Assessment of frailty syndrome using Edmonton frailty scale in Polish elderly sample

Abstract

The aim of the study was to assess the incidence and severity of the frailty syndrome assessed with the Edmonton Frailty Scale. This is a cross-sectional study conducted among 382 patients (236 men and 146 women, mean age 71.9 years). The Edmonton Frailty Scale was administered during the patient’s admission to the hospital. The Polish adaptation was performed using the standard methodology. The Cronbach’s alpha coefficient for the whole Edmonton Frailty Scale was 0.709. The mean correlation between positions and the overall result was r = 0.180. There were no statistically significant differences between women and men in the area of Edmonton Frailty Scale mean score (p < 0.05). The socio-clinical analysis, showed statistically significant differences in the age of respondents, educational attainment, occupational activity, number of drugs taken and co-occurrence of chronic diseases. A higher values of the Edmonton Frailty Scale were indicated for individuals >70 years than for those <70 years (p < 0.001). The Edmonton Frailty Scale proved to be a reliable tool which may be used in the Polish population. The use of this questionnaire for frailty syndrome may be helpful in everyday clinical practice.

Keywords:

• Diagnostic tools
• Edmonton Frailty Scale
• elderly patients
• frailty syndrome
• Polish adaptation

1. Introduction

In spite of advanced therapies among the elderly, the risk of complications related to the treatments is increased, along with the length of hospitalization and the need for institutionalization. Frailty syndrome (FS) is a frequently occurring clinical syndrome in the elderly. It is defined not only as a biological state but also as a multidimensional concept, in which there are many interrelated factors that can disrupt the physiological balance of the elderly [1].

“Frailty syndrome” is associated with the elderly and explained as the syndrome of the reserve deficiency or the syndrome of the fragility. There is no uniform definition of this syndrome, but the most citable is the definition presented by a consensus group consisting of delegates from 6 major international, European, and US societies created four major consensus points on a specific form of frailty: physical frailty [2]. This definition says that FS is: “A medical syndrome with multiple causes and contributors that is characterized by diminished strength, endurance and reduced physiologic function that increases an individual’s vulnerability for developing increased dependency and/or death.” The other consensus says about the prevention of the frailty syndrome and the implementing of the suitable interventions, such as the suitable exercises and the motor rehabilitation, the protein and calorific supplementation, Vitamin D supplementation and the decrease of the polypragmasy [3].

The important role of the FS preventions plays the early selection of the group od so called “pre frail” due to simple screening tools. The fourth important element for the purposes of optimally managing individuals with physical frailty, all persons older than 70 years and all individuals with significant weight loss (≥5%) due to chronic disease should be screened for frailty [2].

Recognizing and explaining the causes of FS is crucial for identifying high-risk groups for its prevention and treatment [4]. Some researchers have identified three stages of the frailty process: the prefrail process, the frailty state and frailty complications. The prefrail stage is clinically silent, where physiological reserves are sufficient to allow the body to respond adequately to a stressor, trauma or acute illness which can be completely cured. The frail stage is characterized by slow and incomplete recovery after stress and injury. Complications of the frailty process are related to the weakening of the physiological capacity, which significantly reduces the body’s ability to deal with stress factors [5]. An increased risk of falls, adverse events, functional impairment, the appearance of polypharmacy, the tendency to suffer infection, risk of hospitalization and the need for hospitalization can all be seen [6,7].

The number of elderly people is constantly increasing, so healthcare professionals should be focused on increasing the ability to recognize, prevent and reduce frailty.

Identification of group of patients at risk of frailty syndrome and conducting a comprehensive geriatric assessment (CGA) will allow to plan-targeted therapeutic intervention [8]. Although there has been considerable progress in recent years in understanding the pathogenesis and essence of the weakness syndrome, while frailty is known to be changeable over time, at this point it is still unclear to what extent the level of frailty can be influenced by intervention [6].

Diagnostic deficiencies in FS represent a lack of uniformly acceptable theoretical and operational definitions and the number of available research tools and questionnaires used on different populations of the elderly. A meta-analysis by Veries et al. [9] conducted in 2011 showed that there are 20 different tools available to measure frailty. However, despite many of these tools having advanced research in frailty, most are impractical for bedside screening by front-line providers because they require the multidimensional clinical data that constitute a comprehensive geriatric assessment. The most frequently used tools include the Fried Frailty Phenotype (FFP), the Clinical Frailty Scale (CFS), the Tilburg Frailty Indicator (TFI) and the Edmonton Frailty Scale (EFS). Despite the fact that the tools for FS identifications are easily accessed, there is still no questionnaires adapted for Polish population.

