Types of studies

We included all RCTs of any design (e.g. parallel, cross‐over, factorial) and controlled clinical studies using a quasi‐randomised method of allocation (such as alternation, use of alternate medical records, date of birth, or other predictable methods). Reports of studies were eligible regardless of the language used or date of publication. We included studies published as full articles or available as a full study report, studies in which results were published on a trial registry, or studies published as conference proceedings.

Types of participants

Types of interventions

We included studies that evaluated potassium binders defined as potassium exchange resins that act to bind potassium within the GI tract including calcium polystyrene sulfonate, sodium polystyrene sulfonate with or without sorbitol, patiromer sorbitex calcium, and ZS‐9. Control of serum potassium levels could be achieved regardless of the route of administration, duration, frequency, or dose of the potassium binders. We included interventions administered orally or rectally.

Comparator treatments could be any of the following.

• Placebo

• Usual care (best supportive care)

• Second potassium binder

• Dietary restriction

• Withdrawal of RAAS inhibition

• Different doses of a potassium binder

• Different routes of administration

• Different frequency of administration.

Types of outcome measures

The outcomes selected included the relevant SONG core outcome sets as specified by the Standardised Outcomes in Nephrology initiative (SONG 2017).

Electronic searches

We searched the Cochrane Kidney and Transplant Register of Studies up to 10 March 2020 through contact with the Information Specialist using search terms relevant to this review. The Register contains studies identified from the following sources.

1. Monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL)

2. Weekly searches of MEDLINE OVID SP

3. Searches of kidney and transplant journals, and the proceedings and abstracts from major kidney and transplant conferences

4. Searching of the current year of EMBASE OVID SP

5. Weekly current awareness alerts for selected kidney and transplant journals

6. Searches of the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov.

Studies contained in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE based on the scope of Cochrane Kidney and Transplant. Details of search strategies, as well as a list of handsearched journals, conference proceedings and current awareness alerts, are available on the Cochrane Kidney and Transplant website.

See Appendix 1 for search terms used in strategies for this review.

Searching other resources

1. Reference lists of review articles, relevant studies, and clinical practice guidelines.

2. Contacting relevant individuals/organisations seeking information about unpublished or incomplete studies.

3. Grey literature sources (e.g. abstracts, dissertations, and theses), in addition to those already included in the Cochrane Kidney and Transplant Register of Studies, were searched.

Selection of studies

The search strategy described was used to obtain titles and abstracts of studies that may be relevant to the review. The titles and abstracts were screened independently by two authors, who discarded studies that were not applicable. Studies and reviews that might include relevant data or information on studies were retained initially. Two authors independently assessed retrieved abstracts and, if necessary, the full text, of these studies to determine which studies satisfy the predetermined inclusion criteria. The review authors resolved discrepancies through discussion or adjudication by a third author.

Data extraction and management

Data extraction was carried out independently by two authors using a standard data extraction form developed for this review. The review authors resolved discrepancies through discussion or adjudication by a third author. Studies reported in non‐English language journals were translated before assessment. Where more than one publication of one study existed, reports were grouped together and the publication with the most complete data was used in the analyses. Where relevant outcomes were only published in earlier versions these data were used. Any discrepancy between published versions was highlighted.

For each included study, we recorded the following:

• Study characteristics including type (e.g. parallel, cross‐over, factorial), country, source of funding, and trial registration status (with registration number recorded if available)

• Participant characteristics including age, sex, stage of CKD, inclusion criteria, exclusion criteria

• Intervention characteristics for each treatment group, and use of co‐interventions

• Outcomes reports including the measurement instrument used and timing of outcome assessment

To prevent/minimise selective inclusion of data based on the results, we used the following a priori defined decision rules to select data from studies.

• Where trial lists report both final values and change from baseline for the same outcome, we extracted final values

• Where trial lists report both unadjusted and adjusted values for the same outcome, we extracted unadjusted values

• Where trial lists report both data analysed on the intention‐to‐treat principle and another sample (e.g. per protocol, as treated), we extracted intention‐to‐treat data

For cross‐over studies, we extracted data from the first period only.

Assessment of risk of bias in included studies

The following items were independently assessed by two authors using the risk of bias assessment tool (Higgins 2011) (see Appendix 2).

• Was there adequate sequence generation (selection bias)?

• Was allocation adequately concealed (selection bias)?

• Was knowledge of the allocated interventions adequately prevented during the study?

  • Participants and personnel (performance bias)

  • Outcome assessors (detection bias)

• Were incomplete outcome data adequately addressed (attrition bias)?

• Are reports of the study free of suggestion of selective outcome reporting (reporting bias)?

• Was the study apparently free of other problems that could put it at a risk of bias?

Measures of treatment effect

For dichotomous outcomes (death (any cause), myocardial infarction, stroke, cardiac arrhythmia, hospitalisation, dialysis vascular access complications, kidney transplant graft outcomes, cancer, infection, withdrawal of blood pressure lowering therapy, adverse events), results were expressed as risk ratio (RR) with 95% confidence intervals (CI). Where continuous scales of measurement were used to assess the effects of treatment (HRQoL, serum potassium, life participation scales, depression, kidney transplant function) the mean difference (MD) was used, or the standardised mean difference (SMD) if different scales have been used.

