Keywords atrial fibrillation - integrated care - ABC pathway - outcomes
Introduction
In the last 10 years, great advancements have been made in the treatment of patients
with atrial fibrillation (AF), in particular regarding stroke prevention by an increasing
use of oral anticoagulant (OAC) drugs.[1 ] As a consequence, rates of stroke and thromboembolic events have decreased markedly,
being very low in more contemporary cohorts.[2 ]
Despite the improvements in reducing thromboembolic events, epidemiological data suggest
that there were no significant temporal changes to the overall age-adjusted risk of
death associated to the presence of AF,[3 ] being particularly related to an increase in the risks of hospitalisation and non-cardiovascular
(non-CV) death,[4 ] also re-emphasising the close relationship between comorbidity, multimorbidity and
AF.[5 ]
[6 ]
In order to address the burden of adverse clinical outcomes beyond thromboembolism,
implementation of a more comprehensive and integrated approach to AF management has
been advociated.[7 ]
[8 ]
[9 ] To streamline the implementation of such a holistic care approach for AF patients,
the ‘Atrial fibrillation Better Care’ (ABC) pathway has been proposed. The ABC pathway
stands on three main pillars: ‘A’: A void stroke (with A nticoagulants); ‘B’: B etter symptom management; ‘C’: C ardiovascular and C omorbidity management.[9 ] The ABC pathway is now recommended in several clinical guidelines, including the
recent European Society of Cardiology (ESC) AF management guidelines.[10 ]
[11 ]
[12 ]
The objective of this article is to present a systematic review of the current evidence
for the use of the ABC pathway on clinical outcomes. We aimed, firstly, to establish
the overall prevalence of adherence to the ABC criteria in the retrospective analyses
available, and secondly, to perform a meta-analysis of ABC pathway compliance on clinical
outcomes.
Methods
This systematic review has been performed according to the Preferred Reporting Items
for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and recommendations (http://www.prisma-statement.org/ ). The protocol was registered into the International Prospective Register of Systematic
Reviews (PROSPERO; N. CRD42020218088). The data underlying this article are available
in the article and in its online Supplementary Material (available in the online version).
Search Strategy
A systematic and comprehensive literature search was performed on PubMed and EMBASE
databases, from inception to December 8, 2020. The search strategy included a combination
of key relevant terms related to the research question, including ‘ABC Pathway’ and
‘Atrial Fibrillation Better Care’. The full search strategy is reported in the Supplementary
Material ([Supplementary Table S1 ], available in the online version).
Table 1
Characteristics of the studies included in the systematic review
Study
Region
Design
N
ABC
(N )
A criterion
(%)
B criterion
(%)
C criterion
(%)
Age
(y)
CHA2 DS2 -VASC
HAS-BLED
OAC
(%)
FU
(y)
Retrospective studies
Gumprecht et al, 2020[16 ]
Europe
Multicentre Observational
2,021
168
55.3
75.1
19.2
56.7
2.3
1.1
56.4
1
Kozieł et al, 2020[17 ]
Europe
Multicentre Observational
2,312
1,013
74.5
90.2
72.2
69.3
3.4
2.0
68.6
NA
Proietti et al, 2018[19 ]
North America
RCT post-hoc
3,169
222
46.3
37.5
32.4
70[a ]
2.9
NA
100
3.7
Proietti et al, 2020[20 ]
Europe
Multicentre Observational
6,646
1,996
75.1
80.2
50.8
68.3
2.9
1.5
89.6
1
Proietti et al, 2021[21 ]
Multinational
RCT post-hoc
3,637
961
46.7
75.9
72.8
72
3.0
3.2
100
1.6
Yang et al, 2020[22 ]
Asia
Nationwide Claim Registry
262,987
49,533
31.4
82.6
54.3
62.2
2.0
1.4
3.2
5.9
Prospective studies
Guo et al, 2020[23 ]
Asia
RCT
3,324
1,646
NA
NA
NA
68.5
3[a ]
1[a ]
29.2
1
Pastori et al, 2020[18 ]
Europe
Single centre Observational
1,157
428
61.8
81.9
70.5
75.2
3.5
1.5
100
1.9[a ]
Abbreviations: ABC, Atrial fibrillation Better Care; DM, diabetes mellitus; FU, follow-up;
NA, not available; OAC, oral anticoagulant; RCT, randomised controlled trial.
a Median value.