Large prospective studies describing the importance of FS as a prognostic factor in acute and chronic disease point out that the assessment of FS can be carried out, not only by people with advanced age, but also those who take care of these patients [10]. The Edmonton questionnaire is easy to use and can be used to assess elderly patients who are not geriatric. That is why we decided to adopt this scale in Polish conditions.

The aim of the study was to assess the incidence and severity of the FS assessed with the EFS. A secondary objective was to adapt and evaluate the psychometric properties of the Polish version of the EFS an instrument that identifies levels of frailty in the elderly population. As part of the project, the original version of the questionnaire was translated into Polish and the psychometric properties of the Polish version of the EFS. The consistency of scale scores was calculated using the Cronbach alpha coefficient. In addition, using the analysis of variance (ANOVA) and multivariate regression analysis, predictors have been found which significantly affect the occurrence and severity of FS.

2. Methods

2.1. Study participants

This is a cross-sectional study conducted among 382 patients (236 men and 146 women) with a mean age of 71.9 years (SD = 7.3 years) who were hospitalized in the Department and Clinic of Internal and Occupational Diseases and Hypertension of the Wroclaw Medical University, Poland. The study was conducted from September 2016 to March 2017. The study was carried out by trained nurses at the stage of patients’ admission to the clinic. The classification of the study participants was conducted by a trained physician, a specialist in geriatrics or internal medicine. Sociodemographic and clinical data were obtained from the hospital register. Participation in the study was anonymous and voluntary. Each patient gave written consent to participate in the study and was informed of the possibility of withdrawing from the study at each stage of the study.

2.2. The course of the trial

During the study 470 patients fulfilling the inclusion criteria such as the age of over 65 years were included. At the primary stage, 16 patients were excluded due to comorbidities which may adversely affect the results of the study. At the next stage, each participant had the cognitive functions evaluated by means of the MMSE questionnaire. As the cutoff point, the 23 points were accepted, which may indicate the cognitive impairments with dementia. So, 65 participants were excluded. During the study, seven patients refused to participate without giving the reason. Finally, 882 patients who met the inclusion criteria were included in this study.

2.3. Patients’ qualification

The criteria for inclusion in the study were: age ≥ 65 years, the lack of mental disorders and cognitive impairment with dementia. Exclusion criteria, except for being aged < 65 years, consisted of the coexistence of severe chronic diseases in the state of exacerbation (cancer, respiratory failure or cardiac decompensation) and the score of the Mini-Mental State Examination (MMSE) showing cognitive impairment with dementia – the cutoff point of 23 points was accepted [11]. All subjects were informed of the purpose, methods and manner of conducting the study. All patients signed a written consent to participate in the study. The participation in the study was voluntary and conscious. At every moment, the participants were able to withdraw their agreement for participation.

2.4. Ethical considerations

The research project was approved by the Bioethics Committee of Wroclaw Medical University (no. KB–221/2017). The study was conducted in accordance with the guidelines of the Helsinki Declaration and the principles of Good Clinical Practice, as well as in respect for the rights and dignity of the other person.

2.5. Edmonton frailty scale (EFS)

The 11 items of the EFS questionnaire were elaborated by Rolfson et al. [12] at the University of Alberta, Edmonton Canada in 2006. The scale in the original version evaluates 9 domains: two domains are tested using performance-based items: the Clock Drawing Test for cognitive impairment and the “Timed Up and Go” test for balance and mobility. The other domains are mood, functional independence, medication use, social support, nutrition, health attitudes, continence, the burden of medical illness and quality of life. The maximum score possible is 17 points, which determines a high level of frailty. The authors of the questionnaire take the following score for the evaluation of FS: 0–4 indicates a lack of frailty, 5–6 denotes apparently vulnerable individuals, 7–8 indicates mild frailty, 9–10 moderate frailty and 11 or more denotes severe frailty [12].

2.6. Polish adaptation

The adaptation was performed using the standard methodology [13]. The Polish adaptation is based on the English-language version of the scale [12]. Having received the authors’ approval, the questionnaire was translated into Polish by two independent translators. Then, the translations were evaluated by a panel of experts, which comprised of a geriatrician and general practitioner, two specialist gerontology nurses and a psychologist. The panel verified the phrasing and meaning of all questions, as well as the clarity and correctness of the instructions. The version selected by the panel subsequently underwent back-translation and the results were submitted for approval by the authors of the original English version. Once approved, the preliminary version was used in a pilot study on a group of 20 patients over 65 years old, who were randomly selected complying with strict ethical guidelines. At this stage, the version of the questionnaire was not approved for testing and assessment of psychometric properties.

2.7. Statistical analysis

The analysis of each quantitative variable was conducted by calculating the mean (M), standard deviation (SD), median (Me), lower (Q1) and upper quartile (Q3), and minimum (Min) and maximum (Max) values. The normality of empirical distribution of quantitative variables (age, BMI, etc.) was verified by the Shapiro–Wilk’s W test and Kolmogorov–Smirnov test for normality.

