The modified heart team protocol facilitated the revascularization decision-making quality in complex coronary artery disease

Abstract

OBJECTIVES

A lack of standardization in heart team implementation potentially leads to suboptimal decision-making quality, and we previously established a modified heart team protocol to improve the decision-making quality. The present trial was to validate the effect of the modified heart team implementation protocol on improving the decision-making quality versus the conventional protocol in complex coronary artery disease (CAD).

METHODS

Eligible interventional cardiologists, cardiac surgeons and non-interventional cardiologists were randomly allocated to the intervention or control arm and established 12 heart teams in each arm. The 12 heart teams in each arm were randomly divided into 6 pairs, and 480 historic cases with complex CAD into 6 sets of 80 cases. In each arm, each set of 80 cases was discussed independently by one pair of heart teams, with each case finally receiving two heart team decisions (‘pairwise decisions’). The intervention arm conducted heart team decision-making according to the previously established protocol and the control arm based on guideline recommendations. The primary outcome was the overall percent agreement of the inter-team pairwise decisions. Decision-making appropriateness was further analysed.

RESULTS

A total of 36 cardiac surgeons, 36 interventional cardiologists and 12 non-interventional cardiologists from 26 centres were enrolled. The overall percent agreement was significantly higher in the intervention arm than the control arm (72.1% vs 65.8%, P = 0.04; kappa 0.51 vs 0.37). Both team-level (19.4% vs 33.0%; P < 0.001) and specialist-level (interventional cardiologists, 19.8% vs 37.7%, P < 0.001; cardiac surgeons, 19.8% vs 28.7%, P < 0.001) inappropriateness rate of decision-making was significantly lower in the intervention arm than the control arm.

CONCLUSIONS

The modified heart team implementation protocol improved the decision-making quality and appropriateness compared with the guideline-based protocol.

Open in new tabDownload slide

INTRODUCTION

Heart Team approach has received a Class 1C/1B recommendation in European and American guidelines on myocardial revascularization in patients with complex coronary artery disease (CAD) to optimize the treatment strategies [1–5]. Clinical guidelines recommended a heart team, consisting of clinical/non-interventional cardiologists (NICs), interventional cardiologists (ICs) and cardiac surgeons (CSs), should take sufficient time to assess all available information of complex cases (hereinafter referred to as ‘the conventional protocol’) [1].

However, there were relatively limited evidence-based data on the heart team implementation in detail [6]. A lack of standardized heart team implementation protocol potentially led to suboptimal decision-making quality [7]. Prior efforts noted poor inter-specialist consistency (kappa 0.26), insufficient intra-team reproducibility (20–24% re-discussion decisions changed after 6 months) and moderate inter-team agreement (kappa 0.58) in heart team revascularization decision-making [8–11].

To solve the above problems and improve the revascularization decision-making quality, we conducted a sequential explanatory mixed method study to establish a modified heart team implementation protocol (hereinafter referred to as ‘the modified protocol’), including specialist selection, specialist training, team composition, team training and meeting process [12]. However, the practical effect of the modified protocol on decision-making quality remained unknown, and a randomized trial for validation was warranted. Therefore, we conducted the present randomized trial to assess the effect of the modified heart team protocol versus the conventional protocol on decision-making stability. Decision-making appropriateness were further assessed.

PATIENTS AND METHODS

Design

The current trial was an open-label, randomized, controlled, two-arm trial involving cardiac specialists (NICs, ICs and CSs) in 26 hospitals from China. The trial design and rationale was published previously [11]. The trial was approved by the ethics committee of Fuwai hospital (protocol no. 2019-1303) and registered on the ClinicalTrials.gov (NCT05039567). The trial was conducted in accordance with the Declaration of Helsinki and the Good Clinical Practice guideline. All participants provided written informed consent.

Participants and recruitment

The inclusion criteria for the cardiac specialists differed from specialties (NICs, ICs and CSs) and was required for specified operator volumes and experience (Supplementary Table S1) [1, 13, 14].

A total of 480 adult cases with stable CAD according to the National Cardiovascular Data Registry (NCDR) CathPCI criteria (stable angina, no or silent myocardial ischaemia) and angiographically confirmed three-vessel disease or left main disease were eligible for inclusion in the study (Supplementary Methods) [15]. Eligible cases were randomly selected from a multicentre registry of consecutive patients who underwent coronary angiography between August 2016 and August 2017 [16].

