Types of studies

We included only RCTs comparing complete revascularisation versus culprit‐only PCI strategy in people with STEMI and MVD. We included studies reported as full‐text, those published as abstract, and unpublished data.

Types of participants

We included adults aged 18 years and above, who underwent P‐PCI for the management of STEMI, with concurrent non‐culprit significant lesions (as defined by the authors of the trial), identified at the time of the index procedure.

Types of interventions

We included studies that looked at a population of people presenting with STEMI and who were initially treated with a P‐PCI and coronary angiography to assess the extent of coronary vessel obstruction in the various branches of the coronary tree. Further, in these people, the culprit and the non‐culprit arteries were identified based on correlation with participants' changes on electrocardiograms. Following initial treatment of the culprit lesion, participants in these RCTs were randomised to receive either complete or culprit vessel only intervention. A comparison was then made between these two strategies, defined as:

1. Complete revascularisation strategy: involving additional revascularisation of the clinically significant stenotic non‐culprit lesion(s) (at least 50% obstruction but less than 100%) at the index procedure or in a second intervention. This was carried out in all eligible participants, except if contraindicated;

2. Culprit vessel‐only intervention: involving no additional invasive approach but rather medical management of all eligible participants, even with evidence of non‐culprit artery disease diagnosed at the time of P‐PCI. Exceptions were made if the participant had clinical symptoms that warranted further testing and intervention (e.g. new‐onset chest pain, dynamic changes on electrocardiogram, or a positive stress test).

Types of outcome measures

Electronic searches

On 4 January 2017, we searched the following sources from their inception and we imposed no restriction on language of publication:

1. Cochrane Central Register of Controlled Trials (CENTRAL): 2016, Issue 11 (Wiley);

2. Database of Abstracts of Reviews of Effects (DARE): Issue 2 of 4, April 2015 (Wiley);

3. Health Technology Assessment Database (HTA): Issue 4 of 4, October 2016 (Wiley);

4. Ovid MEDLINE(R) 1946 to December Week 1 2016, Ovid MEDLINE(R) In‐Process & Other Non‐Indexed Citations 3 January 2017 and Ovid MEDLINE(R) Epub Ahead of Print 3 January 2017;

5. Embase 1974 to 4 January 2017; Embase Classic 1947 to 1973; MEDLINE 1966 to 4 January 2017 (embase.com);

6. Conference Proceedings Citation Index‐Science (CPCI‐S) 1990 to 4 January 2017 (Web of Science).

We applied the Cochrane sensitivity‐maximizing RCT filter (Lefebvre 2011) to MEDLINE (Ovid) but did not apply it to the MEDLINE In‐Process or Epub searches. For Embase, we used the multi‐term Embase filter with the best balance of sensitivity and specificity (Wong 2006) translated from Ovid to embase.com syntax.

For details of terms used in search strategies please see Appendix 1.

Searching other resources

In order to identify articles potentially missed through the electronic searches, we:

1. handsearched reference lists of all included studies and relevant reviews retrieved by electronic searching to identify other potentially eligible trials or ancillary publications;

2. conducted a search on 8 December 2016 for other systematic reviews and Health Technology Assessment reports in Epistemonikos (www.epistemonikos.org);

3. handsearched on 8 December 2016, conference proceedings via their websites for 2011 to 2016 (congress365.escardio.org), ACC Annual Scientific Sessions (www.onlinejacc.org/content/meeting‐abstract‐supplements), AHA Annual Scientific Sessions (circ.ahajournals.org), and Transcatheter Cardiovascular Therapeutics Abstracts (www.tctmd.com);

4. contacted corresponding authors of included studies for any additional published or unpublished data;

5. attempted to contact the authors of trials when information in the study report was lacking or unclear.

We also searched the following trial registers:

1. ClinicalTrials.gov (www.clinicaltrials.gov/);

2. European (EU) Clinical Trials Register (www.clinicaltrialsregister.eu/);

3. WHO International Clinical Trials Registry Platform (apps.who.int/trialsearch/).

Selection of studies

Two review authors (SH and CB) independently screened titles and abstracts for inclusion of all the potential studies that the search identified and coded them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. If there were any disagreements, a third review author arbitrated. We retrieved the full‐text study reports/publication and two review authors (SH and CB) independently screened the full‐text and identified studies for inclusion, and identified and recorded reasons for exclusion of the ineligible studies. We resolved any disagreements through discussion or, if required, we consulted a third review author. We identified and excluded duplicate publications and collated multiple reports of the same study so that each study, rather than each report, became a unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram (Figure 1) and Characteristics of excluded studies table.

Figure 1

---

---

Study flow diagram.

