Randomised controlled trial of standard outpatient cardiac rehabilitation versus personalised cardiac rehabilitation for secondary prevention for all in need (SPAN) after type 1 myocardial infarction

Abstract

Introduction

Cardiac rehabilitation after type 1 myocardial infarction (type 1 MI) is an effective but underutilised method of preventing new cardiac events. This study aims to determine whether a programme of personalised cardiac rehabilitation achieves better completion rates compared with a routine outpatient rehabilitation model post-type 1 MI. A secondary aim is to determine any differences in rates of major adverse cardiovascular events at 12 months.

Methods and analysis

Secondary prevention for all in need (SPAN) is a prospective multisite single-blind comparative effectiveness randomised trial of personalised versus standard outpatient rehabilitation. Five hundred and thirty-two patients with a recent type 1 MI diagnosis will be randomised to either a personalised model of rehabilitation which includes flexibility in the components of rehabilitation and the time and location of contacts, or the standard outpatient model. Exclusion criteria include conditions associated with higher clinical risk to the participant, which introduce confounding to the study or interfere with a participant’s ability to comply.

The primary study endpoint is the proportion of patients who complete the assigned rehabilitation model. Completion of rehabilitation is defined as participation in ≥80% of all scheduled sessions, in either treatment arm. Assuming a power of 90% and an overall type 1 error of 5%, we expect that 50% of those randomised to usual care will meet the primary endpoint compared with 65% of those on personalised rehabilitation. To detect this difference of 15%, we require a sample size of 478 participants (239 per arm). Accounting for a conservative attrition rate of 10%, the total sample size required is 532.

Ethics and dissemination

The trial received ethical approval from the Southern Adelaide Clinical Human Research Ethics Committee (ref. 2022/HRE00071) in September 2022, with an opt-out approach. The findings will be disseminated through peer-reviewed journals.

Trial registration number

ACTRN12622000316707.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This is a large-scale, randomised controlled trial comparing personalised cardiac rehabilitation with usual care.

• The trial utilises an opt-out approach, minimising bias in patient selection.

• Secondary endpoints include patient-reported quality of life.

• 12-months follow-up will allow an understanding of the effects of participation in a personalised cardiac rehabilitation programme.

• The generalisability may be limited given recruitment within Australia.

Introduction

Annually, an estimated 49 329 Australians aged 25 and over experience a non-fatal myocardial infarction (MI) at a rate of 135 events every day.¹ Notwithstanding advances in care, 15% of MI survivors have a new MI, stroke or die within 12 months, with 8.9% of these unexpectedly rehospitalised within a year.

Type 1 myocardial infarction (type 1 MI) national guidelines recommend adopting healthy behaviours, as well as initiation and adherence to pharmacotherapy for preventing new events.² Interventions are best initiated at discharge from hospital through prescription of dual antiplatelet therapy, statins, beta-blockers (when indicated) and a renin–angiotensin antagonist and referral to cardiac rehabilitation. A recent national audit of cardiac rehabilitation found delivery was predominately outpatient hospital-based group exercise, education, counselling and pharmacy services;³ a format essentially unchanged since last century.⁴

Despite the acknowledged benefits of cardiac rehabilitation in reducing new clinical events and bettering quality of life,⁵ participation rates remain low. Outpatient rehabilitation services averaging 11 sessions (range 1–36) over 7 weeks (range 1–12) are vastly underused with low referral, low enrolment, low completion rates and are associated with considerable disparities in access.⁶ Among the patient barriers to participation are transport difficulties, work and social commitments, lack of perceived need and functional impairments, calling into question the relevance of current established approaches.⁷ There have been repeated calls for more flexible, individualised prevention strategies to improve the acceptability, reach and completion of cardiac rehabilitation within the changing social profile of MI survivors.^{8 9}

This has seen the emergence of personalised secondary prevention models, mostly supplementary to or infrequently, instead of outpatient rehabilitation.^{9 10} Yet such innovations require investment. Importantly, robust evidence of effectiveness is needed to justify such a shift in costs and to ensure that personalised models deliver meaningful benefits to patients. Secondary prevention for all in need (SPAN) is a developed, published, flexible framework for providing a more personalised approach to cardiac rehabilitation.⁴ The aim of this study is to test whether an approach of personalised cardiac rehabilitation after type 1 MI results in better completion rates than standard outpatient cardiac rehabilitation. The hypothesis is that those randomised to personalised rehabilitation will have higher completion rates at 3 months compared with those randomised to the standard outpatient rehabilitation model.