Unit of analysis issues

Dealing with missing data

Any further information required from the original author was requested by written correspondence (e.g. emailing corresponding author/s) and any relevant information obtained in this manner was included in the review. Evaluation of important numerical data such as screened, randomised patients as well as intention‐to‐treat, as‐treated and per‐protocol population were carefully performed. Attrition rates, for example drop‐outs, losses to follow‐up and withdrawals were investigated. Issues of missing data and imputation methods (for example, last‐observation‐carried‐forward) were critically appraised (Higgins 2011).

If the number of patients analysed was not presented for each time point, we considered the number of randomised patients in each group at baseline. For continuous outcomes with no SD reported, we calculated SDs from standard errors (SEs), 95% CIs or P‐values using the calculator tool in RevMan. If no measures of variation were reported and SDs could not be calculated, we imputed SDs from other studies in the same meta‐analysis, using the median of the other SDs available. For continuous outcomes presented only graphically, we extracted the mean and 95% CIs from the graphs using plotdigitizer (http://plotdigitizer.sourceforge.net/). For dichotomous outcomes, we used percentages to estimate the number of events or the number of people assessed for an outcome. Where data were imputed or calculated (e.g. SDs calculated from SEs, 95% CIs or P‐values, or imputed from graphs or from SDs in other studies), we reported this in the Characteristics of included studies.

Assessment of heterogeneity

We assessed clinical heterogeneity by determining whether the characteristics of participants, interventions, outcome measures, and timing of outcome measurement are similar across studies. We assessed the heterogeneity by visual inspection of the forest plot. We quantified statistical heterogeneity using the I² statistic, which describes the percentage of total variation across studies that is due to heterogeneity rather than sampling error (Higgins 2003).

We interpreted the I² statistic using the following as an approximate guide:

• 0% to 40%: may not be important heterogeneity

• 30% to 60%: may represent moderate heterogeneity

• 50% to 90%: may represent substantial heterogeneity

• 75% to 100%: considerable heterogeneity.

The importance of the observed value of I² depends on the magnitude and direction of treatment effects and the strength of evidence for heterogeneity (e.g. P‐value from the Chi² test, or a confidence interval for I²) (Higgins 2011).

In the case of considerable heterogeneity, we explored the data further by comparing the characteristics of individual studies and any subgroup analyses.

Assessment of reporting biases

To assess publication bias, we planned to generate funnel plots if at least 10 studies examining the same treatment comparison were included in the review, and comment on whether any asymmetry in the funnel plot was due to publication bias, or methodological or clinical heterogeneity of the studies. To assess for potential small‐study effects in meta‐analysis (i.e. the intervention effect is more beneficial in smaller studies), we planned to compare effect estimates derived from a random‐effects model and a fixed‐effect model of meta‐analysis. In the presence of small‐study effects, the random‐effects model could give a more beneficial estimate of the intervention than the fixed‐effect estimate.

To assess outcome reporting bias, we compared the outcomes specified in study protocols with the outcomes reported in the corresponding study publications.

Data synthesis

Data were pooled using the random‐effects model but the fixed‐effect model was also used to ensure robustness of the model chosen and susceptibility to outliers.

Subgroup analysis and investigation of heterogeneity

We planned to perform subgroup analysis to explore possible sources of heterogeneity (where sufficient data are available).

• Stage of CKD

• Co‐prescribing of RAAS inhibitors

However, subgroup analysis could not be done for heterogeneity owing to insufficient data.

Adverse effects were tabulated and assessed with descriptive techniques, as they were likely to be different for the various agents used. Where possible, the risk difference (RD) with 95% CI was calculated for each adverse effect, either compared to no treatment or to another agent.

Sensitivity analysis

We planned to perform sensitivity analyses in order to explore the influence of the following factors on effect size.

• Repeating the analysis excluding unpublished studies

• Repeating the analysis taking account of risk of bias, as specified

• Repeating the analysis excluding any very long or large studies to establish how much they dominate the results

• Repeating the analysis excluding studies using the following filters: diagnostic criteria, language of publication, source of funding (industry versus other), and country

However, sensitivity analysis could not be done for heterogeneity owing to insufficient data.

'Summary of findings' tables

We presented the main results of the review in 'Summary of findings' tables. These tables present key information concerning the quality of the evidence, the magnitude of the effects of the interventions examined, and the sum of the available data for the main outcomes (Schunemann 2011a). The 'Summary of findings' tables also include an overall grading of the evidence related to each of the main outcomes using the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) approach (GRADE 2008; GRADE 2011). The GRADE approach defines the quality of a body of evidence as the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. The quality of a body of evidence involves consideration of within‐trial risk of bias (methodological quality), directness of evidence, heterogeneity, precision of effect estimates and risk of publication bias (Schünemann 2011b).

We presented the following outcomes in the 'Summary of findings' tables.

• Death (any cause)

• Cardiovascular death

• Cardiac arrhythmia

• HRQoL

• Serum potassium

• Major GI adverse events or minor GI adverse events (constipation)