Comparative effectiveness of initial computed tomography and invasive coronary angiography in women and men with stable chest pain and suspected coronary artery disease: multicentre randomised trialBMJ 2022; 379 doi: https://doi.org/10.1136/bmj-2022-071133 (Published 19 October 2022) Cite this as: BMJ 2022;379:e071133
- DISCHARGE Trial Group
- Correspondence to: M Dewey(https://orcid.org/0000-0002-4402-2733), Charité – Universitätsmedizin Berlin, Freie Universität Berlin and Humboldt-Universität zu Berlin, Department of Radiology, and Berlin Institute of Health, 0117 Berlin, Germany (or @ProfDewey on Twitter)
- Accepted 9 September 2022
Objective To assess the comparative effectiveness of computed tomography and invasive coronary angiography in women and men with stable chest pain suspected to be caused by coronary artery disease.
Design Prospective, multicentre, randomised pragmatic trial.
Setting Hospitals at 26 sites in 16 European countries.
Participants 2002 (56.2%) women and 1559 (43.8%) men (total of 3561 patients) with suspected coronary artery disease referred for invasive coronary angiography on the basis of stable chest pain and a pre-test probability of obstructive coronary artery disease of 10-60%.
Intervention Both women and men were randomised 1:1 (with stratification by gender and centre) to a strategy of either computed tomography or invasive coronary angiography as the initial diagnostic test (1019 and 983 women, and 789 and 770 men, respectively), and an intention-to-treat analysis was performed. Randomised allocation could not be blinded, but outcomes were assessed by investigators blinded to randomisation group.
Main outcome measures The primary endpoint was major adverse cardiovascular events (MACE; cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke). Key secondary endpoints were an expanded MACE composite (cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, transient ischaemic attack, or major procedure related complication) and major procedure related complications.
Results Follow-up at a median of 3.5 years was available in 98.9% (1979/2002) of women and in 99.0% (1544/1559) of men. No statistically significant gender interaction was found for MACE (P=0.29), the expanded MACE composite (P=0.45), or major procedure related complications (P=0.11). In both genders, the rate of MACE did not differ between the computed tomography and invasive coronary angiography groups. In men, the expanded MACE composite endpoint occurred less frequently in the computed tomography group than in the invasive coronary angiography group (22 (2.8%) v 41 (5.3%); hazard ratio 0.52, 95% confidence interval 0.31 to 0.87). In women, the risk of having a major procedure related complication was lower in the computed tomography group than in the invasive coronary angiography group (3 (0.3%) v 21 (2.1%); hazard ratio 0.14, 0.04 to 0.46).
Conclusion This study found no evidence for a difference between women and men in the benefit of using computed tomography rather than invasive coronary angiography as the initial diagnostic test for the management of stable chest pain in patients with an intermediate pre-test probability of coronary artery disease. An initial computed tomography scan was associated with fewer major procedure related complications in women and a lower frequency of the expanded MACE composite in men.
Trial registration NCT02400229ClinicalTrials.gov NCT02400229.
Studies consistently show that women presenting with stable chest pain and suspected coronary artery disease have more symptoms but less severe myocardial ischaemia and less extensive epicardial coronary artery disease compared with men.1234 This phenomenon has come to be known as the “gender paradox” and may lead to misdiagnosis and poorer outcome in women.567 Importantly, in both women and men, coronary artery disease accounts for a similar and high proportion of all cause death in western countries.8910 Nevertheless, the paucity of diagnostic and management trials specifically randomising either women or men and predefining such subgroups as targets for analysis in terms of outcomes precludes informed clinical recommendations.11
The reference standard for the diagnosis of obstructive coronary artery disease is invasive coronary angiography, which additionally allows coronary revascularisation to be done in the same session. However, rare but serious procedural complications can occur, and women undergoing invasive coronary examination and treatment may have a higher risk of bleeding, vascular complications, and stroke than men.12131415
Coronary computed tomography has been shown to be clinically useful for the non-invasive diagnosis of obstructive coronary artery disease in patients with stable chest pain and an intermediate pre-test probability because of a high diagnostic accuracy compared with invasive coronary angiography.16171819 At the same time, computed tomography has been reported to have a comparable or slightly lower accuracy in women than in men, possibly owing to a reduced ability to detect stenosis in smaller coronary branches.172021 However, no large trials have assessed the gender specific comparative effectiveness of computed tomography and invasive coronary angiography as the initial test with regard to avoiding major procedure related complications and other key clinical outcomes.
