Final Report Summary - FOD-CC (Frailty operative definition - consensus conference)
Executive Summary:
Frailty is a relatively new concept trying to explain the predisposition of old people to disability and other adverse outcomes such as death, hospitalisation, institutionalisation and falls.
This concept has been approached several times from different points of view, including a large number of different 'constructs' depending of each expert opinion.
Until now we have many definitions of frailty, but none of them is validated in longitudinal studies or without the bias of a single point of view.
With the aim of providing an inclusive definition of frailty, including the opinions of all the experts interested in it, but also with a useful clinical profile, we have developed the Frailty Operative Definition-Consensus Conference (FOD-CC) project.
The design of this project has taken into account several important points:
1. the need for selecting the core domains to be included in the final definition
2. the need to get these domains from the different fields of interest concerned by frailty, selecting experts in every field of interest of frailty
3. the need to build a definition to be used in daily practice in clinical settings.
To reach our practical consensual definition, we have use a methodology based on the Delphi method. We used three consecutive questionnaires, together with face-to-face meetings between experts.
These experts have been carefully selected of five main groups (focus groups) of professionals interested in frailty, namely geriatricians, non-geriatricians clinicians, non-medical health workers, basic scientists and social and non-governmental organisations (NGOs), each one composed by six to seven members.
Also, external experts have been contacted to participate in filling the questionnaires, until reaching the figure of 83 participants from 111 experts contacted (response rate, 75%), with a good balance between the groups. None group was overrepresented or underrepresented in the final group of participants.
Finally, an advisory board, composed by several renowned scientists in frailty that remained blinded to previous process to avoid contamination, has agreed to give support to our project, sending their comments to refine the final conclusions of the project.
Our proposed definition of frailty is composed by six core domains, but any set of biomarkers reach the consensus to be used as the clinical definition of frailty. In addition, although there was consensus about the potential utility of laboratory biomarkers, anyone has reached enough consensuses to be accepted as a diagnostic tool. There is a strong agreement on the importance and relevance of frailty, its relevance as a clinical syndrome and the necessity to continue making primary research to bridge the lacks of knowledge concerning many issues related to its clinical utility.
Project Context and Objectives:
The concept of frailty tries to increase our understanding of why many older persons develop disability late in life. The concept of frailty has emerged from a wide variety of studies and theoretical discussions and there is, as yet, no clear consensus regarding its definition. The most used accepted approach defines frailty as 'an age-associated biological syndrome' characterised by a decrease of the biological reserve and resistance to stress, due to a decline in several physiological systems, putting the individual at a special risk when facing little stressors and associated with poor outcomes such as disability, death and institutionalisation or hospitalisation' (Campbell and Buchner, 1997; Walston and Fried, 1999; Rockwood et al, 2000). The model proposed by Walston and Fried (1999) places sarcopenia and energy misbalance as the fundamental elements of frailty. Also, Fried et al. identified five criteria to define the frail phenotype, namely weakness, slowness, low physical activity, unintentional loss of weight and exhaustion, in the cardiovascular health study cohort. These researchers suggested that individuals with three or more of these criteria should be characterised as frail. One of the helpful characteristics of this definition is the capacity to determine risk or adverse outcomes including death, disability, impairment in mobility, hospitalisation and institutionalisation (Fried L., 2001; Bandeen-Roche K 2006) and it can also assist in differentiating between frail, pre-frail (individuals who presented only one or two criteria among the five) and non-frail individuals.
Most researchers agree that frailty is a 'syndrome' and a number of definitions have been proposed each of which has its own strengths and weaknesses (de Vries, 2010). With increasing popularity, frailty definitions proliferated; each definition catching a new relevant aspect, proposing novel dimensions or improving its accuracy for the specific context and aims it was created for (epidemiological purposes).
Despite the efforts of several research groups, documenting the predictive validity of various frailty indices remains a challenge, with few prospective population or clinical studies demonstrating the ability to reliably predict the onset of adverse outcomes in frail individuals. Indeed, the Fried criteria are mainly focussed on the physical functioning domain of health. Some authors have claimed for the necessity of including other items (like mental health and accumulated vulnerability) within a frailty definition (Bergman H et al, 2007). In addition, given that lifestyle factors are likely to influence the development and rate of progression of frailty, there is a need for definitions to be validated in a wide variety of cultural, economic, ethnic and clinical settings. The majority of studies have examined frailty in Anglo-American populations and there is a clear need to extend this work to other population groups with different lifestyles and genetic backgrounds. In this regard, two recent papers show in a European population that the risk of adverse outcomes using the Fried criteria of frailty was lower than expected for some outcomes (Avila-Funes JA et al., 2008) and that, cognitive decline, a notion not included in these criteria, was an important element of frailty that increased the predictive validity of the Fried index model in determining risk for adverse outcomes (Avila-Funes JA et al., 2009).
Growing evidence suggests that nutritional status may play a relevant role in the process leading to frailty through a variety of multidimensional effects. In addition, another phenotype strongly associated with frailty and sharing many of its pathophysiologic components has recently emerged, namely sarcopenic obesity. This new concept supports the addition of measures of body mass index (BMI) and waist to hip ratio to the criteria of frailty. In 40 600 women aged 65 to 79 years enrolled in the Woman's Health Initiative Observational Study (Woods NF, 2005), the risk of incident frailty was raised when BMI was over 25 kg/m2 and was decreased when BMI was under 18.5 kg/m2. Consistent with this finding, there is evidence that a state of chronic inflammation in obese people, measured by an increased C-reactive protein (CRP), may be an important marker of the frailty syndrome (Fruhbeck G, 2004). Some authors suggest that the origin of such inflammation could be in the adipose tissue (Guzik TJ, 2006) as it releases many inflammatory substances (leptin, adiponectin, resistin, visfatin, apelin, etc). In addition, it has been found that adipocytes secrete pro-inflammatory cytokines such as IL-6 (Fontana L, 2007). Finally, a growing literature provides evidence of the role of many pathophysiologic components in the syndrome of frailty. Among the components proposed a decrease in sexual hormones (Iannuzzi-Sucich M et al, 2002), dysfunction of the growth hormon-insuline-like growing factor-1 (GH-IGF-1) axis (Waters DL et al, 2003), increase in the ratio of cortisol to Dehidroepiandrosterona (DHEAs) (Walston and Fried 1999, Taylor JM, 2003), testosterone deficit (Kaufman JM and Vermeulen A, 2005; Travison et al., 2007), increases in interleukin-6 (IL-6), IL-1 and tumour necrosis factor (TNF) alpha circulating levels (Penninx BW et al., 2004), CRP and d-dimer (Walston J et al., 2002) and other pro-inflammatory cytokines (Leng S et al.,2002) are included. In addition, data from the cardiovascular health study showed that subclinical cardiovascular disease was related to frailty, and should be considered a risk factor for disability.
The main problem with the majority of the current frailty definitions is their lack of validation in well controlled prospective longitudinal studies, the lack of support for any single approach with regard to the definition of this complex syndrome and the shortage of studies testing the predictive capacity of the current definition in a clinical population (Robinson TN et al, 2009; Sündermann S et al, 2011). There is increasing evidence that multiple biological, clinical and social factors are involved in the development of frailty (Bergman et al. 2007). All these factors interact to produce conditions that place an individual at a high risk for developing disability and a decreased quality of life. The multiplicity of factors leading to frailty has made it difficult to reach a unique, comprehensive and operative definition of the term. Indeed existing definitions tend to be skewed in favour of social, biological or clinical elements depending on the background and expertise of researchers building the definition. On the other hand, frailty markers, individually or in combination may be as important as a model. It is also important to develop biological markers of frailty which are able to support the clinical diagnostic markers in predicting the onset of frailty and its progression to disability and other adverse outcomes.
To conclude, frailty has attracted the attention of the researchers from numerous fields of study. However, despite this interest, the need to develop an operative definition of frailty that can then be translated into a clinical instrument remains paramount if we are to improve the health and quality of life of the older population. In addition, this definition should also be feasible to apply in general population settings, in order to enable tracking over time and identification of populations at greatest risk such that public health and policy interventions can be appropriately targeted and monitored.
Thus, a working group involving experts of each area related to frailty gathered together with the aim of providing the most complete and concrete definition as possible, selected biomarkers of clinical utility and a guideline to allow early diagnosis of frailty to be used in clinical practice. In order to achieve the main objectives, the secondary aims of the project were:
1. to evaluate the rate of consensus between experts regarding the framework of frailty
2. to evaluate the rate of consensus between experts regarding the domains of frailty and their clinical value
3. to evaluate the rate of consensus between experts regarding biomarkers of frailty and their diagnostic value
4. to evaluate the utility of animal models
5. to provide research fields of interest.
