Types of studies

We will include randomized, placebo‐controlled trials (RCTs). Randomization ensures an equal likelihood of receiving either allocated intervention, and should result in a more balanced distribution of confounding characteristics. We will exclude cluster‐randomized studies. For cross‐over RCTs, we will analyze the data produced before the first cross‐over point, but not subsequent data because of the risk of carry‐over effects. We will include studies reported as full‐text and unpublished data.

Types of participants

We will include participants of any age, sex, and ethnicity, with and without evidence of cardiovascular disease. They can have normal lipid parameters or any type of hyperlipidemia (high cholesterol) or dyslipidemia (an abnormal balance between good and bad cholesterol levels without either being out of range). We will accept participants with various comorbid conditions, including type 2 diabetes mellitus, hypertension, metabolic syndrome, or chronic renal failure. We will include studies with subsets of eligible participants, only if data for that subset is reported separately. If not, we will contact study authors to request that data.  

Types of interventions

Lovastatin and placebo must be administered orally at a constant daily dose. We will include any dose of lovastatin, with a treatment duration of three to 12 weeks. We have chosen this time frame as it is long enough to allow sufficient time for a steady‐state effect of lovastatin to occur, but short enough to minimize participants dropping out. We will include studies where lovastatin is administered in the morning or evening, or where timing of administration is not specified. Studies must have a washout baseline dietary stabilization period of at least three weeks, in which all previous lipid‐altering medication is withdrawn. This baseline phase ensures participants follow a standard lipid‐regulating diet and helps to stabilize baseline lipid values prior to treatment. In studies where participants were not receiving lipid‐altering medications or dietary supplements before receiving the test drug, we will not require washout baseline dietary stabilization periods. Participants must not receive other interventions that interfere with lipid metabolism. We will list all treatment arms for each study in the 'Characteristics of included studies' table.

Types of outcome measures

We will include studies that meet the above inclusion criteria regardless of whether they report the following outcomes. For lipid parameters, we will report outcomes at 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12 weeks.  Treatment‐emergent adverse events will be documented with no time limit. We chose LDL cholesterol as our primary outcome measure because a decrease in blood LDL cholesterol is thought to result in a decrease in atherogenesis (the process of forming plaques in the intima layer of arteries)(Taylor 2013).

Electronic searches

The Cochrane Hypertension Information Specialist (IS) will search the following databases without language, publication year, or status restrictions: 

• Cochrane Hypertension Specialised Register via the Cochrane Register of Studies;

• Cochrane Central Register of Controlled Trials (CENTRAL) via the Cochrane Register of Studies;

• Ovid MEDLINE;

• Ovid Embase;

• US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov); 

• World Health Organization International Clinical Trials Registry Platform (https://trialsearch.who.int). 

Our Information Specialist (IS) will model the subject strategies for databases on the search strategy designed for MEDLINE. Where appropriate, the IS will combine the subject strategy adaptations with the highly sensitive search strategy for identifying randomized controlled trials, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021). We present the MEDLINE search strategy in Appendix 1. 

Searching other resources

• We will check the bibliographies of included studies and relevant systematic reviews identified for further references to relevant studies;

• We will check the included studies for retractions and errata via PubMed (https://pubmed.ncbi.nlm.nih.gov), and the Retraction Watch Database (http://retractiondatabase.org [https://We will check the included studies for retractions and errata via PubMed (www.ncbi.nlm.gov/pubmed) and the Retraction Watch Database (http://retractiondatabase.org) and report the search dates in the review.]), and report the search dates in the review;

• We will search Epistemonikos (https://www.epistemonikos.org) for related systematic reviews;

• Where necessary, we will contact experts or organizations in the field to obtain additional information on relevant studies;

• We may contact original authors of included studies for clarification and further data if study reports are unclear;

• We will include grey literature by searching the following additional resources:

  • US Food and Drug Administration (www.fda.gov);

  • European Patent Office (worldwide.espacenet.com).

