Types of studies

We will include randomised trials (RCTs), non‐randomised trials, controlled before‐after studies (CBAs) and interrupted time series studies (ITS) in any language. To be included, controlled before‐after studies require two intervention and two control sites and interrupted time series studies need a clearly‐defined point in time when the intervention occurs and three measuring points before and after the intervention. We do not expect to identify a broad range of evidence based on randomised trials alone, and we therefore choose to include additional study designs that allow us to explore causal relationships. We will exclude reviews from the analysis, but will screen their reference lists for possibly relevant articles, because we want to include primary studies only in this review.

Types of participants

We will include all trials, regardless of the setting, which recruited adult patients (18 years or older), for whom an ethical issue arose and who subsequently received an ethical case intervention.

Types of interventions

We will include ethical case interventions of any form (e.g. moral case deliberation, ethics round, ethics committee). At least one person involved must have had expertise in medical ethics. We will compare the intervention with 'usual care' (e.g. no specific interventions defined) or another active control. We will exclude studies which focus solely on the implementation of ethical case interventions, on the topic of research ethics (e.g. research ethics committees) or policy‐making.

Types of outcome measures

Ethical case interventions are complex interventions. We will therefore take several outcomes into account to measure effectiveness. We will extract any additional outcomes found in the studies.

Main outcomes

Quality of Care:

• Decisional conflict (self‐reported reduction of decisional conflict in patients, relatives or healthcare professionals, as measured with validated scales (e.g. Decisional Conflict Scale (O'Connor 1995)))

• Moral distress (of the stakeholders involved in an ethical conflict in clinical practice (i.e. patients, relatives and healthcare providers) as measured with validated scales (e.g. Moral Distress Scale (Corley 2001)))

• Patient involvement in decision‐making (measured with validated scales (e.g. Dyadic OPTION Scale (Melbourne 2010)))

Patient Outcomes:

• Health‐related quality of life (measured with validated scales (e.g. 36‐Item Short Form Survey (SF36) (Ware 2001)))

Other outcomes

Knowledge

• Ethical competency (of the healthcare providers participating in ethics consultation, measured according to the knowledge domains defined by the American Society for Bioethics and Humanities’ (ASBH) Core Competencies Update Task Force (Tarzian 2013))

Satisfaction

• Satisfaction with care (of the stakeholders (i.e. patients, relatives and healthcare providers), as measured with validated scales (e.g. FamCare Scale (Kristjanson 1993)))

• Satisfaction with ethics consultation (of the stakeholders (i.e. patients, relatives and healthcare providers), as measured with objective outcome measures (e.g. McClung 1996))

Resource Use

• Includes treatment costs, measured as combined costs for treatment

Electronic searches

We will search the following databases:

• Cochrane Central Register of Controlled Trials (CENTRAL), including the EPOC Group Specialised Register, 1997 onwards

• MEDLINE, 1946 onwards, in‐process and other non‐indexed citations, OvidSP

• Embase, 1947 onwards, OvidSP

• CINAHL (Cumulative Index to Nursing and Allied Health Literature), 1980 onwards, EbscoHost

• PsycINFO, 1887 onwards, OvidSP

We performed scoping searches on the ethics‐specific databases Ethxweb and Ethicsweb, leading to our decision not to take them into account, due to the lack of study designs of interest in these databases and their overlap with MEDLINE.

Search strategies are composed of keywords and controlled vocabulary terms. We will not apply language limits. We will use two methodology search filters to limit retrieval to appropriate study designs: a modified version of the Cochrane Highly Sensitive Search Strategy (sensitivity‐ and precision‐maximising version, 2008 revision) (Lefebvre 2011) to identify randomised trials (Higgins 2011); and an EPOC methodology filter to identify non‐randomised designs. See Appendix 1 for the MEDLINE search strategy, which we will adapt for other databases.

We will document the search strategies and processes in a standardised search log (EPOC 2014).

Searching other resources

We will also:

1. search the reference lists of included studies and any systematic reviews identified through the searches for possibly relevant articles;

2. handsearch the Ethik in der Medizin journal for relevant articles;

3. consult with experts; we will contact the first author of each included study, as well members of the European Clinical Ethics Network (ECEN) for further retrieval of relevant studies;

4. search for ongoing studies in the Clinicaltrials.gov website (www.clinicaltrails.gov) and the International Clinical Trials Registry Platform Search Portal (ICTRP) (www.who.int/trialsearch);

5. conduct cited reference searches for all included studies in ISI WEB of Science (Science Citation Index and Social Science Citation Index).