The position that was developed can be summarised in the following points:
1. frailty is a relevant risk factor for disability
2. different definitions including different set of domains are available
3. frailty has been evaluated in epidemiological settings, being a good marker of adverse outcomes in older people
4. the utility of laboratory biomarkers has been scarcely explored.
The project targets are as follows:
1. to establish the basis for a clinical definition
2. to determine domains to include in a clinical definition
3. to determine the utility of several clinical and laboratory biomarkers
4. to determine the gaps to be investigated in the future.
Project Results:
Methodology and scientific results
The aim of the FOD-CC project was to provide an operative definition of 'frailty' which would be useful in daily practice. This definition was constructed from a wide variety of different expert opinions in order to result in the most complete and concrete definition possible.
The Delphi procedure was used to reach this objective. The Delphi method is a structured communication technique, originally developed as a systematic, interactive forecasting method which relies on a panel of experts and which is a well established way to reach consensus between experts from different fields (Linstone HA and Turoff M, 1975).
The Delphi method is based on the principle that forecasts (or decisions) from a structured group of individuals are more accurate than those from unstructured groups. The traditional steps in the Delphi process are formation of a Delphi team, selection of one or more panels to participate (experts in the area), development of a first round Delphi questionnaire, testing the questionnaire for proper wording, transmission of the questionnaire to the panelists, analysis of the first round responses and repeating of the same procedure until a predefined stop criterion is reached (normally the achievement of a consensus) and, finally, the preparation of a report by the analysis team to present the conclusions.
In order to improve this procedure, we introduced some modifications to the classical Delphi method, mostly with the aim of increasing the amount of information available to the experts. Most importantly, we involved all experts in the formulation of the first round questionnaire by asking them four open-ended questions related to frailty and by encouraging communication among them during an initial face-to-face meeting. These modifications were carried out according to the following strategies:
1. after having accepted an offer to participate in the project, all experts were sent a sample of published scientific papers about frailty. This selected bibliography was different for each group depending on their area of specialisation. This was done in order to help them with their answers to the questionnaires.
2. the experts worked together on the open-ended questionnaire about frailty prior to the development of the Delphi questionnaire itself.
3. in addition to extensive e-mail communications between participants, two face-to-face meetings between experts were conducted in order to enable all aspects of the Delphi process to be thoroughly discussed.
4. a website for the project was developed to allow project participants and others to learn more about, and comment on, the goals, methodology and implementation of the project.
The Delphi process consisted of 13 steps.
1. selection of the panel of experts. Five groups of experts were selected with the goal of covering the fields of study that have an interest in the topic of frailty and to provide a multi-disciplinary perspective from all stakeholders. This wide range and variability of fields was very important to our project because this would help to finally ensure that the consensus definition of frailty reached by the project would be valid and applicable to the widest possible number of fields. The focus group were composed, respectively by geriatricians, forming the focus group one (FG1), non-geriatric physicians (FG2), health workers (FG3), basic scientists (FG4) and social and non-governmental workers (FG5). Each FG was composed of about five to seven experts and a chairman. This selection of experts was driven by a steering committee composed by Professor Catherine Feart, Professor José Viña, Professor Giovanni Mann, Dr Somnath Chatterji and Professor Leocadio Rodríguez. The experts were selected not only because of their personal interest in frailty, but also because of their previous work in the field, evaluated by their previous research experience on frailty, publications in scientific journals or daily working experience in managing patients with frailty and related conditions. In addition, a public health perspective also needed to be represented among the experts. These experts were contacted and invited to participate in the project in person or by e-mail. Once they had agreed to participate, all experts received an outline of the project that described the main objectives and tasks.
2. sending of bibliography. To facilitate the participants' immersion in the project, each expert received a compilation of the most relevant references on frailty and was sent a selection of articles related to their own area of interest. Those articles were selected by the leader team, from data bases of medical science publications (Pubmed and Embase), including the keyword 'frailty' and other terms like 'disability', 'diagnosis', 'death', 'hospitalisation' and 'biomarkers'. In addition we made a specific search in the clinical journals included under the section 'geriatrics and gerontology' of the Journal of Citation Report Index. We had an initial selection of 848 articles of which we read the title and make a second selection of the articles with potential interest for our study. This selection was carried out by agreement of two members of the team. After that, those two members read the abstracts and selected the final group of 113 articles. All the members of the FGs received a selection of 27 common articles and around 20 articles specific for each focus group.
3. open-ended preliminary questionnaire. All FG members participated in the development of the first round questionnaire (1RQ). They received an open-ended preliminary questionnaire that included four general and two specific questions about frailty. Their answers were classified according to their similarity. Results were presented during the first face-to-face meeting of the FG in Valencia, Spain.
4. first face-to-face meeting in Valencia, Spain (February 2011). In Valencia, discussions among FG members led to the development of a set of statements for the 1RQ. Participants were then asked to suggest five external experts to whom the 1RQ could be sent. In addition, they were asked to recommend one institution to be invited to support the study and one person for appointment to the advisory board. The advisory board nominated by the steering committee included 11 experts who did not participate in the Delphi process.
5 design of the 1RQ and preliminary analysis (combining steps five and six). After a further review from all FG members, the final version was disseminated. The 1RQ consisted of 107 statements broken down into five blocks, namely framework and definition (52 statements), biomarkers (43 statements), frailty versus disability (7 statements), frailty versus co-morbidity (4 statements) and animal models (1 statement). All experts were asked to score each statement on a 10 point numerical scale ranging from 1 (no agreement) to 10 (full agreement). All the answers were received and analysed with a statistical method whose results would be discussed during the second meeting.
7. second face-to-face meeting in Bordeaux, France (May 2011). During the second face-to-face meeting, participants established criteria to define the strength of agreement and analysed the data collected. The preliminary 1RQ results were discussed. To improve the response rate to the questionnaire, corresponding to 81 of 130, i.e. 62 %, two strategies were adopted, namely to re-contact non-responding experts and to select additional experts for those FG that had the lowest rate of response in the initial analysis (i.e. non-geriatrician clinicians and basic scientists).
8. definitive analysis of the 1RQ. Using the statistical criteria approved by the experts the final rate was 74.5 %. The statistical analyses were performed in accordance with previously established procedures in reviewed literature. Criteria used to classify degree of agreement were positive and negative consensus was defined as percentage of answers rated, respectively, between plus or minus 8 or plus or minus 3 in 80% and more (strong agreement), percent answers between 70 to 80 % (moderate agreement) were selected to enter the 2RQ, statements with low agreement (about 50 to 70 %) or no agreement (less than 50%) were further analysed to determine if this low percentage of consensus was due to heterogeneity in the opinion of experts, assessed using tests for median comparison (Wilcoxon scores rank test) or to dispersion, reflecting a confusing statement, that was assessed by the presence of an inter quartile range (IQR) of 4. All items with heterogeneity or dispersion were reviewed and included in the 2RQ according to the criteria that in light of the experts' opinion a consensus could possibly be reached or appropriate modifications could improve the clarity of the statement.
9. leadership meeting (Oxford, July 2011). During this meeting attended by the chair men of each focus group, and the aim was to decide the members for both the advisory board and the writing committee
10. development of the second round questionnaire (2RQ). The final version of the 2RQ consisted of 52 statements and was sent to all the experts who evaluated the 1RQ. Consistent with the Delphi process, in order to facilitate agreement among experts, respondents were asked to try to either accept (score 8, 9, 10) or reject (score 1, 2, 3) the proposed statements. Median scores and percentage of agreement from the 1RQ were displayed for reference.
11. statistical analysis of the 2RQ. According with the previously accepted methodology, the LTG analysed the results, but this time they only evaluated the accepted or rejected statements.
12. final face-to-face meeting in Madrid, Spain (October 2011). The final results of the Delphi process were presented in Madrid. In addition to the initial classification of the statements into five categories, a second classification in four categories was approved (i.e. alternative classification), consisting in concept, prognosis, diagnosis and prevention and treatment of frailty. The statements finally accepted to enter into the definition of the FOD-CC experts were presented and submitted for approval according to both classifications.
13. final documents. A final document was presented in the Madrid meeting to all of participants, including the methodology and the results of the project.
Results from the open preliminary questionnaire
The open preliminary questionnaire was composed of four general and two specific questions for each focus group. This questionnaire was sent to 30 members of the focus groups and we received 27 answers, a response rate of 90 %.
