Improvement in VO 2peak predicts readmissions for cardiovascular disease and mortality in patients undergoing cardiac rehabilitation

Abstract

Background

Improvement in exercise capacity is a main goal of cardiac rehabilitation but the effects are often lost at long-term follow-up and thus also the benefits on prognosis. We assessed whether improvement in VO_2peak during a cardiac rehabilitation programme predicts long-term prognosis.

Methods and results

We performed a retrospective analysis of 1561 cardiac patients completing cardiac rehabilitation in 2011–2017 in Copenhagen. Mean age was 63.6 (11) years, 74% were male and 84% had coronary artery disease, 6% chronic heart failure and 10% heart valve replacement. The association between baseline VO_2peak and improvement after cardiac rehabilitation and being readmitted for cardiovascular disease and/or all-cause mortality was assessed with three different analyses: Cox regression for the combined outcome, for all-cause mortality and a multi-state model. During a median follow-up of 2.3 years, 167 readmissions for cardiovascular disease and 77 deaths occurred. In adjusted Cox regression there was a non-linear decreasing risk of the combined outcome with higher baseline VO_2peak and with improvement of VO_2peak after cardiac rehabilitation. A similar linear association was seen for all-cause mortality. Applying the multi-state model, baseline VO_2peak and change in VO_2peak were associated with risk of a cardiovascular disease readmission and with all-cause mortality but not with mortality in those having an intermediate readmission for cardiovascular disease.

Conclusion

VO_2peak as well as change in VO_2peak were highly predictive of future risk of readmissions for cardiovascular disease and all-cause mortality. The predictive value did not extend beyond the next admission for a cardiovascular event.

Keywords

Cardiac rehabilitation multi-state model secondary prevention exercise testing exercise capacity

Introduction

Cardiac rehabilitation increases exercise capacity (VO_2peak) and quality of life and reduces the risk of subsequent cardiovascular mortality and morbidity.^1–6 Exercise training is a core component of cardiac rehabilitation and improvement of VO_2peak is a common criterion to measure effect of a cardiac rehabilitation intervention.⁷ VO_2peak is likewise a strong individual predictor of mortality and morbidity.⁸

Improvement in VO_2peak after cardiac rehabilitation differs among patients⁹ and ethnic, psychosocial and physical characteristics such as age, sex and pre-existing fitness level have been shown to predict these differences.^9–11 Few studies have addressed whether difference in improvement of VO_2peak affects future risk of mortality. The prognostic effect of cardiac rehabilitation, in terms of future morbidity and mortality, has mainly been investigated comparing participation versus non-participation in cardiac rehabilitation.^3,4

Risk of readmissions for cardiovascular disease (CVD) and mortality (MACE) is commonly used as a composite endpoint for long-term prognosis after cardiac rehabilitation, as is risk of all-cause mortality.^12,13 Newer statistical methods, multi-state-models, have also made it possible to investigate terminal and non-terminal outcomes separately using a semi-competing risk scenario. Applying such methodology offers a better understanding of the disease process.

We investigated the prognostic value of VO_2peak and change in VO_2peak after cardiac rehabilitation on risk of MACE and all-cause mortality, and, in addition applied a multi-state model to investigate risk of readmissions for CVD and mortality to better understand how VO_2peak affects the progression of disease.¹⁴

Material and methods

Population

We investigated characteristics and clinical outcomes of cardiac patients diagnosed with coronary artery disease (myocardial infarction and patients undergoing percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG)), chronic heart failure or heart valve replacement, who were referred to and completed a cardiac rehabilitation programme between 1 January 2011 and 30 December 2017. The patients were identified using a local clinical database from the cardiac rehabilitation unit at Bispebjerg/Frederiksberg hospital.