Methods

Design

SPAN is a prospective multisite single-blind comparative effectiveness RCT of personalised rehabilitation versus outpatient rehabilitation evaluating the primary endpoint of rehabilitation completion at 3 months. Secondary endpoints include changes in quality of life at 3 months and 5-point major adverse cardiovascular events (MACE; defined as cardiovascular death, non-fatal MI, non-fatal stroke, new/worsening heart failure and unplanned coronary artery revascularisation) at 12 months. An overview of the trial is shown in figure 1.

Figure 1. SPAN trial overview. SPAN, secondary prevention for all in need.

Setting

Cardiology Departments in four Australian Metropolitan public hospitals; Royal Perth Hospital (WA), Flinders Medical Centre (SA), Concord Repatriation and General Hospital (NSW) and Royal Brisbane and Women’s Hospital (Qld).

Patient population

Target patients are aged 18 years or over with a hospital diagnosis of type 1 MI by the fourth Universal definition¹¹ and deemed to be suitable for cardiac rehabilitation. Exclusions have been kept to a minimum and consist of conditions that increase the risk to the participant and would introduce confounding to the study or interfere with a participant’s ability to comply. The full list of inclusion and exclusion criteria is listed in box 1. Patients will be enrolled using an opt-out approach. Eligible patients who do not actively opt out will be included in the trial. The opt-out form is included in the online supplemental file 1

Box 1. SPAN eligibility criteria.

Participant inclusion criteria

• Persons aged >18 years.

• Inpatient diagnosis of suspected ST-segment elevation MI (STEMI), or non-STEMI (NSTEMI) acute coronary syndrome by ACS Guidelines and fourth universal definition of MI; specifically, a rise and/or fall in troponin from at least two samples defined as follows:

  • Hs-Troponin T: a rise and/or fall in the absolute troponin level of 2.5 ng/L/hour (ie, a change of 15 ng/L in 6 hours) between any troponin results before randomisation or a relative change in troponin 20% between earlier and later samples (ie, 100*(hs−TnT [_later] − hs−TnT [_earlier]/hs−TnT [_earlier]) 20%) documented on any troponin results before randomisation.

  • Hs-Troponin I: an absolute elevation of >10 times the upper limit of normal specific for that assay, using gender-specific cut-points if implemented locally or change in troponin 20% between earlier and later samples (ie, 100*(hs−TnI [_later] − hs−TnI [_earlier]/hs−TnI [_earlier]) 20% documented on any troponin results before randomisation.

• Regarded by the treating team to be suitable for cardiac rehabilitation.

Participant exclusion criteria

May include conditions that increase the risk to the participant, that introduce confounding to the study or interfere with a participant’s ability to comply.

• Prior cardiac rehabilitation.

• Clinical diagnosis of uncompensated severe heart failure (NYHA class IV).

• Uncontrolled arrhythmia or angina.

• Severe or symptomatic aortic stenosis.

• Coexisting clinical diagnosis of non-cardiac condition that would rule out participation; for example, advanced dementia, severe rheumatoid arthritis, severe frailty, terminal illness.

Screening, baseline assessment and data collection

Potential participants for SPAN will be identified by their medical team and assessed for trial eligibility using prespecified inclusion/exclusion criteria (see box 1). They will be provided with an opt-out form to consider. Patients who are deemed ineligible for SPAN will not be randomised into SPAN and will instead receive usual care. A screening log at all centres of patients with type 1 MI will be maintained to identify trends in patients approached and screen failures. Individuals will not be identifiable from the screening log. Patients who are deemed to be eligible for the trial and who do not opt out will be randomised. The study will randomise 532 patients with type 1 MI to either allocated standard outpatient rehabilitation (n=266) or personalised rehabilitation (n=266). Participants randomised to the personalised rehabilitation arm may also elect to engage with the standard outpatient services offered by their hospital’s cardiac rehabilitation team.

