Final Report Summary - MID-FRAIL-STUDY (A RANDOMIZED CLINICAL TRIAL TO EVALUATE THE EFFECTIVENESS OF A MULTI-MODAL INTERVENTION IN OLDER PEOPLE WITH TYPE 2 DIABETES ON FRAILTY AND QUALITY OF LIFE: THE MID-FRAIL STUDY)
Executive Summary:
Diabetes is a high prevalence disorder in ageing populations (20% in people ≥ 65 yrs) with a chronic disease time-course and is associated with considerable medical co-morbidity, functional loss and frailty, visual and lower limb disability, multiple medication usage and impaired quality of life and results in a high personal and social health burden, and a significant public health burden.
Diabetes is associated with increased frailty and functional decline in older people and may explain up to 20% of the excess risk of disability in an elderly population, with an annual relative risk of developing any disability around 2.0.
The MID-FRAIL-STUDY project focuses on the use of interventions designed to improve functional status and enhance quality of life rather than traditional treatments such as glucose- and blood pressure- lowering by acting on the mechanisms involved in producing frailty and its progression to adverse outcomes.
The main objective is to evaluate, in comparison with usual clinical practice, the effectiveness of a multi-modal intervention (education, diet and exercise) in frail and pre-frail subjects aged ≥ 70 years with T2D in terms of the difference in function 12 months post randomization, according to changes in score on the SPPB.
As secondary objectives, to evaluate the effect on economic health care costs, symptomatic hypoglycaemia, hospital admissions, permanent institutionalization and carer burden.
Nine hundred and sixty four subjects were included in 74 trial sites along 9 European countries. Baseline characteristics were comparable between both groups (intervention and usual care group).The mean age was 78 years (SD 5.44) and 50.9% were men and 37,7% were fragile.
Regarding the main variable, changes in function after 12 months of follow up measured by SPPB test, the intervention group showed a statistically significant improvement in the SPPB score respect the usual care group.
The results of this preliminary analysis show that the MID-FRAIL intervention induced significant improvement of QoL in older patients with diabetes. Especially the ADDQoL-Senior was sensitive for detecting differences between changes in QoL after 1 year between intervention and control. EQ5D scores were significantly related to the item ‘general health’ of the ADDQoL-Senior, which can be explained by the rather generic QoL domains measured with EQ5D.
The economic evaluation shows that, regardless of the chosen health indicator (SPPB score, percentage of patients with improvement in SPPB score ≥ one point or QALYs), the intervention dominates usual care. That is, the intervention program achieves better health outcomes at the same or lower cost than usual care. Therefore, the intervention program is efficient compared to usual care. The sensitivity analyses performed show that the results are consistent and the conclusions are robust, independently of the changes performed in the parameters and in the variables considered.
Project Context and Objectives:
1. Background
Diabetes is a high prevalent metabolic disorder in ageing populations and is associated with considerable medical co-morbidity, functional loss and frailty, visual and lower limb disability, multiple medication usage and impaired quality of life. The impact on the individual is thus high and the impact on society is costly in terms of increased hospitalisation, community and primary care health and social care costs, and increased care home residency.
Older patients with diabetes have an increased risk of disabling complications associated with lower limb dysfunction and increased falls risk. The MIDFrail Study is a timely innovative and unique study that addresses real clinical issues such as the prevention of disability by its focus on frailty, a pre-disability state. The interventions used (resistance training, nutritional and diabetes education and medical treatment optimisation) provide a novel approach to minimising the progression of pre-frail and frail states to disability and as such as high potential for demonstrating cost-effective interventions in our ageing society.
The MID-FRAIL-STUDY is an innovative clinical trial that may produce the first objective evidence the utility of a multimodal intervention in ageing subjects with type 2 diabetes and features of frailty. The focus of the study is centred around clinical outcomes measured by functional assessment and an enquiry into quality of life dimensions.
The information derived from the MIDFrail Study will enrich the evidence base for studies in diabetic subjects of advanced age, provide meaningful messages about what type of interventions are suitable for older people to undertake with a high degree of confidence and safety, and lead to recommendations in diabetes guidelines that can be implemented widely across the EU and globe.