2. Results from the 1RQ
The 1RQ (107 questions broken into five themes, namely framework, biomarkers, frailty versus disability, frailty versus co-morbidity and animal models), was sent to 130 experts before the meeting of Bordeaux and we received 81 answers giving a response rate of 62 %. According to the criteria described in the statistical methods, 17 statements were accepted for the final definition, 45 statements were excluded, 30 questions had to be reviewed again because the answers were dispersed or not homogeneous among groups of experts and 15 questions had to be included in a 2RQ because the percentage of persons answering above eight or under three was between 70 and 80 %.
After the Bordeaux meeting, we improved the response rate by once again requesting the participation of prior non-responding experts and by recruiting additional experts in the groups with the lowest response rates. In total, we achieved 111 answers, from 149 questionnaires sent, with a final response rate of 74.5 %. In the light of these new answers, the results were slightly modified. Fifteen statements were accepted (14.0 %), 47 excluded statements (43.9 %), 28 questions to be re-reviewed (26.2 %) and 17 questions were to be included in the 2RQ (15.8 %). Depending upon the theme in which a statement was included, there was an agreement in 57.1 % of the statements regarding the difference between frailty and disability, nine statements (17.3 %) regarding the framework of frailty were accepted and an additional 50 % were either included in the 2RQ or entered into the review process. On the other hand, more than 65 % of the statements regarding biomarkers reached a low consensus and were excluded after the first round, one statement was included in the second round and 12 others (27.9 %) entered the review process.
It is important to underline that the rate of answers according to each focus group was very homogeneous except in the basic scientists group that was under-represented, i.e. 21.6 % of the answers came from geriatricians, 20.7 % from non geriatricians and physicians, 23.4 % from health worker professionals, 11.7 % from basic scientists and 22.5 % from social and nongovernmental workers.
Results from the 2RQ
The 2RQ was sent to the same 111 responders of the 1RQ. We received 83 answers, a response rate of 75 %. At this time the distribution of answers according to the group of experts was 27.7 % of geriatricians, 12.0 % from non geriatrician's physicians, 26.5 % from health workers, 12.0 % from basic scientists and 21.7 % from social and nongovernmental workers.
Twenty-four statements were accepted from this 2RQ; almost 60 % of the statements regarding the framework of frailty, 33 % of the statements regarding biomarkers, two statements (40 %) were about the difference between frailty and disability, one statement was about the difference between frailty and co-morbidity.
Final decisions
In total, 132 statements were proposed to the experts, 107 in the 1RQ and 25 more in the 2RQ. Finally, the total number of accepted statements to be included in our definition of frailty was 39, of which 15 were accepted directly after the 1RQ. There was a clear agreement about the difference between frailty and disability with more than 66.6 % of the statements in this area accepted. In addition, more than 38 % of the statements regarding the framework of frailty reached consensus. Only 15.9 % of the proposed statements about biomarkers were accepted.
As mentioned in the methods, we also used an alternative classification of the statements to analyse the information extracted. In this way, we found that 80 % of the statements regarding prevention and treatment of frailty were accepted. In addition, 44.4 % of the statements regarding the concept of frailty were accepted as well as 33.3 % of the statements regarding prognosis. On the other hand, of the 78 statements about the diagnosis of frailty, only 14 (17.9 %) were accepted.
Conclusions
Members of the FG synthesised the 39 accepted statements into a cohesive narrative text. This draft was presented to all participants during the last meeting in Madrid. The final approved version is as follows (the number of the statement supporting the assertion is provided in brackets along with the text).
Frailty is a dynamic, non linear process (24) that involves alterations in multiple domains of functioning (29), impacts multiple body systems (28, 40) and may be considered a clinical syndrome (4). It is different from vulnerability and disability (26, 39, 96, 97) and is often modulated by disease (102) to the extent that the presence of co-morbidities tends to exacerbate the consequences of frailty (104). Definitions of frailty are frequently multidimensional involving a variety of psychological, social, emotional, and spiritual elements in addition to physical components (13). Definitions of frailty should be reproducible over time (9) and be applicable across different clinical and non clinical settings (7, 43).
Frailty is characterised by a decreased reserve and diminished resistance to stressors (6). No single molecular mechanisms or biomarker has been identified that is adequate for the prediction and/or diagnosis of frailty (32, 60). However, variables including assessing nutritional status (8) and physical performance (14), gait speed (16) and mobility (17) have been proposed for use in the diagnosis of frailty. In addition, mental health assessments and cognitive status evaluations are highly recommended as part of the assessment of frailty (9, 64).
The consequences of frailty include an increased vulnerability to stress in which minimal levels of stress can cause functional impairments (6, 21,27). Although disability can occur without prior frailty, frailty is a risk factor for disability (99). A major purpose of diagnosing frailty is to identify and stratify older persons at high risk for disability and/or other adverse outcomes (12, 45). A diagnosis of frailty is of importance because can help to predict a variety of different health outcomes including disability, falls, hospitalisation, permanent institutionalisation, and death (100, 59). The predictive value of frailty depends on its severity and will vary from person to person (101).
The diagnosis of frailty is useful in geriatric medicine (48), primary and community care (46), and is of value in managing older people with chronic diseases (47). Because frailty is a condition that may be amenable to prevention it is mandatory for clinicians and health workers to identify those at risk for frailty as early as possible (23). Among the interventions that show promise for the management or attenuation of frailty are various healthy lifestyles (7), including physical activity (6, 22).
The aim of the FOD-CC project was to reach a consensus definition of frailty that is useful in daily practice, using a Delphi process. Experts agreed on the importance of a more comprehensive definition of frailty that should include assessment of physical performance, including gait speed and mobility, nutritional status, mental health and cognition. Although a consensus was reached on these six domains, the proposed diagnostic paths and procedures needed to achieve an operational definition reached a low level of consensus; only one out of six of the statements related to diagnosis achieved consensus.
Experts agreed that no single biomarker by itself was adequate for the assessment of frailty, suggesting a need for a combination of multiple biomarkers. However, none of the proposed combinations of biomarkers was able to reach the 80 % threshold of agreement required in the Delphi process. The most surprising result regarding clinical biomarkers was the absence of agreement regarding the inclusion of grip strength in the definition. Interestingly, the percentages of agreement for statements about grip strength were very close to achieving consensus (more than 70 % of agreement). We decided not to include the grip strength items in the definition in order to remain faithful to the a-priori criteria established for the acceptance of consensus. It is possible that our inability to reach consensus reflects concerns about the reliability of grip strength as a measure rather than a reluctance to include strength, per se, as one of the domains of frailty. Importantly, among all the laboratory markers suggested for the assessment of frailty, none was accepted.
The low level of consensus regarding the constituent elements to be included in an operational definition of frailty is balanced by the high degree of agreement regarding the underlying conceptual framework of frailty. The experts clearly agreed that frailty is a multidimensional syndrome characterised by decreased reserve and diminished resistance to stressors. The experts established a clear-cut difference between disability and frailty as shown by the percentage of agreement (85 to 95 %) in the related statements. This notion is somewhat different from the World Health Organisation (WHO) conceptualisation of disability, in which disability is considered to be contextual and every human being can experience some degree of disability. Frailty was also differentiated from vulnerability. While everybody is potentially vulnerable, frailty represents a state of extreme vulnerability where minimal stress may cause functional impairment. This conceptual framework has clinical implications. To be very vulnerable with a low reserve capacity provides a high risk of becoming disabled and a low probability for the restoration of function, underscoring the necessity to detect frail people before they become dependent.
Although frailty has a clear conceptual framework, there is no single operational definition of frailty that can satisfy all experts. A possible explanation may rely on the heterogeneity of our expert focus groups. However, this explanation seems unlikely. Heterogeneity between groups in the Delphi questionnaire responses was systematically tested and was only detected for 5 of the non-accepted statements. Another possible explanation may pertain to the lack of agreement in the literature among existing definitions of frailty. Supporting this perspective we found agreement regarding the necessity to combine biomarkers but no agreement regarding which combination of biomarkers to include in the definition; agreement regarding the necessity to assess the severity of frailty but no agreement on identifying specific severity markers; agreement on the relation between age and frailty but little agreement on establishing an age threshold to assess frailty. Finally, there was substantial disagreement about the timeline for assessing clinical and laboratory biomarkers in the diagnostic process. These areas of agreement and disagreement provide a valuable roadmap for future research on frailty.
Our definition of frailty shows some similarities and differences with previous definitions, often establishing a midway position between them. It can be considered more comprehensive and inclusive than Fried's definition because it includes not only physical items but also items related to cognition, mood, as well as social and spiritual factors. However, our definition is different than Rockwood's in that our definition separates frailty from disability as distinct constructs. Moreover, Rockwood's definition embraces other domains not considered by us, such as, self-rated health and the use of health-care services. Gobbens et al. published a definition of frailty that is less precise regarding specific aspects of frailty (i.e. the domains/biomarkers to be included, the description of the risks associated with frailty, etc). However, the Gobbens et al. definition identifies sub-concepts of 'physical frailty', 'psychological frailty' and 'social frailty' as important constituent elements of overall frailty. Finally, our definition shares significant elements with the definition proposed by Bergman et al. which attempts to integrate a life-course approach including psychosocial and societal factors as modulating elements. But it also shows some points of disagreement, the most important being the greater emphasis on biological factors in Bergman et al.