Qualitative variables (nominal and order) are presented in the cross tables in the form of multiplicity (n) and proportion (%). Correlations between sex and socio-demographic and clinical variables were verified by means of a Pearson Chi-square test and, in the case of 2 × 2 table, was performed using the chi-square test with Yates correction or the Fisher’s exact test. The internal consistency of the items of the Polish version of the EFS questionnaire was evaluated by calculating the Cronbach’s alpha coefficient.

Influence of the analyzed socio-clinical parameters on total EFS score was evaluated using one-way analysis of variance (ANOVA). To determine independent predictors of frailty in the one-factor analysis, a stepwise regression with backward elimination was used. All analyses used a significance level; p values of less than 0.05 were interpreted as statistically significant. Statistical analysis was carried out using the STATISTICA software (StatSoft, Dell Inc., USA).

3. Results

3.1. Socio-clinical analysis

The study group consisted of 382 patients, most of whom were female (61.8%), in relationships (68.1%) and rural residents (72%). The mean age of the study group was 71.9 ± 7.3 years (Min-Max: 65–94 years). The level of education of the respondents was most often secondary (39%) and vocational (27.5%). Only 18.6% of the patients were occupationally active. Clinical characteristics showed that 77.2% of respondents at the time of admission were diagnosed with hypertension, 37.7% with diabetes and 27.7% with coronary heart disease. Overall, 61.5% of respondents were chronically treated with 5 or more drugs per day. The detailed characteristics of the study group are presented in Table 1.

Table 1. Characteristics of the study participants.

Variable	Total N = 382	Women N = 236 (61.8%)	Men N = 146 (38.2%)	W vs. M p
Age (years)
M ± SD	71.9 ± 7.3	72.1 ± 7.6	71.5 ± 6.8
Me [Q1; Q3]	70 [65; 77]	70 [65; 77]	70 [65; 76]	0.44
Min–Max	65–94	65–94	65–94
BMI (kg/m^2)
M ± SD	27.4 ± 5.1	27.1 ± 5.4	27.9 ± 4.6
Me [Q1; Q3]	27 [24; 30]	26 [23; 30]	27 [26; 30]	0.04
Min–Max	17–50	17–48	17–50
Habitation, N (%)
City	107 (28.0)	73 (30.9)	34 (23.3)	0.13
Village	275 (72.0)	163 (69.1)	112 (76.7)
Education, N (%)
Basic	71 (18.6)	54 (22.9)	17 (11.6)
Vocational	105 (27.5)	52 (22.0)	53 (36.3)	0.01
Secondary	149 (39.0)	94 (39.8)	55 (37.7)
Higher	57 (14.9)	36 (15.3)	21 (14.4)
Marital status, N (%)
Single	122 (31.9)	87 (36.9)	35 (24.0)	0.01
In relation	260 (68.1)	149 (63.1)	111 (76.0)
Co-morbidities, N (%)
Hypertension	295 (77.2)	188 (79.7)	107 (73.3)	0.19
Coronary disease	106 (27.7)	69 (29.2)	37 (25.3)	0.48
Heart failure	9 (2.4)	6 (2.5)	3 (2.1)	1.00
Diabetes	144 (37.7)	72 (30.5)	72 (49.3)	0.00
Renal failure	68 (17.8)	42 (17.8)	26 (17.8)	0.89
Occupationally active, N (%)
Yes	71 (18.6)	38 (16.1)	33 (22.6)	0.15
Takes over 5 drugs, N (%)
Yes	235 (61.5)	146 (61.9)	89 (61.0)	0.95

Bold values indicate statistical significance (p < 0.05).

3.2. Reliability and consistency of the EFS scale

Scale reliability was measured in a group of 382 people over 65 years of age. The Cronbach’s alpha coefficient for the whole scale was 0.709. The mean correlation between positions and the overall result was r = 0.180. Detailed analysis results of internal consistency of EFS are shown in Table 2.

Table 2. Internal consistency of the Edmonton Frailty Scale.

Frailty domain	Item	Mean if deleted	SD if deleted	Item-Total Correlation	Alpha if deleted
Cognition	1	5.94	3.13	0.314	0.699
General health status (Q1)	2	5.76	3.20	0.300	0.698
(Q2)	3	5.57	3.09	0.413	0.679
Functional independence	4	5.94	2.89	0.641	0.632
Social support	5	6.21	3.26	0.251	0.704
Drugs use (Q1)	6	5.85	3.32	0.249	0.703
(Q2)	7	6.33	3.34	0.214	0.707
Nutrition	8	6.25	3.27	0.337	0.693
Mood	9	6.25	3.23	0.428	0.682
Continence	10	6.30	3.30	0.277	0.700
Functional performance	11	5.49	3.00	0.474	0.668

M = 6.59, SD = 3.46, Cronbach’s alpha α = 0.709, Average inter-item correlation r = 0.180.