A number of key benefits to the health status and well-being of European ageing citizens is expected if the anticipated findings the MIDFrail Study are implemented across the European Union: these are related to significant numbers of older people with diabetes which we have previously calculated to be equivalent to nearly 700,000 fewer cases of disability per year in the EU. A consequence of the reduction in disability levels will be huge savings in social and healthcare expenditure which we have previously calculated to 3 billion euros per year, thus enabling the opportunity to re-invest some of these savings into more sustainable integrated healthcare systems for older people in general and in those with diabetes in particular. We anticipate a greater opportunity to establish networks of support across the EU of groups that have a significant interest in gerontological research who will want to work with the present consortium to develop new initiatives in the area of ageing, frailty and diabetes.
The MIDFrail study is the first of its kind in the world that has targeted a vulnerable and often neglected group of patients and as such the data will be unique and original. It has already brought together a consortium of academics, SMEs and not for profit research organisations, clinical scientists, pharmacologists and health economists that have a major interest in the applied research of older people which provides a tremendous platform for other future EU-wide research initiatives.
Project Results:
2. Objectives:
Main objective:
To evaluate the effectiveness of a multi-modal intervention in frail and pre-frail subjects aged 70 years with T2D in terms of function and quality of life in comparison with usual clinical practice after 12 months post randomization.
Secondary objectives:
To evaluate, in comparison with usual clinical practice, the effectiveness of a multi-modal intervention in any of the following:
(1) Economic costs/healthcare expenditure due to diabetes;
(2) Incidence rate of symptomatic hypoglycemia and hypoglycemic coma;
(3) Incidence of hospital admission;
(4) Incidence of permanent institutionalization, and
(5) Carer burden.
3. Substudies
To evaluate the mechanisms underlying the effects of the intervention:
(1) Studying the changes in the body composition with exercise (SARTRAIN SubStudy)
(2) Studying the effect of increased power in both isometric and dynamic actions (MID-POW SubStudy);
(3) Studying the role of metabolome (MetaboFrail SubStudy)
(4) Studying tha genetic polymorphisms (GeneFrail SubStudy) as determinants of the response to treatment.
(5) To evaluate the efficacy of new therapeutic devices (SENSOLE SubStudy)
(6) To evaluate new ways to measure changes in QoL (QoLFrail SubStudy).
4. Design
This is an open-label cluster randomized multicentre superiority trial, with random allocation by clusters (Trial sites-TS) to a Usual Care Group (UCG) or an Intervention Group (IG). Simple cluster randomization using blocks of size [redacted], stratified by country
The original sample size was based on a dichotomous change in SPPB score of one or more points and resulted in a sample size of 1,718. We since changed the way the primary outcome is measured to use the continuous version of the SPPB and look for a difference between groups of 1 point on average. The new target sample size is 1000 patients. Under the initial assumptions regarding attrition (20%), clustering resulting in a design effect of 1.746875 (an intra cluster coefficient of 0.05 an average cluster size of 15 and a coefficient of variation of cluster size of 0.25 (CONSORT GROUP 2014 and ELDRIDGE), a standard 5% significance threshold (alpha), a two-sided test, we have 97% power to detect a mean difference of one SPPB point between the two groups as statistically significant (nQuery v7.0).
The original study was intended to follow all participants up until 2 years however this was amended to one year in order to extend recruitment.
5. Inclusion and exclusion criteria
The subject can enter the study if ALL of the following apply:
• Subject is willing and able to give written informed consent for participation in the study.
• Subject is aged 70 years or older, with a diagnosis of T2D for at least 2 years.
• Require to fulfil Fried´s criteria for frail or pre-frail individuals.
The subject cannot enter the study if ANY of the following apply:
• Barthel score lower than 60 points.
• Inability to carry out SPPB test (total score=0).
• Mini Mental State Examination score less than 20 points.
• Subjects unwilling or unable to consent or unable to participate safely in intervention program.
• Previous history of myocardial infarction within 6 months, unstable angina or congestive heart failure in III-IV NYHA stage.
• Clinically instable patients in the clinical judgment of the investigator.
• Terminal illness (life expectancy < 6 months).