Our project has several strengths and limitations. Although many scholarly review articles on the topic of frailty are currently available, few of them have employed a systematic approach to establishing a consensus definition of frailty. To reach our consensus we used, for the first time in frailty research, a version of the Delphi protocol. The multidisciplinary research team, the distribution of background readings and the active participation in face-to-face meetings are all very important strengths of this approach. In our work, the contribution of each group of experts in the answers was well balanced. The main limitation of our study is that our results, although representing a step in the right direction on the road to achieving a clinical definition of frailty, still do not allow us to cross from a theoretical framework of frailty to an operational assessment instrument with utility in clinical practice. In this regard, our conclusions highlight the necessity of produce primary data from original research in the field, putting the focus on the issues related to the clinical utility of frailty, namely sensitivity, specificity, predictive values, prognostic value for clinical outcomes, utility in different clinical settings and different types of patients would be the most important ones among many others.
Potential impact:
Defining frailty has become for many investigators and clinicians in the field of aging the quest for the 'holy grail'. The concept of frailty has grown in importance because of a need for a better understanding of the health and functional status of older persons and a need to prevent or at least delay the onset of late-life disability and its adverse consequences. There is to date no clear consensus regarding the definition of frailty. The most frequently used definition of frailty is 'an age-associated biological syndrome characterised by a decreased biological reserve and resistance to stress'. Fried et al. have proposed weakness, slowness, low physical activity, unintentional loss of weight and exhaustion as the five criteria for defining the frail phenotype. Fried and her colleagues suggest that these criteria can be used to identify older persons at elevated risk for numerous adverse outcomes including death, disability, impairments in mobility, hospitalisation and institutionalisation.
We proposed a theoretical definition of frailty with six core domains. It is likely that the assessment of nutritional status, physical performance, mental health and cognitive status are key components in the assessment of frailty. Additional experimental work is needed to find out the specific combination of clinical and laboratory biomarkers that can be used for the diagnosis of frailty. Such studies may well enable us to move beyond a theoretical definition of frailty to a robust, consensual operational definition that can be employed in a variety of settings.
Therefore, this project should be a preliminary step before the implementation of population-based studies in elderly persons at high risk of frailty, because of their nutritional, medical, psychosocial or other characteristics.
The realisation of the present project and the scientific dissemination of the expected results will have the following impacts:
1. Societal impacts. The results of our project need to be reaffirmed with observational studies in elderly European population. We need to develop experimental work which will allow us to confirm that the six domains we have accepted should be of use in daily practice, and to find the best way to cross the line between the theoretical concept and the clinical practice. In this way, the research in biomarkers of frailty related with these domains will be extremely important. If we develop biomarkers for the diagnosis, the prognosis and the stratification of frailty based on the domains that our experts had considered as essentials, we will have the instrument to act on frailty and to prevent disability. This will have an impact on the improvement in the quality of life of older persons and their caregivers in the future, because of the expected interventions on the pathways leading to disability. Future European and national policies related to elderly people social integration and economic impact assessment must be based on studies, definitions and biomarkers surged in projects like FOD-CC. Regarding the future contributions to community societal objectives, frailty definition and the elaboration of strategies to face the frailty associated risks are crucial to ensure the quality of elderly people in Europe, a population that will continue to grow during the next 50 years and which rates of disability increased more than the actual rate of mortality, becoming a future big problem in the sustainability of our health care systems.
2. Economic impacts. As frailty is a previous step before disability, if in the future our definition could be realised in a practical one, it will suppose the possibility to reverse frailty and to prevent disability. We will have an instrument which will provide the domains, in which we have to act to prevent, treat and reverse frailty, and at the same time the way to measure the effectiveness of our interventions. All the suggested measures considered together raise the hope of a prevention of frailty, a decline in the prevalence of frailty related diseases and a drop in the cost of related disability and health care.
Dissemination plan
The dissemination success of our project has been reached because of the plurality of our experts. As we have worked with a large number of experts in very different fields of frailty, this diversity have ensured a ramify effect, spreading the conclusions achieved in the different working meetings. Moreover, as we have contacted experts not only for the focus groups, but also as external experts contributing with their opinions in the Delphi questionnaires, this spread effect has been increased.
The main dissemination activities have consisted in the publishing of our results in international journals with high scientific impact and also in international renowned congresses.
The principal results of our project have been published in Journals of Gerontology, Medical Sciences, with the article 'Searching for an operational definition of frailty: A Delphi method based consensus statement. The frailty operative definition-consensus conference project', that was accepted for publishing on 19 March 2012.
The members of the steering committee and the writing committee have planned another meeting in June 2012, to decide the topics of next publications delivering the data of our project.
The website of the project has been a useful instrument to announce the working meeting, to publish the results achieved in these workshops and to create a discussion forum to enrich the contributions coming from different points of view. This web page was open view for general public.
Moreover, the email communication between experts was very intensive during the whole project, not only regarding the sending and receiving of the questionnaires of the Delphi process, but also including all the news about the progress in the project and the decisions of the steering committee.
In the final meeting of our project the members of the steering committee have decided some actions to improve the dissemination of our conclusions and to continue to work on them. Firstly, it was decided to publicise our results not only in medical specialised magazines, but also general practitioners publications and others general public opinion magazines. It was decided to present our project in several international meeting, congress and workshops:
1. 2011 International Conference on Frailty Research, Taipei, Taiwan, 25 to 27 November 2011. Prof. Finbarr Martin, member of one of the FGs showed the results of the FOD-CC consensus. It was the first time that our results were launched to the scientific community interested in frailty.
2. International Association of gerontology and Geriatrics (IAGG), WHO and Société Française de Gériatrie et Gérontologie (SFGG) third workshop on frailty, Athens, Greece, 20 and 21 January 2012. Prof L. Rodríguez-Mañas presented a summary of the findings of FOD-CC, its clinical implications and its consequences for future research. In this same workshop, Dr Rodriguez-Mañas did a presentation to the executive committee of IAGG and a proposal to release a joint position paper on frailty by WHO-IAGG based upon the FOD-CC conclusions was approved. By the time of the completion of this final report Dr S. Chatterji, Prof. H. Bergman and Dr Rodríguez-Mañas were working in the final refinement of the document. Although the date for launching was not definitive, it would be probably presented in mid-July in Geneva.
3. fifth national meeting of the Spanish Society of Geriatric Medicine, Toledo, Spain, 12 and 13 April 2012. The whole meeting was dedicated to frailty and several round tables and workshops were dedicated to different aspects of frailty with the participation of several members of the FOD-CC FGs. In addition, the first round table named 'Frailty: Concept, domains and measurement' was dedicated exclusively to expose the rationale and the results of the project.
4. joint Federation of European Physiological Societies (FEPS) and Spanish Physiological Society Congress, Santiago de Compostela, Spain, 8 to 11 September 2012. Prof. Giovanni Mann and Prof. Jose Vina, members of the FOD-CC steering group, hosted a European research symposium on the 'Molecular biology of ageing and frailty' at which Prof. Vina delivered a keynote lecture entitled 'Molecular mechanisms underlying frailty in human and animal models'. In addition, other renowned speakers gave keynote lectures on the molecular biology of ageing. Further information was available from 'http://www.seccff.org/ampliacongre.php?cod=VkZod1JsQlJQVDA9'.
5. Oxygen Club of California World Congress, Alba, Italy, 20 to 23 June, 2012. At this world congress, Prof. Jose Vina and Prof. Leocadio Rodriguez-Mañas co-organised a FOD-CC related symposium entitled 'Epigenetics, metabolism and ageing', with specific keynote lectures focussed on 'Modulation of aging genes: Importance in longevity and age-associated frailty' (Prof. Jose Vina) and 'Frailty: A basic and clinical challenge for the future' (Prof. Rodriguez-Mañas). This international conference provided an opportunity of disseminating our FOD-CC findings to colleagues from across the world and of integrating our findings with those in other related disciplines of relevance to frailty. Further information was available from 'http://www.oxyclubcalifornia.org/Alba2011/pdf/Scientific%20Program_Alba.pdf'.
Finally, the project leader was involved in building a new scientific journal focussed thematically on frailty, namely the 'Journal of Frailty'.
