Multidisciplinary Integrated Care in Atrial Fibrillation (Micaf): ASystematic Review and Meta-Analysis

Introduction

Atrial fibrillation (AF) is the commonest cardiac arrhythmia diagnosed in nearly 30 million patients globally [1]. In the United States, an approximately 2.3 million people suffer from atrial fibrillation [2]. It is a global public health problem and is expected to rise in the coming years [3]. It poses a significant economic burden on healthcare systems accounting for a large number of hospital admissions. According to an estimate, hospitalizations due to AF increased by 23% from the year 2000 to 2010 [4]. The patients are at higher risk of developing congestive heart failure (CHF), stroke and systemic thromboembolism [5].

Wagner and colleagues introduced a chronic care model establishing that the chronic disease management requires a different approach in contrast to the standard medical care [6]. Multidisciplinary care represents the comprehensive case management strategy with a greater number of medical and social support personnel. Multidisciplinary care involves a multidisciplinary team that includes the primary care provider, other physicians, nurses, dietitians, pharmacists and social workers that provide long-term care to patients with chronic disease [7]. These programs also integrate a coaching plan, in which the health physician promotes patient empowerment for attaining treatment adherence and behavior modification [8]. For the coordination of integrated care programs, a clinical nurse specialist can play an important role [5].

In patients with heart failure and coronary heart disease, significant improvement has been observed due to the multidisciplinary approach [9,10]. The management of atrial fibrillation is often difficult due to lack of adherence to recommendations [11]. Such multidisciplinary plans are essential for the provision of guideline-based AF management with the collaboration of different physicians [5]. An AF program should include diagnostic tests, heart rate and rhythm control, anticoagulation, management of associated disorders, patient education and counseling for self-management [12].

The integrated healthcare methodology has backgrounds from the chronic healthcare model with the understanding that chronic disease management requires an unusual approach in contrast to standard usual models of healthcare delivery. This prototypical setup gives the patient the principal emphasis, with various significant essentials/components, including a multidisciplinary team and supports from the community, to warrant that the patient population is dynamically engaged regarding their treatment. Augmenting patient-related consequences incorporating this delivery strategy is reached through redesigning usual clinical practice to confirm healthcare is provided tailor-made to the patient’s requirements and founded on existing data [6].

There are five cornerstones of an effective multidisciplinary atrial fibrillation program. These are “comprehensive assessment, systematization of medical care, patient education, coordination of care and evaluation of care plan execution” [5]. A detailed assessment of the patient is essential for making a suitable management plan [13]. Systemization involves coordination of diagnostic work-up, treatment plan and follow-up [5]. Systemization of medical care will improve the excellence of care delivery, patient contentment and the use of means [14]. Patient education leads to dynamic patient involvement and adherence to treatment. An AF program must be effectively coordinated for the implementation of the management plan. The last and the most critical aspect is the evaluation of care plan execution [5]. The cornerstones of multidisciplinary atrial fibrillation program are shown in (Figure 1).

Figure 1: “Cornerstones of multidisciplinary atrial fibrillation program” [5].

This systematic review and meta-analysis were planned to access the effectiveness of multidisciplinary integrated care in the clinical outcomes of atrial fibrillation patients including all-cause mortality, MACE, cardiovascular deaths, cardiovascular &AFrelated hospitalizations, cerebrovascular episodes and major & minor bleeding. Evidence has established the need for integrated systems of healthcare delivery in various chronic cardiovascular afflictions, like heart failure and acute coronary syndrome, there is hardly any for the AF population. Lately, research papers have been probing the need for integrated healthcare in atrial fibrillation patients for further improvement and insight in the management of atrial fibrillation.

Objective

Atrial fibrillation (AF) is linked with substantial morbidity and mortality. Cardiovascular illnesses of chronic settings have shown improved patient specific outcomes from integrated structures of healthcare, the practice of this very methodology in atrial fibrillation patients is a relatively new idea. Current data has recommended that the integrated healthcare methodology may be of advantage in the AF population, still not broadly realized in routine medical management (Figure 2).

Figure 2: “The integrated care approach in atrial fibrillation (AF)” [33].

The aim of the systematic review and meta-analysis was to assess the effect of multidisciplinary integrated care systems of healthcare provision in the AF patient population on outcomes including all-cause mortality, MACE, cardiovascular deaths, cardiovascular & AF-related hospitalizations, cerebrovascular episodes and major & minor bleeding.