• Any other condition that, in the clinical judgment of the investigator, means that it would not be in the best interests of the subject to enter the study.
• Current participation in a clinical trial or any other investigational study.
(See Figure 1)
Data for this study was collected in sites on paper and then uploaded locally via an online data system provided by OnMedic and site based and central data monitoring and querying was undertaking by Niche, with further data cleaning undertaken in Cardiff prior to database lock. This was detailed in the trial protocol and data management plan and included
Potential Impact:
As the data obtained are very sensitive and are subject to restrictions of the scientific journals, only preliminary results will be submitted in this report. We will inform to EC about these definitive results in other separate document when these are available.
1. Participants
The flowchart with subjects included in the study and in the analysis is showed in the following Figure 2.
Ninety six sites were involved in the study and 74 sites were randomized. The following table and figure show the number of trial sites included by country allocated to each arm: (Figure 3)
2. Baseline characteristics
The main baseline characteristics per country of subjects included in the study are showed in the next table. There were some variations between countries in terms of the demographic nature of the sample recruited, particularly by gender. (Figure 4)
The next table shows the main baseline characteristics in each arm of the study. The two randomised groups where broadly similar demographically and on baseline measures of the outcomes.
Baseline demographics by Trial arm (Figure 5)
The next table shows the baseline functional measures by Trial arm ( Figure 6)
Next table shows the baseline medications by trial arm (Figure 7)
Retention of participants has been a major challenge during the study. Withdrawals have exceeded 20%, most of them before the initiation of intervention. Therefore, we have analysed the baseline characteristics between those who had data at follow-up to one year and those who didn’t. Next table provides this comparison. The major difference was that participants in intervention clusters were far more likely to be lost to follow-up. The second factor which was those who were frail at baseline were less likely to be followed at one year. The potentially biases arising from these will be considered in the sensitivity analysis of the primary outcome. (Figure 8)
3. MID Frail main analysis
The analysis of SPPB test shows that the intervention arm had SPPB scores on average about 5/6ths of an SPPB point higher (indicating better function) than the control arm at one year follow-up (95% CI: 0.44 1.26 p-value <0.001).
The compliance of the intervention resulted 80%.
4. MID Frail Secondary analysis
Preliminary results don´t show clinically relevant or statistically significant treatment effects for either IADL, Barthel or EQ-5D-5L. Very few carers were recruited into the study and therefore this comparison is much unpowered and has wide confidence intervals.
5. Report of preliminary analysis of the QoL outcomes
The QoL Frail substudy is focused on the effects of the MID Frail intervention on quality of life (QoL) as part of the secondary outcomes of the main MID-FRAIL study. In addition, QoL Frail investigates the validity of the ADDQoL-Senior instrument which was specifically designed to appraise the impact of diabetes on QoL in older persons.
All subjects included in the MID-FRAIL study were invited to participate in the QoL assessment (see figure). There were no additional in- or exclusion criteria. (Figure 9)
a) ADDQoL-Senior
The ADDQoL-Senior questionnaire consists first in 2 overview items scoring
1/ the ‘present QoL’ (“In general, my present QoL is...”) scored on a 7-point Likert scale ranging from -3=extremely bad to +3=excellent,
2/ the ‘diabetes-dependent QoL’ (“If I did not have diabetes, my QoL would be ...”) scored on a 5-point Likert scale ranging from -1=worse to +3=very much better.
Next, the ADDQoL-Senior contains 17 domain-specific items of QoL for which the impact of diabetes is scored similarly as for the overall ‘diabetes-dependent QoL’. In addition, the importance of these 17 domains are scored on a 4-point Likert scale ranging from 0=not important to +3=very important. The multiplication of impact*importance score provides a weighed score for each of the 17 items.
b) EQ5D
The EQ5D questionnaire consists in 2 parts:
1/ 5 questions concerning 5 domains scored on a 5-point Likert scale ranging from +1=no problems to +5=extreme problems; generating a total sumscore ranging from +5 to +25.
2/ a VAS for QoL ranging from 0=worst imaginable health state to 100=best imaginable health state.