All details regarding study selection, inclusion and exclusion criteria, data extraction,
quality assessment and outcomes are reported in the Supplementary Methods ([Supplementary Material ], available in the online version).
Statistical Analysis
Prevalence of ABC-pathway-adherent management was pooled from each of the studies
included using a random intercept logistic regression model[13 ] with the ‘metaprop’ function in R.
The number of events and the total number of patients of each group of interest were
pooled and compared using a random-effects model. Pooled estimates were reported as
odds ratios (ORs) with 95% confidence intervals (CIs). The inconsistency index (I
2 ) was calculated to measure heterogeneity. According to pre-specified cut-offs, low
heterogeneity was defined as an I
2 of <25%, moderate heterogeneity when I
2 falls between 25 and 75%, and high heterogeneity when I
2 was >75%.
For each outcome, a sensitivity analysis was performed with a ‘leave-one-out’ approach,
in which all studies are removed one at a time to analyse their influence on the pooled
estimate and heterogeneity. We also performed several subgroup analyses: (1) for the
prevalence of ABC-adherent management, according to the geographical location of the
original studies; (2) for outcomes (all-cause death, CV death and ischemic stroke),
according to pre-specified CHA2 DS2 -VASc score groups (i.e., 0–2, 3–5, and 6–9).
To further investigate potential sources of heterogeneity, we performed a meta-regression.
Regarding the prevalence of the ABC-pathway-adherent management, we performed a multivariate
meta-regression using the adherence to each of the ABC pathway criterion as covariates.
For the outcomes, we performed univariate meta-regression with the Knapp–Hartung method[14 ] according to the duration of follow-up and thromboembolic risk factors [i.e., age,
sex, hypertension, diabetes mellitus, coronary artery disease (CAD), history of stroke,
congestive heart failure (CHF)].
Publication bias was assessed for studies reporting outcomes according to the adherence
to the ABC pathway, with the use of funnel plots, which were visually inspected for
asymmetricity. Egger's test was also performed. All the statistical analyses were
performed using R (version 4.0.3, The R Foundation, 2020), with the use of ‘meta’,
‘metafor’ and ‘dmetar’[15 ] packages.
Results
A total of 2,862 results were retrieved from the literature search (761 from PubMed
and 2,101 from EMBASE). After the title and abstract screening, 14 full texts were
evaluated, and eight studies were included in the final systematic review and meta-analysis[16 ]
[17 ]
[18 ]
[19 ]
[20 ]
[21 ]
[22 ]
[23 ] ([Table 1 ] and [Supplementary Fig. S1 ], available in the online version), with a total of 285,253 AF patients included.
Four studies were based in Europe,[16 ]
[17 ]
[18 ]
[20 ] two in Asia,[22 ]
[23 ] one in North America[19 ] and one was multinational.[21 ]
Fig. 1 Pooled prevalence of ABC adherent management. ABC, Atrial fibrillation Better Care;
CI, confidence interval; GLMM, generalised linear mixed model.
Among the eight included studies, two were post-hoc subgroup analyses of previously
performed randomised controlled trials (RCTs),[19 ]
[21 ] four were based on observational registries,[16 ]
[17 ]
[18 ]
[20 ] while the last one was derived from a nationwide claims registry.[22 ] One study did not include a follow-up phase.[17 ] The only RCT, the ‘mobile Atrial Fibrillation Application II’ (mAFA-II), a cluster-randomised
study conducted in China, implemented the ABC pathway through a mobile phone application.
Six out of eight studies[17 ]
[18 ]
[19 ]
[20 ]
[21 ]
[23 ] enrolled both out- and in-patients, while the remaining two only considered hospitalised
patients.[16 ]
[22 ]
Prevalence of ABC-Pathway-Adherent Care
Among the seven observational studies included in the systematic review, we found
a pooled prevalence of a clinical management adherent to the ABC pathway criteria
equal to 21% (95% CI: 13–34%), with a high heterogeneity (I
2 = 100%; [Fig. 1 ]). In order to evaluate the factors accounting for such a high degree of heterogeneity,
we performed a multivariate meta-regression analysis ([Supplementary Table S2 ], available in the online version). Among the factors included in the multivariate
analysis, adherence to the ‘A’, ‘B’ and ‘C’ components of the ABC pathway were found
to be directly associated with the prevalence of ABC-pathway-adherent clinical management
([Supplementary Table S2 ], available in the online version). The final model was able to explain most of the
heterogeneity observed (R
2 = 98.9%, p = 0.004).