Rehabilitation programme

Patients participated in a group-based eight-week outpatient exercise intervention supervised by a physiotherapist. The programme consisted of two weekly training sessions of 1.5 h with high intensity interval training (80% of VO_2peak) and resistance training, in accordance with international recommendations.^15,16 The training sessions were individually adapted to the patient. Strength training was performed as three sets of 15 repetitions for upper and lower extremities and high intensity interval training was performed on a bicycle ergometer for 30 min with various interval lengths. The training sessions was complemented with weekly group-based patient education on lifestyle risk factor management, psychological health and diet counselling. Patients were additionally encouraged to be physically active for at least 30 min per day in their leisure time.

The study was approved by the Danish Data Protection Agency (BFH-2017-117; 06028). According to Danish legislation no ethical approval was required.

Baseline and explanatory variables

VO_2peak was assessed 6–8 weeks post myocardial infarction or post surgery (PCI, CABG or heart valve surgery) before they initiated cardiac rehabilitation, and after cardiac rehabilitation using a cardiopulmonary exercise test (CPET) with a maximal symptom limited bicycle ergometer test (Via Sprint 150P, Ergoline, Germany). Breathing gases were collected and analysed breath-by-breath (Jaeger, Master Screen CPX version 5.21, Cardinal Health, Germany). Each test aimed at physical exhaustion and a respiratory exchange ratio of more than 1.1 to ensure the validity of the CPET.¹⁷ The CPET was performed at baseline and after cardiac rehabilitation. Patient characteristics and explanatory variables were assessed and entered in the local database as the patients entered cardiac rehabilitation.

Primary endpoints

With the use of a unique personal identification number the patients were linked to the Danish national registries, which are continuously updated on hospital admission, death and emigration.

Readmissions for CVD were defined as a composite of myocardial infarction (I21), unstable angina pectoris (I20), heart failure (I50) and stroke (I63) and ascertained through linkage with The Danish National Patient Register.¹⁸ From The Danish Civil Registration System¹⁹ information on all-cause mortality and emigration was ascertained. We defined MACE as the combination of hospital admission for CVD and all-cause mortality.

Potential confounders

The following factors were considered potential confounders: age, sex, working status (employed, unemployed, retired, and being on disability pension), educational attainment (no education (basic school), short-term education (e.g. secondary education or vocational education), medium education (e.g. Bachelor degree), and higher education (Master’s degree or higher), index diagnosis, medication, tobacco use, chronic obstructive pulmonary disorder, diabetes, kidney disease and peripheral artery disease.

Statistical analysis

Differences in clinical characteristics were compared with Chi-square test (and Yates correction with small samples) for categorical variables, and ANOVA for continuous variables when normally distributed and Mann–Whitney when non-normally distributed.

In the basic analysis we used Cox proportional hazards models to determine the associations between VO_2peak before cardiac rehabilitation as well as change after cardiac rehabilitation, both for risk of MACE and risk of mortality. All models were adjusted for potential confounding factors described above. Survival time was calculated from the date of final VO_2peak test to first event or end of follow-up, whichever came first.

Risk of readmission and death was also analysed using a semi-competing risk model, the multi-state model.²⁰ In the multi-state model, patients were not censored if they were readmitted for CVD. When applying the multi-state model, all patients completing cardiac rehabilitation were at risk of being readmitted for CVD (Figure 1, transition 1). If readmitted for CVD, patients stayed in the cohort and stayed at risk of dying (transition 2). The patients not admitted for CVD were similarly at risk of dying (transition 3).

A two-tailed p value < 0.05 was considered as statistically significant. Statistical analyses were carried out using the free statistical software R.²¹ R package mgcv was used to perform the fitting of the Cox proportional hazards models²² including possible non-linear effects of the continuous covariates (i.e. age, baseline VO_2peak, change in VO_2peak), while survival²³ and smoothHR²⁴ packages were used for graphics. BayesX²⁵ was used to fit the multi-state survival models.

Results

Study population

Between 2011 and 2017, 1561 patients participated in and completed the outpatient cardiac rehabilitation programme. Mean age was 63.6 years,¹¹ 1175 (74%) were male, and 1271 (84%) had coronary artery disease, 91 (6%) chronic heart failure and 170 (10%) heart valve replacement. Distribution of baseline characteristics according to readmissions and mortality are given in Table 1.

Table 1.