3.3. Results of EFS and prevalence of FS

In the analyzed group, the average score on the EFS scale was 6.63 points (±3.3). There were no statistically significant differences between women and men in the area of EFS mean score (6.61 and 6.65 points, respectively). Both women and men achieved an average score, indicating an apparent vulnerability.

The presence of FS was found in 41.1% (157 patients). Using the original EFS cutoff points, the same number of patients (157, 41.4%) was classified into the non-frail group. In addition, in 17.8% (68 patients) the result of the apparently vulnerable was observed (Table 3).

Table 3. Edmonton Frailty Scale mean scores and the prevalence of frailty syndrome in the studied sample (N = 382).

Variable	Total N = 382	Women N = 236	Men N = 146	W vs. M p
EFS (score):
M ± SD	6.63 ± 3.33	6.61 ± 3.33	6.65 ± 3.25
Me [Q1; Q3]	6 [4; 9]	7 [4; 9]	6 [4; 9]	0.907
Min–Max	0–17	0–15	0–17
Frailty, N (%)
Non frail	157 (41.1)	100 (57.6)	57 (39.0)
Vulnerable	68 (17.8)	48 (50.0)	26 (50.0)	0.969
Frailty syndrome present	157 (41.1)	102 (65.0)	55 (35.0)

3.4. Socio-clinical characteristics depending on the level of frailty

In the study group, the level of vulnerable frailty was observed in 68 patients (17.8%), with mild frailty indicated in 74 patients (19.4%), moderate in 55 (14.4%) and severe in 28 patients (7.3%) (Table 4).

The analysis of associations between socio-clinical factors and results of EFS questionnaire, showed statistically significant relationships with respondents’ age, education, occupational activity, number of drugs taken and cooccurrence of chronic diseases. People with severe frailty were the oldest, namely: 79.3 ± 8.0 vs. 75.2 ± 7.5 (moderately frail) vs. 74.4 ± 8.1 (mildly frail) vs. 71.3 ± 6.6 (vulnerable) vs. 68.5 ± 4.7 (nonfrail). Similarly, people with severe frailty were most likely to have had a basic education (35.7%), while nonfrail and vulnerable individuals mostly had secondary (43.9% and 44.1%, respectively) and higher education (18.5% and 26.5%, respectively). Among all respondents, the severely frail individuals were mostly occupationally inactive, whereas 28% of the respondents in the nonfrail group were occupationally active. In the study group, there was a relationship between the intensity of the frailty and the number of drugs taken, whereby the higher severity of frailty, the more people will take 5 or more drugs. In the study group, diabetes and hypertension were most commonly diagnosed in people with severe frailty. Heart failure and ischemic heart disease were associated with mild frailty (Table 4).

3.5. The results of FS from the EFS (total score) in relation to socio-demographic and clinical variables

The results of ANOVA analysis of the impact of socio-demographic and clinical factors analyzed on the total score of EFS are presented in Table 5.

Table 4. Characteristics of patients differing in the level of frailty assessed using the Edmonton Frail Scale.