The results of this preliminary analysis show that the MID-FRAIL intervention induced significant improvement of QoL in older patients with diabetes. Especially the ADDQoL-Senior was sensitive for detecting differences between changes in QoL after 1 year between intervention and control. EQ5D scores were significantly related to the item ‘general health’ of the ADDQoL-Senior, which can be explained by the rather generic QoL domains measured with EQ5D. Further analysis of the data for impact of diabetes on the ADDQoL-Senior subdomains will enable us to identify on which specific QoL domains relevant for older diabetes patients the MID-FRAIL intervention had substantial effects.
6. Results from the Health Economic Assessment
As it has been mentioning throughout this report, diabetes mellitus leads to multiple complications and its functional impairment is a major problem affecting individual autonomy and quality of life. Consequently, substantial health-care costs are associated with its management in terms of treatment, hospitalizations, emergency consultations, visits to General Practitioners (GPs) and other specialists, social services and family support. Overall, such disease might lead to a high utilization of resources (both health and non-health resources), considering diabetes as an illness with a significant economic impact. Additionally, as occurs with diabetes, frailty also leads to functional impairment, affecting individual autonomy and social well-being. Consequently, frailty has been associated with an increased use of health and community services (García-Nogueras et al.; Bock et al. 2016). Nevertheless, no evidence exits about the economic impact among frail people suffering from diabetes.
The MID-Frail project focuses on the use of interventions (exercise and nutritional programs) designed to improve functional status and enhance quality of life by acting on the mechanisms involved in producing frailty and its progression to adverse outcomes in people suffering from diabetes and frailty. The evidence about the effectiveness of such type of interventions is ambiguous.
However, due to that resources are scarce and budgets are limited, it is necessary to make decisions about how to invest these resources in the most efficient way. Therefore, economic evaluations are useful tools that provide quality information on the costs and health outcomes of different alternatives in order to include the efficiency dimension in the decision-making process. Although some programs generate savings, a large number of them required an additional investment (higher cost). In fact, several analyses concluded that there was no improvement in health outcomes or it was very small (and cost effectiveness ratios are very high).
The main aim of the health economic assessment is to estimate the incremental cost-effectiveness ratio (ICER) and the incremental cost-utility ratio (ICUR) of the multi-modal intervention in frail and pre-frail subjects aged ≥ 70 years with T2D in comparison with usual care.
More precisely, the main targets of the health economic assessment are:
a. To estimate the cost related to visits to general practitioner, specialist, nurse and emergency service, medical tests, hospitalization, and formal and informal caregiving during the one year-follow up period.
b. To test whether there are statistical differences in costs between intervention (IG) and usual care (UG).
c. To test whether the intervention is a cost-effective option in comparison with usual care during the one year-follow up period.
Results
The health economic assessment showed that both healthcare and non-healthcare costs of the patients who participated in the intervention program were equal to or lower than the costs of patients who participated in the control group (usual care). The cost of the investment in machinery, training and location required by the intervention program was offset by savings in lower hospitalization costs during the first year of follow-up. In terms of non-healthcare costs, lower costs were also observed in patients assigned to the intervention group versus the usual care group, although the differences were not statistically significant. In this sense, the high variability in costs found in both groups should be noted. It is also worth noting the large differences found in the cost of the intervention in each country (machinery costs, training costs and the cost of locating machines). A future challenge is to find the optimal use of resources to take advantage of potential economies of scale in the implementation of the intervention program. This would allow access to the program to a greater number of patients, with the added advantage that the average cost per patient would be reduced consistently.
The economic evaluation shows that, regardless of the chosen health indicator (SPPB score, percentage of patients with improvement in SPPB score ≥ one point or QALYs), the intervention dominates usual care. That is, the intervention program achieves better health outcomes at the same or lower cost than usual care. Therefore, the intervention program is efficient compared to usual care. The sensitivity analyses performed show that the results are consistent and the conclusions are robust, independently of the changes performed in the parameters and in the variables considered.
7. Safety
There have been 334 SAEs since the beginning of the study. Only one classified as related to the intervention. Since the beginning of the study eight SAEs are related to a subject enrolled in Sensole (one in Part 1 and seven in Part 2). All other SAEs related to subjects enrolled in the main MID-frail study.