Impact of ABC-Adherent Care on Outcomes
We performed a meta-analysis on the impact of the ABC pathway on major clinical outcomes
([Fig. 2 ]). Overall, the use of ABC-pathway-adherent care was associated with a significant
reduction of all-cause death compared with non-adherence (OR: 0.42; 95% CI: 0.31–0.56),
with a high between-studies heterogeneity (I
2 = 88%; [Fig. 2A ]). The risk of CV death was significantly lower in patients treated adherent to the
ABC pathway (OR: 0.37; 95% CI: 0.23–0.58) with a high degree of heterogeneity (I
2 = 89%; [Fig. 2B ]).
Fig. 2 Impact of ABC adherent management on outcomes. (A ) All-cause death; (B ) cardiovascular death; (C ) stroke; (D ) major bleeding. ABC, Atrial fibrillation Better Care; CI, confidence interval; MH,
Mantel–Haenszel.
The risk of stroke (OR: 0.55; 95% CI: 0.37–0.82; [Fig. 2C ]) and major bleeding (OR: 0.69; 95% CI: 0.51–0.94; [Fig. 2D ]) were significantly lower in those patients who were treated adherent to the ABC
pathway management, with an overall moderate degree of heterogeneity.
Subgroup Analyses
In order to evaluate the high degree of heterogeneity for the all-cause death outcome,
we performed a subgroup analysis in relation to the geographic location of the patients
in each study ([Fig. 3 ]) (for one study, which was multinational, we analysed the results according to the
regions included). While we did not find a significant difference between European,
Asian and North American patients, this analysis found that the geographic location
accounted for most of the heterogeneity in the main model, with a 40% residual heterogeneity.
Fig. 3 Impact of ABC adherent management on all-cause death according to regions. ABC, Atrial
fibrillation Better Care; CI, confidence interval; MH, Mantel–Haenszel.
Meta-regression Analysis
We performed a univariate meta-regression analysis to examine the relationship between
the clinical variables and the association of adherence to the ABC pathway with the
risk of all-cause death and CV death. In these analyses ([Supplementary Tables S3 ] and [S4 ], available in the online version), we found a direct association between the length
of follow-up and an increase in effectiveness, while conversely an increasing prevalence
of diabetes mellitus, CAD, CHF and stroke was associated with a reduction in effectiveness
of the ABC pathway for both all-cause death and CV death occurrence, all accounting
for most of the heterogeneity for the two outcomes ([Supplementary Tables S3 ] and [S4 ], available in the online version). Furthermore, the meta-regression analysis for
stroke ([Supplementary Table S5 ], available in the online version) and major bleeding ([Supplementary Table S6 ], available in the online version) found a direct association between the length
of follow-up and an increase in effectiveness for both these outcomes, accounting
for a significant proportion of heterogeneity ([Supplementary Tables S5 ] and [S6 ], available in the online version).
Sensitivity Analysis
The sensitivity analysis for the four outcomes according to the ‘leave-one-out’ approach
did not show any significant differences for each study included and any outcome ([Supplementary Figs. S2 ]–[S5 ], available in the online version). In the CHA2 DS2 -VASc-stratified analysis ([Fig. 4 ]), we found that for all-cause death, increasing CHA2 DS2 -VASc strata was associated with a progressively greater reduction of risk amongst
patients adherent to the ABC pathway, being greatest at the highest CHA2 DS2 -VASc strata (OR: 0.30; 95% CI: 0.17–0.54 for CHA2 DS2 -VASc 6-9) ([Fig. 4A ]). No difference in ABC pathway effectiveness was found across CHA2 DS2 -VASc strata for CV death and stroke occurrence ([Fig. 4B, C ]).
Fig. 4 Impact of ABC according to CHA2 DS2 -VASc strata on outcome. (A ) All-cause death; (B ) cardiovascular death; (C ) stroke. ABC, Atrial fibrillation Better Care; CI, confidence interval; MH, Mantel–Haenszel.
Bias Assessment
The risk of bias assessment ([Supplementary Tables S7 ] and [S8 ], available in the online version) showed an overall high quality of studies, with
the exception of Yang et al,[22 ] which was found at high risk of bias for both prevalence and outcomes analysis.