Selection of studies

Review authors involved in the selection and data extraction process (JH, JS) will be trained by an experienced researcher in systematic reviews (SN) and supervised by two experts (CB and MG). We will import the results of the electronic searches to Covidence, removing duplicate records. Two review authors (SN, JH) will independently screen titles and abstracts of the retrieved records for inclusion, using a study selection form in Covidence. We will code all potentially eligible studies as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. We will retrieve the full‐text articles and two review authors (SN, JH) will independently rate each as 'irrelevant', 'relevant' or 'maybe'. We will use a 'maybe' study if it provides insufficient information to make a conclusive decision about study eligibility. In these instances, we will contact the study authors for clarification, resolving any disagreement through consultation with a third review author (JS). We will list excluded studies with reasons for exclusion in the 'Characteristics of excluded studies' tables. We will collate multiple reports for the same study, so that each study rather than each report is the unit of interest. We will report the selection process in sufficient detail to complete a PRISMA Flow diagram (Liberati 2009).

Data extraction and management

We will use a modified data collection form (EPOC 2013a) for study characteristics and outcome data, which has been piloted on at least one study in the review. Two review authors (SN, JH) will independently extract the following study characteristics data from included studies:

1. Methods: study design, number of study centres and location, study setting, withdrawals, date of study, follow‐up

2. Participants: number, mean age, age range, gender, severity of condition, diagnostic criteria, inclusion criteria, exclusion criteria, other relevant characteristics

3. Interventions: intervention components, comparison, fidelity assessment, conceptual framework/theory underlying the intervention

4. Outcomes: main and other outcomes specified and collected, time points reported, outcome measures

5. Notes: funding for trial, notable conflicts of interest of trial authors, ethical approval

6. Outcome Reporting: Outcomes reported in study register (Appendix 2) or trial protocol

Two review authors (SN, JH) will independently extract outcome data from included studies. We will note in the 'Characteristics of included studies' table if outcome data were reported in an unusable way. We will seek relevant missing information on the trial from the corresponding author of the article, if required. We will resolve disagreements by discussion in the review team until we reach consensus.

Assessment of risk of bias in included studies

Two review authors (SN, JH) will independently assess the methodological quality of the included studies, using the 'Risk of bias' criteria and qualifiers outlined in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) and the EPOC guidance (EPOC 2015a) to judge whether a study has a low, high, or unclear risk of bias for each domain. We will provide a quote from the study report together with a justification for our judgement in the 'Risk of bias' table. We will resolve disagreements in our judgements by discussion in the review team.

We will use the following nine standard criteria to assess bias for randomised and non‐randomised trials and controlled before‐after studies:

1. Was the allocation sequence adequately generated?

2. Was the allocation adequately concealed?

3. Were baseline outcome measurements similar?

4. Were baseline characteristics similar?

5. Was the study adequately protected against contamination?

6. Were incomplete outcome data adequately addressed?

7. Was knowledge of the allocated interventions adequately prevented during the study?

8. Was the study free from selective outcome reporting?

9. Was the study free from other risks of bias?

For interrupted time series studies, we will use the following seven criteria to assess bias:

1. Was the intervention independent of other changes?

2. Was the shape of the intervention effect prespecified?

3. Was the intervention unlikely to affect data collection?

4. Was knowledge of the allocated interventions adequately prevented during the study?

5. Were incomplete outcome data adequately addressed?

6. Was the study free from selective outcome reporting?

7. Was the study free from other risks of bias?

We will assess selective reporting through outcome reporting bias (Kirkham 2010) by integrating the data of the 'Matrix of study endpoints (trials register)' (Appendix 2), the outcomes reported in the publication and the results of the 'modified ORBIT study classification table' (Appendix 3)

We will summarise the 'Risk of bias' judgements across different studies for each of the domains listed. Where information on risk of bias relates to unpublished data or correspondence with a trialist, we will note this in the 'Risk of bias' table. When considering treatment effects, we will take into account the risks of bias for the studies that contribute to that outcome.

The risk of bias judgement will contribute to the GRADE process for recommendations in the data synthesis.

