cognitive and mobility problems, as well as polypharmacy (see Tables 1 and 2).
Table 1.
Distribution of clinical characteristics by category in the Clinical Frailty Scale
| Characteristics | 1 Very fit | 2 Well | 3 Managing well | 4 Vulnerable | 5 Mildly frail | 6 Moderately frail | 7 Severely frail | 8 V ery severely frail | 9 Terminally ill | P value |
|---|---|---|---|---|---|---|---|---|---|---|
| Patients, n | 0 | 1 | 18 | 67 | 76 | 79 | 46 | 11 | 9 | − |
| Age, years | − | 72.0 ±- | 79.4±6.2 | 82.6±5.5 | 82.9±6.7 | 83.1±6.4 | 83.8±6.5 | 85.0±7.8 | 85.0±7.1 | 0.083 |
| Female, % | − | 0 | 88.9 | 74.6 | 65.8 | 68.4 | 56.5 | 54.5 | 66.7 | 0.142 |
| BMI, kg/m2 | − | 29.4±- | 28.2±4.5 | 27.5±5.2 | 28.2±5.5 | 27.2±5.8 | 27.1±6.0 | 25.5±6.9 | 28.0±6.0 | 0.452 |
| MMSE, points | − | 28.0±- | 28.1±2.0 | 27.0±3.3 | 26.3±2.9 | 24.6±4.6 | 22.1±7.0 | 25.1±4.2 | 26.0±3.0 | < 0.001 |
| Stroke, % | − | 0 | 5.6 | 11.9 | 19.7 | 26.6 | 26.1 | 45.5 | 0 | 0.035 |
| Heart failure, % | - | 0 | 5.6 | 34.3 | 57.9 | 43.0 | 73.9 | 54.5 | 44.4 | < 0.001 |
| Chronic lung disease, % | - | 0 | 11.1 | 13.4 | 19.7 | 19.0 | 13.0 | 18.2 | 22.2 | 0.921 |
| Cancer, % | - | 0 | 5.6 | 10.4 | 18.4 | 12.7 | 13.0 | 0 | 77.8 | < 0.001 |
| Fall, % | - | 0 | 22.2 | 52.2 | 71.1 | 70.9 | 78.3 | 72.7 | 44.4 | < 0.001 |
| Urinary incontinence, % | - | 0 | 0 | 7.5 | 10.5 | 20.3 | 58.7 | 90.9 | 66.7 | < 0.001 |
| More than 5 medications, % | - | 0 | 83.3 | 100 | 100 | 100 | 100 | 100 | 100 | < 0.001 |
Table 2.
| Characteristics | Robust | Pre-frail | Frail | P value |
|---|---|---|---|---|
| Patients, n | 52 | 122 | 133 | − |
| Age, years | 81.4±5.7 | 82.7±6.5 | 83.7±6.5 | 0.046 |
| Female, % | 75.0 | 72.1 | 60.9 | 0.075 |
| BMI, kg/m2 | 28.1±4.9 | 27.8±5.6 | 27.1±5.8 | 0.179 |
| MMSE, points | 27.5±2.5 | 26.2±3.4 | 24.0±5.5 | < 0.001 |
| Stroke, % | 15.4 | 20.5 | 21.8 | 0.627 |
| Heart failure, % | 34.6 | 45.1 | 54.9 | 0.036 |
| Chronic lung disease, % | 13.5 | 15.6 | 18.8 | 0.629 |
| Cancer, % | 11.5 | 13.1 | 17.3 | 0.503 |
| Fall, % | 36.5 | 67.2 | 72.2 | < 0.001 |
| Urinary incontinence, % | 5.8 | 9.0 | 43.6 | < 0.001 |
| More than five medications, % | 92.3 | 94.3 | 99.2 | 0.037 |
The percentage of patients who showed individual components of the frailty phenotype were as follows: 26.1% showed a weight loss > 10 lb or 5% of body weight in the last 12 months, 41.7% reported a feeling “tired all the time”, 50.2% reported being unable to walk or needed help to walk, 64.5 % had a TUG >19 seconds and 32.2% showed clearly abnormal strength on physical examination.