In the DISCHARGE (Diagnostic Imaging Strategies for Patients With Stable Chest Pain and Intermediate Risk of Coronary Artery Disease) trial, women and men referred for invasive coronary angiography because of stable chest pain and an intermediate pre-test probability of coronary artery disease were allocated to initial computed tomography or initial invasive coronary angiography, with randomisation stratified for gender and centre.22 For the total study population, the trial showed no significant difference between an initial computed tomography guided strategy and an initial invasive coronary angiography guided strategy with regard to the rates of major cardiovascular events (MACE) and found a lower frequency of major procedure related complications for an initial computed tomography strategy. The aim of this prespecified analysis of the DISCHARGE trial was to assess the gender specific comparative effectiveness of computed tomography and invasive coronary angiography.
Study design and patients
The study design, methods, and statistical analysis plan, including the prespecified gender related analysis of the DISCHARGE trial, have previously been published.2223 Briefly, DISCHARGE is an investigator initiated, pragmatic, assessor blinded, parallel group, randomised, multicentre study of the comparative effectiveness of computed tomography guided versus invasive coronary angiography guided management of patients with stable chest pain and a calculated intermediate pre-test probability of obstructive coronary artery disease clinically referred for invasive coronary angiography. The study was funded by the European Union (EU-FP7 Framework Programme) and was conducted at 26 clinical centres in 16 European countries (1 Austria, 1 Czech Republic, 1 Denmark, 1 Finland, 3 Germany, 1 Hungary, 1 Republic of Ireland, 2 Italy, 1 Latvia, 1 Lithuania, 2 Poland, 1 Portugal, 2 Romania, 2 Serbia, 2 Spain, and 4 UK), using standard operating procedures and strict quality control according to good clinical practice. Local or national ethics committees approved the study at each participating centre. The study was conducted and reported according to the CONSORT standards. The checklist is included in web appendix 1.24
Patients were eligible for the study if they were at least 30 years of age, were clinically referred for invasive coronary angiography because of stable chest pain with a calculated intermediate clinical pre-test probability (10-60%) of obstructive coronary artery disease,17 and had no previous coronary revascularisation. The clinical referral for invasive coronary angiography because of suspected stable coronary artery disease was in accordance with the European Society of Cardiology guidelines at the time the study was conducted.252627 We collected self-reported gender information from all patients at baseline and did not biologically determine the sex of participants, following the Sex and Gender Equity in Research (SAGER) guidelines.28 The terms “women” and “men” are used in this publication to acknowledge the absence of information on sex. Equal representation of both genders in the study population was intended, as the pre-test probability accounted for gender differences and randomisation was stratified by gender and centre. The exclusion criteria were haemodialysis, non-sinus rhythm, and pregnancy. All patients provided written informed consent. We estimated the pre-test probability of coronary artery disease after clinical referral of patients for invasive coronary angiography by using an automated calculation tool integrated into a web based system of electronic case report forms. The tool uses a model based on data from the Collaborative Meta-Analysis of Cardiac CT,17 using patients’ age, gender, and the type of stable chest pain.
Patients were randomly allocated to either initial computed tomography or initial invasive coronary angiography. We defined obstructive disease as detection of at least one ≥50% coronary artery luminal diameter stenosis. We categorised obstructive coronary artery disease as high risk anatomy if three vessel coronary artery disease, left main coronary artery stenosis, proximal left anterior descending coronary artery stenosis, or any combination of these was present. The clinical sites were provided with central management recommendations that incorporated contemporary European Society of Cardiology guidelines on the management of stable coronary artery disease in prevention of cardiovascular disease, including the clinical indication for statins in clinical practice and on myocardial revascularisation,2526 as described in the study design and protocol.23 The management decisions on patients were made by local heart team members and referring physicians on the basis of the computed tomography and invasive coronary angiography reports at each study site. Patients randomised to computed tomography as the initial test were subsequently managed according to the DISCHARGE computed tomography guided strategy protocol.22 In both randomisation groups, patients without obstructive coronary artery disease by invasive coronary angiography or computed tomography were discharged back to the referring physician, and patients with obstructive coronary artery disease were managed according to guidelines.252729 Further recommendations included prescription of anti-anginal dugs, additional invasive or non-invasive functional testing, and coronary revascularisation as appropriate. Patients in both randomised groups were advised to be treated to target values for blood pressure, glycaemia, and lipids according to European guidelines on cardiovascular disease prevention.26
Computed tomography and invasive coronary angiography protocols
To ensure adherence to image quality criteria for computed tomography and invasive coronary angiography, we did a pilot study before the randomised study, in which each of the participating clinical centres had to provide imaging datasets of three patients for each imaging strategy. Adherence to the predefined standards for the computed tomography protocol and invasive coronary angiography recommendations were verified at all sites.30
At all clinical centres, trained personnel did computed tomography on at least 64 slice computed tomography scanners following a 10 step acquisition guide and scanner specific recommendations to identify coronary artery obstruction.23 Trained interventional cardiologists did invasive coronary angiography at each participating centre according to contemporary guidelines and routine clinical practice at the individual clinical site.29 Board certified and experienced radiologists with SCCT level II or III certification evaluated computed tomography.