List of websites:
'http://www.fod-cc.com'.
Frailty is a relatively new concept trying to explain the predisposition of old people to disability and other adverse outcomes such as death, hospitalisation, institutionalisation and falls.
This concept has been approached several times from different points of view, including a large number of different 'constructs' depending of each expert opinion.
Until now we have many definitions of frailty, but none of them is validated in longitudinal studies or without the bias of a single point of view.
With the aim of providing an inclusive definition of frailty, including the opinions of all the experts interested in it, but also with a useful clinical profile, we have developed the Frailty Operative Definition-Consensus Conference (FOD-CC) project.
The design of this project has taken into account several important points:
1. the need for selecting the core domains to be included in the final definition
2. the need to get these domains from the different fields of interest concerned by frailty, selecting experts in every field of interest of frailty
3. the need to build a definition to be used in daily practice in clinical settings.
To reach our practical consensual definition, we have use a methodology based on the Delphi method. We used three consecutive questionnaires, together with face-to-face meetings between experts.
These experts have been carefully selected of five main groups (focus groups) of professionals interested in frailty, namely geriatricians, non-geriatricians clinicians, non-medical health workers, basic scientists and social and non-governmental organisations (NGOs), each one composed by six to seven members.
Also, external experts have been contacted to participate in filling the questionnaires, until reaching the figure of 83 participants from 111 experts contacted (response rate, 75%), with a good balance between the groups. None group was overrepresented or underrepresented in the final group of participants.
Finally, an advisory board, composed by several renowned scientists in frailty that remained blinded to previous process to avoid contamination, has agreed to give support to our project, sending their comments to refine the final conclusions of the project.
Our proposed definition of frailty is composed by six core domains, but any set of biomarkers reach the consensus to be used as the clinical definition of frailty. In addition, although there was consensus about the potential utility of laboratory biomarkers, anyone has reached enough consensuses to be accepted as a diagnostic tool. There is a strong agreement on the importance and relevance of frailty, its relevance as a clinical syndrome and the necessity to continue making primary research to bridge the lacks of knowledge concerning many issues related to its clinical utility.
Project Context and Objectives:
The concept of frailty tries to increase our understanding of why many older persons develop disability late in life. The concept of frailty has emerged from a wide variety of studies and theoretical discussions and there is, as yet, no clear consensus regarding its definition. The most used accepted approach defines frailty as 'an age-associated biological syndrome' characterised by a decrease of the biological reserve and resistance to stress, due to a decline in several physiological systems, putting the individual at a special risk when facing little stressors and associated with poor outcomes such as disability, death and institutionalisation or hospitalisation' (Campbell and Buchner, 1997; Walston and Fried, 1999; Rockwood et al, 2000). The model proposed by Walston and Fried (1999) places sarcopenia and energy misbalance as the fundamental elements of frailty. Also, Fried et al. identified five criteria to define the frail phenotype, namely weakness, slowness, low physical activity, unintentional loss of weight and exhaustion, in the cardiovascular health study cohort. These researchers suggested that individuals with three or more of these criteria should be characterised as frail. One of the helpful characteristics of this definition is the capacity to determine risk or adverse outcomes including death, disability, impairment in mobility, hospitalisation and institutionalisation (Fried L., 2001; Bandeen-Roche K 2006) and it can also assist in differentiating between frail, pre-frail (individuals who presented only one or two criteria among the five) and non-frail individuals.
Most researchers agree that frailty is a 'syndrome' and a number of definitions have been proposed each of which has its own strengths and weaknesses (de Vries, 2010). With increasing popularity, frailty definitions proliferated; each definition catching a new relevant aspect, proposing novel dimensions or improving its accuracy for the specific context and aims it was created for (epidemiological purposes).
Despite the efforts of several research groups, documenting the predictive validity of various frailty indices remains a challenge, with few prospective population or clinical studies demonstrating the ability to reliably predict the onset of adverse outcomes in frail individuals. Indeed, the Fried criteria are mainly focussed on the physical functioning domain of health. Some authors have claimed for the necessity of including other items (like mental health and accumulated vulnerability) within a frailty definition (Bergman H et al, 2007). In addition, given that lifestyle factors are likely to influence the development and rate of progression of frailty, there is a need for definitions to be validated in a wide variety of cultural, economic, ethnic and clinical settings. The majority of studies have examined frailty in Anglo-American populations and there is a clear need to extend this work to other population groups with different lifestyles and genetic backgrounds. In this regard, two recent papers show in a European population that the risk of adverse outcomes using the Fried criteria of frailty was lower than expected for some outcomes (Avila-Funes JA et al., 2008) and that, cognitive decline, a notion not included in these criteria, was an important element of frailty that increased the predictive validity of the Fried index model in determining risk for adverse outcomes (Avila-Funes JA et al., 2009).
Growing evidence suggests that nutritional status may play a relevant role in the process leading to frailty through a variety of multidimensional effects. In addition, another phenotype strongly associated with frailty and sharing many of its pathophysiologic components has recently emerged, namely sarcopenic obesity. This new concept supports the addition of measures of body mass index (BMI) and waist to hip ratio to the criteria of frailty. In 40 600 women aged 65 to 79 years enrolled in the Woman's Health Initiative Observational Study (Woods NF, 2005), the risk of incident frailty was raised when BMI was over 25 kg/m2 and was decreased when BMI was under 18.5 kg/m2. Consistent with this finding, there is evidence that a state of chronic inflammation in obese people, measured by an increased C-reactive protein (CRP), may be an important marker of the frailty syndrome (Fruhbeck G, 2004). Some authors suggest that the origin of such inflammation could be in the adipose tissue (Guzik TJ, 2006) as it releases many inflammatory substances (leptin, adiponectin, resistin, visfatin, apelin, etc). In addition, it has been found that adipocytes secrete pro-inflammatory cytokines such as IL-6 (Fontana L, 2007). Finally, a growing literature provides evidence of the role of many pathophysiologic components in the syndrome of frailty. Among the components proposed a decrease in sexual hormones (Iannuzzi-Sucich M et al, 2002), dysfunction of the growth hormon-insuline-like growing factor-1 (GH-IGF-1) axis (Waters DL et al, 2003), increase in the ratio of cortisol to Dehidroepiandrosterona (DHEAs) (Walston and Fried 1999, Taylor JM, 2003), testosterone deficit (Kaufman JM and Vermeulen A, 2005; Travison et al., 2007), increases in interleukin-6 (IL-6), IL-1 and tumour necrosis factor (TNF) alpha circulating levels (Penninx BW et al., 2004), CRP and d-dimer (Walston J et al., 2002) and other pro-inflammatory cytokines (Leng S et al.,2002) are included. In addition, data from the cardiovascular health study showed that subclinical cardiovascular disease was related to frailty, and should be considered a risk factor for disability.
The main problem with the majority of the current frailty definitions is their lack of validation in well controlled prospective longitudinal studies, the lack of support for any single approach with regard to the definition of this complex syndrome and the shortage of studies testing the predictive capacity of the current definition in a clinical population (Robinson TN et al, 2009; Sündermann S et al, 2011). There is increasing evidence that multiple biological, clinical and social factors are involved in the development of frailty (Bergman et al. 2007). All these factors interact to produce conditions that place an individual at a high risk for developing disability and a decreased quality of life. The multiplicity of factors leading to frailty has made it difficult to reach a unique, comprehensive and operative definition of the term. Indeed existing definitions tend to be skewed in favour of social, biological or clinical elements depending on the background and expertise of researchers building the definition. On the other hand, frailty markers, individually or in combination may be as important as a model. It is also important to develop biological markers of frailty which are able to support the clinical diagnostic markers in predicting the onset of frailty and its progression to disability and other adverse outcomes.
To conclude, frailty has attracted the attention of the researchers from numerous fields of study. However, despite this interest, the need to develop an operative definition of frailty that can then be translated into a clinical instrument remains paramount if we are to improve the health and quality of life of the older population. In addition, this definition should also be feasible to apply in general population settings, in order to enable tracking over time and identification of populations at greatest risk such that public health and policy interventions can be appropriately targeted and monitored.
Thus, a working group involving experts of each area related to frailty gathered together with the aim of providing the most complete and concrete definition as possible, selected biomarkers of clinical utility and a guideline to allow early diagnosis of frailty to be used in clinical practice. In order to achieve the main objectives, the secondary aims of the project were:
1. to evaluate the rate of consensus between experts regarding the framework of frailty
2. to evaluate the rate of consensus between experts regarding the domains of frailty and their clinical value
3. to evaluate the rate of consensus between experts regarding biomarkers of frailty and their diagnostic value
4. to evaluate the utility of animal models
5. to provide research fields of interest.