Primary outcome parameter
Death due to any cause
The percentage of patients who died during follow-up was greater in the higher categories of the two frailty measures than in the lower categories of these measures (see Tables 3 and 4).
Table 3.
| Characteristics | 1 Very fit | 2 Well | 3 Managing well | 4 Vulnerable | 5 Mildly frail | 6 Moderately frail | 7 Severely frail | 8 V ery severely frail | 9 Terminally ill | P value |
|---|---|---|---|---|---|---|---|---|---|---|
| Patients, n | 0 | 1 | 18 | 67 | 76 | 79 | 46 | 11 | 9 | - |
| Death due to any cause, % | − | 0 | 0 | 1.5 | 6.6 | 7.9 | 39.1 | 72.7 | 100 | < 0.001 |
| Unplanned admission to hospital during 6 month of follow-up, % | 0 | 16.7 | 29.9 | 38.2 | 29.1 | 52.2 | 18.2 | 55.6 | 0.013 | |
| Fall during 6 month of follow-up, % | − | 0 | 0 | 16.4 | 19.7 | 27.8 | 26.1 | 0 | 22.2 | 0.085 |
Table 4.
Primary and secondary outcomes: Kaplan-Meier estimates at 6 months follow-up by category using the frailty phenotype
| Characteristics | Robust | Pre-frail | Frail | P value |
|---|---|---|---|---|
| Patients, n | 52 | 122 | 133 | − |
| Death due to any cause, % | 1.9 | 4.9 | 30.1 | < 0.001 |
| Unplanned admission to hospital during 6 months of follow-up, % | 34.6 | 34.4 | 34.6 | 0.985 |
| Fall during 6 months of follow-up, % | 17.3 | 20.5 | 21.1 | 0.773 |
The Clinical Frailty Scale, but not the frailty phenotype, was able to discriminate between people who had an unplanned admission to hospital and those patients to whom this did not apply during follow-up (AUC=0.569 (95% CI 0.502-0.636), p=0.046 and AUC=0.500 (95% CI 0.432-0.568), p=0.994, respectively).
The Clinical Frailty Scale and the frailty phenotype were able to discriminate between patients who died and those who survived during follow-up (AUC=0.867 (95% CI 0.807-0.926), p<0.001 and AUC=0.754 (95% CI 0.688-0.821), p<0.001, respectively) (see Figure 1) (the overlap of the confidence intervals of the AUCs indicates that the AUCs were not significantly different).
Figure 1.
Prediction of all-cause mortality during 6 months follow-up using the Clinical Frailty Scale and the frailty phenotype: receiver operation characteristics (ROC) curves
Among individual components of the frailty phenotype, the components feeling “tired all the time”, being unable to walk or needing help to walk, TUG > 19 seconds, and clearly abnormal strength on physical examination, but not the component weight loss >10 lb or 5% of body weight in last 12 month, were able to discriminate between people who died and those who survived during the follow-up period (AUC = 0.668 (95% CI 0.585-0.751), p<0.001; AUC=0.719 (95% CI 0.719 (95% CI 0.648-0.790), p<0.001; AUC=0.609 (95% CI 0.528-0.690), p=0.017; AUC=0.661 (95% CI 0.574-0.749), p<0.001 and AUC=0.547 (95% CI 0.455-0.639), p=0.304; respectively).
Secondary outcome parameters
Unplanned admission to hospital
The percentage of patients with an unplanned admission to hospital during follow-up was greater in the higher categories of the Clinical Frailty Scale than the lower categories, but did not differ among categories of the frailty phenotype (see Tables 3 and 4).
The Clinical Frailty Scale, but not the frailty phenotype, was able to discriminate between people who had an unplanned admission to hospital and those patients to whom this did not apply during follow-up (AUC=0.569 (95% CI 0.502-0.636), p=0.046 and AUC=0.500 (95% CI 0.432-0.568), p=0.994, respectively).