The classification of SAES by disease area is showed in the next table: (Figure 10)
The rest of analysis will be submitted in a separate document when are available.
List of Websites:
http://midfrailstudy.com/
See DataContactList.xls
Diabetes is a high prevalence disorder in ageing populations (20% in people ≥ 65 yrs) with a chronic disease time-course and is associated with considerable medical co-morbidity, functional loss and frailty, visual and lower limb disability, multiple medication usage and impaired quality of life and results in a high personal and social health burden, and a significant public health burden.
Diabetes is associated with increased frailty and functional decline in older people and may explain up to 20% of the excess risk of disability in an elderly population, with an annual relative risk of developing any disability around 2.0.
The MID-FRAIL-STUDY project focuses on the use of interventions designed to improve functional status and enhance quality of life rather than traditional treatments such as glucose- and blood pressure- lowering by acting on the mechanisms involved in producing frailty and its progression to adverse outcomes.
The main objective is to evaluate, in comparison with usual clinical practice, the effectiveness of a multi-modal intervention (education, diet and exercise) in frail and pre-frail subjects aged ≥ 70 years with T2D in terms of the difference in function 12 months post randomization, according to changes in score on the SPPB.
As secondary objectives, to evaluate the effect on economic health care costs, symptomatic hypoglycaemia, hospital admissions, permanent institutionalization and carer burden.
Nine hundred and sixty four subjects were included in 74 trial sites along 9 European countries. Baseline characteristics were comparable between both groups (intervention and usual care group).The mean age was 78 years (SD 5.44) and 50.9% were men and 37,7% were fragile.
Regarding the main variable, changes in function after 12 months of follow up measured by SPPB test, the intervention group showed a statistically significant improvement in the SPPB score respect the usual care group.
The results of this preliminary analysis show that the MID-FRAIL intervention induced significant improvement of QoL in older patients with diabetes. Especially the ADDQoL-Senior was sensitive for detecting differences between changes in QoL after 1 year between intervention and control. EQ5D scores were significantly related to the item ‘general health’ of the ADDQoL-Senior, which can be explained by the rather generic QoL domains measured with EQ5D.
The economic evaluation shows that, regardless of the chosen health indicator (SPPB score, percentage of patients with improvement in SPPB score ≥ one point or QALYs), the intervention dominates usual care. That is, the intervention program achieves better health outcomes at the same or lower cost than usual care. Therefore, the intervention program is efficient compared to usual care. The sensitivity analyses performed show that the results are consistent and the conclusions are robust, independently of the changes performed in the parameters and in the variables considered.
Project Context and Objectives:
1. Background
Diabetes is a high prevalent metabolic disorder in ageing populations and is associated with considerable medical co-morbidity, functional loss and frailty, visual and lower limb disability, multiple medication usage and impaired quality of life. The impact on the individual is thus high and the impact on society is costly in terms of increased hospitalisation, community and primary care health and social care costs, and increased care home residency.
Older patients with diabetes have an increased risk of disabling complications associated with lower limb dysfunction and increased falls risk. The MIDFrail Study is a timely innovative and unique study that addresses real clinical issues such as the prevention of disability by its focus on frailty, a pre-disability state. The interventions used (resistance training, nutritional and diabetes education and medical treatment optimisation) provide a novel approach to minimising the progression of pre-frail and frail states to disability and as such as high potential for demonstrating cost-effective interventions in our ageing society.
The MID-FRAIL-STUDY is an innovative clinical trial that may produce the first objective evidence the utility of a multimodal intervention in ageing subjects with type 2 diabetes and features of frailty. The focus of the study is centred around clinical outcomes measured by functional assessment and an enquiry into quality of life dimensions.
The information derived from the MIDFrail Study will enrich the evidence base for studies in diabetic subjects of advanced age, provide meaningful messages about what type of interventions are suitable for older people to undertake with a high degree of confidence and safety, and lead to recommendations in diabetes guidelines that can be implemented widely across the EU and globe.
