Re: The challenge of ageing populations and patient frailty: can primary care adapt?
The new article by David Reeves and colleagues is a thoughtful analysis of the challenges and opportunities of routine frailty identification and frailty management in the UK primary care settings [1].
The question posed by the authors (“can primary care adapt?”) inevitably brings up a mirroring question – can the concept of “frailty”, however we agree to define it, adapt to primary care? And if yes, what are the tensions which need to be resolved, or at least minimised?
According to the 2017/2018 GP Contract by the NHS England [2], potential frailty should be identified in all patients aged 65 and over using the electronic frailty index (eFI) or another validated tool. “Clinical judgment to confirm or… give further consideration to the tool result” should be used for those with potential moderate or severe frailty. Finally, for those living with severe frailty, action should be taken, including an annual medication review, a falls risk assessment, and seeking informed patient consent to activate the enriched Summary Care Record (SCR).
The “clinical judgement” step is crucial, as simply counting the number of eFI-comprising deficits (out of potential 36) will result in overestimation among those registered with the practice for longer, as Figure 1 in the paper by Reeves et al. [1] clearly shows. On the other hand, frailty will be underestimated in those with limited opportunities to have eFI deficits registered in their primary care electronic health record (due to frequent transfer between practices, limited contact with primary care services, and a multitude of other reasons).
It is encouraging to see that the eFI-defined prevalence of severe and moderate frailty appears very similar across three primary care research databases: 3% and 12% in ResearchOne [3], 4% and 16% in THIN (The Health Improvement Network) [3], and 3% and 10% in CPRD (Clinical Practice Research Datalink) [1], respectively.
However, these reassuringly similar findings provide minimal indication of the proportion of patients who would be diagnosed with severe or moderate frailty after a clinical assessment. Hopefully, this issue is being addressed in the ongoing eFI improvement work by Dr Reeves and colleagues [1].
The scope and scale of the gap between eFI-defined and clinically validated frailty is confirmed by the NHS Digital data for the GP Contract Services for England, 2017/2018 [4]. The recent analysis of these data [5] optimistically emphasizes the rapidly increasing levels of frailty reporting across GP practices.
However, a closer look at the data suggests that this rapid increase occurred mostly in the last quarter of the reporting period. It raises a question whether GP practices interpret frailty reporting as yet another performance management exercise, and do not appear confident in the NHS England promise of not using these data “for performance management or benchmarking purposes” [2].
Even if we concentrate on the aggregate data for all quarters of the reporting period (April 1st 2017 – March 31st 2018), as presented in the recent report [5], multiple problems are obvious. In England overall, the proportion of people aged 65+ who have had a frailty assessment is far from the target 100% level (mean 26%, median 13%).
The proportion of people with a diagnosis of moderate or severe frailty (mean 9%, median 5%) [5] is lower than the eFI-defined proportions in SystemOne (15%) [3], THIN (20%) [3], or CPRD (13%) [1]. It could be due to the fact that after clinical assessment, the “false positive” cases of moderate or severe frailty are dropped. However, the ratios of moderate to severe frailty could be expected to be similar for the eFI-defined levels (3.8-4.0) [1, 3] and for the clinically confirmed levels of frailty. The latter ratio is considerably lower (mean 2.0, median 1.8) [5], which suggests either under-diagnosis of moderate frailty, or over-diagnosis of severe frailty, or both.
Among the indicators reflecting frailty management, the highest levels are reported for the proportion of severely frail patients who have received annual medication review (mean 66%, median 71%) [5]. This is not surprising, as regular medicines reviews are also required by the NICE (National Institute for Health and Clinical Excellence) guidelines, such as those on clinical assessment and management of multimorbidity [6] and older people with social care needs and multiple long-term conditions [7].
Of note, there is a discrepancy between some frailty care indicators (falls risk assessment, falls clinic referral, and seeking consent to activate SCR) originally outlined in the NHS England guidance on frailty and those collected and reported by the NHS Digital. While the former focus on people with severe frailty [2], the latter refer to both moderately and severely frail patients [4]. This discrepancy and potential misinterpretation could have negatively affected the reported levels of frailty care.
For example, the proportion of moderately or severely frail people who have had a fall in the past 12 months (mean 11%, median 9%) [5] is substantially lower than that suggested by the NICE guidelines on falls in older people (“…30% of people older than 65… falling at least once a year”) [8]. Only 25% (median 17%) of people with moderate or severe frailty and a recent fall are referred to a falls clinic [5]. The extremely wide range for the falls clinic referral (0-2300%) suggests that the indicator is misinterpreted. Finally, the consent to activate the enriched SCR has been obtained for only a fraction of moderately or severely frail people (mean 15%, median 5%).
