Background

Cardiovascular disease (CVD) is the leading cause of death globally, with coronary heart disease (CHD) accounting for nearly one-fifth of total deaths [1]. About 10% of people who survive a myocardial infarction experience another acute cardiac event within the first year [2], most of which are largely preventable [3, 4]. Comprehensive cardiac rehabilitation (CR) including structured exercise, psychosocial care, and timely patient education focussed on support for a healthy lifestyle and medication adherence is a cornerstone of care for patients with CHD [5, 6]. Patients who complete CR reduce their risk of repeat myocardial infarction (RR 0.72, 95% CI 0.55–0.93) and all-cause hospital admissions (RR 0.58, 95% CI 0.43–0.77) [7] accompanied by clinically meaningful improvements in functional capacity and health-related quality of life (HRQL) [7, 8]. Despite these benefits and strong recommendation of CR for the secondary prevention of CHD in international guidelines (Class 1 A evidence) [9], program participation rates continue to be suboptimal [10,11,12].

Poor attendance and completion of CR programs mean that many eligible patients fail to achieve the potential benefits of these programs. For instance, studies have demonstrated a dose-response benefit between exercise sessions attended and reductions in all-cause mortality [13] and risk of major adverse cardiovascular events (MACE) [14]. To achieve these benefits, it is recommended that CR programs offer a sufficient dose of exercise and efforts should focus on enabling patients to adhere to all prescribed sessions. Despite this, CR participation rates remain poor worldwide, with average rates ranging from 26.9% [10] to 48% [11]. Barriers to CR participation have been identified and include a combination of factors, both organisational (such as lack of adequate space and equipment) [15], and patient (such as not prioritising CR attendance amongst other life demands) [16]. There is growing evidence that suggests the quality of the CR program likely influences patient participation and thus outcomes [17]. National audits using quality indicators, performance measures, and key performance indicators (KPIs) developed by international peak bodies provide a means to evaluate the implementation of guideline-recommended therapies [18,19,20,21]. For instance, an audit of 170 programs in the United Kingdom found only 30.6% were considered high-quality according to their pre-determined minimum standards for CR (defined as having multidisciplinary team, offered to priority patient groups, and of adequate duration) [22]. In this audit, 5.3% does not meet any of the minimum standards and thus unlikely to facilitate patient participation or optimise outcomes [22]. Improving CR program quality may therefore be one of the keys to addressing low patient participation and completion.

One way to improve patient participation and completion of CR is by drawing on a Collaborative Quality Improvement Methodology, where various stakeholders work together toward a common goal of improving performance using a defined set of quality measures [23]. This collaborative approach enables rapid and sustainable changes in service delivery and, thus, has the potential to leverage the collective power of CR programs working simultaneously on shared problems and using data to drive performance [23]. Collaborative quality improvement has been successfully used in general practice [24], as well as in the management of asthma [25], and chronic heart failure [26], but to our knowledge, has not been tested for the optimal delivery of CR. Although a strategy of using purposeful and planned quality improvement initiatives demonstrated significant increase in patient attendance and completion in CR [27], this study was single-centre and non-randomized. Therefore, more evidence on the benefits of these initiatives is needed using robust research designs.

The Quality Improvement in Cardiac Rehabilitation (QUICR) trial aims to determine whether the implementation of a 12-month data-driven collaborative approach, relative to usual care:

  1. 1)

    increases CR program completion of patients with CHD (primary outcome),

  2. 2)

    reduces unplanned hospital admissions, emergency department presentations and deaths, and associated costs, and.

  3. 3)

    improves the proportion of patients receiving guideline-indicated CR according to national and international benchmarks, and

Furthermore, QUICR will include a process evaluation that runs parallel to the trial. This process evaluation will investigate the implementation of the collaborative approach and explore whether it is feasible and sustainable for CR staff to implement routinely, and identify enablers and barriers to implementation.

Methods

This trial protocol follows the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) checklist [28] and QUICR will be conducted in accordance with the Consolidated Standards of Reporting Trials (CONSORT) extension for cRCTs [29].

Study Design

A type-2 hybrid effectiveness-implementation study [24], which uses a 2-arm, multi-centre, cluster-randomized controlled trial (cRCT) design will be conducted over 12 months (Fig. 1). The cRCT design enables accurate confirmatory causal inference [30]. The type-2 hybrid effectiveness-implementation study design, which has a dual focus on effectiveness and implementation outcomes, allows for simultaneous evaluation of the implementation intervention/strategy during an effectiveness trial [31].

Fig. 1
figure 1

The QUICR Trial flow diagram

Full size image

Study Setting/Recruitment

CR programs across two Australian states (New South Wales and Victoria) will be identified through our existing professional networks, societies, and available CR directories [32]. The research team will approach potential CR program leaders and provide information about the trial via email and/or phone call and during state-based in-person/online CR events. CR program leaders will be invited to ask questions and discuss any potential barriers to participation. Research team members will coordinate expressions of interest to participate, confirm eligibility, and commence formal recruitment processes.

