Types of studies

We included published and non‐published clinical trials involving the randomised (RCT) or quasi‐randomised (Q‐RCT) allocation (double‐blind, single‐blind or open‐label) of a uric acid (UA)‐lowering agent or placebo, with parallel or cross‐over study designs (with an adequate washout period of at least two weeks before cross‐over).

We only included studies with a minimum duration of four weeks of treatment with a UA‐lowering agent.

We did not include cluster‐randomised trials in this review.

Types of participants

We included participants of any gender, age, or ethnicity, with primary hypertension or prehypertension, plus hyperuricaemia. We defined primary hypertension as systolic blood pressure (BP) ≥ 140 mmHg, diastolic BP ≥ 90 mmHg, or both; prehypertension as systolic BP 120 mmHg to 139 mmHg or diastolic BP 80 mmHg to 89 mmHg in adults, and as average systolic or diastolic BP levels ≥ 90th percentile but < 95th percentile in children, or ≥ 120/80 mmHg in adolescents (Falkner 2004); and hyperuricaemia as serum UA greater than 6 mg/dL in women, 7 mg/dL in men, and 5.5 mg/dL in children and adolescents. In comparison to a previous version of this review, we expanded the inclusion criteria to cover studies enrolling people with prehypertension (Gois 2013a).

Types of interventions

UA‐lowering agents: either a xanthine‐oxidase inhibitor (allopurinol, febuxostat, topiroxostat or a pharmaceutically acceptable salt thereof) or a uricosuric agent (benzbromarone, benziodarone, probenecid, lesinurad, sulfinpyrazone, ethebencid, zoxazolamine, or a pharmaceutically acceptable salt thereof). We excluded ticrynafen, an uricosuric agent, because of its known direct effect on blood pressure.

Comparison: placebo

Types of outcome measures

Electronic searches

The Cochrane Hypertension Information Specialist conducted systematic searches in the following databases for randomised controlled trials, without language, publication year, or publication status restrictions:

• the Cochrane Hypertension Specialised Register via the Cochrane Register of Studies (CRS‐Web; searched 6 May 2020);

• the Cochrane Central Register of Controlled Trials (CENTRA) (2020, Issue 4) via the Cochrane Register of Studies (CRS‐Web; searched 11 May 2020);

• MEDLINE Ovid, MEDLINE Ovid Epub ahead of print, and MEDLINE Ovid In‐process & other non‐indexed citations (from 1946; searched 6 May 2020);

• Embase Ovid (from 1974; searched 6 May 2020);

• ClinicalTrials.gov (www.clinicaltrials.gov; searched 6 May 2020);

• World Health Organization International Clinical Trials Registry Platform (www.who.int/trialsearch; searched 6 May 2020);

• LILACS (Latin American and Caribbean Health Science Information database)(searched 6 May 2020).

The Information Specialist (IS) modelled subject strategies for databases on the search strategy designed for MEDLINE. Where appropriate, the IS combined them with subject strategy adaptations of the highly sensitive search strategy designed by Cochrane for identifying randomised controlled (as described in the Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0, Box 6.4.b. (Higgins 2011)). Please see Appendix 1 for the search strategies for major databases.

Searching other resources

• The Cochrane Hypertension Information Specialist searched the Hypertension Specialised Register segment (which includes searches of MEDLINE and Epistemonikos for systematic reviews) on 6 May 2020 to retrieve existing systematic reviews relevant to this review, so that we could scan their reference lists for additional trials. The Specialised Register also includes searches of Commonwealth Agricultural Bureau (CAB) Abstracts & Global Health, CINAHL (Cumulative Index to Nursing and Allied Health Literature), ProQuest Dissertations & Theses, and Web of Science.

• We searched the Database of Abstracts of Reviews of Effects (DARE) on 22 May 2020 for related reviews.

• We checked the bibliographies of included studies and any relevant systematic reviews identified for further references to relevant trials.

• Where necessary, we contacted authors of key papers and abstracts to request additional information about their trials.

Selection of studies

The two review authors independently selected the studies for evaluation by analysing titles and abstracts identified from the databases searched. We itemised the selection criteria in a customised checklist, which we used to select studies and exclude irrelevant reports. We retrieved the full text of potentially relevant reports for evaluation, and linked multiple reports of the same study under one study ID.

Data extraction and management

We created a data collection and abstraction form, and trained ourselves to use it. We performed a pilot test using 336 abstracts retrieved for the first version of this review. If a study was reported in more than one publication, we used only one data collection form for that study.

We encountered no disagreements about which studies to include in the review.

Assessment of risk of bias in included studies

Both review authors used the Cochrane 'Risk of bias' tool to assess the risk of bias in included studies (Higgins 2011). We evaluated the following items: randomisation; allocation concealment; blinding; incomplete outcome data reporting; selective reporting; and other biases (e.g. industry sponsorship).