Level of frailty, N (%)	Non Frail 157 (41.1)	Vulnerable 68 (17.8)	Mild 74 (19.4)	Moderate 55 (14.4)	Severe 28 (7.3)
Age (years)
M ± SD	68.5 ± 4.7	71.3 ± 6.6	74.4 ± 8.1	75.2 ± 7.5	79.3 ± 8.0
Me [Q1; Q3]	66 [65; 70]	69 [65; 78]	74 [67; 79]	75 [69; 81]	82 [74; 87]	0.00
Min–Max	65–87	65–89	65–94	65–94	65–89
BMI (kg/m^2)
M ± SD	27.6 ± 4.5	27.4 ± 5.1	27.9 ± 6.1	26.7 ± 5.6	26.2 ± 4.5
Me [Q1; Q3]	28 [24; 30]	26 [24; 31]	27 [24; 30]	25 [22; 31]	25 [24; 29]	0.22
Min–Max	18–41	17–45	17–50	21–40	18–38
Sex, N (%)
Male	57 (36.3)	34 (50.0)	26 (35.1)	15 (27.3)	14 (50.0)	0.06
Female	100 (63.7)	34 (50.0)	48 (64.9)	40 (72.7)	14 (50.0)
Habitation, N (%)
City	119 (75.8)	50 (73.5)	54 (73.0)	32 (58.2)	20 (71.4)	0.17
Village	38 (24.2)	18 (26.5)	20 (27.0)	23 (41.8)	8 (28.6)
Education, N (%)
Basic	16 (10.2)	9 (13.2)	21 (28.4)	15 (27.3)	10 (35.7)
Vocational	43 (27.4)	11 (16.2)	21 (28.4)	20 (36.4)	10 (35.7)	0.00
Secondary	69 (43.9)	30 (44.1)	28 (37.8)	17 (30.9)	5 (17.9)
Higher	29 (18.5)	18 (26.5)	4 (5.4)	3 (5.5)	3 (10.7)
Occupationally active
Yes	44 (28.0)	13 (19.1)	11(14.9)	2 (3.6)	1 (3.6)	0.00
Takes over 5 drugs, N (%)
Yes	83 (52.9)	39 (57.4)	53 (71.6)	41 (74.5)	19 (67.9)	0.01
Co-morbidities, N (%)
Hypertension	116 (73.9)	47 (69.1)	60 (81.1)	46 (83.6)	26 (92.9)	0.05
Coronary disease	31 (19.7)	19 (27.9)	32 (43.2)	17 (30.9)	7 (25.0)	0.01
Heart failure	1 (0.6)	0 (0.0)	6 (8.1)	1 (1.8)	1 (3.6)	0.01
Diabetes	43 (27.4)	22 (32.4)	35 (47.3)	22 (40.0)	22 (78.6)	0.00
Renal failure	21 (13.4)	13 (19.1)	19 (25.7)	11 (20.0)	4 (14.3)	0.22

Bold values indicate statistical significance (p < 0.05).

Table 5. Analysis of one–way variance (ANOVA) of the influence of analysed socio-clinical variables on the overall ESF scores.

Factor	EFS - Total	ANOVA
Gender
Women	6.61 ± 3.33	p = 0.91
Men	6.65 ± 3.25
Age
≤70 years	5.48 ± 2.88	p = 0.00
>70 years	8.07 ± 3.22
Habitation:
City	6.96 ± 3.45	p = 0.21
Village	6.49 ± 3.23
Education
Basic	8.11 ± 3.47	p = 0.00
Vocational	6.90 ± 3.29
Secondary	6.21 ± 3.00
Higher	5.37 ± 3.13
Marital status
Free	7.55 ± 3.08	p = 0.00
In relationship	6.19 ± 3.31
Hypertension
No	5.80 ± 3.07	p = 0.01
Yes	6.87 ± 3.32
Coronary disease
No	6.34 ± 3.33	p = 0.01
Yes	7.38 ± 3.08
Heart failure
No	6.58 ± 3.29	p = 0.06
Yes	8.67 ± 2.87
Diabetes
No	5.98 ± 3.06	p = 0.00
Yes	7.69 ± 3.40
No diseases
No	6.82 ± 3.26	p = 0.00
Yes	4.71 ± 3.00
Occupationally active
No	6.97 ± 3.32	p = 0.00
Yes	5.11 ± 2.70
Takes over 5 drugs
No	5.85 ± 3.30	p = 0.00
Yes	7.11 ± 3.20

Bold values indicate statistical significance (p < 0.05).

The results of the analysis showed that higher scores of the EFS questionnaire and thus higher exposure to FS were indicated for individuals > 70 years than for those < 70 years (8.07 ± 3.22 vs. 5.48 ± 2.88, respectively; p < 0.001). In addition, the relationship between education and the occurrence of FS was observed. The higher the education, the lower the severity of FS; people with higher education had an EFS of 5.37 ± 3.13, while for people with a basic education, this was 8.11 ± 3.47 (p < 0.001). Similarly, people who were in relationships had a lower EFS score than those who were not (6.19 ± 3.31 vs. 7.55 ± 3.08; p < 0.001). In addition, the relationship between coexistence of chronic diseases and frailty was noted: people with hypertension had a higher score than people without (6.87 ± 3.32 vs. 5.80 ± 3.07; p = 0.008), people with diabetes compared to people without (7.69 ± 3.40 vs. 5.98 ± 3.06; p < 0.001) and persons with ischemic heart disease compared to those without (7.38 ± 3.08 vs. 6.34 ± 3.33; p = 0.006). People who were occupationally active had a lower score than nonactive individuals (5.11 ± 2.70 vs. 6.97 ± 3.32; p < 0.001). Differences in EFS levels were also observed, depending on the number of drugs taken. Higher scores and greater frailty were observed in people taking 5 or more drugs (7.11 ± 3.20) compared to those taking fewer than 5 drugs (5.85 ± 3.30; p < 0.001) (Table 5).

3.6. Multiple regression analysis of independent predictors of frailty

To determine which of the factors of relevance in the one-factor analysis is the independent predictor of frailty on the EFS scale, multivariate analysis using multiple regression was performed. The results are presented in Table 6, where the values of linear regression coefficients and β multiple regression were contained.