Program Eligibility Criteria

CR programs will be eligible to participate if they:

  1. a)

    enrol at least 70 patients eligible for CR per year,

  2. b)

    have internet-enabled computers, and.

  3. c)

    have staff with access to and proficiency in using the REDCap electronic collection data platform.

CR programs will be excluded if they are:

  1. a)

    unwilling to provide written agreement to participate in the quality improvement intervention or.

  2. b)

    already participating in a structured research project that entails changes to their usual CR program delivery.

The patient cohort for QUICR will comprise a dataset of all consecutive patients enrolled in participating CR programs who are ≥ 18 years old with a documented diagnosis of CHD in the medical record.

Study outcomes

Table 1 outlines the QUICR KPIs. The primary outcome is CR program completion, defined as participation in ≥ 80% of sessions of the usual 6-8-week program. Secondary outcomes include: (1) unplanned hospital admissions, emergency department presentations and deaths, and associated costs via analysis of linked administrative data, (2) proportion of patients receiving guideline-indicated CR according to national and international benchmarks (e.g. assessment of risk factors, exercise capacity, medication adherence, and HRQL at entry and completion, and documented discharge transition plan), and (3) enablers and barriers to implementation of the QUICR intervention.

Table 1 QUICR Key Performance Indicators
Full size table

Randomization

Each CR program will be randomized 1:1 to intervention (QUICR) or control group, via a computer-generated sequence allocation using simple randomization. CR programs that closely interact with each other were treated as a single cluster to circumvent the likelihood of contamination. Allocation will be stratified 50:50 by location of the programs (urban and regional areas) using SAS v9.4.

Blinding and allocation concealment

An independent statistician will generate the randomisation schedule so that the trial statistician remains blinded to group allocation for analysis. Given the design and nature of the quality improvement intervention where CR program staff enter data and use these to improve care, it is not possible to conceal the group allocation from the program staff themselves or the research team delivering the intervention. However, where possible, data collection is conducted blinded to treatment allocation (e.g. data linkage) as well as the statistical analyses.

Intervention and control groups

Intervention Group

CR programs allocated to the intervention arm will participate in a multicomponent, data-driven quality improvement intervention (Table 2). The intervention is designed to use local CR data to drive small, progressive changes using Plan-Do-Study-Act (PDSA) cycles [33] aligned to pre-determined KPIs (Table 1). These indicators have been selected using national and international benchmarks of quality [18,19,20,21] and were developed and approved through an iterative process involving the trial investigators, CR clinicians, quality improvement experts, and consumers. The intervention will be guided by the Model for Improvement [34] using streamlined data entry and regular visual reports to programs via purpose-built databases using software systems REDCap and Microsoft PowerBI and underpinned by principles of collaboration and support. CR program leaders will collaborate in an ongoing way over 12 months supported and facilitated by the research team.

Table 2 Components of the QUICR collaborative quality improvement intervention
Full size table

Control Group

CR programs allocated in the control arm are to provide usual care to their patients. For the purpose of measuring change in primary and secondary outcomes of intervention sites relative to control sites, staff working in CR at control sites are required to enter patient data into the purpose-built REDCap data collection platform. They will not have access to any of the abovementioned QUICR resources. These materials will be provided to the control group following trial completion.

Fig. 2
figure 2

QUICR PDSA Diagram

Full size image

Data Collection and Management

All data will be collected at baseline, 12, and 24 months. The primary endpoint is 12 months. Individual patient-level data will be collected by CR program staff as part of routine care. De-identified demographic and CR patient outcomes data including costs will be entered into a purpose-built login-protected REDCap data collection platform hosted by the University of Sydney [35]. The master coding sheet with full patient identifiers will remain at each CR site’s server and will be the responsibility of the data custodian nominated at each program. Coding is crucial for this trial to enable access to and analyse administrative data for linkage. Data extracted to the QUICR Dashboard will be aggregated and automatically converted to visual form and no individual data point will be traceable. The QUICR Dashboard will also be login-protected, to which only a select number of the research team will have access.

Unplanned hospital admissions, emergency department presentations, and deaths will be collected via state-based linked administrative data. Probabilistic matching will be used to link records and the estimated proportion of invalid and missed links using data linkage is expected to be very low [36].

Sample size

A sample of 40 CR programs (20/arm) with 33 patients with CHD in each program (660 patients/arm), will achieve 80% power to detect an increase in the CR completion rate of 22% from 59%. The control proportion estimated is 59% from our Australian National Cardiac Rehabilitation Quality Benchmark study (n = 2,436; 39 programs) [37]. The estimated change in completion rate is based on US data (n = 1103) reporting 25% increase in completions from quality improvement interventions [27]. We anticipate a slightly reduced effect as we are not using their small financial incentives. The intraclass correlation is 0.29, based on a cross-sectional study of cardiac rehabilitation [27]. The significance level of the test is 0.05.

