Background
Increasingly frail patients are being to be referred for invasive cardiac interventions and cardiac surgery. We aimed to evaluate the utility of a quick clinical assessment of frailty against a validated frailty assessment tool in an acute cardiology setting.
Methods
Forty-seven cardiology in-patients ≥70 years were recruited in this prospective study. All patients were first assessed by a senior cardiology registrar as either not-frail or frail. This was based on general observation and brief discussions. Following this, patients were administered the Reported Edmonton Frail Scale (REFS) questionnaire. After a registrar assessment, the foot-of-the bed frailty assessment was independently repeated by one or two consultant cardiologists.
Results
None of the three clinicians showed satisfactory similarity to the REFS score. When the two consultants were compared with the registrar, and with each other, the Cohen's kappa was only above 0.7 for the comparison between Consultant 1 and the registrar. Consultant 1 and the registrar were also significantly more likely to disagree at higher REFS score with a mean REFS score of 8.8.
Conclusion
A quick foot-of-the-bed clinical assessment is not a reliable way to determine frailty.
Keywords
Introduction
The assessment of frailty is increasingly topical in both cardiac surgery and cardiology. Frail patients are more vulnerable to the stresses of acute illnesses and are at increased risk of surgical complications, recurrent hospital admissions, eventual institutionalisation and death [
1
, 2
, 3
, 4
, 5
]. Increasingly frail patients are presenting to be referred for invasive cardiac interventions and cardiac surgery. It is therefore important to identify frail patients who are unlikely to benefit from such procedures or whom may in fact come to harm.There is no gold standard in the assessment of frailty [
6
, 7
, 8
, 9
, 10
]. In an acute setting where a comprehensive assessment by a geriatrician is seldom practical, patient's frailty assessment is often done at the foot-of the bed based on visual appearance and a quick clinical judgment [9
, 10
]. Various frailty assessment tools have been developed to make frailty assessment more objective and to make the decision-making more transparent. Most of these are also time consuming and have not been formally assessed in the acute cardiology setting. Many frailty assessment tools assess around 30-70 domains of frailty and these tools are usually poorly understood by non-geriatricians [9
, 11
, 12
].Simplified tools have been developed for use in settings such as acute care clinical practice [
13
, 14
]. One such tool is the Edmonton Frail Scale [[14]
]. This scale uses 11 items to assess physical and psychosocial features of frailty and incorporates some performance measures. It has been validated against a geriatrician's comprehensive assessment, the Geriatric Clinical Impression of Frailty (GCIF). In an acute care setting, however, performance based measures may be confounded by performance limitation related to the acute illness. The Reported Edmonton Frail Scale (REFS) which was adapted from the Edmonton Frail Scale, uses participants’ self-reported function overcoming the limitations of performance assessment [[13]
]. It is a scale that can be readily completed in a few minutes by staff without specific geriatric training. The REFS has been performed by non-geriatrician researchers and has been cross-validated against the GCIF in an Australian acute care hospital and was found to correlate moderately well (R=0.61) with the GCIF with an excellent inter-rater reliability (kappa=0.83) [[13]
].The aim of this study was to evaluate the utility of a quick clinical assessment against this validated frailty assessment tool to determine if an elderly patient is frail or not. We hypothesised that a traditional foot-of-the-bed frailty assessment is closely related to a frailty assessment tool with little inter-observer variability.
The secondary aim of this study was to evaluate the frailty status of elderly patients who have been offered coronary intervention or cardiac surgery at the Christchurch Hospital. We hypothesised that based on current practice, patients who are offered either coronary intervention or cardiac surgery were more likely to be non-frail.
Methods
This prospective study was conducted in Christchurch Hospital. Ethical approval for the study was obtained from the Health and Disabilities Ethics Committees (HDEC). We recruited cardiology in-patients 70 years or older. Patients admitted for elective procedures were excluded. Patients who met inclusion criteria were identified using the hospital electronic database and were approached. Patients who declined consent or who were unable to provide consent were not assessed.
Patients were first assessed by a Senior Cardiology Registrar. The registrar was blinded to their past medical history, investigation results, current diagnosis and treatment. After obtaining verbal consent, an initial foot-of-the-bed frailty assessment was made by the Registrar. Patients were assessed as either not-frail or frail. The assessment was made based on general observation and brief discussions with the patients. This process took only a few minutes with the intention that this mimicked a typical foot-of- the-bed assessment made on ward rounds.
After the initial assessment, a structured interview was undertaken to obtain baseline demographics which included age, gender, ethnicity, living circumstances and mobility status. Following this, patients were administered the REFS questionnaire (Table 1). After the Registrar assessment and REFS scoring, the foot-of-the-bed frailty assessment was independently repeated by one or two Consultant Cardiologists who would spend no more than five minutes at the patient's bedside.