Table 6. Multiple regression of independent frailty predictors of the Edmonton Frail Scale score.

	One-variable analysis		Multivariate analysis
Predictor	b	p	B	p	β	SEβ
Age > 70 years	2.592	0.00	1.98	0.00	0.299	0.048
Basic education	1.826	0.00	0.85	0.05	0.101	0.052
Vocational education	0.372	0.32	–	–	–	–
Secondary education	−0.047	0.05	−1.15	0.69	−0.022	0.054
Higher education	−1.478	0.00	−1.19	0.01	−0.128	0.050
Lives in relationship	−1.357	0.00	−0.89	0.01	−0.126	0.044
Hypertension	1.063	0.01	0.40	0.34	0.051	0.054
Coronary disease	1.040	0.01	0.40	0.22	0.055	0.045
Type 2 diabetes	1.704	0.00	1.73	0.00	0.255	0.046
No diseases	−2.104	0.00	−0.47	0.46	−0.042	0.056
Occupationally active	−1.858	0.00	−0.71	0.07	−0.084	0.046
Takes over 5 drugs	1.260	0.00	0.89	0.00	0.132	0.044

Bold values indicate statistical significance (p < 0.05).

The independent predictors of the severity of FS in the EFS scale were: age over 70 years (β = 0.299; SE_β = 0.048), occurrence of diabetes (β = 0.255; SE_β = 0.044), and polypharmacy (intake of more than 5 drugs) (β = 0.132; SE_β = 0.044). Lowering FS is influenced by: higher education (β = −0.128; SE_β = 0.050), and life in a relationship (β = −0.126; SE_β = 0.044) (Table 6).

Model: $$EFS=\text{ 4}.\text{98}+\text{1}.\text{98}\times Age+\text{1}.\text{73}\times Diabetes+0.\text{89}\times Drugs-\text{1}.\text{19}\times Education-0.\text{89}\times Relationship\pm \text{2}.\text{83},R=0.\text{521},F\left(\text{4},\text{ 377}\right)=\text{35}.\text{1},\mathrm{p }<0.00\text{1}$$

4. Discussion

The main aim of this study was to determine the incidence of FS among elderly patients hospitalized in the Clinic of Internal and Occupational Diseases and Hypertension and to assess the psychometric properties of the EFS questionnaire in the Polish population.

The evaluation of FS in daily practice should be routinely performed and taken into account when planning therapeutic activities [14,15]. In the study by Dasgupta et al. [16], it was demonstrated that the frailty assessment is helpful in the stratification of global surgical risk in older adults undergoing mainly orthopedic, elective surgery.

This study is the first attempt to translate and evaluate the EFS questionnaire in Polish population. At present, except for the original version, only validation of the Portuguese [17] and Turkish versions [18] has been performed.

In our own studies, the Polish language version of the EFS questionnaire showed satisfactory psychometric characteristics. Cronbach’s alpha coefficient was 0.709 for total EFS score and it was closer and even higher than in the validation of the original version performed by Rolfson et al. [12], where the Cronbach’s alpha coefficient for the whole EFS scale was 0.62. A similarly high Cronbach’s alpha coefficient of 0.75 was obtained in the study assessing the Turkish version of EFS among 130 patients in a nursing home. Available publications indicate the good characteristics of the EFS questionnaire in different populations of the elderly. Fabricio-Wehbe et al. [17] documented satisfactory psychometric properties of the Portuguese language version performed in the community population. In other studies, Hilmer et al. [15] conducted an assessment using a modified scale, the Reported Edmonton Frail Scale, which is particularly designed for the evaluation of frailty for older acute inpatients. The Cronbach’s alpha coefficient for REFS was 0.68. The results of this scale showed a high correlation with other geriatric evaluation tools.

The results of this study have confirmed the reliability of the Polish version of the EFS, meaning that it can be used in research in the Polish population.

In the presented analysis, FS was found in 157 patients (41.1%), with a further 68 subjects (17.8%) being found to be vulnerable, and therefore in a risk of FS. In the remaining 157 patients (41.1%) no exponents of FS were found. The results are consistent with those of other studies on the incidence of FS. In the literature of the subject, a number of studies were performed to investigate the incidence of FS using various diagnostic tools, as well as investigating the incidence of FS in diverse groups of the elderly, depending on the place of residence (own environment, residents of the nursing home) or clinical situation (patients hospitalized due to acute illness), which obviously affected the observed prevalence of FS. This syndrome is one of the reasons for the increased risk of institutionalization as well as hospitalization in cases of worsening of health [19]. Since there is still no single standard tool for recognizing FS in the literature, there have been many publications exploring its incidence, as diagnosed by different diagnostic tools; the most frequently used criterion is the phenotype of the weakness syndrome by Fried [20]. In the systematic review by Collard et al. [21], who analyzed 21 publications assessing the incidence of FS in community-dwelling individuals, it has been shown that the frequency of occurrence was within the range from 4.0% to 59.1%, in 10.7% of respondents on average, while there were 41.6% in the prefrail phase. In most of the studies analyzed in the cited review, the authors used the Fried phenotype to identify the FS.

