Before you instruct a medical record screening provider, there are nine questions worth asking: who reviews the records, whether the provider is accredited, what the realistic turnaround is, how the work is priced, where the data sits, whether any public AI tools are involved, what you actually receive, whether gaps in disclosure are flagged, and whether the same set carries through to a court bundle. The answers tell you far more than a brochure will.
We are a screening provider ourselves, so treat this as a partial source and check our answers against anyone else you are considering. That said, these are the questions our clients ask us, and the ones we would ask in their position. If a provider cannot answer them plainly, that is useful information in itself.
Last reviewed: August 2026.
Screening sits at the point in a claim where a small amount of money buys a large amount of certainty. Get it right and you decline weak enquiries in days rather than months, and back the good ones with confidence. Get it wrong and you have paid for an opinion you cannot rely on, which is worse than having no opinion at all.
The difficulty is that screening services look similar on paper. Everyone says fast, accurate and secure. The differences only surface when you ask about the mechanics: who does the reading, what happens when records are missing, and what you are left holding at the end.
This is the question that separates a genuine clinical opinion from a summary of the paperwork. A non-clinician can tell you what a record says. Only a clinician can tell you whether the care described fell below a reasonable standard, and that is the judgement you are paying for.
A good answer names the profession and the relevant experience. A weak answer talks about processes and technology without ever saying who reaches the conclusion.
Accreditation matters less for the badge than for what it implies: an external standard someone else is checking. It is also worth asking whether the people doing the work receive structured, accredited training, or whether they were simply given the files and left to it.
Be sceptical of vague assurances. "Fully compliant with industry standards" means nothing without a named standard and a named assessor.
Turnaround is where optimistic marketing shows up most often. The honest answer has a shape to it: a normal range, the factors that extend it, and how you will be told if a case falls outside the range.
Ask specifically who contacts you if a deadline is at risk, and how quickly. Silence until the due date is the failure mode that actually costs you.
Fixed-fee matters for screening more than for most services, because the whole point is a low-risk look at an uncertain case. If the cost of finding out is itself uncertain, the proposition weakens considerably.
Compare the fee against the fee earner time a manual review of an unstructured file would consume, not against zero. That is the real alternative.
Medical records are special-category data under the UK GDPR, so this is a compliance question as much as a commercial one. Your firm remains accountable for what happens to those records after you send them.
Two practical follow-ups. First, how are completed files delivered: through a controlled portal, or as email attachments and generic cloud links? Second, can you get a record of who accessed what, if you are ever asked?
This question has become essential rather than technical. AI is genuinely useful in this work: sorting a thousand pages into chronological order takes minutes rather than hours. But there is a real difference between a private, purpose-built platform and someone pasting extracts into a consumer chatbot.
Ask what the AI is actually doing, too. Sorting, indexing and searching are appropriate uses. Forming the clinical opinion is not.
A sample tells you more than any conversation. You are looking for something a fee earner can act on in ten minutes: a conclusion up front, reasoning that can be followed, and specifics rather than hedging.
Also ask whether you receive the organised records themselves, or only the report. If the sorted bundle stays with the provider, you are paying for the same sorting work twice when the case proceeds.
This is the most underrated question on the list. Disclosure is rarely complete first time. Pages go missing, whole episodes of care are omitted, and results referenced in one letter never appear anywhere else.
A provider who tells you only what the records contain has done half the job. You need to know what should be there and is not, while there is still time to request it and before limitation planning is affected.
Screening and bundle preparation are often bought separately, from different suppliers, which means the same thousands of pages get organised twice. That is pure duplicated cost, and it introduces the risk of two slightly different versions of the file existing at once.
If a provider cannot continue the work, ask whether the output is at least in a format someone else can pick up without starting again.
Some responses are more revealing than others. If you hear any of the following, ask a second question.
Screening is worth buying properly. The questions above take ten minutes to ask and they surface the things that actually determine whether an early opinion is reliable: a qualified clinician forming the judgement, an honest view of turnaround, sensible handling of sensitive data, and an output your team can act on straight away.
MRC is AI-driven and clinically led. MRC AI sorts and structures the records, typically ordering around 1,000 pages in roughly 15 minutes, and a qualified GP then reviews them and reaches the recommendation. The AI does the heavy lifting; the clinician brings the judgement, and never the other way round. Completed work is delivered through the secure MRC Portal, and records that proceed flow into MRC Pagination with a clinical chronology.
If you would like to put these questions to us directly, or see a sample MRC Screening report, we are happy to talk. Website: mrcgroup.uk. Phone: 0161 928 1636. Email: info@mrcgroup.uk.