Selection of studies

We will download all titles and abstracts retrieved by electronic searching to a reference management database (Covidence). After removing duplicates, two review authors (SA and NA) will independently screen titles and abstracts for inclusion. We will retrieve the full‐text reports and two review authors will independently apply the eligibility criteria to the full texts, identify studies for inclusion, and identify and record reasons for exclusion of ineligible studies. We will resolve any disagreement through discussion or, when required, through consulting a third review author (JMW). We will list studies that initially appear to meet the inclusion criteria but that we later exclude in the ‘Characteristics of excluded studies’ table, with the reasons for their exclusion. We will collate multiple reports of the same study so that each study, rather than each report, is the unit of interest in the review. We will also provide any information we can obtain about ongoing studies. We will record the selection process in sufficient detail to complete a PRISMA flow diagram (Lefebvre 2020; Liberati 2009).

Data extraction and management

Two review authors (SA and NA) will independently extract data using a form piloted in a similar Cochrane Review (Adams 2020b). We will check for disagreements before entry into Review Manager (RevMan Web 2020). We will use a program called WebPlotDigitizer to extract numerical data from graphs. We will contact study authors for clarification and further data if required. We will collect characteristics of the included studies in sufficient detail to populate a table of 'Characteristics of included studies' in the full review. If data are not reported in a format that can be entered directly into a meta‐analysis, we will convert them to the required format using the information in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021).

We will extract the following data:

• identification: sponsorship source, country, study setting, contact details;

• methods: study design, treatment group (lovastatin group, placebo group), additional methods data;   

• population: inclusion criteria, exclusion criteria, group differences, baseline characteristics (number, gender, age, LDL cholesterol, co‐morbidities), co‐interventions (lifestyle coaching, diet, exercise);

• interventions and comparisons: various doses of lovastatin measured between 3 to 12 weeks for interventions and placebo for comparison;

• outcomes: percentage change from baseline for each dose in total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides for treatment period of 3 to 12 weeks and withdrawals due to adverse effects (WDAEs) with no time restriction. We will attempt to identify specific adverse effects such as statin‐induced myopathy if possible.

Assessment of risk of bias in included studies

Two review authors (SPA and NA) will assess all studies using the original Cochrane risk of bias tool (RoB 1), as described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We will use judgments and supports for those judgments across the following domains of bias:

• random sequence generation;

• allocation concealment;

• blinding of participants and personnel (performance bias);

• blinding of outcome assessment (detection bias);

• selective reporting (reporting bias);

• attrition bias;

• other bias.

We will resolve any disagreement through discussion or, when required, through consulting a third review author (JMW). We will complete a risk of bias table for each included study using the risk of bias tool in RevMan Web 2020.

Measures of treatment effect

We will directly extract the mean percentage change or, if this is not reported, we will calculate it from the baseline and endpoint values. If there is disagreement over a value, we will reach consensus by recalculation to determine the correct value. We will extract standard deviations (SDs) and standard errors (SEs), or if these are not reported, we will calculate them when possible. We will resolve any disagreement through discussion or, when required, through consulting a third review author (JMW). We will present treatment effects for continuous data as mean differences (MD), with 95% confidence intervals (CIs) for blood total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides.  We will present the dichotomous outcome WDAEs as a risk ratio (RR) with 95% CI.

Unit of analysis issues

The unit of analysis will be the number of participants completing each study. We will categorize studies with multiple treatment arms with different doses of lovastatin, into intervention groups. We will divide the placebo group equally across intervention arms to avoid double counting. When there are data at different times, we will calculate a single weighted mean from all the time points as, in a previous review, we established that the lipid‐lowering effect of
atorvastatin was stable over time (Adams 2015).

Dealing with missing data

We expect follow‐up to be reasonably high for these short‐term studies. We will contact study authors if outcome data are missing from study reports. The most common type of value that is not reported is the SD of the change.

If study authors are not able to provide the SD for the change in lipid parameters, we will impute the SD using the following methods (listed in order of preference).

• SD calculated either from the t statistics corresponding to the exact P value reported or from the 95% CI of the MD between treatment groups, based on methods outlined in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021).

• Average weighted SD of the change from other studies in the review (Furukawa 2006).

It is common for the SD to be miscalculated. In order not to overweight studies where the SD is inaccurately calculated and much lower than expected, we will impute the value using the method described by Furukawa 2006, when SD values are less than 40% of the average weighted SDs.