Table 1The Reported Edmonton Frail Scale
[13]
.| Frailty domain | Item | 0 point | 1 point | 2 points |
|---|---|---|---|---|
| Cognition | Please imagine that this pre-drawn circle is a clock. I would like you to place the numbers in the correct positions then place the hands to indicate a time of ‘ten after eleven’ | No errors | Minor spacing errors | Other errors |
| General health status | In the past year, how many times have you been admitted to a hospital? | 0 | 1-2 | ≥2 |
| In general, how would you describe your health? | Excellent/Very good/Good | Fair | Poor | |
| Functional independence | With how many of the following activities do you require help? (meal preparation, shopping, transportation, telephone, housekeeping, laundry, managing money, taking medications) | 0–1 | 2–4 | 5-8 |
| Social support | When you need help, can you count on someone who is willing and able to meet your needs? | Always | Sometimes | Never |
| Medication use | Do you use five or more different prescription medications on a regular basis? | No | Yes | |
| At times, do you forget to take your prescription medications? | No | Yes | ||
| Nutrition | Have you recently lost weight such that your clothing has become looser? | No | Yes | |
| Mood | Do you often feel sad or depressed? | No | Yes | |
| Continence | Do you have a problem with losing control of urine when you don’t want to? | No | No | |
| Self-reported performance | Two weeks ago were you able to: | |||
| (1) Do heavy work around the house like washing windows, walls or floors without help? | Yes | No | ||
| (2) Walk up and down stairs to the second floor without help? | Yes | No | ||
| (3) Walk 1 km without help? | Yes | No | ||
Scoring the Reported Edmonton Frail Scale (/18):
Not Frail 0–5
Apparently Vulnerable 6–7
Mild Frailty 8–9
Moderate Frailty 10–11
Severe Frailty 12–18
Secondary Study
A separate observational study was conducted to assess frailty status as determined by REFS in 15 cardiology in-patients and 15 cardiothoracic in-patients ≥ 70yrs who would have undergone invasive coronary intervention and cardiac surgery respectively. Patients who met the criteria were approached by the Senior Cardiology Registrar and after obtaining consent were assessed using the REFS questionnaire.
Statistical Analysis
The REFS classifies patients with a score of 0-5 as non-frail, 6-7 as vulnerable and 8-18 as frail. However, for the purpose of this study, a REFS score of 0-7 was classified as non-frail and 8-18 as frail. Comparison between non-frail and frail groups were carried out using the Wilcoxon Rank Sum test for non-parametric continuous variable and the Fisher's exact test for categorical variables with small cell sizes. Cohen Kappa was used to assess REFS-observer and inter-observer variability.
Results
Due to the acute nature of cardiology inpatient care 50 patients were approached for consent during a study period of 45 days. Three patients declined to participate. Patient demographics of the 47 cardiology inpatients we studied are shown in Table 2.
Table 2Patient characteristics.
| All participants | N=47 |
|---|---|
| Mean age (years) | 78 |
| Male | 55% |
| Reason for admission | |
| Chest pain | 53% |
| Arrhythmia | 28% |
| Heart failure | 11% |
| Valvular | 6% |
| Hypotension | 2% |
Based on their REFS, patients were divided into non-frail (0-7) and frail (8-18). Participants’ baseline characteristics, are shown in Table 3. There were no differences between the two groups in terms of age, gender, and living circumstances. However, the use of mobility aids significantly correlated with frailty. Patients needing walking frames being more likely to be classified as frail and those independently mobile as non-frail (p<0.05).
Table 3Participant characteristics stratified by frailty on REFS score.
| Characteristics | Frail 9 (19.2%) | Non-frail 38 (80.8%) | P-value | |
|---|---|---|---|---|
| Age (SE) | 76 (4.1) | 78 (6.4) | 0.29 | |
| Gender | Male | 5 | 21 | 1.00 |
| Female | 4 | 17 | ||
| Mobility aids | Frame | 4 | 3 | 0.02 |
| Stick | 1 | 8 | ||
| None | 4 | 27 | ||
| Living at | Home | 8 | 36 | |
| Other | 1 | 2 | 0.47 | |
| Living with | Spouse | 7 | 26 | 0.06 |
| Alone | 0 | 10 | ||
| Other | 2 | 2 |
In this table, ‘non-frail’ is defined as REFS 0-7 and ‘frail’ is defined as REFS 8-18. All variables are reported as mean (standard deviation) or number (% within category). Comparisons between frail and non-frail participants are carried out using Wilcoxon Rank Sum test for nonparametric continuous variables and Fisher's exact test for categorical variables with small cell sizes.