There are well-developed mechanisms at all centres for identifying patients with type 1 MI and channelling them into cardiovascular studies. Participants will be identified from daily admissions to coronary care and general medical units.

Randomisation

There is a well-controlled randomisation for allocation to either standard outpatient rehabilitation or personalised rehabilitation. Randomisation is at the level of the individual. It will be performed centrally utilising a computer-generated random allocation sequence in a uniform 1:1 allocation ratio and will be concealed from the primary analyst throughout the life of the trial.

Intervention

Outpatient cardiac rehabilitation

Patients randomised into the outpatient rehabilitation arm will follow the standard programme offered at their participating site. Outpatient rehabilitation typically commences within 2–4 weeks of leaving hospital, providing on average 11 sessions (range 1–42) over 7 weeks (range 1–12);³ however, the start times, duration and services offered will be variable across participating sites. Outpatient rehabilitation typically incorporates disease education, generic lifestyle counselling, psychosocial support and supervised exercise. Outpatient rehabilitation services will be performed by usual hospital cardiac rehabilitation staff.

Personalised rehabilitation (includes access to full outpatient model where requested)

Trial sites will employ a specialised nurse to administer personalised rehabilitation for the trial.

Personalised rehabilitation is bookended by a face-to-face, telephone or video call induction contact (median 45 min, IQR 30–60 min) and a face-to-face/telephone/video call close-out contact (median duration 20 min; range 5–50 min). In between, participants will be able to receive optional choice of disease and lifestyle education and psychosocial support.

Participants will be offered modules targeting the four key risk factors of elevated blood pressure, elevated low-density lipoprotein cholesterol, physical inactivity and cigarette smoking, as applicable. They may choose to target any combination or all modules, depending on their individual circumstances.

At the induction contact, the participant and site personalised rehabilitation coordinator will review the participant’s medical history and lifestyle behaviours. They will then mutually agree with the participant on suitable modules to address and goals to set. The participant will assign a personal confidence rating for achieving each goal. The strategy for addressing each module is also flexible. In SPAN, strategies are categorised as medical, flexible, home-based or self-directed and range from primary supervision by their usual general practitioner (medical) to a totally self-directed approach (see table 1 for details of each strategy).

Table 1. Description of SPAN strategies for addressing selected risk factor modules.

SPAN strategy	Approach for addressing selected modules
Medical	Visiting usual doctor and following their advice (doctors fee may apply/Medicare)
Flexible	Taking part in free facilitated sessions to address risk factor online/in the local community/at hospital site
Home-based	Participating in a home-based programme to address risk factor supported by the site personalised rehabilitation co-ordinator
Self-directed	Managing own risk factors with or without support of personalised rehabilitation co-ordinator

SPAN strategies for addressing the selected modules of the personalised cardiac rehabilitation modules (lowering blood pressure, reducing LDL cholesterol levels, increasing activity levels and quitting smoking).

LDL, low-density lipoprotein; SPAN, secondary prevention for all in need.

Regardless of the approach chosen, participants can elect to engage in up to three 20 min phone calls with the personalised rehabilitation co-ordinator outside of induction and close-out contacts, to provide support and monitor progress. Additional ad-hoc contacts may also take place. The personalised rehabilitation co-ordinator may also provide assistance with sourcing and/or co-ordination of hospital, community or online services.

At the induction visit and throughout the rehabilitation process, participants will nominate specific contacts they plan to make to address the selected risk factors. Examples of this may include calling the Quitline, visiting a fitness facility or attending a scheduled education session. These planned contacts will be recorded on a rehabilitation plan. Throughout the rehabilitation process and at the close-out visit, details of the specific contacts undertaken will be recorded on a rehabilitation attendance log. By comparing the planned activities with those performed, a participant’s completion rate for their rehabilitation will be ascertained.