Demographic characteristics of each subgroup of the population: no events, death and readmission.

Variables	Level	No events	Death	Readmission
Variables	Level	n = 1342	n = 52	n = 167	p-value
Age, years	Mean (SD)	63 (11)	72 (9)	65 (11)	<0.001
Sex	Males	1002 (75%)	39 (75%)	134 (80%)	0.289
BMI	Mean (SD)	27.5 (5)	27.5 (5)	27.8 (4)	0.513
Co-habitant	Yes	809 (60%)	24 (46%)	89 (53%)	0.002
Ethnicity	Western European	1126 (84%)	46 (88%)	142 (85%)	0.827
Index diagnosis	ACS	764 (59%)	30 (59%)	96 (59%)	0.219
	Stable CAD	325 (25%)	11 (22%)	45 (27%)
	CHF	72 (6%)	5 (10%)	14 (9%)
	Heart valve replacement	156 (11%)	5 (10%)	9 (5%)
Revascularization	PCI	767 (59%)	35 (70%)	113 (70%)	0.062
	CABG	296 (23%)	8 (16%)	29 (18%)
	None	227 (18%)	7 (14%)	19 (12%)
Working status	Employed	460 (35%)	3 (6%)	39 (6%)	<0.001
	Unemployed	74 (6%)	1 (2 %)	12 (2%)
	Retired	694 (53%)	46 (92%)	100 (63%)
	Disability	79 (6%)	0 (0%)	7 (4%)
Educational attainment	No education	187 (16%)	7 (18%)	32 (24%)	<0.001
	Short term	502 (37%)	21 (55%)	60 (44%)
	Medium	195 (16%)	5 (13%)	15 (11%)
	Higher	303 (26%)	5 (13%)	29 (21%)
Smoking status	Non-smoker	835 (62%)	27 (52%)	101 (60%)	0.109
	Previous smoker	274 (20%)	11 (21%)	31 (19%)
	Current smoker	182 (14%)	8 (15%)	24 (14%)
Diabetes	Yes	205 (16%)	12 (24%)	33 (21%)	0.08
COPD	Yes	120 (9%)	9 (18%)	20 (13%)	0.06
PAD	Yes	79 (6%)	9 (19%)	18 (12%)	<0.001
Kidney disease	Yes	66 (5%)	8 (19%)	9 (6%)	<0.001
Beta blockers	Yes	976 (77%)	31 (74%)	123 (82%)	0.290
Statins	Yes	1103 (87%)	35 (81%)	138 (92%)	0.009
VO_2peak baseline	Median (IQR)	21.7 (6.8)	17.8 (4.6)	20.8 (7.1)	<0.001
Change in VO_2peak	Median (IQR)	2.6 (3.9)	1.2 (4.9)	0.5 (3.1)	<0.001

BMI: body mass index; ACS: acute coronary syndrome; CAD: coronary artery disease; CHF: chronic heart failure; PCI: percutaneous coronary intervention; CABG: coronary artery bypass grafting; COPD: chronic obstructive pulmonary disease; PAD: peripheral artery disease

Median follow-up time was 2.3 years (interquartile range 1.3–4.1 years). There were 167 readmissions for CVD and 77 deaths, of which 25 occurred after suffering a new hospital admission.

The patients who died, whether with or without readmission, were older, lived alone, were more likely to be retired, had lower educational attainment, more co-morbidity and had lower VO_2peak than patients who survived throughout the follow-up.

Influence of VO_2peak on MACE

Baseline VO_2peak was highly predictive of suffering MACE after cardiac rehabilitation (see Figure 2(a) with a linear association which remained statistically significant after full adjustment. The hazard ratio for MACE was 0.94 (0.91–0.97) for each ml/kg per min change in baseline VO_2peak. Change in VO_2peak showed a highly significant curvilinear association with risk of MACE (Figure 2(b)). Results of the regression analyses are given in Table 2. Males were at higher risk than females (hazard ratio 1.98, p = 0.003). CHF, having been revascularized with PCI and suffering from peripheral artery disease were also associated with increased risk of MACE (see Table 2).