4.1. Frequency of frailty using EFS

In the papers where the EFS was used as a diagnostic tool, the frequency of FS was also varied. In the study by Sutorius et al. [22], within community-dwelling persons (study conducted among people under primary care), the frequency of FS identified by the EFS was 5% (with a cutoff of 6 points). In studies conducted among community-dwelling individuals in Taiwan, the prevalence of the syndrome was higher (14.9%), with the cutoff point adopted by the authors being 7 points on the EFS scale, which could have exacerbated the outcome [23]. It is worth noting that in the cited study the study group comprised of persons under the age of 80, which could also have had a significant impact on the obtained results. In the paper presented by Perna et al. [24], the frequency of FS among patients hospitalized in the rehabilitation ward was assessed in people whose somatic condition was stable and did not present any significant acute dysfunction. The prevalence of the FS was high and exceeded 80% (80.7%), with the authors highlighting a group of patients with severe FS, which included 13.9%, whereas most of the patients (66.4%) achieved an EFS score of 6 to 11 points.

As noted in the present study, the frequency of FS was 41.1%. The higher prevalence – 64% – was noted by Hilmer et al. [15] in a similar group of patients as in our analysis. These were patients hospitalized in several different departments of the University Hospital in Sidney using a modified form of the Edmonton Frail Scale (Reported Edmonton Frail Scale). The scale was modified by replacing the “Timed Up and Go” test with three questions about taking specific physical activity within 2 weeks before hospitalization, including washing windows or walking 1 km without help, and the FS is recognized by obtaining 8 or more points. It appears test modifications made by authors regarding physical fitness may have influenced the higher percentage of frailty in the cited study. Walking a distance of 1 kilometer or from a ground floor to the 2nd floor without assistance requires greater physical fitness and higher functional efficiency than the “Timed Up and Go” test, even if the period of 2 weeks before hospitalization is taken into consideration; therefore, a higher percentage of patients could gain higher scores in this area, which affected the higher overall scores in the scale [15].

4.2. Dependence of FS on socio-demographic factors

The results showed that the frequency of FS was higher among women and it was also increasing with age, as confirmed in our study. From a mentioned systematic review by Collard et al. [21], it appears that frailty and prefrailty are significantly more common among women, and this rate increases with age. Similar results were presented by Theou et al. [25], who compared the incidence of FS using 7 different diagnostic tools (including EFS); with each of them, the frequency of FS among women was higher and increased with age. Similar data can be found in the papers of many other authors evaluating the presence of FS using various diagnostic tools [14,15,26–33]. It should be noted that in the present study the participants were under 80 years of age, which may be related to the results obtained. However, our results confirmed the association of age with FS – the older the individuals, the stronger the predisposition to FS. In addition, in the regression analysis, age below 70 years was an independent predictor of less severe FS.

Another socio-demographic factor that influences the frequency of FS is education. Lower levels of education are significantly more common in people with FS, which was also confirmed in our own analysis. In the regression analysis, a higher level of education had a significant effect on lower FS scores.

The descripted dependency is reflected in the published results and is noticeable irrespective of socio-cultural conditions. Studies from different European countries and even from Asia or Mexico present similar conclusions in this regard [23,26,29,30].

Among people who are particularly susceptible to the occurrence of FS are single individuals who have been hospitalized repeatedly with depression [34]. Also, in our own studies, there was a significant correlation between the occurrence of FS and the civil status and occupational activity. In the one-factor analysis, occupational activity reduced the occurrence of FS while being in a relationship was an independent determinant associated with limiting the occurrence of this syndrome in the elderly.

Among the clinical variables that were noted to be associated with the existence of FS, the incidence of co-morbidities (diabetes, hypertension, ischemic heart disease) must be noted. Each of the comorbidities (one-factor analysis) and diabetes (multivariate analysis) affected the increase of the severity of frailty. Likewise, the polypharmacy issue in our own study was an independent predictor of severity of FS. In a study by Vu et al. [35], as in the current study, there was a negative association between cardiovascular disease and malnutrition with FS, while in the study by Carneiro [34], symptoms of depression, cardiovascular disease, previous hospitalization, and falls in the past 12 months were associated with an increased risk of frailty. In the studies by González-Vaca at al. [36] and Abizanda et al. [37], female gender, the score on Barthel Index, the risk of depression and comorbidities were factors that increased the risk of FS. In the study by Jha et al. [38], the occurrence of FS was related to the occurrence of an adverse outcome, increased risk of co-morbidities, hospitalization, and mortality.