REFS-observer and Inter-observer Agreement
Table 4 compares the REFS with the foot-of-the bed frailty status as determined by the three clinicians. A Cohen's kappa of greater than 0.70 is generally considered to represent satisfactory similarity between indices. None of the three clinicians showed satisfactory similarity to the REFS score. When the two Consultants were compared with the Registrar and when the two Consultants were compared with each other, the Cohen's kappa was only above 0.7 for the comparison between Consultant 1 and the Registrar. When this association was looked at more closely, Consultant 1 and the Registrar were significantly more likely to disagree at higher REFS score with a mean REFS score of 8.8 (Table 5).
Table 4Inter-observer agreement.
| Observer | Frail by REFS | Non-frail by REFS | Cohen's kappa | |
|---|---|---|---|---|
| Registrar | Frail | 5 (10.6%) | 13 (27.7%) | 0.1545 |
| Non-frail | 4 (8.5%) | 25 (53.2%) | ||
| Consultant 1 | Frail | 3 (9.7%) | 9 (29.0%) | 0.0428 |
| Non-frail | 4 (12.9%) | 15 (48.4%) | ||
| Consultant 2 | Frail | 1 (6.3%) | 4 (25.0%) | 0.2558 |
| Non-frail | 0 (0%) | 11 (68.7%) | ||
| Registrar | Frail | Non-frail | Cohen's kappa | |
| Consultant 1 | Frail | 10 (32.3%) | 2 (6.45%) | 0.7281 |
| Non-frail | 2 (6.45%) | 17 (54.8%) | ||
| Consultant 2 | Frail | 3 (18.75%) | 2 (12.5%) | 0.2131 |
| Non-frail | 4 (25.0%) | 7 (43.75%) | ||
| Consultant 2 | Frail | Non-frail | Cohen's kappa | |
| Consultant 1 | Frail | 2 (18.2%) | 1 (9.1%) | 0.5417 |
| Non-frail | 1 (9.1%) | 7 (63.6%) |
In this table, all variables are reported as number (% within total). Cohen's kappa is used to assess REFS-observer agreement, as well as inter-observer agreement. Larger kappa indicates greater similarity. Generally, a Kappa > 0.70 is considered satisfactory.
Table 5Association with REFS score for Registrar and Consultant 1.
| N | REFS Mean (SD) | P-value | |
|---|---|---|---|
| Agree | 27 | 5.0 (2.8) | 0.0230 |
| Disagree | 4 | 8.8 (3.3) |
In this table, student t-test was used to assess statistical significance as normality assumption holds.
REFS in Elderly Patients who Underwent PCI or Cardiac Surgery
We recruited 15 cardiology in-patients who had percutaneous coronary intervention (PCI) and nine cardiothoracic in-patients who had cardiac surgery (67% coronary artery bypass surgery and 33% valvular surgery). The mean age in both groups were comparable. Mean age in those who received PCI was 78.2 years (range: 73-90 years) and those who underwent cardiac surgery had a mean age of 77.6 years (range: 72-86 years).
The mean REFS in the PCI group was 2.7 (median 3, range: 0-6). Mean REFS was higher in the cardiac surgical group with a score of 4.4 (median 2, range: 1-11). Two patients in the surgical group scored 8 and one scored 11 on their REFS questionnaire. This finding is counter to our initial hypothesis. On further analysis, we found that one of the patients who scored an 8 has a complex history of incomplete tetraplegia with chronic back pain. The other had been waiting for his mitral valve surgery for some months until his symptoms got worse and had to be admitted acutely. The patient who scored 11 was a rest home resident with a complex medical history of end-stage renal failure on haemodialysis and chronic airways disease with declining health over some years. The first two patients had uncomplicated postoperative recovery. The other patient had postoperative right pneumothorax on day-4 requiring a chest drain. All three patients were seen again either as in-patient or at their clinic reviews at around one month. They were found to be doing reasonably well apart from minor infections (one with sternal wound cellulitis and two lower respiratory tract infections) responding well to antibiotics.