Table 2.

Hazard ratios (and 95% confidence intervals) from Cox-regression models for MACE and death from all causes.

	Risk of MACE		Risk of death
Variables	HR	95% CI	HR	95% CI
Age, per year	1.01	(0.99–1.03)	1.06**	(1.02–1.10)
Sex, male	1.98**	(1.25–3.14)	2.19	(0.92–5.22)
Index diagnosis
ACS	ref		ref
Stable CAD	0.97	(0.68–1.39)	1.10	(0.56–2.16)
Heart valve disease	1.56	(0.57–4.23)	2.19	(0.38–12.68)
CHF	2.71*	(1.13–6.49)	5.39*	(1.37–21.15)
Revascularization
PCI	Ref.		Ref.
CABG	0.59*	(0.38–0.90)	0.68	(0.31–1.49)
None	0.32**	(0.14–0.72)	0.57	(0.16–2.06)
PAD	1.98**	(1.22–3.22)	2.23*	(1.04–4.76)
Baseline VO_2peak, per ml/kg per min	0.94***	(0.91–0.97)	0.89**	(0.83–0.95)
Change in VO_2peak, per ml/kg per min	0.79***^,a	(0.70–0.90)	0.87***	(0.80–0.96)

p < 0.05,

p < 0.01

***

p < 0.001

Hazard ratio for the linear effect between −2 and 2 ml/kg per min change in VO_2peak.

MACE: major adverse cardiovascular event; HR: hazard ratio; CI: confidence interval; ACS: acute coronary syndrome; CAD: coronary artery disease; CHF: chronic heart failure; PCI: percutaneous coronary intervention; Ref.: reference; CABG: coronary artery bypass grafting; PAD: peripheral artery disease

Influence of VO_2peak on all-cause mortality

Baseline VO_2peak was highly predictive of future risk of mortality with a linear association, hazard ratio 0.89 (0.83–0.95) per ml/kg per min, as was change in VO_2peak, hazard ratio 0.87 (0.80–0.96) (Figure 2(c) and (d)). Having CHF and peripheral artery disease was also associated with higher mortality risk but the association between VO_2peak and outcome remained largely unaffected by adjustment (Table 2).

Multi-state model: influence of VO_2peak on readmissions for CVD and/or all-cause mortality

Transition 1: risk of being readmitted for CVD

The number of patients readmitted for CVD was 167. There was a significant non-linear association between baseline VO_2peak and a linear association between change in VO_2peak and readmission for the range of change in VO_2peak from −2 ml/kg per min to 2 ml/kg per min. Here the hazard ratio was 0.76 per ml/kg per min change, p < 0.001. The hazard did not increase nor decrease significantly with greater changes of VO_2peak. See Figure 3(a) and (b), transition 1.

Transition 2: risk of dying after a readmission for CVD

Twenty-five patients died after a readmission. The risk of dying after readmission was not associated with baseline VO_2peak or change in VO_2peak (Figure 3(c) and (d)). Having CHF at baseline was the only significant predictor of mortality in this transition, although statistical power was limited.

Transition 3: risk of dying without readmission for CVD

Fifty-two patients died without an intermittent hospital admission. Both baseline and change in VO_2peak were highly predictive of all-cause mortality (Figure 3(e) and (f)). Males had higher risk of dying (hazard ratio 3.85, p = 0.031). See Table 3.

Table 3.

Transitional hazard ratios for multi-state prediction model.