The evidence from the NHS Digital 2017/2018 GP Contact dataset suggests a limited engagement of primary care with the currently used frailty indicators. It is encouraging to hear that Dr Reeves and his colleagues involve GPs in their ongoing work on eFI improvement [1]. Only with active engagement of clinicians and patients can we achieve the integration of frailty identification and management into primary care practice.
Time will tell whether the use of a revised eFI, or an equivalent new screening tool, can translate into the primary care buy-in, as reflected in the frailty indicators across all levels of reporting, improved health outcomes, and reduced professional burden.
REFERENCES
[1] Reeves D, Pye S, Ashcroft DM, Clegg A, Kontopantelis E, Blakeman T, van Marwijk. The challenge of ageing populations and patient frailty: can primary care adapt? BMJ 2018; 362:k3349.
[2] NHS England. Updated guidance on supporting routine frailty identification and frailty care through the GP Contract 2017/2018. 2017. Available at: https://www.england.nhs.uk/wp-content/uploads/2017/04/supporting-guidanc...
[3] Clegg A, Bates C, Young J, Ryan R, Nichols L, Teale EA, Mohammed MA, Parry J, Marshall T. Development and validation of an electronic frailty index using routine primary care electronic health record data. Age Ageing 2016;358:353-60.
[4] NHS Digital GP Contract Services, England, 2017-18. 2018. Available at: https://digital.nhs.uk/data-and-information/publications/statistical/gp-...
[5] Seymour D. One small step for older people with frailty, one giant leap for frailty care? An analysis of GP Contract Services data for routine frailty identification and frailty care through the GP Contract 2017/2018. 2018. Available at: http://fusion48.net/uploads/documents/Giant_leap_for_frailty_care_-_main...
[6] National Institute for Health and Clinical Excellence. Multimorbidity: clinical assessment and management (NICE guideline NG56). 2016. Available at: https://www.nice.org.uk/guidance/ng56
[7] National Institute for Health and Clinical Excellence. Older people with social care needs and multiple long-term conditions (NICE guideline NG22). 2015. Available at: https://www.nice.org.uk/guidance/ng22
[8] National Institute for Health and Clinical Excellence. Falls in older people: assessing risk and prevention (NICE guideline NG161). 2013. Available at: https://www.nice.org.uk/guidance/cg161
Competing interests:
No competing interests
31 August 2018
Olga Vikhireva
Research Fellow
Institute of Applied Health Research, University of Birmingham
Rapid Response:
Re: The challenge of ageing populations and patient frailty: can primary care adapt?
The new article by David Reeves and colleagues is a thoughtful analysis of the challenges and opportunities of routine frailty identification and frailty management in the UK primary care settings [1].
The question posed by the authors (“can primary care adapt?”) inevitably brings up a mirroring question – can the concept of “frailty”, however we agree to define it, adapt to primary care? And if yes, what are the tensions which need to be resolved, or at least minimised?
According to the 2017/2018 GP Contract by the NHS England [2], potential frailty should be identified in all patients aged 65 and over using the electronic frailty index (eFI) or another validated tool. “Clinical judgment to confirm or… give further consideration to the tool result” should be used for those with potential moderate or severe frailty. Finally, for those living with severe frailty, action should be taken, including an annual medication review, a falls risk assessment, and seeking informed patient consent to activate the enriched Summary Care Record (SCR).
The “clinical judgement” step is crucial, as simply counting the number of eFI-comprising deficits (out of potential 36) will result in overestimation among those registered with the practice for longer, as Figure 1 in the paper by Reeves et al. [1] clearly shows. On the other hand, frailty will be underestimated in those with limited opportunities to have eFI deficits registered in their primary care electronic health record (due to frequent transfer between practices, limited contact with primary care services, and a multitude of other reasons).
It is encouraging to see that the eFI-defined prevalence of severe and moderate frailty appears very similar across three primary care research databases: 3% and 12% in ResearchOne [3], 4% and 16% in THIN (The Health Improvement Network) [3], and 3% and 10% in CPRD (Clinical Practice Research Datalink) [1], respectively.
However, these reassuringly similar findings provide minimal indication of the proportion of patients who would be diagnosed with severe or moderate frailty after a clinical assessment. Hopefully, this issue is being addressed in the ongoing eFI improvement work by Dr Reeves and colleagues [1].
The scope and scale of the gap between eFI-defined and clinically validated frailty is confirmed by the NHS Digital data for the GP Contract Services for England, 2017/2018 [4]. The recent analysis of these data [5] optimistically emphasizes the rapidly increasing levels of frailty reporting across GP practices.
However, a closer look at the data suggests that this rapid increase occurred mostly in the last quarter of the reporting period. It raises a question whether GP practices interpret frailty reporting as yet another performance management exercise, and do not appear confident in the NHS England promise of not using these data “for performance management or benchmarking purposes” [2].