Among the individual components of the frailty phenotype, none was able to discriminate between patients who had an unplanned admission to hospital and those who were not during follow-up (weight loss > 10 lb or 5% of body weight in last 12 month: AUC=0.503 (95% CI 0.435-0.571), p=0.937; feeling “tired all the time”: AUC=0.484 (95% CI 0.4160.552), p=0.649; being unable to walk or needing help to walk: AUC=0.520 (95% CI 0.452-0.588), p=0.557; TUG > 19 seconds: AUC=0.519 (95% CI 0.451-0.587), p=0.584 and clearly abnormal strength on physical examination: AUC=0.499 (95% CI 0.431-0.567), p=0.970; respectively).
Fall
The percentage of patients who experienced a fall during follow-up tended to be greater in the higher categories of the Clinical Frailty Scale compared to lower categories of that scale, but did not differ among categories of the frailty phenotype (see Tables 3 and 4).
The Clinical Frailty Scale tended to be able to discriminate between patients who had a fall and those who did not during follow-up (AUC=0.574 (95% CI 0.501-0.647), p=0.071). The frailty phenotype was not able to discriminate between patients who experienced and those who did not experience a fall during follow-up (AUC=0.518 (95% CI 0.439-0.598), p=0.658).
Among the individual components of the frailty phenotype, the component being unable to walk or needing help to walk (AUC=0.610 (95% CI 0.533-0.687), p=0.007), but none of the other components (weight loss > 10 lb or 5% of body weight in last 12 month: AUC=0.529 (95% CI 0.447-0.610), p=0.485, feeling “tired all the time”: AUC=0.451 (95% CI 0.372-0.530), p=0.233, TUG > 19 seconds: AUC=0.541 (95% CI 0.4620.620), p=0.324, and clearly abnormal strength on physical examination: AUC=0.490 (95% CI 0.410-0.570), p=0.807; respectively) were able to discriminate between individuals who had a fall and those who did not during the follow-up period.
Discussion
The major finding of the current study is that both the Clinical Frailty Scale and the frailty phenotype were predictors of the primary outcome mortality during 6 months of follow-up. Consequently, the predictive accuracy of the two frailty instruments did not differ in terms of statistical difference, as was shown by the overlap of the 95% confidence intervals of the AUCs of the two frailty instruments. However, the overlap of the 95% confidence intervals was only minimal. The Clinical Frailty Scale, but not the frailty phenotype, was a predictor of the secondary outcome unplanned admission to hospital during 6 months of follow-up. Neither the Clinical Frailty Scale nor the frailty phenotype accurately predicted the other secondary outcome fall during 6 months of follow-up. Interestingly, some individual components of the frailty phenotype predicted the primary and some of the secondary outcome measures. Thus,
Primary and secondary outcomes: Kaplan Meier-Estimates at 6 months follow-up by category on the Clinical Frailty Scale our data indicate that not only the frailty phenotype per se, but indeed individual phenotypic components manifested a role as predictors of at least some of the outcome measures in the current study.
Data are scarce regarding the accuracy of short frailty tools as predictors of mortality and other clinical outcomes in older patients who were hospitalized in geriatric wards. Joosten et al (20) found that the frailty phenotype, but not the Study of Osteoporotic Fracture frailty index (FI-SOF), which is, as the frailty phenotype, also a short tool to assess physical frailty (9), predicted 6-month mortality in a cohort of 220 patients who were hospitalized in geriatric wards and 70 years of age or older. A study by Pilotto et al (21), that compared to the aforementioned study by Joosten et al (20), had much greater power because of a much larger sample size, found that the FI-SOF predicted 1-month and 1-year mortality in a cohort of 2033 patients who were hospitalized in geriatric wards and who were at least 65 years or older.