Significant publication bias was found for all-cause death (Egger's test, p = 0.021) and stroke (Egger's test, p = 0.008, [Supplementary Table S9 ], available in the online version). Visual inspection of the funnel plots ([Supplementary Fig. S6A ], [C ], available in the online version) revealed that, in both cases, asymmetricity was
caused by a void in the left side of the funnel plot, in which one would expect to
find studies with positive results. The addition of these potential studies may lead
to lower pooled ORs for both all-cause death and stroke.
Discussion
Firstly, in this systematic review and meta-analysis, clinical management adherent
to the ABC pathway was suboptimal, being adopted in one of every five AF patients.
Secondly, meta-analysis regarding clinical events showed that adherence to the ABC
pathway was associated with a significant reduction in the risk of major adverse outcomes
([Fig. 5 ]). Thirdly, adherence to the ABC pathway was largely driven by the implementation
of adequate antithrombotic therapy, adequate symptom control and by optimal control
of CV risk factors and comorbidities. Lastly, the meta-regression analyses regarding
outcomes showed that the increasing clinical complexity directly affects the effectiveness
of an integrated management strategy, while a longer follow-up was associated to a
greater reduction in risk.
Fig. 5 Graphical synopsis of the main study results. CI, confidence intervals; OR, odds
ratio.
Epidemiologically, the worldwide impact of AF has increased in the last 40 years.
Despite a significant reduction in age-standardised prevalence and incidence, the
absolute number of AF patients has almost doubled, being significantly increased in
countries with middle and low socio-demographic levels.[3 ] Additionally, observational studies have shown that the clinical risk profile of
AF patients has worsened over time, due to an increase in the prevalence of comorbidities.[24 ] This public health burden of AF has a major impact on mortality, where the total
number of attributable deaths has more than doubled, almost reaching 300,000 in 2017,
especially in middle and low socio-demographic countries.[3 ] Furthermore, an increasing effect on health-care-associated costs has been found.[25 ]
[26 ]
Given these concerns, a more holistic approach to AF management was needed. In 2018,
the 6th AFNET/EHRA Consensus Conference defined integrated care as ‘a coordinated
patient-centred approach by interdisciplinary specialists to improve AF outcomes’,
by improving all the specific domains related to AF management.[7 ]
[27 ] A model of care was proposed in which any AF patient should ideally be managed by
the AF Heart Team (specifically addressing the specific electrophysiology/cardiology
issues) and the Integrated Care AF Clinic, where several specialists could provide
the specific expertise to handle any aspect of the patient's care.[7 ]
[8 ]
[28 ]
In 2017, the ABC pathway was proposed to streamline and simplify the implementation
of integrated management in AF patients.[9 ] This simple model focuses on three main components, which are all essential to reduce
the risk of major adverse outcomes in AF. The ‘A’ criterion (A void stroke) refers to the management of thromboembolic and bleeding risks by appropriate
prescription and use of OAC drugs; the ‘B’ criterion (
B etter symptom management) aims to reduce and control symptom burden and patient-centred,
symptom-directed decisions on rate or rhythm control therapy; the ‘C’ criterion (C ardiovascular and C omorbidity risk optimisation) refers to the optimised management of any concomitant
comorbidity or CV risk factor.
Our systematic review demonstrates that a significant amount of evidence has already
been produced regarding the potential role of the ABC pathway in mitigating the risk
of major adverse outcomes. The retrospective analyses showed that the level of adherence
to the ABC pathway was low, with just one-fifth of the patients being managed optimally.
In particular, our meta-regression results suggest that more efforts are needed to
obtain more optimal adherence to all the ABC pathway components: for example, implementation
of adequate antithrombotic therapy and by more optimal control of CV risk factors
and comorbidities. However, we cannot exclude that some factors not considered in
this analysis may affect the integrated care of AF patients, such as education level,
health perception, household income, availability of a public health care system,
distance from health care services/hospitals and presence of caregiver for patients
with disability.[28 ]
[29 ] Furthermore, while it is important to underline that a clear heterogeneity exists
in the various definitions of ‘ABC-pathway-adherent care’ across the retrospective
studies (as further reported below), the importance of our results stands in the fact
that irrespective of how the studies defined components of the ABC pathway, good control
of anticoagulation quality, improved control of symptoms burden and the proper management
of the most relevant comorbidities require an effort from the treating physician which
is ‘integrated or holistic care’. This article shows how few patients are clinically
managed in this way.