Measures of treatment effect

For pooled continuous variables, we presented BP data (clinic‐measured and 24‐hour ambulatory systolic and diastolic BP) and serum UA as mean differences (MDs) with a 95% confidence interval (CI). For the dichotomous variable (adverse events), we presented the data as a risk ratio (RR) with a 95% CI.

Unit of analysis issues

All included trials used an intention‐to‐treat analysis, including all randomised participants.

For cross‐over trials that met the inclusion criteria, we aimed to use the first parallel‐group period in the analysis, whenever possible.

Dealing with missing data

We contacted Dr Daniel Feig, contact author of two included trials (Feig 2008; Soletsky 2012), requesting standard deviations for all changes in BP (clinical and 24‐hour ambulatory), and data from the first parallel‐group period in the cross‐over trial (Feig 2008). As we did not obtain the requested data, we used Review Manager (RevMan 5) software to calculate the standard deviations from the CIs (Review Manager 2014). For the same reason, we combined the data from both periods of Feig 2008 in our analysis, using the generic inverse variance method.

Assessment of heterogeneity

We calculated the heterogeneity of treatment effects between the trials in RevMan 5, using a standard Chi² test and the I² statistic. According to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), we used these thresholds to interpret the I²: 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; 75% to 100% indicates considerable heterogeneity.

Assessment of reporting biases

There were no unpublished trials with complete reporting to be included in the analysis.

Data synthesis

We carried out statistical analysis according to the guidelines referenced in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We conducted data synthesis and analysis using RevMan 5 software. Where studies were sufficiently similar in terms of inclusion criteria, interventions and/or outcomes, we pooled the data using meta‐analysis. We anticipated at least moderate heterogeneity among trials given different age groups of participants, UA‐lowering drugs and trial duration. For 24‐hour ambulatory and clinic‐measured BP, we used a random‐effects model to obtain summary statistics for estimates of effect given the substantial heterogeneity among trials (I² > 50%). For serum UA and adverse events, we used fixed‐effect model as the heterogeneity among trials were respectively moderate and not important.

Subgroup analysis and investigation of heterogeneity

We had planned to conduct subgroup analysis for adolescents and for different UA‐lowering drugs (e.g. allopurinol vs. placebo; febuxostat vs. placebo; and probenecid vs. placebo). We performed subgroup analysis for the outcome change in 24‐hour ambulatory systolic and diastolic BP in adolescents. For the outcome change in clinic‐measured systolic and diastolic BP, data was only available in adolescents. There were too few studies to perform subgroup analysis for different UA‐lowering drugs.

Sensitivity analysis

We were unable to perform sensitivity analysis in this review due to lack of data.

Summary of findings and assessment of the certainty of the evidence

Both review authors assessed the overall certainty of the evidence for each outcome, using the GRADE approach, as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). The certainty of the evidence on a specific outcome is based on the assessment of the studies against the following criteria: study design and limitations (risk of bias of a study); directness; consistency; precision of results; and publication bias. The overall GRADE rating starts at high when RCTs with a low risk of bias provide results for the outcome, and it is reduced by one level for each of the following criteria not met.

• High‐certainty evidence: at least 75% of RCTs present no limitations of study design, consistency, directness, and precision of the result. There are no known or suspected publication biases. Further research is unlikely to change the estimate or the confidence in the results.

• Moderate‐certainty evidence: one of the domains is not met. Further research is likely to have an important impact on the confidence in the estimate of effect, and may change the estimate.

• Low‐certainty evidence: two of the domains are not met. Further research is very likely to have an important impact on our confidence in the estimate of effect, and is likely to change the estimate.

• Very low‐certainty evidence: three of the domains are not met. We are very uncertain about the results.

• No evidence: No RCTs were identified that addressed this outcome.

For assessments of the overall certainty of evidence for each outcome that included pooled data from RCTs only, we downgraded the evidence from ’high certainty’ by one level for serious (or by two for very serious) study limitations (risk of bias), indirectness of evidence, serious inconsistency, imprecision of effect estimates, or potential publication bias. We did not assess the evidence for any outcomes as moderate‐ or high‐certainty, since no pooled estimates revealed a large magnitude of effect.

We used GRADEpro GDT software to import data from RevMan 5 and create a 'Summary of findings’ table (GRADEpro GDT; Higgins 2011). This table provides outcome‐specific information concerning the overall certainty of evidence from studies included in the comparison, the magnitude of effect of the interventions examined, and the sum of available data for the outcomes evaluated. We included the following outcomes in the 'Summary of findings' table: 24‐hour systolic BP; 24‐hour diastolic BP; clinic‐measured systolic BP; clinic‐measured diastolic BP; serum UA; and withdrawals due to adverse effects (summary of findings Table 1).