Statistical analyses

A study statistician blinded to the group allocation will lead all analyses according to a prespecified statistical analysis plan following intention-to-treat principles and consider CR program clustering. Descriptive statistics will be calculated and used to characterise patients, programs and KPIs. Mixed-effects linear and logistic regression models will be used for continuous and binary outcomes, respectively, accounting for clusters within the trial. Mixed-effects regression analysis adjusting for imbalances in patients at baseline will be conducted as sensitivity analysis. Prespecified subgroup analyses will be performed for gender, urban/regional CR sites, program size and duration, and private, public and community-funded programs. A detailed analysis plan will be developed and signed off prior to unblinding.

Ethics

This trial will adhere to the National Health and Medical Research Council (NHMRC) ethical guidelines for human research [38] and the Ottawa Statement on the Ethical Design and Conduct of Cluster Randomized Trials [39]. Ethical approval has been obtained from the Northern Sydney Local Health District Human Research Ethics Committee (HREC) (2023/ETH01093). A waiver of patient consent has been granted because requiring individual consent introduces potential selection bias. Furthermore, patients in this trial will not be contacted for follow up and data linkage outcomes will be ascertained via data extraction from a reliable clinical software following robust governance processes overseen by national and State-based linkage units. This trial is registered with the Australian New Zealand Clinical Trials Registry (ANZCTR) (ACTRN12623001239651, Registered 30th November 2023) (anzctr.org.au). Any modifications to the trial protocol will be reported to the relevant HREC and ANZCTR.

Economic evaluation of QUICR compared to Usual Care

Pertinent healthcare and other costs incurred in delivering the QUICR intervention will be identified by the research team and counted and valued using widely recognised methods [40, 41]. Key cost items include workforce, infrastructure and technology, equipment, and consumables; travel costs related to receiving CR and labour productivity costs to patients; and patient inpatient hospital admissions, emergency department presentations (and their respective stay duration and costs) during the intervention period. A specially built cost database will be developed by the research team to document the type of resources, quantity of each, and the value of these items. In addition, the trial will evaluate information on changes in healthcare utilisation and costs (hospital admissions and emergency department presentations) through data-linkage to State health administrative data. All these cost data along with effectiveness data such as CR completion, HRQL (EQ-5D-5 L) at program completion and reduction in unplanned hospital admissions will enable quantification of a broad range of costs and health outcomes for patients in the QUICR intervention and those in usual care at each site (and in total) as well as incorporating them in an assessment of cost-effectiveness.

Process evaluation

A process evaluation for QUICR will be conducted parallel to the cRCT (Fig. 1). We will follow the process evaluations for cluster-randomised trials of complex interventions framework by Grant (2013), and unpack processes involving clusters (recruitment, delivery, responses) and individuals including CR staff and patients, as well as maintenance, effectiveness, and unintended consequences [42]. We will develop a logic model [43] to evaluate the implementation of the QUICR trial in terms of fidelity [44], processes and mechanisms, contexts [45], and theory [46] that contribute to causal pathways to intended/observed impacts. We will also identify barriers and enablers to implementation of the quality improvement intervention using a mixed methods approach.

We will collect quantitative and qualitative data throughout the QUICR intervention period using four sources:

  1. 1.

    Pre–post surveys of CR staff knowledge, beliefs, and attitudes towards collaborative quality improvement that will examine the uptake, feasibility, acceptability, utility, and sustainability of QUICR;

  2. 2.

    Learning workshop attendance and quality improvement engagement records to understand response to the QUICR intervention;

  3. 3.

    PDSA cycle completion, quality, and application into programs that indicate how the quality improvement is implemented and maintained;

  4. 4.

    Focus group interviews of CR program staff representatives with detailed field notes (n ≈ 20) at the start and end of the QUICR intervention. Focus group interviews at the start will assess the requisites for participating in QUICR including motivation, time commitment, staff skills and capacity, as well as learning needs and expectations. Information gathered in this focus group interviews will guide the content and delivery of succeeding workshops. The end-of-trial focus group interview will explore overall experiences with the QUICR intervention package including perceived barriers and enablers to implementation, sustainability, and unintended consequences. Recommendations for future enhancements of the QUICR intervention will also be sought.

For the process evaluation, descriptive statistics will be used to summarise quantitative data for the process evaluation. Maximum variation purposive sampling will be sought for the focus group interview participants for age, sex, and profession and to ensure data richness and saturation of the qualitative data [47]. Thematic analysis will be conducted [48] and will use context-method-outcome configuration to understand contextual influences [49].

Conclusion

To the authors’ knowledge, this is the first RCT of this scale to evaluate the effectiveness of a CR program-level data-driven collaborative quality improvement intervention in CR aimed at improving completion in patients with CHD. Increasing the number of CR programs that provide best practice care promotes improved CR completion and patient outcomes at scale. The use of linked administrative data will enable evaluation of QUICR as a potentially efficient strategy for improving important secondary endpoints including reductions in hospital admissions, emergency department presentations, and deaths. This trial will also produce a robust health economic analysis and a process evaluation, specifically identifying barriers and enablers to the implementation of QUICR needed for optimisation of resources and sustainment beyond the trial completion.

Dissemination

At the conclusion of the trial, results will be presented to local, national, and international scientific conferences and manuscripts submitted to peer-reviewed international journals following a publication protocol.