Measures of treatment effect

We will estimate the effect of the intervention using risk ratios (RRs) with their 95% confidence intervals (CIs) for dichotomous data, and standardised mean differences (SMDs) with their 95% CIs in case of continuous data. We will ensure that an increase in scores for continuous outcomes can be interpreted in the same way for each outcome, explaining the direction to the reader and reporting where the directions were reversed if this proves necessary. As part of the structured analysis, we aim to provide (along with a possible meta‐analysis) subgroup analyses for possible gender‐ and age‐associated effects. For the measurement of treatment effect for ITS studies, we will conduct regression analysis with time trends adjusted for autocorrelation before and after the intervention. We plan to use the change in the level of outcome at the first point of measurement after implementing the intervention and the post‐intervention slopes minus the pre‐intervention slopes.

Unit of analysis issues

Analyses for clustered studies should be done at the same level as allocation. Clustered studies in which individuals are randomised or allocated to intervention groups have to take into account correlation of observations within the clusters. This is done with the intra‐cluster correlation (ICC) analysis. If this is not reported, we will analyse the correlation (ICC) by contacting the study's authors for the necessary information. If this is not possible we will report the effect size with a 'unit of analysis' error.

Dealing with missing data

We will contact investigators by e‐mail to verify key study characteristics and obtain missing outcome data where possible (e.g. when a study is reported as an abstract only). We will evaluate important numerical data such as screened, randomised participants and intention‐to‐treat, as‐treated and per‐protocol populations. We will also investigate attrition rates. If data are still missing, we will report them as such. We will use available results for analysis, if results are not biased by that fact.

Assessment of heterogeneity

We will assess heterogeneity graphically with forest plots and statistically with the I² statistic. I² values of 25%, 50% and 75% correspond respectively to low, medium and high levels of heterogeneity (Higgins 2003). If we identify substantial clinical heterogeneity, we will explore it by prespecified subgroup analysis.

Assessment of reporting biases

We will try to contact study authors for missing outcome data. Where this is not possible, and the missing data are thought to introduce serious bias, we will explore the impact of including such studies in the overall assessment of results by a sensitivity analysis. If we can pool more than 10 trials, we will create and examine a funnel plot to explore possible publication biases, interpreting the results with caution (Sterne 2011).

Data synthesis

We will undertake meta‐analysis only where this is meaningful, for example if the models of ethical case interventions, participants, the underlying clinical question and outcome measurement are similar enough for pooling to make sense. A common way for trialists to indicate skewed data is by reporting medians and interquartile ranges. When we encounter this, we will note that the data are skewed and consider the implication of this. Where a single trial reports multiple trial arms, we will include only the relevant arms. If two comparisons (e.g. intervention A versus usual care and intervention B versus usual care) are entered into the same meta‐analysis, we will halve the control group to avoid double‐counting.

We will use Review Manager 5 (RevMan 2014) software for meta‐analysis.

We expect heterogeneity and variability between the studies due to different outcomes measures, interventions and population of the complex ethical case intervention. We will therefore assume that the true effect is related but not necessarily the same in different studies, and will use a random‐effects model for analysis. We will apply the Mantel‐Haenszel method for dichotomous outcomes and generic inverse variance for continuous outcomes. We will treat ordinal outcomes as dichotomous or continuous data, depending on the way the study authors performed the original analysis (Higgins 2011).

In a structured analysis focusing on how and why ethical case interventions work, following the Template for Intervention Description and Replication (TiDieR) rationale (Hoffmann 2014) and the elements of the conceptual frameworks for ethics consultation (Schildmann 2016), we will present a preliminary synthesis of findings from the included studies, an exploration of relationships within and between studies, and an assessment of the robustness of the synthesis. We will also produce a graphical synthesis, in the form of a map, to illustrate different outcomes measured in the studies. Subsequently we will analyse the appropriateness of the outcomes and measures with respect to the underlying intervention and theoretical assumptions.

Subgroup analysis and investigation of heterogeneity

We will conduct subgroup analyses, if the structured analysis yields sufficient studies. Possible subgroups are partly derived from our conceptual framework (Schildmann 2016) and will consist of different diseases (e.g. cancer, stroke), different settings (e.g. psychiatric care, hospice care), different models of ethical case interventions (e.g. ethical rounds, ethical committee etc.), studies with active control versus no active control. We will explore this visually (e.g. box‐plots or bubble plots, or both) or through meta‐regression for main outcomes only. If meta‐analysis is not possible, we will conduct a visual table analysis for effect measures and explanatory factors, guided by the conceptual frameworks.

Sensitivity analysis

We will perform meta‐analysis with and without studies with outlying results, with and without unpublished results, large studies only, with and without active control, and including all studies or only those at low risk of bias. We will perform a sensitivity analysis for different assumptions about missing values with imputation,