Data extraction and management

We used a data collection form for study characteristics and outcome data that had been piloted on at least one study in the review. One review author (CB) extracted study characteristics from included studies including:

1. methods: study design, total duration of study, details of any 'run‐in' period, number of study centres and location, study setting, withdrawals, and date of study;

2. participants: number, mean age, age range, gender, severity of the condition, diagnostic criteria, inclusion criteria, and exclusion criteria;

3. interventions: intervention, comparison, concomitant medications, and excluded medications;

4. outcomes: primary and secondary outcomes specified and collected, and time points reported;

5. notes: funding for trial and notable conflicts of interest of trial authors.

Two review authors (CB and SH) independently extracted outcome data from included studies. We resolved any disagreements by consensus or by involving a third review author. One review author (CB) transferred data into the Review Manager 5 (RevMan 2014). We double‐checked that data were entered correctly by comparing the data presented in the systematic review with those of the study reports. A second review author (SH) spot‐checked study characteristics for accuracy against the trial report.

Assessment of risk of bias in included studies

Two review authors (SH and CB) independently assessed the risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We resolved any disagreements by discussion or by involving a third review author. We assessed the risk of bias according to the following domains (Wood 2008).

1. Random sequence generation.

2. Allocation concealment.

3. Blinding of participants and personnel.

4. Blinding of outcome assessment.

5. Incomplete outcome data.

6. Selective outcome reporting.

7. Other bias (e.g. industry funding).

We graded each potential source of bias as high, low, or unclear and provide a quote from the study report together with a justification for our judgement in the 'Risk of bias' table. We summarised the risk of bias judgements across different studies for each of the domains listed. Wherever the information on the risk of bias related to unpublished data or correspondence with a trialist, we noted the same in the 'Risk of bias' table.

We conducted the review according to the published protocol (Hirji 2015), and reported any deviations from it under the Differences between protocol and review. When considering treatment effects, we took into account the risk of bias for the studies that contributed to such an outcome.

Measures of treatment effect

We analysed dichotomous data as risk ratios (RR) with 95% confidence intervals (CI) and continuous data as mean difference (MD) or standardised mean difference (SMD) with 95% CI. SMD will be used if the outcome is measured in a variety of ways, for instances with different scales and MD will be used if the outcomes is measured with the same method. We entered data presented as a scale with a consistent direction of effect. We described skewed data reported as medians and interquartile ranges narratively.

Unit of analysis issues

All included trials were randomised at the individual participant level.

Dealing with missing data

We contacted investigators to verify key study characteristics and obtain missing numerical outcome data wherever possible (e.g. when a study was published as abstract only). Where this was not feasible, and the missing data were thought to introduce serious bias, we explored the impact of including such studies on the overall assessment of results using a sensitivity analysis.

Assessment of heterogeneity

We examined heterogeneity using the I² statistic, which quantifies inconsistency across studies to assess the impact of heterogeneity on the meta‐analysis, with an I² statistic of 50% or more indicative of a considerable level of inconsistency.

If we identified substantial heterogeneity, we reported it and explored possible causes by prespecified subgroup analysis and performed the analysis using a random‐effects model. In the event of substantial clinical, methodological, or statistical heterogeneity, we decided not to report study results as pooled effect estimates.

Assessment of reporting biases

We were unable to assess reporting bias as the number of included studies was insufficient for an informative funnel plot (Higgins 2011).

Data synthesis

If there was evidence for homogeneous effects across studies, we analysed the data using RR and summarised all data using a fixed‐effect model (Riley 2011; Wood 2008). In addition, we performed statistical analyses according to the statistical guidelines contained in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

Subgroup analysis and investigation of heterogeneity

We carried out the following subgroup analyses and investigated interaction.

1. Drug‐eluting stent (DES) compared to bare‐metal stents (BMS).

2. Sex.

3. People with diabetes mellitus compared to people without diabetes mellitus.

4. Non‐culprit and culprit intervention during the same procedure compared to in separate interventions (staged).

5. Low risk of bias articles compared to high risk of bias articles.

6. Participants in cardiogenic shock compared to participants not in cardiogenic shock.

We used the following outcomes in subgroup analyses.

1. Long‐term all‐cause mortality.

2. Long‐term cardiovascular mortality.

3. Long‐term non‐fatal myocardial infarction.

We used the formal test for subgroup interactions in Review Manager 5 (RevMan 2014).

Sensitivity analysis

We performed sensitivity analyses to explore the influence of the following factors on effect sizes.

1. Restricting the analysis to published studies.

2. Restricting the analysis to trials at low risk of bias, as specified in Assessment of risk of bias in included studies.

3. Restricting the analysis to trials using the following filters: language of publication, source of funding (industry), and country.

4. Restricting the analysis to published trials that utilised mostly DES.