	T1: end CR to readmission		T2: readmission to death		T3: end CR to death
Variables	HR	95% CI	HR	95% CI	HR	95% CI
Age, per year^a	1.01	(0.99–1.03)	1.08*	(1.01–1.17)	1.05*	(1.00–1.1)
Sex, male	2.01*	(1.24–3.26)	0.54	(0.12–2.31)	3.85*	(1.12–13.15)
Index diagnosis	ACS		ACS		ACS
Stable CAD	1.07	(0.72–1.59)	2.30	(0.77–6.88)	0.69	(0.28–1.72)
Heart valve disease	0.61	(0.24–1.54)	5.05	(0.47–54.44)	0.67	(0.08–5.47)
CHF	2.32	(0.7–7.65)	18.63	(0.84–415.1)	3.82	(0.6–24.22)
Revascularization	PCI		PCI		PCI
CABG	0.53*	(0.33–0.84)	1.78	(0.41–7.65)	0.47	(0.17–1.31)
None	0.38	(0.15–0.95)	0.78	(0.04–14.28)	0.47	(0.07–3.03)
PAD	2.04*	(1.24–3.36)	2.61	(0.74–9.27)	2.51	(0.90–7.03)
Baseline VO_2peak^a	0.95**	(0.92–0.99)	0.99	(0.87–1.12)	0.86***	(0.79–0.94)
Change in VO_2peak, ml/kg per min
–4.00	2.00	(1.20–3.34)	1.03	(0.44–2.37)	2.45	(1.39–4.32)
–3.00	1.96	(1.3–2.97)	1.02	(0.55–1.89)	1.95	(1.28–2.97)
–2.00	1.72	(1.27–2.32)	1.02	(0.68–1.52)	1.55	(1.17–2.06)
–1.00	1.35	(1.15–1.58)	1.01	(0.83–1.23)	1.25	(1.08–1.43)
0.00 (reference)	1.00	–	1.00	–	1.00	–
1.00	0.75	(0.64–0.88)	0.99	(0.82–1.20)	0.80	(0.70–0.92)
2.00	0.61	(0.46–0.81)	0.99	(0.68–1.45)	0.65	(0.49–0.85)
3.00	0.55	(0.38–0.8)	1.00	(0.57–1.74)	0.52	(0.34–0.79)
4.00	0.54	(0.35–0.83)	1.01	(0.48–2.10)	0.42	(0.24–0.73)

P < 0.05, **P < 0.01, ***P < 0.001.

T: transition; CR: cardiac rehabilitation; HR: hazard ratio; CI: confidence interval; ACS: acute coronary syndrome; CAD: coronary artery disease; CHF: chronic heart failure; PCI: percutaneous coronary intervention; CABG: coronary artery bypass grafting; PAD: peripheral artery disease

Indicates that the presented hazard ratio was calculated considering the variable effect as linear.

Discussion

Change in VO_2peak following cardiac rehabilitation was highly predictive for the risk of readmissions for CVD and all-cause mortality, even after full adjustment for baseline VO_2peak, cardiac diagnosis and co-morbidities. This result seems to emphasize the importance of achieving significant improvements in VO_2peak during a cardiac rehabilitation programme. The results of the multi-state model, however, indicate that the mortality benefits of baseline and improvement in VO_2peak do not extend beyond the first readmission for CVD.

Use of VO_2peak as endpoint for success in cardiac rehabilitation

We found that both baseline VO_2peak and change in VO_2peak after cardiac rehabilitation were inversely associated with risk of readmissions for CVD and mortality. The association we found between baseline VO_2peak and risk of CVD or mortality is coherent with previous research findings.^26–29 Change in VO_2peak is the main measure used when evaluating the impact of a cardiac rehabilitation programme both with regard to individual changes and for comparison with no cardiac rehabilitation or alternative cardiac rehabilitation delivery modes such as home-based or telemonitored cardiac rehabilitation. To our knowledge, only one previous study of patients undergoing cardiac rehabilitation has investigated the effect of change in VO_2peak on mortality. This was a retrospective study of 5641 patients with CAD participating in cardiac rehabilitation with a one year follow-up.³⁰ This study also found an inverse association between change in VO_2peak and risk of mortality. There are, however, important methodological differences from our study. We performed a gold standard CPET, whereas the comparative study applied a graded exercise test, which is not validated to estimate change in effect over time, something they also pointed out as a limitation in their study. In addition to mortality, the present study addresses CVD outcomes over an extended period (median 2.3 years versus one year).