Although data from studies that were carried out in a different hospital setting or included a patient group that differed from that used in the current study might not be directly comparable to the data derived from our study, we think that in light of the findings of the current study these data are important for discussion. Khandelwal et al (22) found that the frailty phenotype could predict in-hospital mortality in a cohort of 250 patients who were at least 60 years of age and hospitalized and cared for in medical wards. Le Maquet et al (23) found that the Clinical Frailty Scale, but not the frailty phenotype, predicted 6-month mortality and that the frailty phenotype, but not the Clinical Frailty Scale, predicted mortality at the intensive care unit in a cohort of 196 patients who were 65 years of age or older and were hospitalized and cared for in an intensive care unit for longer than 24 hours. Ekerstad et al (24) found that the 7-scale Clinical Frailty
Scale that was recently expanded into the now 9-scale Clinical Frailty Scale, predicted the primary composite outcome of that study (death from any cause, myocardial reinfarction, revascularization due to ischemia, hospitalization for any cause, major bleeding, stroke/transient ischemic attack, and a need for dialysis for up to 1 month after inclusion), in-hospital mortality, and 1-month mortality in a cohort of 307 patients who were diagnosed with a non-ST-segment elevation myocardial infarction who were hospitalized on either cardiologic, acute medicine, geriatric wards or a ward for other areas of internal medicine and aged 75 years or older.
In our study a greater frequency of co-morbid conditions, such as stroke, heart failure, cancer, urinary incontinence and being on more than five medications, was associated with the Clinical Frailty Scale or the frailty phenotype. Rockwood et al have used co-morbid conditions, due to their ability to predict premature mortality, among other aspects to grade patients in relation to frailty (10, 11, 13). In terms of the Rockwood et al’s Clinical Frailty Scale co-morbidity is an aspect among others that has to be considered when grading a patient into one of the nine categories of this scale (18). Co-morbid conditions have been suggested as etiologic risk factors for phenotypic frailty (14). In this respect the findings of the current study are in line with a previous analysis by Fried et al (14) in which an association between frequencies of co-morbid diseases and phenotypic frailty were found. Co-morbid diseases and polypharmacy have been found to be associated with adverse outcomes including mortality in older people (25, 26).
Our study has some strengths, including its prospective design, the analysis of two different frailty tools that relate to the two different major frailty constructs together in one cohort, an evaluation of the predictive accuracy of the frailty instruments for both mortality and other clinical outcomes, and the completeness of follow-up.
Our study also has some limitations. The study participants were older patients hospitalized and cared for in geriatric wards. It could be misleading to extrapolate the findings of our study to older patients in other hospitalized settings or other patient groups. The operationalization of the individual phenotypic components in the current study differed from that used in the initial version by Fried et al (14) in the Cardiovascular Health Study with a cohort of 5317 community dwelling older people. This might have reduced the ability of the frailty phenotype to predict outcomes in the current study. However, in contrast to the initial operationalization of the individual phenotypic components by Fried et al (14), the operationalization implemented in our study was not dependent on gender, body size, or other variables. Therefore, the operationalization of the individual phenotypic components might be easier to perform and therefore constitute a feasible option for routine daily clinical practice in a busy hospital setting. In addition, the operationalization of the phenotypic components as was performed in the current study was previously validated using a large cohort of 2305 older people who were community dwelling or residents in long-term care institutions, in the second phase of the Canadian Study on Health and Aging (CSHA-2) by Rockwood et al (19).
In conclusion, the Clinical Frailty Scale and the frailty phenotype predicted mortality during a 6 month follow-up period. The Clinical Frailty Scale, but not the frailty phenotype, predicted at least some of the second outcomes, i.e., the outcome unplanned admission to hospital during 6 months of follow-up. In addition, individual components of the frailty phenotype had a prognostic value in relation to the primary outcome mortality and secondary outcome “fall”, but not the secondary outcome “unplanned admission to hospital” during the 6 month follow-up period.
Acknowledgements
The study was supported by a grant from the Robert Bosch Foundation to MR.
Conflicts of interest
MR, CS, VK, AD, LCB, KGG and CCS have no conflicts of interest to declare.
Disclosure statements
MR reports grants from the Robert Bosch Foundation, during the conduct of the study. CS, VK, AD, LCB, KGG and CCS have nothing to disclose.
