Dr Gaurav Solanki | Hair Transplant Surgeon

How We Intend to Read and Explain Clinical Evidence

Methodology version 2.0 · Effective 1 August 2026

Policy owner: Gaurav Solanki

Policy-owner disclosure: Gaurav Solanki owns or controls this website and may be professionally affiliated with services discussed elsewhere on it. This methodology page is not evidence of any credential, treatment outcome or service claim.

This editorial method applies prospectively only to articles that explicitly identify the method version, the evidence-search date and a named, verified reviewer. It does not certify, validate or retrospectively review existing pages on this website. It is not a clinical guideline and is not individual medical advice.

1. Frame the question before searching

For each evidence-led article, we intend to record the population or condition, intervention or exposure, comparison, outcomes, setting, time horizon and search date. Conclusions should remain within that recorded scope.

2. Declare the type and limits of the search

An article should label its search as either a systematic review or systematic search—only when a predefined adopted standard is fully met, including a protocol-led and reproducible attempt to identify all eligible evidence within a defined scope—or a targeted or narrative search, which is narrower and is not presented as exhaustive.

For either type, the article record should name the databases and other sources searched, reproduce the search strategy, state the last search date, list eligibility criteria, explain study selection and record material exclusions. The scope label should remain visible so that a targeted search is not mistaken for a systematic review.

3. Appraise sources using explicit criteria

Guidelines and systematic reviews should not be accepted merely because they come from a recognized organization or are recent. We intend to assess their scope, search date, methods, conflict disclosures, applicability and underlying studies. Primary studies may be examined when they answer a gap or help readers understand the source of a conclusion.

Randomized trials can be informative for intervention effects. Observational studies may be necessary for uncommon adverse events, long-term outcomes or questions that cannot reasonably be randomized. Study design alone does not establish reliability.

4. Keep reporting quality separate from methodological quality

Reporting guidelines help show whether authors disclosed what they did. CONSORT applies to randomized-trial reporting; the core STROBE Statement applies to cohort, case-control and cross-sectional studies; and PRISMA applies to systematic-review reporting. Other designs require a relevant extension or alternative guideline. A checklist does not by itself show that a study was well designed, free from bias or correct.

5. Use a design- and result-specific risk-of-bias method

When risk of bias could affect an article’s conclusion, the article record should identify the assessment tool and version, study design, specific result or outcome assessed and reasons for each judgement. For example, Cochrane RoB 2 may be suitable for a result from a randomized trial, while ROBINS-I may be suitable for an eligible non-randomized study of an intervention. Tools should not be treated as interchangeable or reduced to an unsupported universal score.

Relevant concerns may include randomization, deviations from intended interventions, missing outcome data, outcome measurement, selective reporting and evidence that eligible results are missing from the available record.

6. Describe certainty at the outcome and body-of-evidence level

GRADE domains may be used to assess the certainty of a body of evidence for a specific outcome. They are not a quality label for a single study or an entire article. No formal rating of high, moderate, low or very low certainty should be shown unless an outcome-level assessment and its reasons are documented.

Certainty of evidence and strength of a recommendation are different judgements. This editorial method does not convert an evidence summary into a treatment recommendation. Statistical significance should not be treated as clinical importance, and absence of statistical significance should not be treated as proof of equivalence. Where data permit, effect size, uncertainty and absolute as well as relative effects should be reported.

7. Test applicability

We intend to compare the participants, intervention protocols, comparators, follow-up, care settings and measured outcomes with the question being answered. Important indirectness should be stated, and group averages should not be converted into predictions for an individual.

8. Present benefits, reported harms and alternatives within scope

An evidence summary should not present potential benefits while omitting reported harms identified within the stated evidence scope, uncertainty, relevant alternatives or the option of no intervention where appropriate. If definitions are inconsistent or follow-up is short, those limits should remain visible in the conclusion.

9. Record funding and conflicts

Study funding and declared author conflicts should be recorded when available. Funding does not automatically invalidate evidence, but it is relevant context for design, analysis and reporting. Website authors and reviewers should disclose material conflicts relevant to an article.

10. Match language to the evidence

We intend to distinguish association from causation and use qualified language when uncertainty remains. Guarantees, universal claims and superiority claims should be excluded unless the cited evidence directly supports the stated population, comparison and outcome.

11. Make versions, updates and corrections visible

Each article using this method should display its publication or substantive-update date, methodology version and a named, verified final reviewer. Reviewer verification means confirming identity, permission to publish the name, relevant qualification or competence, role, scope, review date and conflict status. It does not imply physician or specialist review unless that is factually true.

A substantive update should say what changed. A material correction should retain the original publication date and add a dated correction note describing the change rather than silently replacing a conclusion.

Methodology sources

The sources below were accessed on 1 August 2026. The applicable edition and update metadata should be rechecked when this methodology or an article using it is updated.

Methodological safeguards

Formal RoB 2, ROBINS-I or GRADE assessments require a documented procedure and a reviewer with verified methodological competence for that tool and scope. A template alone is insufficient. Before reproducing any checklist or assessment-tool content, the applicable licence or permission should be checked.

No MedicalWebPage, Physician, MedicalProcedure, Review, AggregateRating or Offer type—and no credential, outcome, reviewedBy or lastReviewed property—should be inferred from this methodology page. Any proposed structured-data type and each property must be independently verified and visibly supported by the page.