Discussion
Frailty is a difficult topic with there being no single definition of frailty. Traditionally the five phenotypes of frailty are: weakness, sarcopenia, weight loss, physical inactivity and slowness [
5
, 10
]. However, there exists an overlap between frailty and other syndromes associated with ageing [1
, 8
, 11
, 12
]. Therefore the inclusion of disability and comorbidity has been recommended when assessing frailty.A comprehensive geriatric assessment depends on clinical judgement and would include a personal and informant history, physical exam, functional performance assessment, and a mental state exam. The results of such an assessment will vary between clinicians, and while most geriatricians can accurately identify a frail patient, there is no consensus regarding its definition [
6
, 7
, 8
, 9
, 10
].In this prospective study, we compared a quick foot-of-the bed frailty assessment to the REFS. The 47 cardiology inpatients in our study population were reflective of typical elderly patients admitted under cardiology in the Christchurch Hospital. From our study, we found that there was poor agreement between the clinicians and the REFS, and also poor agreement amongst the clinicians. Clinicians were more likely to agree in patients who were clearly “not frail” on their REFS and more likely to disagree in patients with REFS in the borderline zone of “not frail” and “frail”. The findings highlight the difficulties in frailty assessment. Experienced cardiologists may only satisfactorily agree on frailty in cases where the patient is clearly not frail. In borderline cases clinicians should be aware of the limitations of quick assessment and adopt a more tailored approach. Repeated patient contact or longer discussion and assessment may be means of increasing the validity of frailty assessments by individual clinicians in the acute inpatient care setting. Current practice is that on occasion a clinician may ask an interventionalist or surgeon to see a patient to assess frailty before they are finally listed for a procedure. Our data supports using that approach frequently in the elderly and highlights that second opinions should also be sought in some cases where initial individual assessment is that the patient is too frail to consider intervention or surgery.
In the second part of the study we applied the REFS to small groups of patients undergoing PCI and cardiac surgery. The mean REFS was higher in the cardiac surgical group with no patients in the PCI group being classified as frail. A detailed analysis of the three patients who were classified as frail in the surgical group revealed that these patients had chronic debilitating symptoms prior to their surgery and this may have contributed to their higher REFS. Our sample size in both groups is small but this data highlights again the limitations of using a frailty score. The patients had been deemed suitable for the operation by both their Cardiologist and the Cardiac Surgeon.
Study Limitations
The sample size of this prospective study was small. This was mainly due to early patient discharges and the Registrar not being able to reach these patients before discharge. These patients were likely to be well, have non-cardiac or non-acute cardiac issues and would likely have low REFS. Despite being a small sample size, our patient population was sufficiently reflective of the typical elderly patients admitted under cardiology. A longer study period with a larger study population may be useful to emphasise the findings of our study.
Secondly, the REFS which was chosen as a quick and easy to perform Frailty assessment tool, although it correlates moderately well to the GCIF, would not be useful to accurately determine if a person was frail or not. This would likely contribute to the poor agreement between the clinicians and the REFS when assessing frailty.
Conclusion
Data from our study suggest that a quick foot-of-the-bed clinical assessment is not a reliable way to determine frailty. There was poor agreement between clinicians and in particular patients who were considered borderline frail on their REFS. In addition, the role for REFS or other frailty indices in setting of acute cardiology remains highly uncertain. Therefore, we would recommend that when making major cardiac interventional or surgical decisions, it is important to spend time with the patients and adopt a consultative approach.
Conflict of Interest
None to declare
References
- Frailty in the older surgical patient: a review.Age Ageing. 2012; 41: 142-147
- Frailty is associated with postoperative complications in older adults with medical problems.Arch Gerontol Geriatr. 2009; 48: 78-83
- Frailty as a predictor of surgical outcomes in older patients.J Am Coll Surg. 2010; 210: 901-908
- Comprehensive assessment of frailty for elderly high-risk patients undergoing cardiac surgery.Eur J Cardiothorac Surg. 2011; 39: 33-37
- Frailty in older adults: evidence for a phenotype.J Gerontol A Biol Sci Med Sci. 2001; 56: M146-M156
- Frailty: An Emerging Research and Clinical Paradigm - Issues and Controversies.J Gerontol A Biol Sci Med Sci. 2007; 62: 731-737
- Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care.J Gerontol Biol Sci Med Sci. 2004; 59A: 255-263
- What would make a definition of frailty successful.Age and Ageing. 2005; 34: 432-434
- Role of frailty assessment in patients undergoing cardiac interventions.Open Heart. 2014; 1: e000033
- Frailty assessment in the cardiovascular care of older adults.J Am Coll Cardiol. 2014; 63: 747-762
- Frailty defined by deficit accumulation and geriatric medicine defined by frailty.Clin Geriatr Med. 2011; 27: 17-26
- A standard procedure for creating a frailty index.BMC Geriatrics. 2008; 8
- The assessment of frailty in older people in acute care.Australasian Journal on Ageing. 2009; 28: 182-188
- Validity and reliability of the Edmonton Frail Scale.Age and Ageing. 2006; 35: 526-529
Article info
Publication history
Published online: December 18, 2014
Accepted:
November 25,
2014
Received in revised form:
November 17,
2014
Received:
September 16,
2014
Identification
Copyright
© 2014 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier Inc. All rights reserved.