Primary and secondary endpoints
The primary endpoint of this prespecified gender analysis was major adverse cardiovascular events (MACE), defined as a composite of cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke. Cardiovascular death was diagnosed as defined by the Cardiac Safety Research Consortium, myocardial infarction was determined according to the third universal definition of myocardial infarction, and stroke was determined according to the updated definition for the 21st century as previously described.22 Possible adverse cardiovascular events were collected in the electronic case report form, and events were adjudicated by independent assessors blinded to the study group assignment.
Key secondary endpoints included an expanded MACE composite (cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, transient ischaemic attack, or major procedure related complication occurring during or within 48 hours after computed tomography or invasive coronary angiography or related tests as prespecified in table 3 of the statistical analysis plan). Furthermore, the following additional composite endpoints were recorded: vascular death or myocardial infarction, cardiac death or myocardial infarction, all cause death, and myocardial infarction or stroke.
Major procedure related complications included death, non-fatal myocardial infarction, non-fatal stroke, further complications prolonging hospital admission by at least 24 hours, dissection (coronary, aortic), cardiogenic shock, cardiac tamponade, retroperitoneal bleeding, cardiac arrhythmia (ventricular tachycardia, ventricular fibrillation), or cardiac arrest. Complications were classified according to the NCDR®CathPCI Registry®v4.4 Coder’s Data Dictionary. The CONSORT extension for harms was adhered to.31
Randomisation and masking
Randomisation used a web based system (SecuTrial) to confirm eligibility criteria and record data for individual calculation of pre-test probability of disease. This web based system was used to randomly assign patients (1:1) to either computed tomography or invasive coronary angiography as the initial test. Block randomisation used computer generated and randomly permuted blocks with lengths of four, six, or eight stratified according to centre with central assignment. Randomisation of patients at each clinical centre was stratified according to gender. Patients and physicians could not be blinded to the study assignment.
The sample size estimation for the trial has been previously published22; it was based on the comparison of invasive coronary angiography versus computed tomography regarding the primary outcome, first occurrence of a MACE. According to the statistical analysis plan, a total of 3546 patients would provide 80% power to detect a relative reduction in the annual primary outcome rate from 1.4% for the invasive coronary angiography group to 0.8% for the computed tomography group.22 In this study, we analysed the effect of gender on primary and secondary endpoints of the DISCHARGE trial as well as interactions between gender and study arm. A post-hoc power analysis using group size and number of events is presented in supplementary table A in web appendix 2.
We compared categorical variables by using χ2 tests (or Fisher’s exact test for small datasets). We compared continuous variables between groups by using the independent samples Student’s t test for normally distributed data or the Mann-Whitney U test for non-normally distributed data.
For the primary endpoint, we calculated cumulative incidences of MACE. Major adverse cardiovascular events included cardiovascular death, non-fatal stroke, and non-fatal myocardial infarction, and non-cardiovascular death and unknown cause of death were considered as competing risks. Follow-up was defined as the period from randomisation until the occurrence of the outcome or otherwise censored at death (non-cardiovascular events and unknown causes of death), loss to follow-up, or the end of the study. We also analysed secondary time to event endpoints such as the composite of MACE and major procedure related complications by using cumulative incidences adjusted for competing risks.32 We used the sub-distribution Cox proportional hazard model of Fine and Gray to analyse the effect of the study group (computed tomography versus invasive coronary angiography) stratified for gender, as well as the interaction between gender and group. For sensitivity analysis, we used a multivariate sub-distribution Cox proportional hazard model to evaluate adjusted differences in hazard risk of MACE and secondary endpoints by using the covariates gender, randomisation group (invasive coronary angiography/computed tomography), angina type, age, and interaction of gender and group. We report results as hazard ratios and 95% confidence intervals. We evaluated the proportional hazard assumption of the Cox model by using “log-log” plot curves and Schoenfeld residuals.
We used SAS software version 9.4, SPSS for Windows version 26, and the statistical programming language R version 4.0.3 for statistical analyses. In this secondary analysis, we defined statistical significance as a two sided P value of less than 0.05, whereas in the primary analysis a P value of 0.048 was considered significant following a prespecified interim analysis of the primary outcome. We did not adjust for multiple testing. All analyses were done on the intention-to-treat population.