References
- 1.Vellas B, Cestac P, Morley JE. Implementing frailty into clinical practice: we cannot wait. J Nutr Health Aging. 2012;16(7):599–600. doi: 10.1007/s12603-012-0096-y. 10.1007/s12603-012-0096-y PubMed PMID: 22836699. [DOI] [PubMed] [Google Scholar]
- 2.McMillan GJ, Hubbard RE. Frailty in older inpatients: what physicians need to know. Q J Med. 2012;105:1059–1065. doi: 10.1093/qjmed/hcs125. 10.1093/qjmed/hcs125 [DOI] [PubMed] [Google Scholar]
- 3.Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. Lancet. 2013;381(9868):752–762. doi: 10.1016/S0140-6736(12)62167-9. 10.1016/S0140-6736(12)62167-9 PubMed PMID: 23395245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rockwood K, Bergman H. Frailty: a report from the 3(rd) joint workshop fo the IAGG/WHO/SFGG, Athens, january 2012. Can Geriatr J. 2012;15(2):31–36. doi: 10.5770/cgj.15.35. 10.5770/cgj.15.35 PubMed PMID: 23259017, PMCID 3516241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.O’Connell MDL, Tajar A, O’Neill TW, et al. Frailty is associated with impaired quality of life and falls in middle-aged and older european men. J Frailty Aging. 2013;2(2):77–83. doi: 10.14283/jfa.2013.12. [DOI] [PubMed] [Google Scholar]
- 6.Abizanda P, Romero L, Sanchez-Jurado PM, Martinez-Reig M, Alfonso-Silguero SA, Rodriguez-Manas L. Age, frailty, disability, institutionalization, multimorbidity or comorbidity. Which are the main targets in older adults? J Nutr Health Aging. 2014;18(6):622–627. doi: 10.1007/s12603-014-0033-3. 10.1007/s12603-014-0033-3 PubMed PMID: 24950154. [DOI] [PubMed] [Google Scholar]
- 7.Lombard K, van Steijn J, Schuur T, et al. Compliance of energy-dense, small volume oral nutritional supplements in the daily clinical paractice on a geriatric ward-an observational study. J Nutr Health Aging. 2014;18(7):649–653. doi: 10.1007/s12603-014-0496-2. 10.1007/s12603-014-0496-2 PubMed PMID: 25226102. [DOI] [PubMed] [Google Scholar]
- 8.Cesari M, Landi F, Vellas B, Bernabei R, Marzetti E. Sarcopenia and physical frailty: two sides of the same coin. Front Aging Neurosci. 2014;6:192. doi: 10.3389/fnagi.2014.00192. PubMed PMID: 25120482, PMCID 4112807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Morley JE, Vellas B, van Kan GA, et al. Frailty consensus: A call to action. J Am Med Dir Assoc. 2013;14(6):392–397. doi: 10.1016/j.jamda.2013.03.022. 10.1016/j.jamda.2013.03.022 PubMed PMID: 23764209, PMCID 4084863. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Mitnitski A, Mogilner AJ, Rockwood K. Accumulation of deficits as a proxy measure of ageing. ScientificWorldJournal. 2001;1:323–336. doi: 10.1100/tsw.2001.58. 10.1100/tsw.2001.58 PubMed PMID: 12806071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.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. 10.1007/BF03327413 PubMed PMID: 16485864. [DOI] [PubMed] [Google Scholar]
- 12.Rockwood K, Song X M, Knight C, et al. A global clinical measure of fitness and frailty in elderly people. CMAJ. 2005;5(173):489–495. doi: 10.1503/cmaj.050051. 10.1503/cmaj.050051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Rockwood K, Rockwood MRH, Mitnitski A. Physiological redundancy in older adults in relation to the change with age in the slope of a frailty index. J Am Geriatr Soc. 2010;58:318–323. doi: 10.1111/j.1532-5415.2009.02667.x. 10.1111/j.1532-5415.2009.02667.x PubMed PMID: 20370858. [DOI] [PubMed] [Google Scholar]
- 14.Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: Evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56A:M146–MM56. doi: 10.1093/gerona/56.3.m146. 10.1093/gerona/56.3.M146 [DOI] [PubMed] [Google Scholar]