Material and Methods

Preferred Reporting Items for Systematic Reviews and Metaanalyses (PRISMA) guidelines were adhered for this systematic review and meta-analysis. The study is registered with PROSPERO International Prospective Register of Systematic Reviews (PROSPERO registration number CRD42018110613).

Study design

Systematic review and meta-analysis

Eligibility criteria

Studies were selected as per the following criteria:
Studies included were randomized controlled trials (RCTs), including cluster RCTs.
We excluded controlled (non-randomized) clinical trials (CCTs) or cluster trials, cross-sectional studies, case series, case reports, interrupted time series (ITS) studies, prospective and retrospective comparative cohort studies, and case-control or nested casecontrol studies.

Participants

Studies including the general adult human population (18 years or older) of either gender diagnosed with and treated for Atrial Fibrillation; defined as “recurrent paroxysmal, persistent, or permanent atrial fibrillation”.

Interventions

Multidisciplinary integrated healthcare intervention, emphasis on all-inclusive and far-reaching AF treatment and management with at least a 6-month follow-up period. Intervention is defined as “a coordinated patient-centered approach by interdisciplinary specialists to improve AF outcomes” [15].

“Integrated care facilitates treatment of AF patient’s population in all five spheres of management: acute stabilization, detection and management of primary cardiovascular co-morbidities and risk dynamics, apposite oral anticoagulation for stroke prevention, and treatment with rate and/or rhythm control therapy” [11,16] (Figure 3).

Figure 3: “Integrated AF care and AF heart team” [15].

Comparators

The patients in the comparator group were treated according to the usual standards of clinical care, regular medical care, outpatient management and pharmacotherapy. The management was Not Integrated or Multidisciplinary.

Outcomes

The major outcomes/endpoints for meta-analysis were MACE; major adverse cardiovascular event(s) and/or NACE; net adverse clinical event(s)
a. All-cause mortality
b. Cardiovascular mortality
c. AF- related hospitalization
d. Cardiovascular hospitalization
e. MACE
f. Major Bleeding & Minor Bleeding
g. Cerebrovascular accidents(stroke)

Timing

Studies inclusion was based on the follow-up of outcomes. For all endpoint outcomes, RCTs had a follow-up time of at least 06 months with further analysis at follow-up of 01 year.

Setting

Any kind of setting with no restrictions

Language

Included RCTs reported in the English language only

Information sources

Medical Literature search strategies were devised using medical subject headings (MeSH) and text words linked to atrial fibrillation. The articles were searched on MEDLINE (PubMed and OVID interface, 1948 onwards till 2017), EMBASE (OVID interface, 1974 onwards till 2017), and the Cochrane Central Register of Controlled Trials (Wiley interface). The scientific literature search was limited to the English language and human subjects.

Search strategy

Medline, EMBASE, and the CENTRAL trials registry of the Cochrane Collaboration were searched for keywords, including ‘atrial fibrillation’, ‘integrated healthcare’, ‘multidisciplinary’, ‘outpatient’, ‘interdisciplinary’, ‘treatment outcome’, ‘treatment adherence, ‘death’, ‘mortality’, ‘fatal’, ‘hospitalization’, ‘hospital admissions’, ‘quality of life’ and ‘symptom burden’.

Data management, selection process and data collection

Investigators individually evaluated all pertinent articles to categorize studies meeting criteria for inclusion. Any inconsistencies were deliberated and a consensus verdict was obtained. The search results from each database were saved in EndNote X9 and duplicates were removed.

Data items

Data was extracted into a standard recording form (Microsoft EXCEL datasheet) which was initially tried to confirm clarity and uniformity between authors. This empowers the authors to evaluate the quality of studies and integrate the findings. Data identified and tabulated from appropriate articles included: author, data collection year, publication year, participants, gender, mean age, follow-up duration and outcomes studied.

Outcomes and prioritizations

Outcomes that were extracted from the selected studies included all-cause mortality, cardiovascular mortality, MACE, cerebrovascular accidents, major and minor bleeding.

Data synthesis

A quantitative synthesis of data and meta-analysis was conducted for selected studies with similar study design and sufficient available outcome parameters to perform a statistical analysis using the STATA/SE 15.0 statistical program (StataCorp, Texas, USA).