A number of key benefits to the health status and well-being of European ageing citizens is expected if the anticipated findings the MIDFrail Study are implemented across the European Union: these are related to significant numbers of older people with diabetes which we have previously calculated to be equivalent to nearly 700,000 fewer cases of disability per year in the EU. A consequence of the reduction in disability levels will be huge savings in social and healthcare expenditure which we have previously calculated to 3 billion euros per year, thus enabling the opportunity to re-invest some of these savings into more sustainable integrated healthcare systems for older people in general and in those with diabetes in particular. We anticipate a greater opportunity to establish networks of support across the EU of groups that have a significant interest in gerontological research who will want to work with the present consortium to develop new initiatives in the area of ageing, frailty and diabetes.
The MIDFrail study is the first of its kind in the world that has targeted a vulnerable and often neglected group of patients and as such the data will be unique and original. It has already brought together a consortium of academics, SMEs and not for profit research organisations, clinical scientists, pharmacologists and health economists that have a major interest in the applied research of older people which provides a tremendous platform for other future EU-wide research initiatives.
Project Results:
2. Objectives:
Main objective:
To evaluate the effectiveness of a multi-modal intervention in frail and pre-frail subjects aged 70 years with T2D in terms of function and quality of life in comparison with usual clinical practice after 12 months post randomization.
Secondary objectives:
To evaluate, in comparison with usual clinical practice, the effectiveness of a multi-modal intervention in any of the following:
(1) Economic costs/healthcare expenditure due to diabetes;
(2) Incidence rate of symptomatic hypoglycemia and hypoglycemic coma;
(3) Incidence of hospital admission;
(4) Incidence of permanent institutionalization, and
(5) Carer burden.
3. Substudies
To evaluate the mechanisms underlying the effects of the intervention:
(1) Studying the changes in the body composition with exercise (SARTRAIN SubStudy)
(2) Studying the effect of increased power in both isometric and dynamic actions (MID-POW SubStudy);
(3) Studying the role of metabolome (MetaboFrail SubStudy)
(4) Studying tha genetic polymorphisms (GeneFrail SubStudy) as determinants of the response to treatment.
(5) To evaluate the efficacy of new therapeutic devices (SENSOLE SubStudy)
(6) To evaluate new ways to measure changes in QoL (QoLFrail SubStudy).
4. Design
This is an open-label cluster randomized multicentre superiority trial, with random allocation by clusters (Trial sites-TS) to a Usual Care Group (UCG) or an Intervention Group (IG). Simple cluster randomization using blocks of size [redacted], stratified by country
The original sample size was based on a dichotomous change in SPPB score of one or more points and resulted in a sample size of 1,718. We since changed the way the primary outcome is measured to use the continuous version of the SPPB and look for a difference between groups of 1 point on average. The new target sample size is 1000 patients. Under the initial assumptions regarding attrition (20%), clustering resulting in a design effect of 1.746875 (an intra cluster coefficient of 0.05 an average cluster size of 15 and a coefficient of variation of cluster size of 0.25 (CONSORT GROUP 2014 and ELDRIDGE), a standard 5% significance threshold (alpha), a two-sided test, we have 97% power to detect a mean difference of one SPPB point between the two groups as statistically significant (nQuery v7.0).
The original study was intended to follow all participants up until 2 years however this was amended to one year in order to extend recruitment.
5. Inclusion and exclusion criteria
The subject can enter the study if ALL of the following apply:
• Subject is willing and able to give written informed consent for participation in the study.
• Subject is aged 70 years or older, with a diagnosis of T2D for at least 2 years.
• Require to fulfil Fried´s criteria for frail or pre-frail individuals.
The subject cannot enter the study if ANY of the following apply:
• Barthel score lower than 60 points.
• Inability to carry out SPPB test (total score=0).
• Mini Mental State Examination score less than 20 points.
• Subjects unwilling or unable to consent or unable to participate safely in intervention program.
• Previous history of myocardial infarction within 6 months, unstable angina or congestive heart failure in III-IV NYHA stage.
• Clinically instable patients in the clinical judgment of the investigator.
• Terminal illness (life expectancy < 6 months).
• Any other condition that, in the clinical judgment of the investigator, means that it would not be in the best interests of the subject to enter the study.
• Current participation in a clinical trial or any other investigational study.