The pivotal role of a holistic approach to AF care is underlined by the results of
this meta-analysis, showing that all the major adverse outcomes are significantly
and consistently reduced in the ABC-pathway-adherent group of patients. Indeed, ABC-pathway-adherent
care was associated with a 40 to 60% risk reduction for all the outcomes considered.
Moreover, the positive results of the only RCT about the ABC pathway produced strengthen
the view that more effort should be put into translating this evidence-based approach
into daily clinical practice. Our evidence that a longer observation is associated
with a greater reduction in risk for all the outcomes corroborates the evidence for
the effectiveness of the intervention, where there is a ‘dose–response’ effect, extending
the evidence from the mAFA-II trial secondary analyses.[30 ] The negative impact of a higher prevalence of several comorbidities on the effectiveness
of ABC-pathway-adherent care emphasises the role of comorbidities and increased clinical
complexity in influencing the clinical course. Indeed, an increasing prevalence of
diabetes mellitus, CAD, CHF and stroke was associated with a reduction in effectiveness.
Such evidence is reinforced by several studies illustrating how an increasing level
of multimorbidity is independently associated with an increased risk of outcomes,
also determining a differential approach in OAC prescription.[5 ]
[6 ] In our study, the results of the CHA2 DS2 -VASc-stratified analysis showed a greater risk reduction for all-cause death in patients
with the highest thromboembolic risk; looking at this evidence and given prior evidence
regarding the specific impact of ABC-pathway-adherent care in reducing the risk of
outcomes in ‘clinically complex’ AF patients, for example those with multimorbidity,[31 ] we would suggest that such a holistic approach is even more needed in those with
the highest risk profiles.
This evidence, together with other data generated by secondary analyses of the studies
included in this systematic review, which showed a significant impact of the ABC-pathway-adherent
care in reducing the risk of dementia[32 ] and a significant reduction in health-care-associated costs,[33 ] strongly supports the recent changes introduced in the 2020 ESC AF clinical guidelines.[12 ] In these guidelines, there is a paradigm shift in approach, placing the patient
at the centre of the physicians' action, not the disease itself. With the aim of managing
the patient holistically, the application of the ABC pathway is central to the guideline
recommendations.
Limitations
Our article has some limitations. First, the observational and retrospective nature
of most of the included studies inherently limits the generalisability of the results.
Even though we performed several meta-regression analyses, unmeasured residual confounders
may have influenced our results given that the data were mostly from observational
studies. Furthermore, since two of the studies included were originally performed
more than 10 years ago, the different clinical practices could have impacted the overall
rate of adverse outcomes. Another major limitation which we can recognise is related
to an inevitable heterogeneity in the ABC pathway criterion definition, particularly
in relation to the ‘B’ criterion, which varied significantly between the studies.
Notwithstanding this, it should be taken in mind that in the spirit of the original
ABC pathway proposal, the point is related to the best control of the particular criterion
in the individual studies, irrespective of the methods used to obtain the control
or compliance with uniform targets, and to evaluate the effect on outcomes. Even though
there was heterogeneity of assessments used in each study, the evaluation of the ABC
criteria aimed to identify patients who were best managed to obtain the best control
possible for each criterion. Lastly, for the evaluation of the ‘C’ criterion, most
of the studies limited the evaluation to the main CV risk factors and comorbidities.
Conclusion
In this systematic review and meta-analysis, clinical management adherent to the ABC
pathway was sub-optimally applied, being adopted in one in every five AF patients.
Adherence to the ABC pathway was associated with a significant reduction in the risk
of major adverse outcomes, with a significantly reduced risk of all-cause death, CV
death, stroke and major bleeding.
What is known about this topic?
In atrial fibrillation (AF) patients, multimorbidity and clinical complexity increase
the risk of death and hospitalisation.
In recent years the need for a more comprehensive and holistic approach to AF patients
has been recognised.
The ‘Atrial fibrillation Better Care’ (ABC) pathway has been proposed to streamline
the application of integrated or holistic care in AF patients.
What does this paper add?
In this systematic review and meta-analysis, we show that a clinical management adherent
to the ABC pathway was sub-optimally applied in AF patients.
ABC-pathway-adherent care was associated with a significant reduction of all-cause
death, cardiovascular death, stroke and major bleeding risk in AF patients.
Increased clinical complexity decreases the effectiveness of the ABC pathway in reducing
risk of death, while a longer follow-up time maximises the effect of an integrated
care approach.