Patient and public involvement
Patient interest groups at all sites were involved in the planning of the study, and all participating patients will be informed about the results of this analysis.
Between 3 October 2015 and 12 April 2019, 2052 women and 1615 men were recruited at 26 European centres (fig 1). A total of 1031 women were randomised to the computed tomography group (10 withdrew consent and two were randomised in error) and 1021 were randomised to the invasive coronary angiography group (32 withdrew consent and six were randomised in error); the corresponding numbers in men were 802 randomised to the computed tomography group (10 withdrew consent and three were randomised in error) and 813 randomised to the invasive coronary angiography group (37 withdrew consent and six were randomised in error). Overall, 3561 patients (women: 2002 (computed tomography 1019; invasive coronary angiography 983); men: 1559 (computed tomography 789; invasive coronary angiography 770)) were included in the intention-to-treat analysis (fig 1). At 3.5 years, follow-up for the primary endpoint was available in 98.9% (computed tomography 1010; invasive coronary angiography 969) of women and 99.0% (computed tomography 782; invasive coronary angiography 762) of men.
Baseline patient characteristics
Table 1 provides baseline characteristics separately for women and men and stratified by gender and initial test strategy (computed tomography versus invasive coronary angiography); reproductive factors in women are given in supplementary table B in web appendix 2. Women were slightly older (61.8 (SD 9.7) years) than men (58.1 (10.3) years), more often had typical angina as the type of chest pain, and more frequently had arterial hypertension and asthma than did men. Furthermore, smoking was less frequent and the pre-test probability of coronary artery disease was lower in women. Overall, the characteristics of women and men were similar in the two randomised study groups.
Initial test findings and revascularisation
Table 2 shows initial test findings by either computed tomography or invasive coronary angiography, frequency of computed tomography and coronary artery disease tests done during initial management, invasive access site information, and coronary revascularisation. The median time from enrolment to the initial test was longer in the invasive coronary angiography group than in the computed tomography group both for women (10 (interquartile range 1-35) days v 3 (0-15) days) and for men (13 (1-41) days v 4 (0-13) days).
The frequency of detection of obstructive coronary artery disease was similar with either a computed tomography strategy or an invasive coronary angiography strategy, both in women (201 (19.7%, 95% confidence interval 17.4% to 22.4%) and 179 (18.2%, 16.1% to 21.1%)) and in men (264 (33.5%, 30.4% to 37.2%) and 272 (35.8%, 32.4% to 39.4%)) (table 2). Non-obstructive coronary artery disease was more frequently recorded by computed tomography than by invasive coronary angiography in both genders. With both tests, women more frequently had no signs of coronary artery disease and were less frequently found to have obstructive and/or high risk anatomy coronary artery disease compared with men. In the computed tomography group, the frequency of a non-diagnostic test was similar in women and men (63 (6.2%, 4.8% to 7.9%) v 40 (5.1%, 3.7% to 6.9%); table 2).
Invasive procedures were most commonly done using radial access (table 2). Coronary revascularisation was less frequent in women than in men and, for both genders, less frequent in the computed tomography group.
Primary and secondary endpoints
The primary endpoint (MACE) occurred with similar frequency in women and men (supplementary table C in web appendix 2). The interaction between gender and study group for MACE was not significant (hazard ratio 1.58, 95% confidence interval 0.68 to 3.66; P=0.29). Subgroup analysis stratified by gender showed a similar risk of having a MACE in the computed tomography and invasive coronary angiography groups in women (22 (2.2%) v 24 (2.4%)) and men (16 (2.0%) v 28 (3.6%)) (fig 2). The results for the individual components of the primary endpoint MACE are shown in table 3 and supplementary figures A-C in web appendix 2.
The expanded MACE composite of cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, transient ischaemic attack, or major procedure related complication occurred with similar frequency in women and men (supplementary table C in web appendix 2). The interaction between gender and study group for the expanded MACE composite was not significant (hazard ratio 1.31, 0.65 to 2.67; P=0.45). We did further subgroup analysis stratified by gender. In men, the expanded MACE composite endpoint occurred less frequently in the computed tomography group than in the invasive coronary angiography group (22 (2.8%) v 41 (5.3%); hazard ratio 0.52, 0.31 to 0.87) (table 3 and fig 3). In women, the secondary expanded MACE composite was similar in the computed tomography and invasive coronary angiography groups (table 3 and fig 3). Rates of secondary additional composite endpoints were similar for the two test strategies in both women and men (table 3 and supplementary figures D-F in web appendix 2).