(See Figure 1)
Data for this study was collected in sites on paper and then uploaded locally via an online data system provided by OnMedic and site based and central data monitoring and querying was undertaking by Niche, with further data cleaning undertaken in Cardiff prior to database lock. This was detailed in the trial protocol and data management plan and included
Potential Impact:
As the data obtained are very sensitive and are subject to restrictions of the scientific journals, only preliminary results will be submitted in this report. We will inform to EC about these definitive results in other separate document when these are available.
1. Participants
The flowchart with subjects included in the study and in the analysis is showed in the following Figure 2.
Ninety six sites were involved in the study and 74 sites were randomized. The following table and figure show the number of trial sites included by country allocated to each arm: (Figure 3)
2. Baseline characteristics
The main baseline characteristics per country of subjects included in the study are showed in the next table. There were some variations between countries in terms of the demographic nature of the sample recruited, particularly by gender. (Figure 4)
The next table shows the main baseline characteristics in each arm of the study. The two randomised groups where broadly similar demographically and on baseline measures of the outcomes.
Baseline demographics by Trial arm (Figure 5)
The next table shows the baseline functional measures by Trial arm ( Figure 6)
Next table shows the baseline medications by trial arm (Figure 7)
Retention of participants has been a major challenge during the study. Withdrawals have exceeded 20%, most of them before the initiation of intervention. Therefore, we have analysed the baseline characteristics between those who had data at follow-up to one year and those who didn’t. Next table provides this comparison. The major difference was that participants in intervention clusters were far more likely to be lost to follow-up. The second factor which was those who were frail at baseline were less likely to be followed at one year. The potentially biases arising from these will be considered in the sensitivity analysis of the primary outcome. (Figure 8)
3. MID Frail main analysis
The analysis of SPPB test shows that the intervention arm had SPPB scores on average about 5/6ths of an SPPB point higher (indicating better function) than the control arm at one year follow-up (95% CI: 0.44 1.26 p-value <0.001).
The compliance of the intervention resulted 80%.
4. MID Frail Secondary analysis
Preliminary results don´t show clinically relevant or statistically significant treatment effects for either IADL, Barthel or EQ-5D-5L. Very few carers were recruited into the study and therefore this comparison is much unpowered and has wide confidence intervals.
5. Report of preliminary analysis of the QoL outcomes
The QoL Frail substudy is focused on the effects of the MID Frail intervention on quality of life (QoL) as part of the secondary outcomes of the main MID-FRAIL study. In addition, QoL Frail investigates the validity of the ADDQoL-Senior instrument which was specifically designed to appraise the impact of diabetes on QoL in older persons.
All subjects included in the MID-FRAIL study were invited to participate in the QoL assessment (see figure). There were no additional in- or exclusion criteria. (Figure 9)
a) ADDQoL-Senior
The ADDQoL-Senior questionnaire consists first in 2 overview items scoring
1/ the ‘present QoL’ (“In general, my present QoL is...”) scored on a 7-point Likert scale ranging from -3=extremely bad to +3=excellent,
2/ the ‘diabetes-dependent QoL’ (“If I did not have diabetes, my QoL would be ...”) scored on a 5-point Likert scale ranging from -1=worse to +3=very much better.
Next, the ADDQoL-Senior contains 17 domain-specific items of QoL for which the impact of diabetes is scored similarly as for the overall ‘diabetes-dependent QoL’. In addition, the importance of these 17 domains are scored on a 4-point Likert scale ranging from 0=not important to +3=very important. The multiplication of impact*importance score provides a weighed score for each of the 17 items.
b) EQ5D
The EQ5D questionnaire consists in 2 parts:
1/ 5 questions concerning 5 domains scored on a 5-point Likert scale ranging from +1=no problems to +5=extreme problems; generating a total sumscore ranging from +5 to +25.
2/ a VAS for QoL ranging from 0=worst imaginable health state to 100=best imaginable health state.
The results of this preliminary analysis show that the MID-FRAIL intervention induced significant improvement of QoL in older patients with diabetes. Especially the ADDQoL-Senior was sensitive for detecting differences between changes in QoL after 1 year between intervention and control. EQ5D scores were significantly related to the item ‘general health’ of the ADDQoL-Senior, which can be explained by the rather generic QoL domains measured with EQ5D. Further analysis of the data for impact of diabetes on the ADDQoL-Senior subdomains will enable us to identify on which specific QoL domains relevant for older diabetes patients the MID-FRAIL intervention had substantial effects.