In the standard rehabilitation arm, data will be extracted from the local system used to record cardiac rehabilitation. Completion rates will be determined through attendance records based on the standard programme provided.

Patients will typically receive lipid-lowering and other evidence-based drugs (subject to contraindication and intolerance); however, this would be at the discretion of their treating clinician and independent of their trial intervention.

At baseline, participants in both treatment arms will complete the EuroQol 5 Dimension 5 Level (EQ-5D-5L) quality of life questionnaire and the Specific Activity Questionnaire (SAQ). The EQ-5D-5L is a reasonably valid, reliable and responsive quality of life measure for patients with acute coronary syndrome,¹² while the SAQ provides a moderately good measure of functional capacity in cardiac patients.¹³ Both questionnaires are low-burden. Participants will repeat these measures at the conclusion of the rehabilitation intervention. Both questionnaires are included in the onlinesupplemental file 2 (EQ-5D-5L) file 3 (Specific Activity Questionnaire). A timeline of intervention is provided in figure 2.

Figure 2. Timeline of interventions in SPAN. EQ-5D-5L, EuroQol 5 Dimension 5 Level; MACE, major adverse cardiovascular event.

Outcomes

The primary outcome is the proportion of patients completing at least 80% of all scheduled contacts for their assigned rehabilitation strategy at 3 months follow-up. In both treatment arms, this will be calculated from attendance records of planned activities, determined at induction visit or pre-outpatient rehabilitation contact.

Secondary outcomes are 5-point MACE (composite of non-fatal stroke, non-fatal MI, cardiovascular death, heart failure hospitalisation and coronary/peripheral artery revascularisation). These will be collected directly in-trial by the study trial coordinator at each site and thereafter indirectly from linkage of administrative health data collections at completion of 12 month follow-up for all participants.

Responses to the EQ-5D-5L and SAQ will also be used to calculate changes in quality of life and physical activity from baseline and compared across the two treatment arms.

Statistical considerations

The trial will be reported according to Consolidated Standards of Reporting Trials (CONSORT) guidelines.¹⁴ Summary statistics including means and SDs (or medians and IQRs where appropriate) or counts and percentages will be calculated for all baseline characteristics by treatment arm, as well as for the number of scheduled and attended contacts for each study arm.

The primary analysis will be based on the intention-to-treat principle. The primary outcome of meeting at least 80% participation of all scheduled sessions (yes/no) for personalised rehabilitation versus standard outpatient rehabilitation will be assessed using logistic regression with stratification by site.

For all time to event secondary endpoints, including the composite of readmission to hospital for MACE, treatment groups will be compared using a two-sided log-rank test with stratification by site and visualised using cumulative incidence curves. Subsequently, Cox proportional hazards regression models, with stratification by site, will be performed. HRs of personalised rehabilitation versus standard outpatient rehabilitation, with corresponding 95% CIs, will be calculated.

All other binary and continuous outcomes will be analysed using logistic regression and linear regression, respectively, with stratification by site. A modified intent-to-treat analysis will also be performed where only those who achieved a minimum exposure to outpatient rehabilitation, defined as attendance at one or more sessions after pre-outpatient rehabilitation contact and a minimum exposure to personalised rehabilitation, defined as one contact; that is, induction session, with the personalised co-ordinator will be included.

Two-tailed p-values<0.05 will be considered statistically significant; however, given the limited number of secondary endpoints and therefore the chance of increasing the family-wise error rate is low and they are exploratory in nature, we contend formal adjustments for multiple comparisons are not necessary. However, results will be interpreted with caution.

A detailed statistical analysis plan will be finalised prior to closure of the trial and before any data analysis begins. Once finalised, this will be made publicly available.

Sample size

The participating tertiary centres report aggregated annual admissions for MI, ranging from 720 to 950/hospital. All eligible participants will be identified by study personnel from daily admissions to coronary care and general medical units.