Major procedure related complications occurred with similar frequency in women and men (supplementary table C in web appendix 2). We found no significant interaction between gender and study group for major procedure related complications (hazard ratio 0.28, 0.06 to 1.33; P=0.11). In women, the risk of having a major procedure related complication was lower in the computed tomography group than in the invasive coronary angiography group (3 (0.3%) v 21 (2.1%); hazard ratio 0.14, 0.04 to 0.46) whereas no such difference was noted in men (table 3 and fig 4). Most major procedure related complications were observed in relation to invasive coronary angiography procedures, and the frequency was highest in women allocated to the invasive coronary angiography group (supplementary tables D-H in web appendix 2). Major procedure related complications related to invasive coronary angiography with percutaneous intervention were lower in women when invasive coronary angiography followed initial computed tomography (supplementary table G in web appendix 2).
In the sensitivity analysis including, for instance, age and chest pain as covariates, we found a smaller risk for expanded MACE in women. The interactions between gender and study group for MACE, expanded MACE, and major procedure related complications were also not significant (supplementary tables I and J in web appendix 2).
In this predefined gender subgroup analysis of the computed tomography and invasive coronary angiography groups in the DISCHARGE trial, in which randomisation was stratified by gender and centre, we found no evidence for a difference between women and men in the benefit of using computed tomography rather than invasive coronary angiography as the initial diagnostic test for the management of stable chest pain in patients with an intermediate pre-test probability of coronary artery disease. Initial computed tomography was associated with fewer major procedure related complications in women and a lower frequency of the expanded MACE composite endpoint in men. A sensitivity analysis showed a smaller risk for expanded MACE in women but again no significant interactions between gender and study group for the primary and secondary endpoints.
Comparison with other studies
Our results corroborate and further extend the findings of previous post-hoc gender specific investigations conducted in the PROMISE, ISCHEMIA, and SCOT-HEART diagnostic strategy trials.234 In accordance with these investigations, we found that women with chest pain had less obstructive epicardial coronary artery disease and more frequently had normal coronary arteries compared with men. Importantly, the proportion of patients in whom obstructive coronary artery disease was found was similar with either a computed tomography strategy or an invasive coronary angiography strategy, in both women (19.7% and 18.2%) and men (33.5% and 35.3%), suggesting comparable diagnostic accuracy of the two investigations to identify obstructive disease. Interestingly, non-obstructive coronary artery disease was more frequently detected by computed tomography than by invasive coronary angiography, and the frequency of detected non-obstructive disease was similar in women and men. This higher sensitivity of computed tomography to detect non-obstructive coronary artery disease compared with invasive coronary angiography may have important clinical implications in terms of treatment because non-obstructive disease is associated with poorer long term outcome.33 As recently highlighted by Al-Lamee and colleagues in The BMJ, aggressive risk factor modification with medical therapy in patients with stable coronary artery disease has the greatest effect on the long term risk of myocardial infarction and death.34 That, in both genders, preventive medical therapy could be improved when guided by computed tomography seems plausible. The frequency of non-diagnostic computed tomography was similar in women and men, ranging from 5% to 6%, in accordance with similar large scale studies using computed tomography scanners with 64 slices and higher.3536 On the other hand, a frequency of non-diagnostic computed tomography of 3.7% was previously reported when exclusively 320 slice computed tomography technology was used, suggesting that the frequency of non-diagnostic computed tomography might be lower if more advanced computed tomography technology was used consistently.36
With one exception, all major procedure related complications were attributable to invasive coronary angiography without or with subsequent coronary revascularisation. One man undergoing computed tomography experienced an episode of bradycardia, which prolonged hospital admission by at least 24 hours and was probably attributable to pre-treatment with a β blocker (see supplementary table E in appendix 2). The overall risk of having a major procedural complication was not significantly different when we compared all women and all men in the DISCHARGE cohort. This finding is consistent with reports suggesting an overall decline in invasive coronary procedural complications over the past decade, especially in women. This trend reduces the previously reported gender difference,1315 and it likely reflects the more common general use of radial access.37 Nevertheless, women randomised to an initial computed tomography strategy had a significantly lower risk of a major procedure related complication than did women randomised to an initial invasive coronary angiography strategy. Most likely, this observation reflects the lower prevalence of coronary artery disease needing invasive investigation and coronary revascularisation, especially in women, highlighting the positive contribution of computed tomography as a gatekeeper for invasive investigation and treatment in this setting, as well as a lower risk of invasive coronary angiography and percutaneous coronary intervention in patients with an earlier computed tomography scan.