Randomization

Eligible specialties (NICs, ICs and CSs) were randomly assigned to ‘the modified arm’ or ‘the conventional arm’ (2:1 ratio for ICs and CSs; non-ICs were only added to the modified protocol group) to randomly establish 12 heart teams in each arm (Fig. 1). The 12 heart teams in each arm were then randomly divided into 6 pairs, and 480 historic cases with complex CAD into 6 sets of 80 cases. Each set of 80 cases were randomly assigned to one pair of heart teams for decision-making, with each case finally receiving two heart team decisions (‘pairwise decisions’).

Research staff was informed of the randomization and organized the allocated specialists to establish heart teams. Participating specialists were unaware of the randomization.

Intervention

The modified arm

The modified arm conducted heart team decision-making according to a previously published protocol, which covered the whole process of heart team meeting organization and implementation, including specialist selection, specialist training, team composition, team training and meeting process (Supplementary Fig. S1) [11]. A heart team organized by the modified protocol consisted of two ICs and two CSs, and the technical and administrative position were balanced. Before the formal meeting, a unified training was conducted to make consensus on main factors influencing revascularization decision-making. After the unified training, a pilot discussion following the formal meeting process was conducted for running-in. The formal meeting process was derived from the standardized meeting process widely used in previous studies (Supplementary Fig. S2) [17–19]. All specialists were required to choose from among five treatment categories (percutaneous coronary intervention (PCI), coronary artery bypass grafting (CABG), PCI/CABG equipoise, medical therapy or further testing) before (Round I) and after (Round II) the heart team discussion. The final treatment strategy was determined by majority decision [20]. All meetings were conducted online using a heart team meeting assistance system.

The conventional arm

The conventional arm conducted heart team according to the principles mentioned in guideline and expert consensus, which only included specialist selection, team composition and meeting process. Each team was consisted of an IC, a CS and a NIC. The formal meeting process also followed the standardized meeting process (Supplementary Fig. S2).

Data collection and measurements

We collected demographic and professional status of participating specialists. Case data were obtained from our registry database and reported previously [12]. All the clinical informations were obtained according to the NCDR CathPCI data definitions [15]. An independent angiographic core laboratory took responsibility for all angiogram images screening and risk score evaluation by using a computer-based automatic calculator (Supplementary Methods).

Outcomes and definition

The primary outcome was the overall percent agreement (OPA) of the inter-team pairwise decisions. The ‘pairwise decisions’ were generated according to the way mentioned above in the ‘Randomization’ section. The secondary outcomes included (i) Fleiss’s (more than two raters) and Cohen’s (two raters) kappa coefficients to evaluate inter-team, intra-team, inter-specialist, intra-specialist and inter-round (before-after discussion) agreement for treatment decisions; (ii) inappropriate decision rate adjudicated for appropriateness using the 2018 ESC/EACTS Guidelines on myocardial revascularization [1]. The ‘inappropriate decision’ referred to as the treatment that received IIb or III classes of recommendations. Analysis of inappropriateness did not need to separate groups into pairs.

Statistical analysis

Categorical variables were expressed as frequency and percentage. Continuous variables were expressed as mean ± standard deviation or median and interquartile range. Categorical variables were analysed with the likelihood ratio χ² test or Fisher’s exact test if more than 25% of the cells had an expected frequency smaller than 5. Continuous variables were computed with the two-sample t-test when data following a normal distribution or the Wilcoxon rank sum test for a non-normal distribution. Normal distributions were tested by Kolmogorov–Smirnov test; 95% confidence intervals were computed for all measurements. The inter-/intra-team, inter-/intra-specialist and inter-round agreement were assessed using OPA or Fleiss’s (more than 2 raters)/Cohen’s (2 raters) kappa coefficients whenever applicable. Kappa values ≤0 were considered as no agreement and 0.01–0.20 as none to slight, 0.21–0.40 as fair, 0.41–0.60 as moderate, 0.61–0.80 as substantial and 0.81–1.00 as almost perfect agreement.

The primary outcomes were analysed according to different cases stratified by gender, age, body mass index, left ventricular ejection fraction, diabetes, stroke history, SYNTAX stratification, left main (LM), three-vessel disease and SYNTAX II recommendations. The comparisons in these analyses were not powered for hypothesis testing and are descriptive in nature. All the analyses were performed at a significance level of two-sided 0.05. All tests were performed using SAS software, version 9.4 (SAS Institute, Cary, NC).