In line with these findings, a recent study of 421 patients with CHF also found that patients who improved VO_2peak after cardiac rehabilitation had a decreased risk of readmissions and all-cause mortality.³¹

Application of the multi-state model in a semi-competing risk setting

In addition to the Cox-model we applied a multi-state model to analyse a semi-competing risk scenario of being readmitted for CVD and/or all-cause mortality. This prediction model provides a new dimension to the extent to which VO_2peak can predict future outcomes. The model has been applied previously in cardiovascular research^32–34 but not in cardiac rehabilitation. Upshaw et al. proposed the use of multi-state models to predict risk of heart failure hospitalizations and all-cause mortality, arguing the strength of not censoring patients in the case of non-terminal events.³⁴

The multi-state analysis found the same inverse association between both baseline VO_2peak and change in VO_2peak and risk of all-cause mortality. In addition, the model added a level to the risk prediction, transition 2; the risk of all-cause mortality after being readmitted for CVD. There was no association between baseline VO_2peak and risk of mortality in the patients readmitted for CVD after cardiac rehabilitation. This result would indicate that the protective effect of improved VO_2peak is no longer present if a patient is readmitted to hospital, may imply a need for a new cardiac rehabilitation intervention and emphasizes the need to promote and maintain a long-term physically active lifestyle, which is the most important contributor of VO_2peak.³⁵ However, only 25 patients died after readmission for CVD, so this result should be confirmed in a larger cohort.

Limitations

There are some limitations to this study. One is that we assessed only patients who completed cardiac rehabilitation, introducing the risk of selection bias. Hence, these results may not be generalizable to cardiac patients not attending or completing cardiac rehabilitation. In a previous publication on cardiac rehabilitation, out of the present cohort, we found that patients who did not complete cardiac rehabilitation both had lower baseline VO_2peak and more co-morbidity.³⁶

Some confounding factors were missing at large, such as left ventricular ejection fraction and other co-morbidity, making the study prone to some residual confounding. This clinical database contained the most important clinical variables, so residual confounding was considered to be minimal. However, physical activity post cardiac rehabilitation was not accounted for in the database.

Patients with CHF and heart valve disease constituted approximately 16% of the population, and we did not have statistical power for analysis of potential interaction between measures of VO_2peak and index event.

Conclusion

Both baseline VO_2peak and improvement in VO_2peak during a cardiac rehabilitation programme are strong predictors of subsequent prognosis in cardiac patients. Patients who are not able to improve their exercise capacity by cardiac rehabilitation have increased risk of readmissions for CVD and mortality. Future studies should address how prognosis can be improved in these patients. Also, our results indicate that the protective factor of improved VO_2peak is restricted to the time until new CVD admissions occur, and this may indicate the need for repeated participation in cardiac rehabilitation.

Footnotes

Author contribution

NM conceived and drafted the manuscript. EP contributed to conception, design analysis and interpretation and critically revised; HR contributed to acquisition and analysis and critically revised; CC, CD, OL and JH contributed to design and statistical methods and analyses; CG contributed with design and critically revised.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Hung

Al-Mallah

McEvoy

, et al. Prognostic value of exercise capacity in patients with coronary artery disease: The FIT (Henry Ford ExercIse Testing) project. Mayo Clin Proc 2014; 89: 1644–1654.

Kodama

Saito

Tanaka

, et al. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: A meta-analysis. JAMA 2009; 301: 2024–2035.

Heran

Chen

Ebrahim

, et al. Exercise-based cardiac rehabilitation for coronary heart disease. Cochrane Database Syst Rev 2011; 7: CD001800–CD001800.

Anderson

Oldridge

Thompson

, et al. Exercise-based cardiac rehabilitation for coronary heart disease: Cochrane systematic review and meta-analysis. J Am Coll Cardiol 2016; 67: 1–12.

Doyle

Indraratna

Tardo

, et al. Safety and efficacy of aerobic exercise commenced early after cardiac surgery: A systematic review and meta-analysis. Eur J Prev Cardiol 2019; 26: 36–45.