In both women and men, a strategy of initial computed tomography resulted in a rate of MACE that was not significantly different from that with a strategy of initial invasive coronary angiography during a median follow-up period of 3.5 years. However, we noted a borderline trend to improved outcome in men, which reached statistical significance for the predefined secondary expanded MACE composite endpoint of cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, transient ischaemic attack, or major procedural complications. The clinical benefit for most patients with stable chest pain undergoing guideline defined examination and treatment is relief of chest pain, and some patients also have improved clinical outcomes.3839 The strategy of initial computed tomography in the DISCHARGE trial comprised a stringent clinical triage based on computed tomography angiography findings, including referral for an invasive procedure for coronary revascularisation, optimal medical therapy, or search for non-cardiac causes.22 As previously reported, after a median follow-up of 3.5 years, most women and men in the DISCHARGE cohort had achieved symptom relief, and the rates were similar for an initial computed tomography strategy and an initial invasive coronary angiography strategy.22 Evidently, the important goal of relief of chest pain is safely achieved with either strategy and in women and men alike.
Both women and men were included in our trial if they had a low to intermediate pre-test probability of 10-60% of disease, taking into account factors such as age, gender, and type of chest pain symptoms.17 We defined his approach according to clinical guidelines in order to optimise the diagnostic pathway in patients with chest pain and coronary artery disease.27 In this context, we note that women included in our trial more frequently had typical angina pectoris and were slightly older than men, which probably reflects selection as a consequence of the features of the pre-test probability tool used. Nevertheless, the prevalence of coronary artery disease predicted by the tool in women and men was only moderately above the rates we found with either computed tomography or invasive coronary angiography (approximately 10% in men and 13% in women).
Strengths and weaknesses of study
Strengths of our study include the multicentre pragmatic design, in which women and men were randomised in a stratified fashion to either an initial computed tomography strategy or an initial invasive coronary angiography strategy. Furthermore, women and men were almost equally represented, with a high adherence to follow-up in both genders, facilitating a statistically plausible comarison.40 A limitation of the study is, as previously noted, a lower than expected event rate during the course of the trial.22 This might reflect a general temporal trend towards fewer procedural complications related to invasive diagnosis and treatment, optimised medical treatment, and a generally improved adherence to lifestyle recommendations in participating countries.41 The natural history of coronary atherosclerosis may differ between males and females. Patients in our study were categorised according to self-reported gender (women and men) not sex, and the extent of heterogeneity within each gender group with regard to sex is therefore not known. The results should therefore be interpreted with caution.
We found no evidence for a difference between women and men in the benefit of using computed tomography rather than invasive coronary angiography as the initial diagnostic test for the management of stable chest pain in patients with an intermediate pre-test probability of coronary artery disease. Computed tomography as the initial diagnostic test in this patient population resulted in a lower frequency of procedural complications in women and a lower frequency of cardiovascular death, non-fatal myocardial infarction, non-fatal stroke, transient ischaemic attack, or major procedure related complications in men compared with a strategy of initial invasive coronary angiography. Both men and women benefitted from computed tomography as the initial diagnostic test.
What is already known on this topic
Compared with men, women presenting with chest pain have more symptoms but less severe myocardial ischaemia and less extensive epicardial coronary artery disease
Computed tomography (CT) can rule out obstructive coronary artery disease in patients with low to intermediate pre-test probability of CAD, but possibly less accurately in women than men
The comparative effectiveness of CT and invasive coronary angiography in the diagnosis and clinical management of CAD in women versus men in terms of clinical outcome is uncertain
What this study adds
The proportion of patients in whom obstructive CAD was found was similar with either a CT or ICA strategy, both in women (19.7% and 18.2%) and in men (33.5% and 35.3%)
In men, CT as the initial diagnostic test instead of initial ICA was associated with a lower risk of the expanded MACE composite (2.8% v 5.3%)
In women, initial CT was associated with a lower frequency of major procedure-related complications compared with initial ICA (0.3% v 2.1%)
This study was approved by the ethics committee of the Charité –Universitätsmedizin Berlin (EA1/294/13) and by the German Federal Office for Radiation Protection (Z5–2246/2–2014-001). On the basis of this, all other clinical partners received ethical approval as required by national law. All patients provided written informed consent.
Data availability statement
Individual participant data that underlie the results reported in the published article after de-identification will be shared with researchers who provide a sound proposal approved by the Dissemination Committee of the DISCHARGE trial. Proposals should be directed to firstname.lastname@example.org.
We thank the clinical staff supporting the trial in the clinical centres and at the core laboratories at Charité, Harold C Sox for his leadership of the external advisory board, the members of the external advisory board, the data safety and monitoring board, and the clinical events committee, as well as all participants who took part in the study.