Assessment of heterogeneity

The Chi² test to identify heterogeneity is not appropriate as it has low power when there are few studies but excessive power to detect clinically unimportant heterogeneity when there are many studies. The I² is a better statistic, as it measures the proportion of total variation in the estimate of the treatment effect that is due to heterogeneity between studies. This statistic is also independent of the number of studies in the analysis (Higgins 2002). 

We will interpret I² with reference to the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2020), using the following thresholds:

• 0% to 40%: might not be important;

• 30% to 60%: may represent moderate heterogeneity;

• 50% to 90%: may represent substantial heterogeneity;

• 75% to 100%: considerable heterogeneity.

Where heterogeneity is substantial or considerable, we will investigate, and conduct sensitivity analyses, if appropriate.

Assessment of reporting biases

We will assess reporting biases as part of the risk of bias assessment in studies included in the review. We will assess publication bias using funnel plots, where there are at least 10 studies included in each analysis, as outlined in Chapter 13 of the Cochrane Handbook for Systematic Reviews of Interventions (Page 2021). We will quantify funnel plots by Egger’s test (Rothstein 2005)

Data synthesis

We will undertake meta‐analysis using standard Cochrane methodology (Deeks 2020). We will analyze data using RevMan Web 2020. We will use a fixed‐effect model with 95% CIs to determine the weighted treatment effect for blood total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides. If an I² is equal to or greater than 50%, we will use the random‐effects model to assess whether the pooled effect is statistically significant. We will extract data from each study and put them into (GraphPad Prism), to yield a weighted least squares analysis based on the inverse of the square of the SE for each lipid parameter, and generate weighted log dose response curves (Tallarida 1987). We will use the percentage reduction from baseline of the following surrogate markers to describe the dose‐response relationship of the effect of lovastatin: total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides (Boekholdt 2012).

Subgroup analysis and investigation of heterogeneity

 We will conduct subgroup analyses for the primary outcome LDL cholesterol. If we find at least 10 trials for each subgroup, we will analyze subgroups based on the following factors.

• Men versus women (effect may be greater in women due to smaller average size).

• Morning administration versus evening administration (effect may be greater when lovastatin is administered in the evening).

• Drug industry‐funded versus non‐drug‐industry‐funded studies (drug industry‐funded studies may overestimate the efficacy of statins). 

• Twice‐daily versus once‐daily (effect may be greater with twice‐daily dosing).

• Cardiovascular disease versus no cardiovascular disease (effect may be greater in those with cardiovascular disease due to better adherence).

• Chronic renal failure versus no chronic renal failure (effect may be diminished in people with chronic renal failure).

• Familial hypercholesterolemia versus non‐familial hypercholesterolemia (effect appears to be reduced in familial hypercholesterolemia in previous reviews) (Adams 2015; Adams 2018).

• Asian (China, Japan and Korea) versus non‐Asian participants (Asians have a higher incidence of poor metabolizers) (Liao 2017).

Sensitivity analysis

We will conduct sensitivity analyses in trials assessed as having low risk of bias across all bias domains. We will also conduct sensitivity analyses that exclude trials where we identify substantial or considerable heterogeneity (I² ≥ 50%). We will assess the influence of low adherence and the effect of our approach of combining results at different time points.

Summary of findings and assessment of the certainty of the evidence

We will use the GRADE approach to assess the certainty of the supporting evidence behind each estimate of treatment effect using the GRADEprofiler Guideline Development Tool software (GRADEpro GDT,) and according to the guidelines provided in the CochraneHandbook for Systematic Reviews of Interventions (Schünemann 2021a; Schünemann 2021b). We will use the five GRADE considerations (overall risk of bias, consistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of the body of evidence for each outcome.

The GRADE system uses the following criteria for assigning grade of evidence:

• high certainty: we are very confident that the true effect lies close to that of the estimate of the effect;

• moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of effect, but there is a possibility that it is substantially different;

• low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect; and

• very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

Two review authors (SA and JMW) will independently judge the certainty of evidence. Any disagreement will be resolved by discussion or by consulting a third review author (NA). We will justify, document, and incorporate judgments into reporting of results for each outcome, using footnotes to aid the reader's understanding. We will present key findings of the review in a summary of findings table for the LDL‐C lowering effect of each dose of lovastatin as this will provide the dose–response relationship for the drug. We will also present a separate summary of findings table for WDAEs for all doses as the best measure of harm.