4.3. Study limitations

We are well aware of the potential limitations of this study. The most important of these stem from the fact that our study sample consisted of inpatients recruited from a single clinical center. Another limitation of our study is the use of only one tool for evaluating frailty and the lack of comparative analysis of the Edmonton questionnaire with another diagnostic and research tool. It would be interesting to continue the study among elderly patients in a primary care setting, which would allow comparative analysis and evaluation of the frequency of frailty in the general population of elderly people (nonhospitalized and noninstitutionalized).

4.4. The practical implications

The Polish version of the EFS proved to be a valid and reproducible tool for assessment of Frailty Syndrome for the Polish population. We would recommend to be used as the screening tool to assess frailty and helping with the geriatric assessment in clinical practice. The complex evaluation of the frailty syndrome may have important practical implications for public health by implementation of the strategies preventing the profession of FS and by education of the patients in order to increase the self-care.

5. Conclusions

The EFS questionnaire proved to be a reliable tool which may be used in the Polish population. Our study points out the wide dissemination of the frailty syndrome in population with hypertension (42% and 20% for the prefrail status).

The use of this questionnaire in order to identify and assess the severity of FS may be helpful in everyday clinical practice when making therapeutic decisions.

Factors that negatively affect the severity of FS were age >70 years, diabetes and polypharmacy. In turn, the factors that reduce the severity of frailty were higher education and being in a relationship. Knowledge of the factors associated with frailty allows the development of health actions aimed at the elderly.

Ethics approval

The study protocol was approved by the Independent Bioethics Committee of the Wroclaw Medical University (decision no. KB–388/2017). All participants gave written informed consent after thorough explanation of the procedures involved. The study was carried out in accordance with the tenets of the Declaration of Helsinki.

Acknowledgements

There were no other contributors to the article than the authors.

Disclosure statement

The authors report no conflicts of interest. All authors of this manuscript meet the authorship criteria according to the latest guidelines of the International Committee of Medical Journal Editors (ICMJE), as well as all authors have seen and approved the manuscript being submitted and published.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Availability of data and materials

All data and materials used in this research are freely available. References have been provided.

Additional information

Notes on contributors

Beata Jankowska-Polańska

Beata Jankowska-Polańska is a Registered Nurse, PhD; she is an assistant professor at the Wroclaw Medical University, Poland; her specialisations are nursing procedures in elderly, cardiovascular medicine, quality of life in chronic diseases, and evidence-based nursing procedures. Dr. Jankowska-Polańska is also a principal investigator of the project “Effect of Frailty Syndrome on Treatment Compliance in Hypertensive Elderly Patients.”

Bartosz Uchmanowicz

Bartosz Uchmanowicz is a Doctor of Medicine, PhD; he is an assistant professor at the Wroclaw Medical University, Poland; his specialisations are internal medicine and medical education. Dr. Uchmanowicz is also a Director of the European Center for Postgraduate Education in Wroclaw, Poland.

Hanna Kujawska-Danecka

Hanna Kujawska-Danecka is a Doctor of Medicine, PhD, she is an assistant professor at the Medical University of Gdansk, Poland; her specialisations are clinical geriatrics and internal medicine. Dr. Kujawska-Danecka is also a Deputy Head of Internal Medicine, Connective Tissue Diseases and Geriatrics Clinic at the University Clinical Center in Gdansk, Poland.

Katarzyna Nowicka-Sauer

Katarzyna Nowicka-Sauer is a Master of Psychology, PhD; she is an assistant professor at the Medical University of Gdansk, Poland; her specialisations are clinical psychology, mental illness, and neuropsychological assessment.

Anna Chudiak

Anna Chudiak is a Registered Nurse, PhD, she is assistant at the Wroclaw Medical University, Poland; her specialisations are surgical nursing and cardiovascular diseases. She is also a nurse in Department of Thoracic Surgery at the Lower Silesian Centre of Lung Diseases in Wroclaw, Poland.

Krzysztof Dudek

Krzysztof Dudek is a Master of Science in Engineering, PhD; he is an assistant professor at the Wroclaw University of Technology, Poland; his specialisations are statistical analysis and mathematical modelling.

Joanna Rosińczuk

Joanna Rosińczuk is a Registered Nurse, PhD; she is a full professor at the Wroclaw Medical University, Poland; her specialisations are nursing care in elderly and evidence-based nursing procedures. Prof. Rosińczuk is also a Dean of the Faculty of Health Sciences and Head of Department of Clinical Nursing at the Wroclaw Medical University.