Study power and significance

Assuming a power of 90% and an overall type 1 error of 5%, we expect that 50% of those randomised to usual care will meet the primary endpoint compared with 65% of those on personalised rehabilitation. To detect this difference of 15%, we require a sample size of 478 participants (239 per arm). Accounting for a conservative attrition rate of 10%, the total sample size required is 532. The sample size calculation for the primary outcome was performed using PS: Power and Sample Size software (V.3.1.6).

As indicated in the protocol, the primary outcome will be tested at a 5% significance level; however, given the limited number of secondary endpoints and therefore the chance of increasing the family-wise error rate is low and they are exploratory in nature, we contend formal adjustments for multiple comparisons are not necessary. However, results will be interpreted with caution.

Trial oversight and governance

The SPAN trial is sponsored by the University of Western Australia (UWA). The trial received ethical approval from the Southern Adelaide Clinical Human Research Ethics Committee on 19 September 2022 (Project ID 2022/HRE00071).

Patient and public involvement and engagement

The trial was developed with extensive input from two community consumers with lived cardiac experience. Recommendations were implemented into the final trial design, such as the inclusion of patient-reported outcome measures, as well as the development of participant-facing documentation.

Trial progress

The first participant was randomised on 23 August 2023. As of the date of submission, there have been 253 participants randomised to the trial, across four sites. The trial is planned to continue to recruit until July 2026 and complete all cardiac rehabilitation follow-ups by 31 December 2026.

Discussion

The SPAN trial will address the established evidence-practice gap that 7/10 MI survivors in Australia are not accessing guideline-advocated secondary prevention measures.⁴ Barriers to improving reach, completion and outcomes are complex and well documented but can be addressed by studying more personalised ways of delivering prevention. This trial, where patients have an equal chance of receiving preventive treatment options, will evaluate the implementation of a published and flexible framework for improving secondary prevention among MI survivors.⁴

By taking a disruptive approach to innovation in this area of care, the SPAN strategy seeks to establish the effectiveness of moving to a personalised precision model of rehabilitation compared with the traditional model of outpatient rehabilitation after type 1 MI. Pilot work^{15 16} has demonstrated potential benefits in reach (uptake by >50% of non-outpatient rehabilitation attenders); high completion (93%), sustainability (68% at 4 years) and estimated cost-savings of $30 million annually in Australia alone, by avoiding potentially preventable hospital admissions and procedures through more personalised and simplified models of post-type 1 MI care.^{17 18}

The SPAN trial involves comparing personalised rehabilitation with standard outpatient, group-based rehabilitation. The SPAN framework offers a flexible, targeted and sustainable approach for risk factor control and disease management through shared decision-making and goal attainment. The trial will report on the differences between the two rehabilitation strategies in terms of completion defined as participation in ≥80% of scheduled contacts as established at the initial engagement, and subsequently all admissions to hospital and death identified from administrative records at 12 months.

SPAN will determine if there is an absolute increase of at least 15% in the proportion of MI survivors who complete personalised versus outpatient rehabilitation, that is, 65% versus 50% at 3 months follow-up, respectively. If successful, this personalised rehabilitation strategy could be rapidly and widely implemented and would be expected to reduce the risk of recurrent cardiovascular events and unplanned readmissions to hospital nation-wide.

Limitations and contingencies

Given the subjective nature of the primary effectiveness outcome, we have mitigated against detection bias and between group contamination by collection of comprehensive workflow records, trial monitoring and exclusion of patients with prior cardiac rehabilitation experience. The possibility of a cluster randomised trial to address the potential for between-group contamination was not pursued due to the larger sample size requirements for this approach and the modest number of participating sites in the SPAN trial.¹⁹

All available participant data, including from those who withdraw from the trial, will be used to derive the primary endpoint of participation in ≥80% of all scheduled sessions according to the intention-to-treat principle. Finally, should the proposed number of sites fail to recruit the required number of participants, a reserve list of potential sites will be engaged.

Ethics and dissemination

The trial received ethical approval from the Southern Adelaide Clinical Human Research Ethics Committee (ref. 2022/HRE00071) in September 2022, with an opt-out approach. The findings will be disseminated through peer-reviewed journals.