Pratesi

Baldasseroni

Burgisser

, et al. Long-term functional outcomes after cardiac rehabilitation in older patients. Data from the Cardiac Rehabilitation in Advanced aGE: EXercise TRaining and Active follow-up (CR-AGE EXTRA) randomised study. Eur J Prev Cardiol 2019; 26: 1470–1478.

Mitchell

Lock

Davison

, et al. What is the effect of aerobic exercise intensity on cardiorespiratory fitness in those undergoing cardiac rehabilitation? A systematic review with meta-analysis. Br J Sports Med 2018; 53: 1341.

Lee

D-C

Sui

Ortega

, et al. Comparisons of leisure-time physical activity and cardiorespiratory fitness as predictors of all-cause mortality in men and women. Br J Sports Med 2011; 45: 504–510.

Uddin

Zwisler

Lewinter

, et al. Predictors of exercise capacity following exercise-based rehabilitation in patients with coronary heart disease and heart failure: A meta-regression analysis. Eur J Prev Cardiol 2016; 23: 683–963.

10.

Daullxhiu

Haliti

Poniku

, et al. Predictors of exercise capacity in patients with chronic heart failure. J Cardiovasc Med (Hagerstown) 2011; 12: 223–225.

11.

Mikkelsen

Dall

Frederiksen

, et al. Depression, socioeconomic factors, and ethnicity as predictors of cardiorespiratory fitness before and after cardiac rehabilitation. J Cardiopulm Rehabil Prev 2018; 39: E1–E6.

12.

Goel

Lennon

Tilbury

, et al. Impact of cardiac rehabilitation on mortality and cardiovascular events after percutaneous coronary intervention in the community. Circulation 2011; 123: 2344–2352.

13.

Pack

Goel

Lahr

, et al. Participation in cardiac rehabilitation and survival after coronary artery bypass graft surgery: A community-based study. Circulation 2013; 128: 590–597.

14.

Meira-Machado

de Una-Alvarez

Cadarso-Suarez

, et al. Multi-state models for the analysis of time-to-event data. Stat Methods Med Res 2009; 18: 195–222.

15.

Vanhees

Rauch

Piepoli

, et al. Importance of characteristics and modalities of physical activity and exercise in the management of cardiovascular health in individuals with cardiovascular disease (Part III). Eur J Prev Cardiol 2012; 19: 1333–1356.

16.

Mezzani

Hamm

Jones

, et al. Aerobic exercise intensity assessment and prescription in cardiac rehabilitation: A joint position statement of the European Association for Cardiovascular Prevention and Rehabilitation, the American Association of Cardiovascular and Pulmonary Rehabilitation and the Canadian Association of Cardiac Rehabilitation. Eur J Prev Cardiol 2013; 20: 442–467.

17.

Guazzi

Adams

Conraads

, et al. EACPR/AHA Joint Scientific Statement. Clinical recommendations for cardiopulmonary exercise testing data assessment in specific patient populations. Eur Heart J 2012; 33: 2917–2927.

18.

Lynge

Sandegaard

Rebolj

. The Danish National Patient Register. Scand J Public Health 2011; 39(7 Suppl): 30–33.

19.

Pedersen

. The Danish Civil Registration System. Scand J Public Health 2011; 39(7 Suppl): 22–25.

20.

Haneuse

Lee

. Semi-competing risks data analysis: Accounting for death as a competing risk when the outcome of interest is nonterminal. Circ Cardiovasc Qual Outcomes 2016; 9: 322–331.

21.

Team RC. Computer software manual. R: A language and environment for statistical computing, 2016.

22.

Wood S. Mixed GAM computation vehicle with automatic smoothness estimation cran.r-project.org: R, updated 6 February 2019, https://cran.r-project.org/web/packages/mgcv (2019, accessed 1 May 2019).

23.

Therneau TM and Lumley T. survival: Survival analysis cran.r-project: R, updated 2018. R package version 2.43-3, https://cran.r-project.org/web/packages/survival (2018, accessed 1 May 2019).

24.

Meira-Machado L and Araujo A. smoothHR: Smooth hazard ratio curves taking a reference value cran.r R, updated 2015, R package version 1.0.2, http://CRAN.R-project.org/package=smoothHR (2015, accessed 21 February 2019).