Assuming a variation of the true effect sizes in the selected studies - due to different study populations and different integrated care models (as obvious in many integrated care models), we chose to perform a random-effect meta-analysis model to gain the pooled estimates of effects. For the summary statistics, we used exclusively dichotomous data and results were expressed by using odds ratio (OR) with 95% confidence interval (CI). A two-tailed value of p<0.05 was considered statistically significant. The presence of publication bias was visually assessed using funnel plots of effect size against SE. Non-quantitative data was presented descriptively. To test statistical heterogeneity between the studies, the chi-square (χ2) test and the I² statistic were applied. If a high level of heterogeneity was detected (I² >= 50%), a narrative synthesis of studies will be appropriate (Figure 4).

Figure 4: PRISMA flow diagram.

Risk of bias individual studies

Cochrane Collaboration tool (The RoB 2.0 tool http://www.riskofbias.info/) was used to assess the risk of bias for every RCT, which screens: randomization process, deviation from intended intervention, missing outcome data and selective result reporting (Table 1).

Table 1: Risk of bias of studies included in the meta-analysis.

Results

The search generated an aggregate of 506 articles that were evaluated by title and abstract. Four eighty four articles were excluded as they did not fulfill the inclusion criteria. Nine articles were examined for full-text assessment with five fulfilling the criteria for inclusion in the meta-analysis. Primary and secondary outcomes, as shown in table 2 below, are tabulated for easy reference and the selection for qualitative and quantitative analysis marked as per the formal inclusion/exclusion criteria.
Key characteristics of the various studies screened, shortlisted and selected are tabulated in (Tables 2 and 3).
Parameters that were obtained from the studies included for meta-analysis were all-cause mortality, MACE, cardiovascular mortality, cardiovascular hospitalizations, AF-related hospitalizations, cerebrovascular events (stroke) and major & minor bleeding episodes.

Table 2: Characteristics of studies included in the systematic review and meta-analysis.

Table 3: Parameters of included studies.

All-cause mortality

Data of all-cause mortality was provided in five studies included in the meta-analysis with an acceptable heterogeneity among the studies (I² =45.5%, p=0.119) which showed a significant 35% reduction of all-cause mortality in the integrated care model (OR 0.65, 95% CI 0.44 to 0.95; p=0.028) (Figure 5).

Figure 5: Integrated care outcome on all-cause mortality.

Major adverse cardiovascular event (MACE)

Five studies reported on this outcome. There was a significant reduction of MACE in favor of the integrated care approach (OR 1.39, 95% CI 1.19-1.62, p-value 0.000; I-squared > 0.05). No evidence of significant heterogeneity (I²=32%, p=0.208) (Figure 6).

Cardiovascular mortality

Two studies reported on this outcome, favoured integrated care (OR 0.49, 95%CI 0.21 to 1.17, p=0.109). But the result was not statistically significant as only 2 studies reported this outcome. Significant heterogeneity not established (I²=47.1%, p=0.169) (Figure 7).

Figure 6: Impact of integrated care on MACE.

Figure 7: Integrated care on cardiovascular death.

AF-related hospitalization

Three studies reported about AF-related hospitalization. AFrelated hospitalization had not been significantly influenced by the integrated care model, however, there was a trend favoring the integrated care approach (OR 1.36, 95% CI 0.99 to 1.87; p=0.056), no evidence for heterogeneity could be observed (I² =0.0%, p=0.414) (Figure 8).

Cardiovascular hospitalization

Three studies reported cardiovascular hospitalization. The integrated care approach led to a significant 34% reduction of CV hospitalization compared to standard care (OR 0.66, 95% CI 0.49 to 0.89; p=0.007), see Figure 4. No heterogeneity between studies was observed (I² =0.0%, p=0.966) (Figure 9).

Figure 8: Integrated care on AF-related hospitalization.

Figure 9: Integrated care on cardiovascular hospitalization.

Major bleeding

Four studies reported on this outcome didn’t favour integrated care or the standard care (OR 1.02, 95%CI 0.59 to 1.75, p=0.945). No significant heterogeneity (I²=0%, p=0.610) (Figure 10).

Minor bleeding

Four studies reported on this outcome, didn’t favour integrated care or standard care (OR 0.90, 95%CI0.66 to 1.24, p=0.52). Evidence of significant heterogeneity not established (I²=0%, p=0.691) (Figure 11).

Figure 10: Integrated care on major bleeding.

Figure 11: Integrated care on minor bleeding.