Sample size

Case sample size

The OPA of the conventional arm was estimated as 66.3% based on our pilot study (unpublished data). We speculated that the OPA of the modified arm was 76% according to intra-team reproducibility rate in previous literature [9, 10]. Using a 5% level of two-side significance and a confidence level with 90%, combined with assignment convenience, it was estimated that a total number of 480 ‘pairwise decisions’ for each arm would be necessary to meet the study acceptance criterion.

Heart team and specialist sample size

Considering the feasibility of implementation, we assigned 80 cases to each heart team. Thus, 480 cases were divided into 6 sets of 80 cases, and 6 pairs of heart teams in each arm were needed. In total, 36 CSs, 36 ICs, and 12 NICs were needed.

RESULTS

Characteristics of specialists and cases

During January 2022 and October 2022, 125 heart team meetings were conducted (Fig. 1). The characteristics of ICs and CSs were similar in both arms (Table 1). The mean age of the 480 historic cases was 61.2 ± 9.0 years, and 363 (75.6%) were male (Table 2).

Decision-making agreement

The distribution of heart team decisions (PCI or equipoise, CABG and MT and further testing) were 57.4%, 34.5% and 7.6% in the conventional arm, and 42.5%, 49.5% and 5.3% in the modified arm. Decision distributions were different across different tertiles of SYNTAX score (Supplementary Table S2). The primary and secondary outcomes of decision-making were listed in Table 3. The OPA of inter-team pairwise decisions in the modified arm was significantly increased compared with the conventional arm (72.1% vs 65.8%; odds ratio 1.16, 95% confidence interval 1.01–1.34; P = 0.04). The inter-team agreement and intra-specialist agreement in the modified arm were moderate (Kappa 0.51 and 0.41–0.60), while that in the conventional arm were fair (Kappa 0.37 and 0.21–0.40) (Supplementary Tables S3 and S4).

Decision-making appropriateness

The inappropriateness rate of inter-team decision-making in the modified protocol was significantly lower than the conventional arm (19.4% vs 33.0%; odds ratio 0.59, 95% confidence interval 0.50–0.69; P < 0.001). The decision-making inappropriateness of both CSs (P < 0.001) and ICs (P < 0.001) significantly decreased in the modified arm (Table 3) and the decrease of inappropriateness after heart team discussion (Round I vs Round II) was larger in the modified arm than the conventional arm (Supplementary Tables S5 and S6).

Subgroup analysis

The OPA of heart team pairwise decisions were significantly higher in the modified arm, especially in patients >65 years, body mass index ≥30 kg/m², left ventricular ejection fraction ≤40%, with intermediate or higher risk of complexity, with three-vessel disease or the SYNTAX II score recommend CABG as the favourable treatment (Fig. 2).

Explain why the decision-making quality improved

We analysed the potential influence of different steps in the modified protocol on decision-making quality (Supplementary Table S7). Every step in the modified protocol played the anticipating role in improving the decision quality.

DISCUSSION

The present randomized controlled trial provided evidence that a modified heart team protocol improved the inter-team agreement of coronary revascularization decision-making and reduced both team-level and specialist-level decision inappropriateness, compared with the conventional protocol for patients with complex CAD.

Previous clinical guidelines and expert statements have recommended several basic principles of heart team implementation, including team composition and standardized meeting process [1, 2, 17–19]. The current guideline recommended at least 1 CS, 1 IC and 1 NIC in a team. During the standardized meeting process, an elaborate patient data sheet and angiography were first reviewed. Then team members conducted an open discussion, including evidence-based decision-making and technical feasibility of PCI and CABG. Finally, each specialist independently made decisions, and the final strategy was determined based on the majority. The present trial supported to modify the current guideline-recommended heart team implementation protocol in the following two aspects. First, appropriate training was recommended, including (i) reviewing state-of-art evidence; (ii) discussing decision thought, key variables understanding (i.e. age, left ventricular ejection fraction) and the latest PCI/CABG technique to finally make a consensus on decision-making; (iii) pilot heart team meeting. Second, the team composition was recommended to be modified as non-routine inclusion of NIC and at least two participants in both CS and IC.