The position that was developed can be summarised in the following points:
1. frailty is a relevant risk factor for disability
2. different definitions including different set of domains are available
3. frailty has been evaluated in epidemiological settings, being a good marker of adverse outcomes in older people
4. the utility of laboratory biomarkers has been scarcely explored.
The project targets are as follows:
1. to establish the basis for a clinical definition
2. to determine domains to include in a clinical definition
3. to determine the utility of several clinical and laboratory biomarkers
4. to determine the gaps to be investigated in the future.
Project Results:
Methodology and scientific results
The aim of the FOD-CC project was to provide an operative definition of 'frailty' which would be useful in daily practice. This definition was constructed from a wide variety of different expert opinions in order to result in the most complete and concrete definition possible.
The Delphi procedure was used to reach this objective. The Delphi method is a structured communication technique, originally developed as a systematic, interactive forecasting method which relies on a panel of experts and which is a well established way to reach consensus between experts from different fields (Linstone HA and Turoff M, 1975).
The Delphi method is based on the principle that forecasts (or decisions) from a structured group of individuals are more accurate than those from unstructured groups. The traditional steps in the Delphi process are formation of a Delphi team, selection of one or more panels to participate (experts in the area), development of a first round Delphi questionnaire, testing the questionnaire for proper wording, transmission of the questionnaire to the panelists, analysis of the first round responses and repeating of the same procedure until a predefined stop criterion is reached (normally the achievement of a consensus) and, finally, the preparation of a report by the analysis team to present the conclusions.
In order to improve this procedure, we introduced some modifications to the classical Delphi method, mostly with the aim of increasing the amount of information available to the experts. Most importantly, we involved all experts in the formulation of the first round questionnaire by asking them four open-ended questions related to frailty and by encouraging communication among them during an initial face-to-face meeting. These modifications were carried out according to the following strategies:
1. after having accepted an offer to participate in the project, all experts were sent a sample of published scientific papers about frailty. This selected bibliography was different for each group depending on their area of specialisation. This was done in order to help them with their answers to the questionnaires.
2. the experts worked together on the open-ended questionnaire about frailty prior to the development of the Delphi questionnaire itself.
3. in addition to extensive e-mail communications between participants, two face-to-face meetings between experts were conducted in order to enable all aspects of the Delphi process to be thoroughly discussed.
4. a website for the project was developed to allow project participants and others to learn more about, and comment on, the goals, methodology and implementation of the project.
The Delphi process consisted of 13 steps.
1. selection of the panel of experts. Five groups of experts were selected with the goal of covering the fields of study that have an interest in the topic of frailty and to provide a multi-disciplinary perspective from all stakeholders. This wide range and variability of fields was very important to our project because this would help to finally ensure that the consensus definition of frailty reached by the project would be valid and applicable to the widest possible number of fields. The focus group were composed, respectively by geriatricians, forming the focus group one (FG1), non-geriatric physicians (FG2), health workers (FG3), basic scientists (FG4) and social and non-governmental workers (FG5). Each FG was composed of about five to seven experts and a chairman. This selection of experts was driven by a steering committee composed by Professor Catherine Feart, Professor José Viña, Professor Giovanni Mann, Dr Somnath Chatterji and Professor Leocadio Rodríguez. The experts were selected not only because of their personal interest in frailty, but also because of their previous work in the field, evaluated by their previous research experience on frailty, publications in scientific journals or daily working experience in managing patients with frailty and related conditions. In addition, a public health perspective also needed to be represented among the experts. These experts were contacted and invited to participate in the project in person or by e-mail. Once they had agreed to participate, all experts received an outline of the project that described the main objectives and tasks.
2. sending of bibliography. To facilitate the participants' immersion in the project, each expert received a compilation of the most relevant references on frailty and was sent a selection of articles related to their own area of interest. Those articles were selected by the leader team, from data bases of medical science publications (Pubmed and Embase), including the keyword 'frailty' and other terms like 'disability', 'diagnosis', 'death', 'hospitalisation' and 'biomarkers'. In addition we made a specific search in the clinical journals included under the section 'geriatrics and gerontology' of the Journal of Citation Report Index. We had an initial selection of 848 articles of which we read the title and make a second selection of the articles with potential interest for our study. This selection was carried out by agreement of two members of the team. After that, those two members read the abstracts and selected the final group of 113 articles. All the members of the FGs received a selection of 27 common articles and around 20 articles specific for each focus group.
3. open-ended preliminary questionnaire. All FG members participated in the development of the first round questionnaire (1RQ). They received an open-ended preliminary questionnaire that included four general and two specific questions about frailty. Their answers were classified according to their similarity. Results were presented during the first face-to-face meeting of the FG in Valencia, Spain.
4. first face-to-face meeting in Valencia, Spain (February 2011). In Valencia, discussions among FG members led to the development of a set of statements for the 1RQ. Participants were then asked to suggest five external experts to whom the 1RQ could be sent. In addition, they were asked to recommend one institution to be invited to support the study and one person for appointment to the advisory board. The advisory board nominated by the steering committee included 11 experts who did not participate in the Delphi process.
5 design of the 1RQ and preliminary analysis (combining steps five and six). After a further review from all FG members, the final version was disseminated. The 1RQ consisted of 107 statements broken down into five blocks, namely framework and definition (52 statements), biomarkers (43 statements), frailty versus disability (7 statements), frailty versus co-morbidity (4 statements) and animal models (1 statement). All experts were asked to score each statement on a 10 point numerical scale ranging from 1 (no agreement) to 10 (full agreement). All the answers were received and analysed with a statistical method whose results would be discussed during the second meeting.
7. second face-to-face meeting in Bordeaux, France (May 2011). During the second face-to-face meeting, participants established criteria to define the strength of agreement and analysed the data collected. The preliminary 1RQ results were discussed. To improve the response rate to the questionnaire, corresponding to 81 of 130, i.e. 62 %, two strategies were adopted, namely to re-contact non-responding experts and to select additional experts for those FG that had the lowest rate of response in the initial analysis (i.e. non-geriatrician clinicians and basic scientists).
8. definitive analysis of the 1RQ. Using the statistical criteria approved by the experts the final rate was 74.5 %. The statistical analyses were performed in accordance with previously established procedures in reviewed literature. Criteria used to classify degree of agreement were positive and negative consensus was defined as percentage of answers rated, respectively, between plus or minus 8 or plus or minus 3 in 80% and more (strong agreement), percent answers between 70 to 80 % (moderate agreement) were selected to enter the 2RQ, statements with low agreement (about 50 to 70 %) or no agreement (less than 50%) were further analysed to determine if this low percentage of consensus was due to heterogeneity in the opinion of experts, assessed using tests for median comparison (Wilcoxon scores rank test) or to dispersion, reflecting a confusing statement, that was assessed by the presence of an inter quartile range (IQR) of 4. All items with heterogeneity or dispersion were reviewed and included in the 2RQ according to the criteria that in light of the experts' opinion a consensus could possibly be reached or appropriate modifications could improve the clarity of the statement.
9. leadership meeting (Oxford, July 2011). During this meeting attended by the chair men of each focus group, and the aim was to decide the members for both the advisory board and the writing committee
10. development of the second round questionnaire (2RQ). The final version of the 2RQ consisted of 52 statements and was sent to all the experts who evaluated the 1RQ. Consistent with the Delphi process, in order to facilitate agreement among experts, respondents were asked to try to either accept (score 8, 9, 10) or reject (score 1, 2, 3) the proposed statements. Median scores and percentage of agreement from the 1RQ were displayed for reference.
11. statistical analysis of the 2RQ. According with the previously accepted methodology, the LTG analysed the results, but this time they only evaluated the accepted or rejected statements.
12. final face-to-face meeting in Madrid, Spain (October 2011). The final results of the Delphi process were presented in Madrid. In addition to the initial classification of the statements into five categories, a second classification in four categories was approved (i.e. alternative classification), consisting in concept, prognosis, diagnosis and prevention and treatment of frailty. The statements finally accepted to enter into the definition of the FOD-CC experts were presented and submitted for approval according to both classifications.
13. final documents. A final document was presented in the Madrid meeting to all of participants, including the methodology and the results of the project.
Results from the open preliminary questionnaire
The open preliminary questionnaire was composed of four general and two specific questions for each focus group. This questionnaire was sent to 30 members of the focus groups and we received 27 answers, a response rate of 90 %.