Cerebrovascular events (stroke)

Five studies reported on this outcome. There was no significant evidence in favour of integrated care (OR 1.38, 95%CI 0.84 to 2.27, p=0.20) and there was no evidence of significant heterogeneity (I²=0%, p=0.573) (Figure 12).

Multidisciplinary integrated care versus educational integrated care

Studies reporting on this outcome, didn’t favour significantly multidisciplinary integrated care over educational integrated care (OR 1.38, 95%CI 1.22 to 1.56, Chi-Square p<0.05). No heterogeneity (I²=0.0%, p=0.832) (Figure 13).

MACE according to different studies characteristics

Studies reporting on this outcome; parameters used were age, heart failure population, diabetes, use of oral anticoagulants; favoured integrated care (OR 4.00, 95%CI 2.23 to 7.17, p=0.015). There was obvious heterogeneity due to subgroup analysis (I²=97.6%, p=0.000) (Figure 14).

Figure 12: Integrated care on stroke.

Figure 13: Multidisciplinary integrated care versus educational integrated care.

Figure 14: MACE According to Different Studies Characteristics.

Discussion

The meta-analysis of research studies estimating the importance of a multidisciplinary integrated care management in atrial fibrillation patients establishes that this intervention forms the basis of:
a) Enhanced patient health consequences, resulting from a decrease in all-cause mortality, cardiovascular mortality & hospitalization, AF-related hospitalization and MACE
b) Insignificant influence of integrated care on the incidence of major or minor bleeding and cerebrovascular accidents
c) Educational and multidisciplinary integrated approach had similar outcomes
d) There are subpopulations (older AF patients with heart failure and diabetes) who can benefit from an integrated approach.

These outcomes have meaningful inference on integrated healthcare delivery in atrial fibrillation patients and advocate its practice as an applicable and effectual intervention. Nevertheless, several queries remain unrequited and additional exploration is requisite to report how provision of this methodology is applied in the best possible manner and approach.

Integration and collaboration between outpatient & inpatient departments and physicians play a vital role in the provision of excellent patient care [17]. The multidisciplinary approach has become increasingly important in the last few decades as a result of advancements in diagnostic and therapeutic techniques [18]. The purpose of integrated care is to strengthen the coordination in healthcare organizations, enhance the patients’ experience, clinical outcomes and upgrade the efficiency of health systems [19].

A meta-analysis was done by Gallagher et al. to evaluate the role of integrated care in atrial fibrillation patients. The inclusion criteria were randomized and non-randomized studies on integrated care in AF patients with the control group and a minimum follow-up of 6 months. Three studies conducted by Hendriks et al., Stewart et al. and Carter et al. were included. The outcomes assessed were allcause mortality, cardiovascular and AF-related hospitalizations and cerebrovascular events. The results of the meta-analysis showed that there was 49% and 42% decrease in all-cause mortality and cardiovascular hospitalizations, respectively with integrated care. However, the integrated care had no effect on AF-related hospitalizations and cerebrovascular episodes. Further analysis revealed that 19 patients need to be treated by the integrated approach to prevent one death. With aspect to hospitalizations, the treatment of 18 patients will decrease one hospital admission [20].

In a randomized controlled trial conducted in the Netherlands, 712 patients with atrial fibrillation were randomly allocated to the usual care and nurse-led care group. Each group had 356 patients. The follow-up period was a minimum 1 year. The patients with AF of age ≥18 years were eligible. Cardiovascular hospitalizations occurred in 48(13.5%) and 68(19.1%) patients in the nurse-led and usual care group, respectively. The death occurred in 4(1.1%) patients receiving integrated care in contrast to 14(3.9%) patients receiving usual care. In the usual care group, 3 patients died from a stroke whereas none of the patients died from stroke in the nurseled group [12].

A multicenter randomized controlled trial was conducted in Australia in which 335 patients were enrolled. The minimum follow-up period was 24 months with 168 patients in the SAFETY intervention group and 167 in the standard management group. 264(79%) participants had an unplanned admission or death, out of which 127 were in the SAFETY group and 137 in the standard group. 19(11%) patients died in the SAFETY group and 30(18%) in the standard group. The median event-free survival was greater in the SAFETY group as compared to the standard group. The length of hospital stay in each admission was less in the SAFETY participants (2.8 days) in contrast to standard management patients (3.6 days). There was no statistical difference for readmissions for atrial fibrillation (54[32%] in SAFETY group versus 57[34%] in standard group). The cerebrovascular events were reported in 8(5%) patients in the SAFETY group and 6(4%) patients in the standard group [22].