Authors of study (see supplementary file for full list of author affiliations):
Klaus F Kofoed, Maria Bosserdt, Pál Maurovich-Horvat, Nina Rieckmann, Theodora Benedek, Patrick Donnelly, José Rodriguez-Palomares, Andrejs Erglis, Cyril Štěchovský, Gintarė Šakalytė, Filip Ađić, Matthias Gutberlet, Jonathan D Dodd, Ignacio Diez, Gershan Davis, Elke Zimmermann, Cezary Kępka, Radosav Vidakovic, Marco Francone, Małgorzata Ilnicka-Suckiel, Fabian Plank, Juhani Knuuti, Rita Faria, Stephen Schröder, Colin Berry, Luca Saba, Balazs Ruzsics, Christine Kubiak, Kristian Schultz Hansen, Jacqueline Müller-Nordhorn, Bela Merkely, Tem S Jørgensen, Imre Benedek, Clare Orr, Filipa Xavier Valente, Ligita Zvaigzne, Vojtěch Suchánek, Laura Zajančkauskienė, Milenko Čanković, Michael Woinke, Stephen Keane, Iñigo Lecumberri, Erica Thwaite, Michael Laule, Mariusz Kruk, Aleksandar N Neskovic, Massimo Mancone, Donata Kuśmierz, Gudrun Feuchtner, Mikko Pietilä, Vasco Gama Ribeiro, Tanja Drosch, Christian Delles, Bruno Loi, Michael Fisher, Bálint Szilveszter, Per E Sigvardsen, Mihaela Ratiu, Stephanie Kelly, Bruno Garcia del Blanco, Ainhoa Rubio, Zsófia D Drobni, Charlotte Kragelund, Ioana Rodean, Susan Regan, Hug Cuéllar Calabria, Melinda Boussoussou, Thomas Engstrøm, Roxana Hodas, Adriane E Napp, Robert Haase, Sarah Feger, Mahmoud M A Mohamed, Lina M Serna-Higuita, Konrad Neumann, Henryk Dreger, Matthias Rief, Viktoria Wieske, Melanie Estrella, Peter Martus, Marc Dewey.
Contributors: The DISCHARGE Trial Group investigators contributed to data collection, approved the manuscript after reviewing the findings and interpretation, provided critical revisions of the manuscript for important intellectual content, gave final approval of the version to be published, and are accountable for the work. RH and MD developed the concept and design of the study. KFK, MBosserdt, PM, and MD acquired, analysed, and interpreted the data. KFK, MBosserdt, and MD drafted the manuscript. PM and LMS-H did the statistical analysis. AEN, MBosserdt, ME, and MD supervised the trial. KFK and MBosserdt contributed equally. MD is the guarantor. The corresponding author affirms that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.
Funding: This study was funded by grants from the EU-FP7 Framework Program (FP 2007-2013, EC-GA 603266) to MD and others (Berlin Institute of Health (grant from Digital Health Accelerator); British Heart Foundation (Centre of Research Excellence RE/18/6/34217); Rigshospitalet, University of Copenhagen (grant and non-financial support); and German Research Foundation (grants from Radiomics Priority Programme: DE 1361/19-1  and 20-1  in SPP2177/1) and grants from graduate program BIOQIC (GRK 2260/1 )). The funder had no role in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; the preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/disclosure-of-interest/ and declare: funding from the EU-FP7 Framework Program. KFK, MBosserdt, PMH, NR, TB, PD, JRP, AE, CS, GS, FA, MG, JDD, ID, GD, EZ, CKępka, RV, MFrancone, MIS, FP, JK, RF, SS, CB, LS, BR, CKubiak, KSH, JMN, BM, TSJ, IB, FXV, LZ, VS, LZ, MC, MW, SK, IL, ET, MK, ANN, MM, GF, MP, VGR, TD, CD, BL, MFischer, BS, PES, MRatiu, SK, BGB, CKragelund, IR, SR, HCC, MBoussoussou, TE, RH, AEN, RH, SF, MMAM, KN, HD, MR, VW, ME, and MD received grants from the European Commission during the conduct of the study. KFK received grants and non-financial support from Rigshospitalet, University of Copenhagen, during the conduct of the study and grants from AP Møller og hustru Chastine McKinney Møllers Fond, the Danish Heart Foundation, the Danish Agency for Science, Technology and Innovation by the Danish Council for Strategic Research, and the Health Insurance Company Denmark and unrestricted research grants from Canon Medical Corporation and GE Healthcare outside the submitted work. PMH is a share holder of Neumann Medical Ltd outside the submitted work. NR is principal investigator for grants (no own salary) from the German Ministry of Education and Research (NAVICARE 01GY1911) outside the submitted work. TB has received grants from the Romanian Ministry of European Funds the Romanian Government and the European Union outside the submitted work. GS has received payments or honorariums from Amgen, Astra Zeneca, Bayer, Berlin Chemie Menarini Baltic, Clinical Financial Service, IQVIA, Novartis Baltics, Sanofi Aventis, and Servier Pharma; received support for attending meetings or travel from Servier and Novartis; and participated on data safety monitoring boards or advisory boards for Boehringer Ingelheim outside the submitted work. GS is fellow of the ECS and is involved in the Lithuanian Society of Cardiology, the Lithuanian Heart Association, and the Lithuanian