Even if we concentrate on the aggregate data for all quarters of the reporting period (April 1st 2017 – March 31st 2018), as presented in the recent report [5], multiple problems are obvious. In England overall, the proportion of people aged 65+ who have had a frailty assessment is far from the target 100% level (mean 26%, median 13%).
The proportion of people with a diagnosis of moderate or severe frailty (mean 9%, median 5%) [5] is lower than the eFI-defined proportions in SystemOne (15%) [3], THIN (20%) [3], or CPRD (13%) [1]. It could be due to the fact that after clinical assessment, the “false positive” cases of moderate or severe frailty are dropped. However, the ratios of moderate to severe frailty could be expected to be similar for the eFI-defined levels (3.8-4.0) [1, 3] and for the clinically confirmed levels of frailty. The latter ratio is considerably lower (mean 2.0, median 1.8) [5], which suggests either under-diagnosis of moderate frailty, or over-diagnosis of severe frailty, or both.
Among the indicators reflecting frailty management, the highest levels are reported for the proportion of severely frail patients who have received annual medication review (mean 66%, median 71%) [5]. This is not surprising, as regular medicines reviews are also required by the NICE (National Institute for Health and Clinical Excellence) guidelines, such as those on clinical assessment and management of multimorbidity [6] and older people with social care needs and multiple long-term conditions [7].
Of note, there is a discrepancy between some frailty care indicators (falls risk assessment, falls clinic referral, and seeking consent to activate SCR) originally outlined in the NHS England guidance on frailty and those collected and reported by the NHS Digital. While the former focus on people with severe frailty [2], the latter refer to both moderately and severely frail patients [4]. This discrepancy and potential misinterpretation could have negatively affected the reported levels of frailty care.
For example, the proportion of moderately or severely frail people who have had a fall in the past 12 months (mean 11%, median 9%) [5] is substantially lower than that suggested by the NICE guidelines on falls in older people (“…30% of people older than 65… falling at least once a year”) [8]. Only 25% (median 17%) of people with moderate or severe frailty and a recent fall are referred to a falls clinic [5]. The extremely wide range for the falls clinic referral (0-2300%) suggests that the indicator is misinterpreted. Finally, the consent to activate the enriched SCR has been obtained for only a fraction of moderately or severely frail people (mean 15%, median 5%).
The evidence from the NHS Digital 2017/2018 GP Contact dataset suggests a limited engagement of primary care with the currently used frailty indicators. It is encouraging to hear that Dr Reeves and his colleagues involve GPs in their ongoing work on eFI improvement [1]. Only with active engagement of clinicians and patients can we achieve the integration of frailty identification and management into primary care practice.
Time will tell whether the use of a revised eFI, or an equivalent new screening tool, can translate into the primary care buy-in, as reflected in the frailty indicators across all levels of reporting, improved health outcomes, and reduced professional burden.
REFERENCES
[1] Reeves D, Pye S, Ashcroft DM, Clegg A, Kontopantelis E, Blakeman T, van Marwijk. The challenge of ageing populations and patient frailty: can primary care adapt? BMJ 2018; 362:k3349.
[2] NHS England. Updated guidance on supporting routine frailty identification and frailty care through the GP Contract 2017/2018. 2017. Available at: https://www.england.nhs.uk/wp-content/uploads/2017/04/supporting-guidanc...
[3] Clegg A, Bates C, Young J, Ryan R, Nichols L, Teale EA, Mohammed MA, Parry J, Marshall T. Development and validation of an electronic frailty index using routine primary care electronic health record data. Age Ageing 2016;358:353-60.
[4] NHS Digital GP Contract Services, England, 2017-18. 2018. Available at: https://digital.nhs.uk/data-and-information/publications/statistical/gp-...
[5] Seymour D. One small step for older people with frailty, one giant leap for frailty care? An analysis of GP Contract Services data for routine frailty identification and frailty care through the GP Contract 2017/2018. 2018. Available at: http://fusion48.net/uploads/documents/Giant_leap_for_frailty_care_-_main...
[6] National Institute for Health and Clinical Excellence. Multimorbidity: clinical assessment and management (NICE guideline NG56). 2016. Available at: https://www.nice.org.uk/guidance/ng56
[7] National Institute for Health and Clinical Excellence. Older people with social care needs and multiple long-term conditions (NICE guideline NG22). 2015. Available at: https://www.nice.org.uk/guidance/ng22
[8] National Institute for Health and Clinical Excellence. Falls in older people: assessing risk and prevention (NICE guideline NG161). 2013. Available at: https://www.nice.org.uk/guidance/cg161
Competing interests: No competing interests