A before-and-after study done in Nova Scotia included 413 patients with atrial fibrillation with age ≥18 years. They included 185 patients in the AF clinic group and 228 patients were enrolled in the usual care group. The follow-up time was 21.5 months in the AF clinic and 28 months in the usual care group. Death was reported in 4(1.8%) patients in the usual care group whereas no patient died in the AF clinic group. 20(8.8%) patients had a hospital admission for a cardiovascular cause in the usual care group and 11(5.9%) in the AF clinic group. 25(13.5%) and 54(23.7%) patients had AF-related emergency (ED) visits in the AF clinic group and usual care group, respectively. Lower rates of stroke and bleeding were seen in both groups and the difference was not statistically significant. Stroke occurred in 4(2.2%) patients in the AF clinic group and 8(3.5%) patients in the usual care group. A greater percentage of patients were prescribed oral anticoagulants in the AF clinic group (57.7%) as compared to the usual care group(39.3%) [16].

In a randomized controlled study by Fuenzalida et al., 240 atrial fibrillation patients with age ≥18 years were recruited. One hundred sixteen patients were included in the intervention group and 124 were enrolled in the control group. The follow-up period was 1 year. The death occurred in 18(10.43%) and 26(20.96%) patients in the intervention and control group, respectively. A total of 116 patients had ED visits with 51(43.96%) patients in the intervention group and 65(51.61%) patients in the control group. Seventy four patients had hospital admissions with 31(26.72%) in the intervention group and 43(34.67%) in the control group. The most common complication was heart failure occurring in 19(16.37%) and 33(26.61%) patients in the intervention and control group, respectively. The difference in other complications between the two groups was not significant [23].

A cluster-randomized, controlled trial was conducted by Vinereanu et al. in which 48 clusters were enrolled from 5 countries. The follow-up period was 12 months. The patients with atrial fibrillation who were ≥18 years and had an indication of oral anticoagulation were included. The intervention and control group included 1184 and 1092 patients, respectively. In the intervention group, oral anticoagulant use increased from 804(68%) patients at baseline to 943(80%) patients whereas, in the control group, the oral anticoagulant use increased from 703(64%) patients to 732(67%) patients after 1 year. The death occurred in 5% of the patients in both the intervention and control group. There was a marked decrease in stroke in the intervention group (11%) in contrast to the control group (21%). There was no difference in other secondary outcomes between the two groups [24].

In another randomized study, 31 patients were randomly allocated to the intervention group and 34 patients in the usual care group. The follow-up period was 5 years. The patients in the intervention group had a longer event-free survival of 34 months as compared to 17 months in the control group. The patients in the intervention group had fewer hospital readmissions (1.9%) as compared to 2.5% in the usual care group. Similarly, the associated hospital stay was 16.3% in the intervention group in contrast to 20.3% in the usual care group [25].

In the ATHERO-AF cohort study, Pastoriet al. evaluated the risk reduction in cardiovascular events by implementing the Atrial fibrillation Better Care (ABC) pathway. The ABC pathway is an integrated care management program and includes prevention of stroke with anticoagulation, better management of symptoms and cardiovascular risk management. The study recruited 907consecutive patients with AF using vitamin K antagonists. The follow-up duration was 36.9 months. The patients managed using the ABC pathway had a lower risk of cardiovascular events [28]. Pastoriet al. also estimated the effect of ABC pathway on healthcare costs showing significantly lower health-related costs [29]. A posthoc analysis of the atrial fibrillation follow-up investigation of rhythm management (AFFIRM) trial was done between the ABC and non-ABC group of patients. The analysis showed that the patients managed with the ABC pathway had a lower risk of all-cause mortality, cardiovascular mortality, major bleeding, stroke and first hospitalization [30]. Yoon et al. included 204,842 nonvalvular AF patients in Korea with a follow-up duration of 6.2 ± 3.5years.The patients were allocated into the ABC and non-ABC groups. The results reported lower rates of all-cause mortality, cardiovascular mortality, stroke, major bleeding, myocardial infarction and major bleeding in the ABC group in contrast to the non-ABC group [31]. The IMPACT-AF study was a randomized cluster trial conducted in Canada to assess the improvement in care and outcomes in atrial fibrillation patients using the electronic Clinical Decision Support System (CDSS). It showed that that the integrated management program run by CDSS can be more cost-efficient as compared to usual AF care [32].