25.

Belitz C, Brezger A, Klein N, et al. BayesX: Software for Bayesian inference in structured additive regression models, version 3.0.2, http://www.BayesX.org (2015, accessed 1 May 2019).

26.

Grazzi

Myers

Bernardi

, et al. Association between VO(2) peak estimated by a 1-km treadmill walk and mortality. A 10-year follow-up study in patients with cardiovascular disease. Int J Cardiol 2014; 173: 248–252.

27.

Myers

Prakash

Froelicher

, et al. Exercise capacity and mortality among men referred for exercise testing. N Engl J Med 2002; 346: 793–801.

28.

De Lannoy

Sui

Blair

, et al. All-cause mortality risk among active and inactive adults matched for cardiorespiratory fitness. Eur J Prev Cardiol 2019; 26: 554–556.

29.

Grazzi

Mazzoni

Myers

, et al. Improved walking speed is associated with lower hospitalisation rates in patients in an exercise-based secondary prevention programme. Heart 2016; 102: 1902–1908.

30.

Martin

Arena

Haykowsky

, et al. Cardiovascular fitness and mortality after contemporary cardiac rehabilitation. Mayo Clin Proc 2013; 88: 455–463.

31.

Sabbag

Mazin

Rott

, et al. The prognostic significance of improvement in exercise capacity in heart failure patients who participate in cardiac rehabilitation programme. Eur J Prev Cardiol 2018; 25: 354–361.

32.

Bakal

McAlister

Liu

, et al. Heart failure re-admission: Measuring the ever shortening gap between repeat heart failure hospitalizations. PLoS One 2014; 9: e106494–e106494.

33.

Ieva

Jackson

Sharples

. Multi-state modelling of repeated hospitalisation and death in patients with heart failure: The use of large administrative databases in clinical epidemiology. Stat Methods Med Res 2017; 26: 1350–1372.

34.

Upshaw

Konstam

Klaveren

, et al. Multistate model to predict heart failure hospitalizations and all-cause mortality in outpatients with heart failure with reduced ejection fraction: Model derivation and external validation. Circ Heart Fail 2016; 9: e003146–e003146.

35.

Fihn

Gardin

Abrams

, et al. 2012 ACCF/AHA/ACP/AATS/PCNA/SCAI/STS Guideline for the diagnosis and management of patients with stable ischemic heart disease: A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines, and the American College of Physicians, American Association for Thoracic Surgery, Preventive Cardiovascular Nurses Association, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol 2012; 60: e44–e164.

36.

Mikkelsen

Dall

Frederiksen

, et al. Depression, socioeconomic factors, and ethnicity as predictors of cardiorespiratory fitness before and after cardiac rehabilitation. J Cardiopulm Rehabil Prev 2019; 39: E1–E6.

Improvement in VO 2peak predicts readmissions for cardiovascular disease and mortality in patients undergoing cardiac rehabilitation

Abstract

Background

Methods and results

Conclusion

Keywords

Introduction

Material and methods

Population

Rehabilitation programme

Baseline and explanatory variables

Primary endpoints

Potential confounders

Statistical analysis

Results

Study population

Influence of VO2peak on MACE

Influence of VO2peak on all-cause mortality

Multi-state model: influence of VO2peak on readmissions for CVD and/or all-cause mortality

Transition 1: risk of being readmitted for CVD

Transition 2: risk of dying after a readmission for CVD

Transition 3: risk of dying without readmission for CVD

Discussion

Use of VO2peak as endpoint for success in cardiac rehabilitation

Application of the multi-state model in a semi-competing risk setting

Limitations

Conclusion

Footnotes

Author contribution

Declaration of conflicting interests

Funding

References

Influence of VO_2peak on MACE

Influence of VO_2peak on all-cause mortality

Multi-state model: influence of VO_2peak on readmissions for CVD and/or all-cause mortality

Use of VO_2peak as endpoint for success in cardiac rehabilitation