- 15.Ensrud KE, Ewing SK, Taylor BC, et al. Comparison of 2 frailty indexes for prediction of falls, disability, fractures, and death in older women. Arch Intern Med. 2008;168(4):382–389. doi: 10.1001/archinternmed.2007.113. 10.1001/archinternmed.2007.113 PubMed PMID: 18299493. [DOI] [PubMed] [Google Scholar]
- 16.Ensrud KE, Ewing SK, Cawthon PM, et al. A comparison of frailty indexes for prediction of falls, disability, fractures and mortality in older men. J Am Geriatr Soc. 2009;57(3):492–498. doi: 10.1111/j.1532-5415.2009.02137.x. 10.1111/j.1532-5415.2009.02137.x PubMed PMID: 19245414, PMCID 2861353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Moorhouse P, Rockwood K. Frailty and its quantitative clinical evaluation. J R Coll Physicians Edinb. 2012;42:333–340. doi: 10.4997/JRCPE.2012.412. 10.4997/JRCPE.2012.412 PubMed PMID: 23240122. [DOI] [PubMed] [Google Scholar]
- 18.http://geriatricresearch.medicine.dal.ca/clinical_frailty_scale.html. Accessed 13 November 2014
- 19.Rockwood K, Andrew M, Mitnitski A. A comparison of two apporaches to measuring frailty in elderly people. J Gerontol A Biol Sci Med Sci. 2007;62A:738–743. doi: 10.1093/gerona/62.7.738. 10.1093/gerona/62.7.738 [DOI] [PubMed] [Google Scholar]
- 20.Joosten E, Demuynck M, Detroyer E, Milisen K. Prevalence of frailty and its ability to predict in hospital delirium, falls, and 6-month mortality in hospitalized older patients. BMC Geriatrics. 2014;14:1. doi: 10.1186/1471-2318-14-1. 10.1186/1471-2318-14-1 PubMed PMID: 24393272, PMCID 3905102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Pilotto A, Rengo F, Marchionni N, et al. Comparing the prognostic accuracy for allcause mortality of frailty instruments: a multicentre 1-year follow-up in hospitalized older patients. PLoS ONE. 2011;7(1):e29090. doi: 10.1371/journal.pone.0029090. 10.1371/journal.pone.0029090 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Khandelwal D, Goel A, Kumar U, Gulatj V, Narang R, Dey AB. Frailty is associated with longer hospital stay and increased mortality in hospitalized older patients. J Nutr Health Aging. 2012;16(8):732–735. doi: 10.1007/s12603-012-0369-5. 10.1007/s12603-012-0369-5 PubMed PMID: 23076517. [DOI] [PubMed] [Google Scholar]
- 23.Le Maguet P, Roquilly A, Lasocki S, et al. Prevalence and impact of frailty on mortality in elderly ICU patients: a prospective, multicenter, observational study. Intensive Care Med. 2014;40(5):674–682. doi: 10.1007/s00134-014-3253-4. PubMed PMID: 24651884. [DOI] [PubMed] [Google Scholar]
- 24.Ekerstad N, Swahn E, Janzon M, et al. Frailty is independently assoicated with short-term outcomes for elderly patients with non-ST-segment elevation myocardial infarction. Circulation. 2011;124:2397–2404. doi: 10.1161/CIRCULATIONAHA.111.025452. 10.1161/CIRCULATIONAHA.111.025452 PubMed PMID: 22064593. [DOI] [PubMed] [Google Scholar]
- 25.Chen C, Sia I, Ma H-M, et al. The synergistic effect of functinoal status and comorbidity burden on mortality: a 16-year survivial analysis. PLoS ONE. 2014;9(8):e106248. doi: 10.1371/journal.pone.0106248. 10.1371/journal.pone.0106248 PubMed PMID: 25170612, PMCID 4149557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gnjidic D, Hilmer SN, Blyth FM, et al. Polypharmacy cutoff and outcomes: five or more medications were used to identify community-dwelling older men at risk of different adverse outcomes. J Clin Epidemiol. 2012;65(9):989–995. doi: 10.1016/j.jclinepi.2012.02.018. 10.1016/j.jclinepi.2012.02.018 PubMed PMID: 22742913. [DOI] [PubMed] [Google Scholar]