2. Results from the 1RQ
The 1RQ (107 questions broken into five themes, namely framework, biomarkers, frailty versus disability, frailty versus co-morbidity and animal models), was sent to 130 experts before the meeting of Bordeaux and we received 81 answers giving a response rate of 62 %. According to the criteria described in the statistical methods, 17 statements were accepted for the final definition, 45 statements were excluded, 30 questions had to be reviewed again because the answers were dispersed or not homogeneous among groups of experts and 15 questions had to be included in a 2RQ because the percentage of persons answering above eight or under three was between 70 and 80 %.
After the Bordeaux meeting, we improved the response rate by once again requesting the participation of prior non-responding experts and by recruiting additional experts in the groups with the lowest response rates. In total, we achieved 111 answers, from 149 questionnaires sent, with a final response rate of 74.5 %. In the light of these new answers, the results were slightly modified. Fifteen statements were accepted (14.0 %), 47 excluded statements (43.9 %), 28 questions to be re-reviewed (26.2 %) and 17 questions were to be included in the 2RQ (15.8 %). Depending upon the theme in which a statement was included, there was an agreement in 57.1 % of the statements regarding the difference between frailty and disability, nine statements (17.3 %) regarding the framework of frailty were accepted and an additional 50 % were either included in the 2RQ or entered into the review process. On the other hand, more than 65 % of the statements regarding biomarkers reached a low consensus and were excluded after the first round, one statement was included in the second round and 12 others (27.9 %) entered the review process.
It is important to underline that the rate of answers according to each focus group was very homogeneous except in the basic scientists group that was under-represented, i.e. 21.6 % of the answers came from geriatricians, 20.7 % from non geriatricians and physicians, 23.4 % from health worker professionals, 11.7 % from basic scientists and 22.5 % from social and nongovernmental workers.
Results from the 2RQ
The 2RQ was sent to the same 111 responders of the 1RQ. We received 83 answers, a response rate of 75 %. At this time the distribution of answers according to the group of experts was 27.7 % of geriatricians, 12.0 % from non geriatrician's physicians, 26.5 % from health workers, 12.0 % from basic scientists and 21.7 % from social and nongovernmental workers.
Twenty-four statements were accepted from this 2RQ; almost 60 % of the statements regarding the framework of frailty, 33 % of the statements regarding biomarkers, two statements (40 %) were about the difference between frailty and disability, one statement was about the difference between frailty and co-morbidity.
Final decisions
In total, 132 statements were proposed to the experts, 107 in the 1RQ and 25 more in the 2RQ. Finally, the total number of accepted statements to be included in our definition of frailty was 39, of which 15 were accepted directly after the 1RQ. There was a clear agreement about the difference between frailty and disability with more than 66.6 % of the statements in this area accepted. In addition, more than 38 % of the statements regarding the framework of frailty reached consensus. Only 15.9 % of the proposed statements about biomarkers were accepted.
As mentioned in the methods, we also used an alternative classification of the statements to analyse the information extracted. In this way, we found that 80 % of the statements regarding prevention and treatment of frailty were accepted. In addition, 44.4 % of the statements regarding the concept of frailty were accepted as well as 33.3 % of the statements regarding prognosis. On the other hand, of the 78 statements about the diagnosis of frailty, only 14 (17.9 %) were accepted.
Conclusions
Members of the FG synthesised the 39 accepted statements into a cohesive narrative text. This draft was presented to all participants during the last meeting in Madrid. The final approved version is as follows (the number of the statement supporting the assertion is provided in brackets along with the text).
Frailty is a dynamic, non linear process (24) that involves alterations in multiple domains of functioning (29), impacts multiple body systems (28, 40) and may be considered a clinical syndrome (4). It is different from vulnerability and disability (26, 39, 96, 97) and is often modulated by disease (102) to the extent that the presence of co-morbidities tends to exacerbate the consequences of frailty (104). Definitions of frailty are frequently multidimensional involving a variety of psychological, social, emotional, and spiritual elements in addition to physical components (13). Definitions of frailty should be reproducible over time (9) and be applicable across different clinical and non clinical settings (7, 43).
Frailty is characterised by a decreased reserve and diminished resistance to stressors (6). No single molecular mechanisms or biomarker has been identified that is adequate for the prediction and/or diagnosis of frailty (32, 60). However, variables including assessing nutritional status (8) and physical performance (14), gait speed (16) and mobility (17) have been proposed for use in the diagnosis of frailty. In addition, mental health assessments and cognitive status evaluations are highly recommended as part of the assessment of frailty (9, 64).
The consequences of frailty include an increased vulnerability to stress in which minimal levels of stress can cause functional impairments (6, 21,27). Although disability can occur without prior frailty, frailty is a risk factor for disability (99). A major purpose of diagnosing frailty is to identify and stratify older persons at high risk for disability and/or other adverse outcomes (12, 45). A diagnosis of frailty is of importance because can help to predict a variety of different health outcomes including disability, falls, hospitalisation, permanent institutionalisation, and death (100, 59). The predictive value of frailty depends on its severity and will vary from person to person (101).
The diagnosis of frailty is useful in geriatric medicine (48), primary and community care (46), and is of value in managing older people with chronic diseases (47). Because frailty is a condition that may be amenable to prevention it is mandatory for clinicians and health workers to identify those at risk for frailty as early as possible (23). Among the interventions that show promise for the management or attenuation of frailty are various healthy lifestyles (7), including physical activity (6, 22).
The aim of the FOD-CC project was to reach a consensus definition of frailty that is useful in daily practice, using a Delphi process. Experts agreed on the importance of a more comprehensive definition of frailty that should include assessment of physical performance, including gait speed and mobility, nutritional status, mental health and cognition. Although a consensus was reached on these six domains, the proposed diagnostic paths and procedures needed to achieve an operational definition reached a low level of consensus; only one out of six of the statements related to diagnosis achieved consensus.
Experts agreed that no single biomarker by itself was adequate for the assessment of frailty, suggesting a need for a combination of multiple biomarkers. However, none of the proposed combinations of biomarkers was able to reach the 80 % threshold of agreement required in the Delphi process. The most surprising result regarding clinical biomarkers was the absence of agreement regarding the inclusion of grip strength in the definition. Interestingly, the percentages of agreement for statements about grip strength were very close to achieving consensus (more than 70 % of agreement). We decided not to include the grip strength items in the definition in order to remain faithful to the a-priori criteria established for the acceptance of consensus. It is possible that our inability to reach consensus reflects concerns about the reliability of grip strength as a measure rather than a reluctance to include strength, per se, as one of the domains of frailty. Importantly, among all the laboratory markers suggested for the assessment of frailty, none was accepted.
The low level of consensus regarding the constituent elements to be included in an operational definition of frailty is balanced by the high degree of agreement regarding the underlying conceptual framework of frailty. The experts clearly agreed that frailty is a multidimensional syndrome characterised by decreased reserve and diminished resistance to stressors. The experts established a clear-cut difference between disability and frailty as shown by the percentage of agreement (85 to 95 %) in the related statements. This notion is somewhat different from the World Health Organisation (WHO) conceptualisation of disability, in which disability is considered to be contextual and every human being can experience some degree of disability. Frailty was also differentiated from vulnerability. While everybody is potentially vulnerable, frailty represents a state of extreme vulnerability where minimal stress may cause functional impairment. This conceptual framework has clinical implications. To be very vulnerable with a low reserve capacity provides a high risk of becoming disabled and a low probability for the restoration of function, underscoring the necessity to detect frail people before they become dependent.
Although frailty has a clear conceptual framework, there is no single operational definition of frailty that can satisfy all experts. A possible explanation may rely on the heterogeneity of our expert focus groups. However, this explanation seems unlikely. Heterogeneity between groups in the Delphi questionnaire responses was systematically tested and was only detected for 5 of the non-accepted statements. Another possible explanation may pertain to the lack of agreement in the literature among existing definitions of frailty. Supporting this perspective we found agreement regarding the necessity to combine biomarkers but no agreement regarding which combination of biomarkers to include in the definition; agreement regarding the necessity to assess the severity of frailty but no agreement on identifying specific severity markers; agreement on the relation between age and frailty but little agreement on establishing an age threshold to assess frailty. Finally, there was substantial disagreement about the timeline for assessing clinical and laboratory biomarkers in the diagnostic process. These areas of agreement and disagreement provide a valuable roadmap for future research on frailty.
Our definition of frailty shows some similarities and differences with previous definitions, often establishing a midway position between them. It can be considered more comprehensive and inclusive than Fried's definition because it includes not only physical items but also items related to cognition, mood, as well as social and spiritual factors. However, our definition is different than Rockwood's in that our definition separates frailty from disability as distinct constructs. Moreover, Rockwood's definition embraces other domains not considered by us, such as, self-rated health and the use of health-care services. Gobbens et al. published a definition of frailty that is less precise regarding specific aspects of frailty (i.e. the domains/biomarkers to be included, the description of the risks associated with frailty, etc). However, the Gobbens et al. definition identifies sub-concepts of 'physical frailty', 'psychological frailty' and 'social frailty' as important constituent elements of overall frailty. Finally, our definition shares significant elements with the definition proposed by Bergman et al. which attempts to integrate a life-course approach including psychosocial and societal factors as modulating elements. But it also shows some points of disagreement, the most important being the greater emphasis on biological factors in Bergman et al.