The two above modifications played an anticipating role in improving the team decision-making quality. The present trial proved that the appropriate training and modified team composition may improve decision-making quality in the following ways. For the training, evidence review and multidisciplinary discussion helped to rectify the wrong perceptions of specialists and make revascularization decisions tend to be more evidence-based. Compared with the conventional arm, the decision inappropriateness decreased by more than 40% in the modified arm (Table 3). For the team composition, we excluded NIC who were considered as an unstable factor in team decision-making and added the number of IC and CS to form an intra-discipline discussion mechanism for improving decision stability (Supplementary Table S7). Our data demonstrated that the decision difference before and after the heart discussion was more likely to be observed in NIC (Kappa value 0.68, Supplementary Table S4). And the intra-discipline decision agreement was found higher in the modified arm (2 IC and 2 CS, Kappa value 0.50–0.54) compared with the conventional arm (1 IC and 1 CS, Kappa value 0.31–0.33). The potential influence of the above modifications was consistent with the data from our previous mixed-method exploratory study [11]. And the present study further confirmed the effectiveness of the series of the modifications on decision-making quality.

The present study mainly aimed to point out the direction of the heart team implementation protocol to improve decision-making quality, rather than establishing a detailed heart team workflow. Although the present study provided evidence on adding appropriate training and modified team composition to the current conventional protocol, each detailed measure in the above two aspects might not be appropriate in real-world practice. Taking the team composition as an example, the training system of NIC was different among countries. Although the decision stability of NICs was fair (Kappa value 0.38) in our study, these results may not be promoted to other regions with more serious NIC training [21]. Thus, further studies were needed to refine the detailed measures under the two new principles (appropriate training and modified team composition) validated by our trial to further optimize the protocol for better real-world practice.

Limitations

The study had several limitations. First, we only involved the cardiac centres in China, and the sample size of discussed cases and participating specialists were not extensive enough to cover hospitals at all levels, which might diminish the generalizability of the modified protocol in global centres. However, it did not affect to reach the conclusion that the modified protocol improved the decision-making quality and appropriateness compared with the conventional protocol. International multicentre randomized controlled trials are expected in future design. Second, the cases discussed in this trial were retrospectively enrolled, and no clinical outcomes can be reported. Further trials are warranted to assess the feasibility and effectiveness of the modified protocols on clinical outcomes in real-world practice. Third, this is a specialist-level interventional trial, which might be hard to understand the trial design for readers. A better designed patient-level trial to assess the effect of the modified protocols is imperative.

CONCLUSION

A modified heart team implementation protocol improved the decision-making agreement and reduced the inappropriateness rate compared with the conventional protocol. The conventional heart team practice based on guideline recommendations needs to be refined and standardized.

SUPPLEMENTARY MATERIAL

Supplementary material is available at ICVTS online.

FUNDING

This work was supported by the grants from the Capital’s Funds for Health Improvement and Research (CFH; No. 2022-1-4031) and the Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences (CIFMS, 2021-I2M-1-063).

Conflict of interest: none declared.

ACKNOWLEDGEMENTS

The authors thank Drs. Shuo Yuan, Zhiwei Zeng, Xiaoting Su, Runchen Sun, Juntong Zeng, Xiaohong Huang, Boshizhang Peng and Lihua Xie from the Chinese Academy of Medical Sciences and Peking Union Medical College Fuwai hospital for their contribution to the collection of standardized case information and assessment of clinical scores. We also thank all the participating specialists from 26 sites who assisted with the trial processes (Supplementary Acknowledgement).

DATA AVAILABILITY

The data underlying this article cannot be shared publicly due to the privacy of individuals that participated in the study. The data will be shared on reasonable request to the corresponding author.

Author contributions

Hanping Ma: Conceptualization; Data curation; Formal analysis; Methodology; Software; Writing—original draft; Writing—review & editing. Shen Lin: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Writing—original draft; Writing—review & editing). Xi Li: Methodology; Writing—review & editing. Yang Wang: Methodology; Writing—review & editing. Weixian Yang: Writing—review & editing. Kefei Dou: Writing—review & editing. Sheng Liu: Writing—review & editing. Zhe Zheng: Conceptualization; Funding acquisition; Resources; Supervision; Writing—review & editing.

Reviewer information

Interdisciplinary CardioVascular and Thoracic Surgery thanks Giuseppe Biondi-Zoccai, Suvitesh Luthra, Haralabos Parissis and the other anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• BMI

  Body mass index

 
• CABG

  Coronary artery bypass grafting

 
• CAD

  Coronary artery disease

 
• CS

  Cardiac surgeon

 
• IC

  Interventional cardiologist

 
• NIC

  Non-interventional cardiologist

 
• OPA

  Overall percent agreement

 
• PCI

  Percutaneous coronary intervention

Author notes