Everything researchers ask us before they start.

ClinStat Research provides methodological and statistical support across the clinical research lifecycle: study design and protocol development, sample-size calculation, data validation and management, statistical analysis, systematic review and meta-analysis, manuscript preparation and publication support, and structured research mentorship. We work with individual clinicians, research teams, academic departments and institutions, remotely and worldwide.

Message us on WhatsApp or call, and we will arrange a free 15-minute consultation. Bring the research question, a description of the data you have or plan to collect, and your deadline. At the end of that call you will have an honest assessment of what is achievable, what is not, and what it would involve — whether or not you go ahead with us.

Cost depends on the design, the size and state of the dataset, the methods required and the timeline. We quote per project after the initial consultation, so you know the scope and the price before anything begins. We do not quote before understanding the project, because a quote given without seeing the data is a guess that gets revised later.

A focused statistical analysis on a clean, well-documented dataset can be completed in one to two weeks. A large multicentre dataset requiring validation, or a systematic review with meta-analysis, takes considerably longer. Tight deadlines are workable — we have delivered a validated analysis of an ~1,800-patient multicentre dataset in two weeks by running validation and analysis scripting in parallel — but we will tell you when a deadline forces a genuine trade-off rather than accept it silently.

Yes. We work with researchers and institutions worldwide and operate remotely. Communication is in English or Arabic, over WhatsApp, email and video calls, scheduled around your time zone.

Yes. We work with de-identified data, we do not share datasets or findings with anyone outside the project team, and we are happy to sign a confidentiality agreement or your institution's data-handling agreement before receiving anything. Data are retained only as long as the project requires and deleted on request.

Reliability is a process, not a claim. On every project we work to a standard covering eight core checks: high-quality validated data; a study design and sample size appropriate to the question; a statistical analysis matched to the design and variable types; formal testing of model assumptions; sensitivity and robustness analyses; appropriate handling of missing data; independent verification of the analysis by a second statistician; and full reproducibility from raw data to final table. Where the study calls for it we also address confounding, effect estimation with confidence intervals, multiplicity, risk of bias and consistency with prior evidence. The full standard is published here.

Three things. First, a highly qualified and research-active team — every ClinStat team member has at least 50 scientific publications and an h-index of 5 or higher, which means sustained experience in methodology, scientific writing, peer review and publishing, not just clinical practice. Second, an extensive and proven track record — more than 1,000 researchers and clients supported across more than 5,000 research projects, spanning many specialties, designs and stages. Third, real quality-control infrastructure — a structured workflow with data-quality checks, appropriate methodology, independent review of analyses and reproducible practices applied before anything is delivered.

R and Stata for most work, with SPSS and Python where an institution requires them. You receive the analysis scripts, so every reported number can be reproduced by your team or by a reviewer.

Both, and the choice is driven by the question rather than by what is impressive. Our team works routinely with propensity scores and IPTW, mixed-effects models and GEE, survival and competing-risk analysis, multiple imputation, restricted cubic splines, mediation analysis, Bayesian methods, network and IPD meta-analysis, and clinical prediction modelling including machine learning. The methods index lists all 18 method families with when each is appropriate.

We tell you, before you spend money on an analysis that cannot succeed. Sometimes the question can be reframed into one the data can answer; sometimes the honest answer is that additional data are needed. Discovering this at the analysis stage is unwelcome; discovering it after a reviewer does is far worse.

No. We will make sure the analysis is correct, the assumptions hold, the effect estimate is precise as the data allow, and the reporting is complete. If the result is null, we help you publish a well-conducted null study — which is publishable, useful, and increasingly valued. We do not run analyses until something reaches p < 0.05.

Most desk rejections come from a small, avoidable set of problems: the paper is out of scope for the journal, statistical reporting is incomplete (effect sizes without confidence intervals, missing-data handling not described, methods that do not match the results), ethics approval and consent statements are missing or vague, the relevant reporting guideline was not followed, or authorship and conflict declarations are unclear. All of them are fixable in a single structured review before submission. Our 12-point checklist is the list we run internally.

Yes, and it is one of the things we do most. We diagnose the manuscript across four dimensions — scientific, methodological, statistical and presentational — and separately reassess whether the journals it has been sent to were ever appropriate for its scope and methodology. One manuscript that reached us after more than ten rejections was accepted by the first journal the revised version was submitted to. See the case study.

No, and no honest provider can — acceptance is decided by editors and reviewers. What we control is scientific quality, statistical correctness, reporting-standard compliance and a realistic journal strategy. In a long-running institutional partnership producing roughly 100 manuscripts a quarter, that combination has yielded an 80% publication rate in peer-reviewed journals.

We follow ICMJE authorship criteria. Methodological and statistical contribution is credited in the way you and your institution consider appropriate — authorship where the criteria are met, acknowledgement where they are not. We never accept authorship we have not earned, and we do not produce work for submission as someone's unaided output.

We prepare point-by-point responses to statistical and methodological comments, run any additional analyses reviewers request, and revise the manuscript accordingly. Clinical and interpretive responses are drafted with you, because those are yours to make.

Five questions separate a statistician who will strengthen your study from one who will simply run tests on it:

1. Do they ask about your research question before they ask for your data? The question determines the analysis, not the other way round.
2. Can they explain the method in plain language rather than jargon? If they cannot explain it to you, you cannot defend it to a reviewer.
3. Do they review your study design, or only analyse what has already been collected? Design problems cannot be fixed by analysis.
4. Do they document every step so that someone else could reproduce it? Reproducibility is what separates a defensible result from a hopeful one.
5. What does their own publication record look like? At ClinStat, every team member has 50+ publications and an h-index of at least 5.

Software competence and methodological judgement are different skills. Running a test is easy; knowing whether that test answers your question, whether its assumptions hold in your data, whether your confounder set is sufficient, and whether the result survives a reasonable alternative specification is the part that decides whether the finding is true. Most of the manuscripts we rescue were analysed by someone who knew the software perfectly well.

Often, yes. Fifteen focused minutes is usually enough to establish where you actually stand and what the next logical step is — whether the design can answer the question, whether the dataset is analysable, and whether the deadline is realistic. It is free and carries no obligation.