Our project has several strengths and limitations. Although many scholarly review articles on the topic of frailty are currently available, few of them have employed a systematic approach to establishing a consensus definition of frailty. To reach our consensus we used, for the first time in frailty research, a version of the Delphi protocol. The multidisciplinary research team, the distribution of background readings and the active participation in face-to-face meetings are all very important strengths of this approach. In our work, the contribution of each group of experts in the answers was well balanced. The main limitation of our study is that our results, although representing a step in the right direction on the road to achieving a clinical definition of frailty, still do not allow us to cross from a theoretical framework of frailty to an operational assessment instrument with utility in clinical practice. In this regard, our conclusions highlight the necessity of produce primary data from original research in the field, putting the focus on the issues related to the clinical utility of frailty, namely sensitivity, specificity, predictive values, prognostic value for clinical outcomes, utility in different clinical settings and different types of patients would be the most important ones among many others.
Potential impact:
Defining frailty has become for many investigators and clinicians in the field of aging the quest for the 'holy grail'. The concept of frailty has grown in importance because of a need for a better understanding of the health and functional status of older persons and a need to prevent or at least delay the onset of late-life disability and its adverse consequences. There is to date no clear consensus regarding the definition of frailty. The most frequently used definition of frailty is 'an age-associated biological syndrome characterised by a decreased biological reserve and resistance to stress'. Fried et al. have proposed weakness, slowness, low physical activity, unintentional loss of weight and exhaustion as the five criteria for defining the frail phenotype. Fried and her colleagues suggest that these criteria can be used to identify older persons at elevated risk for numerous adverse outcomes including death, disability, impairments in mobility, hospitalisation and institutionalisation.
We proposed a theoretical definition of frailty with six core domains. It is likely that the assessment of nutritional status, physical performance, mental health and cognitive status are key components in the assessment of frailty. Additional experimental work is needed to find out the specific combination of clinical and laboratory biomarkers that can be used for the diagnosis of frailty. Such studies may well enable us to move beyond a theoretical definition of frailty to a robust, consensual operational definition that can be employed in a variety of settings.
Therefore, this project should be a preliminary step before the implementation of population-based studies in elderly persons at high risk of frailty, because of their nutritional, medical, psychosocial or other characteristics.
The realisation of the present project and the scientific dissemination of the expected results will have the following impacts:
1. Societal impacts. The results of our project need to be reaffirmed with observational studies in elderly European population. We need to develop experimental work which will allow us to confirm that the six domains we have accepted should be of use in daily practice, and to find the best way to cross the line between the theoretical concept and the clinical practice. In this way, the research in biomarkers of frailty related with these domains will be extremely important. If we develop biomarkers for the diagnosis, the prognosis and the stratification of frailty based on the domains that our experts had considered as essentials, we will have the instrument to act on frailty and to prevent disability. This will have an impact on the improvement in the quality of life of older persons and their caregivers in the future, because of the expected interventions on the pathways leading to disability. Future European and national policies related to elderly people social integration and economic impact assessment must be based on studies, definitions and biomarkers surged in projects like FOD-CC. Regarding the future contributions to community societal objectives, frailty definition and the elaboration of strategies to face the frailty associated risks are crucial to ensure the quality of elderly people in Europe, a population that will continue to grow during the next 50 years and which rates of disability increased more than the actual rate of mortality, becoming a future big problem in the sustainability of our health care systems.
2. Economic impacts. As frailty is a previous step before disability, if in the future our definition could be realised in a practical one, it will suppose the possibility to reverse frailty and to prevent disability. We will have an instrument which will provide the domains, in which we have to act to prevent, treat and reverse frailty, and at the same time the way to measure the effectiveness of our interventions. All the suggested measures considered together raise the hope of a prevention of frailty, a decline in the prevalence of frailty related diseases and a drop in the cost of related disability and health care.
Dissemination plan
The dissemination success of our project has been reached because of the plurality of our experts. As we have worked with a large number of experts in very different fields of frailty, this diversity have ensured a ramify effect, spreading the conclusions achieved in the different working meetings. Moreover, as we have contacted experts not only for the focus groups, but also as external experts contributing with their opinions in the Delphi questionnaires, this spread effect has been increased.
The main dissemination activities have consisted in the publishing of our results in international journals with high scientific impact and also in international renowned congresses.
The principal results of our project have been published in Journals of Gerontology, Medical Sciences, with the article 'Searching for an operational definition of frailty: A Delphi method based consensus statement. The frailty operative definition-consensus conference project', that was accepted for publishing on 19 March 2012.
The members of the steering committee and the writing committee have planned another meeting in June 2012, to decide the topics of next publications delivering the data of our project.
The website of the project has been a useful instrument to announce the working meeting, to publish the results achieved in these workshops and to create a discussion forum to enrich the contributions coming from different points of view. This web page was open view for general public.
Moreover, the email communication between experts was very intensive during the whole project, not only regarding the sending and receiving of the questionnaires of the Delphi process, but also including all the news about the progress in the project and the decisions of the steering committee.
In the final meeting of our project the members of the steering committee have decided some actions to improve the dissemination of our conclusions and to continue to work on them. Firstly, it was decided to publicise our results not only in medical specialised magazines, but also general practitioners publications and others general public opinion magazines. It was decided to present our project in several international meeting, congress and workshops:
1. 2011 International Conference on Frailty Research, Taipei, Taiwan, 25 to 27 November 2011. Prof. Finbarr Martin, member of one of the FGs showed the results of the FOD-CC consensus. It was the first time that our results were launched to the scientific community interested in frailty.
2. International Association of gerontology and Geriatrics (IAGG), WHO and Société Française de Gériatrie et Gérontologie (SFGG) third workshop on frailty, Athens, Greece, 20 and 21 January 2012. Prof L. Rodríguez-Mañas presented a summary of the findings of FOD-CC, its clinical implications and its consequences for future research. In this same workshop, Dr Rodriguez-Mañas did a presentation to the executive committee of IAGG and a proposal to release a joint position paper on frailty by WHO-IAGG based upon the FOD-CC conclusions was approved. By the time of the completion of this final report Dr S. Chatterji, Prof. H. Bergman and Dr Rodríguez-Mañas were working in the final refinement of the document. Although the date for launching was not definitive, it would be probably presented in mid-July in Geneva.
3. fifth national meeting of the Spanish Society of Geriatric Medicine, Toledo, Spain, 12 and 13 April 2012. The whole meeting was dedicated to frailty and several round tables and workshops were dedicated to different aspects of frailty with the participation of several members of the FOD-CC FGs. In addition, the first round table named 'Frailty: Concept, domains and measurement' was dedicated exclusively to expose the rationale and the results of the project.
4. joint Federation of European Physiological Societies (FEPS) and Spanish Physiological Society Congress, Santiago de Compostela, Spain, 8 to 11 September 2012. Prof. Giovanni Mann and Prof. Jose Vina, members of the FOD-CC steering group, hosted a European research symposium on the 'Molecular biology of ageing and frailty' at which Prof. Vina delivered a keynote lecture entitled 'Molecular mechanisms underlying frailty in human and animal models'. In addition, other renowned speakers gave keynote lectures on the molecular biology of ageing. Further information was available from 'http://www.seccff.org/ampliacongre.php?cod=VkZod1JsQlJQVDA9'.
5. Oxygen Club of California World Congress, Alba, Italy, 20 to 23 June, 2012. At this world congress, Prof. Jose Vina and Prof. Leocadio Rodriguez-Mañas co-organised a FOD-CC related symposium entitled 'Epigenetics, metabolism and ageing', with specific keynote lectures focussed on 'Modulation of aging genes: Importance in longevity and age-associated frailty' (Prof. Jose Vina) and 'Frailty: A basic and clinical challenge for the future' (Prof. Rodriguez-Mañas). This international conference provided an opportunity of disseminating our FOD-CC findings to colleagues from across the world and of integrating our findings with those in other related disciplines of relevance to frailty. Further information was available from 'http://www.oxyclubcalifornia.org/Alba2011/pdf/Scientific%20Program_Alba.pdf'.
Finally, the project leader was involved in building a new scientific journal focussed thematically on frailty, namely the 'Journal of Frailty'.
List of websites:
'http://www.fod-cc.com'.