Hypertension Society. MG has received payments or honorariums from Bayer, Siemens, Bracco, and the German Roentgen Society (DRG) to the institution outside the submitted work. MG is a member of the Scientific Committee of the ESCR and member of the Committee of the Working Group on Cardiovascular Imaging of the German Roentgen Society (DRG). EZ has received grants from the Deutsche Forschungsgemeinschaft (Radiomics in SPP 2177/1) outside the submitted work. JK has received speaker fees from GE Healthcare, Merck, Lundbeck, Bayer, Boehringer-Ingelheim, and Pfizer and personal fees for study protocol review from AstraZeneca and GE Healthcare outside the submitted work. JK is chair of the European Society of Cardiology guidelines on chronic coronary syndrome; the work is on a voluntary basis and only remuneration of travel expenses occurs. The University of Glasgow, which employs CB, who has no personal contracts and receives no payments outside of his employment, has received in kind support for research from AstraZeneca, grant and in kind support for research and R&D consultancy funding from Abbott Vascular, a grant from GSK, in kind support for research and R&D consultancy funding from HeartFlow, educational funding from Menarini, R&D consultancy funding from Novartis, and in kind support for research from Siemens, outside the submitted work; furthermore, the University of Glasgow received research funding from the British Heart Foundation (RE/18/6134217). BM has received personal fees from Biotronik, Medtronic, and Abbott and grants from Boston Scientific outside the submitted work. IB has received grants from the Romanian Ministry of European Funds, the Romanian Government, and the European Union outside the submitted work. MC has received direct payments from Abbott Vascular as lecturer, workshop participant as operator, and lecturer, Astra Zeneca as lecturer, and Boehringer Ingelheim, as well as payment to the institution from the National Science Center (Poland) outside the submitted work. MK has a granted patent (EP3157444B1) and pending patents (WO2015193847A1, WO2013060883A4). CD has received grants from the British Heart Foundation (Centre of Research Excellence RE/18/6/34217) during the conduct of the study. PES has received consulting fees from Novo Nordisk outside the submitted work. MR has received grants from the Deutsche Forschungsgemeinschaft (Radiomics in SPP 2177/1) outside the submitted work. TE has received personal fees from Abbott and Bayer outside the submitted work. MD has received grants from the Berlin Institute of Health (grant from Digital Health Accelerator), the German Research Foundation (grants from Radiomics Priority Programme: DE 1361/19-1  and 20-1  in SPP2177/1) and grants from the graduate program BIOQIC (GRK 2260/1 )) during the conduct of the study and grants from the Heisenberg Programme of the German Research Foundation, grants from fractal analysis of myocardial perfusion of the German Research Foundation, and a grant from the Berlin University Alliance (GC_SC_PC 27) outside the submitted work. In addition, MD holds a joint patent with Florian Michallek on dynamic perfusion analysis using fractal analysis (PCT/EP2016/071551 and USPTO 2021 10 991 109 approved) and is European Society of Radiology (ESR) research chair (2019-22) and elected ESR publications chair (2022-25); the opinions expressed in this article are the author’s own and do not represent the view of ESR; per the guiding principles of ESR, the work as research and publications chair is on a voluntary basis and only remuneration of travel expenses occurs. MD is also the editor of Cardiac CT, published by Springer Nature, and offers hands-on courses on computed tomography imaging (www.ct-kurs.de); institutional master research agreements exist with Siemens, General Electric, Philips, and Canon; the terms of these arrangements are managed by the legal department of Charité – Universitätsmedizin Berlin.
The lead author (KFK) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.
Dissemination to participants and related patient and public communities: We plan to share this work through multiple social media channels—for example, on Twitter: @team_dewey, @PalMaurovich, @kkofoed1, @RigsHeart, @JRodriPalomares, @lucasabaITA, @ColinBerryMD, @JuhaniKnuuti, @gudrunfeuchtner, @zsofidrobni, @Balint_CCTA, @AdrianeNapp, @UCDProfDodd, and @ProfDewey. On publication of the manuscript, appropriate communications will be made with the press and the trial website (https://www.dischargetrial.eu/) will be used to share results with appropriate citations to the work.
Provenance and peer review: Not commissioned; externally peer reviewed.
This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.