6. Results from the Health Economic Assessment
As it has been mentioning throughout this report, diabetes mellitus leads to multiple complications and its functional impairment is a major problem affecting individual autonomy and quality of life. Consequently, substantial health-care costs are associated with its management in terms of treatment, hospitalizations, emergency consultations, visits to General Practitioners (GPs) and other specialists, social services and family support. Overall, such disease might lead to a high utilization of resources (both health and non-health resources), considering diabetes as an illness with a significant economic impact. Additionally, as occurs with diabetes, frailty also leads to functional impairment, affecting individual autonomy and social well-being. Consequently, frailty has been associated with an increased use of health and community services (García-Nogueras et al.; Bock et al. 2016). Nevertheless, no evidence exits about the economic impact among frail people suffering from diabetes.
The MID-Frail project focuses on the use of interventions (exercise and nutritional programs) designed to improve functional status and enhance quality of life by acting on the mechanisms involved in producing frailty and its progression to adverse outcomes in people suffering from diabetes and frailty. The evidence about the effectiveness of such type of interventions is ambiguous.
However, due to that resources are scarce and budgets are limited, it is necessary to make decisions about how to invest these resources in the most efficient way. Therefore, economic evaluations are useful tools that provide quality information on the costs and health outcomes of different alternatives in order to include the efficiency dimension in the decision-making process. Although some programs generate savings, a large number of them required an additional investment (higher cost). In fact, several analyses concluded that there was no improvement in health outcomes or it was very small (and cost effectiveness ratios are very high).
The main aim of the health economic assessment is to estimate the incremental cost-effectiveness ratio (ICER) and the incremental cost-utility ratio (ICUR) of the multi-modal intervention in frail and pre-frail subjects aged ≥ 70 years with T2D in comparison with usual care.
More precisely, the main targets of the health economic assessment are:
a. To estimate the cost related to visits to general practitioner, specialist, nurse and emergency service, medical tests, hospitalization, and formal and informal caregiving during the one year-follow up period.
b. To test whether there are statistical differences in costs between intervention (IG) and usual care (UG).
c. To test whether the intervention is a cost-effective option in comparison with usual care during the one year-follow up period.
Results
The health economic assessment showed that both healthcare and non-healthcare costs of the patients who participated in the intervention program were equal to or lower than the costs of patients who participated in the control group (usual care). The cost of the investment in machinery, training and location required by the intervention program was offset by savings in lower hospitalization costs during the first year of follow-up. In terms of non-healthcare costs, lower costs were also observed in patients assigned to the intervention group versus the usual care group, although the differences were not statistically significant. In this sense, the high variability in costs found in both groups should be noted. It is also worth noting the large differences found in the cost of the intervention in each country (machinery costs, training costs and the cost of locating machines). A future challenge is to find the optimal use of resources to take advantage of potential economies of scale in the implementation of the intervention program. This would allow access to the program to a greater number of patients, with the added advantage that the average cost per patient would be reduced consistently.
The economic evaluation shows that, regardless of the chosen health indicator (SPPB score, percentage of patients with improvement in SPPB score ≥ one point or QALYs), the intervention dominates usual care. That is, the intervention program achieves better health outcomes at the same or lower cost than usual care. Therefore, the intervention program is efficient compared to usual care. The sensitivity analyses performed show that the results are consistent and the conclusions are robust, independently of the changes performed in the parameters and in the variables considered.
7. Safety
There have been 334 SAEs since the beginning of the study. Only one classified as related to the intervention. Since the beginning of the study eight SAEs are related to a subject enrolled in Sensole (one in Part 1 and seven in Part 2). All other SAEs related to subjects enrolled in the main MID-frail study.
The classification of SAES by disease area is showed in the next table: (Figure 10)
The rest of analysis will be submitted in a separate document when are available.
List of Websites:
http://midfrailstudy.com/
See DataContactList.xls
