OpenEvidence has moved further into the work that follows a clinical decision. Its new Access Intelligence offer brings a practical question into focus: once a clinician has identified an appropriate treatment, what still stands between that decision and the patient receiving it?
That is a substantial problem for clinical software to address. A convincing evidence summary can leave the clinician with an unresolved coverage question, an assistance application and a letter to write. Those tasks have different sources, owners and completion criteria. Putting them closer to the clinical conversation could be useful, provided the output makes clear which part of the process has actually been completed.
Access Intelligence was announced on 11 November 2026. This analysis examines the published offer and the work it could support; the examples are fictional workflows, rather than results from testing an account. Our view is that the launch makes OpenEvidence a more serious workflow competitor while leaving a clear role for iatroX in source-aware reference, clinical reasoning and professional learning.
What is available at launch
OpenEvidence’s release lists four current functions: information about eligibility and availability across more than 2,500 patient-support programmes, updated daily; copay savings and discount sharing through Patient TakeHome; pre-populated forms or direct programme enrolment; and guideline-grounded prior-authorisation requests, appeals and medical-necessity letters with checklists and relevant ICD-10 codes. Letters can be edited, with interactive follow-up on Sackett and Snow. The company describes the offer as free for clinicians. OpenEvidence’s Access Intelligence announcement
The same release reports more than 500,000 prior-authorisation documents already written on the platform. That is a company usage figure. It does not measure approvals, treatment starts or improved affordability.
The important change is the proposed continuity between a clinical question and the administrative work that follows it. The value of that continuity should be assessed through completed, appropriate work and the effort needed to review it.
The word access describes several different outcomes
Access can mean finding information, identifying an applicable service, submitting a complete application or receiving a favourable decision. These are separate outcomes even when one interface helps with all of them.
A clinician who discovers a relevant programme has made progress. The patient may still need to provide information, meet eligibility criteria and complete an application. An apparently complete form can still be awaiting submission. A submitted request can still be awaiting review.
That distinction belongs in the working interface and in any evaluation of the product. The clinician should be able to identify what has happened, what remains uncertain and who owns the next step.
For example, “a supporting letter has been prepared” is a useful, bounded statement. “Access has been secured” is a stronger claim requiring a different observation. A good workflow preserves that difference throughout the process.
This is why the launch should be assessed as a series of connected tasks. An attractive demonstration of the first task cannot establish that the final one will succeed.
Information about eligibility is not an eligibility determination
Programme discovery has a clear potential benefit: reduce the effort required to find the relevant rules. Its usefulness then depends on matching those rules to the actual circumstances.
Consider a fictional clinician investigating assistance for a patient who has encountered a cost barrier. The first question is what information is needed to establish eligibility. Some of that information may be unknown, incomplete or subject to change.
The clinician needs a route to the applicable programme documentation and an intelligible account of the criteria. If the answer depends on an assumption, that assumption should be visible before anyone relies on it.
The appropriate evaluation therefore separates retrieval from application. Did the tool find a relevant programme? Did it preserve the qualifications? Did the review identify the missing facts? Could the clinician verify the decisive rule?
A system can be useful even when it cannot settle the question immediately. Identifying the information still needed can save an application from proceeding on a mistaken premise.
A form, a submission and a decision need separate records
Administrative automation becomes more consequential as it moves from preparing text towards acting on it. The difference is operational, not merely a change in how polished the output looks.
A prepared form needs checking for accuracy and completeness. An external submission also needs a known destination, appropriate authority and a record of what was sent. A final decision requires confirmation from the party responsible for making it.
Those stages should have different status labels. The reader should not have to infer from a successful-looking screen whether an application is saved, submitted or approved.
The Access Intelligence announcement does not establish automatic payer approval or universal submission coverage. A service evaluating the product should ask which actions are supported in the particular arrangement and how exceptions are handled.
Useful questions include what happens when a required field is missing, an attachment cannot be transmitted or an external service is unavailable. The proposed fallback matters because the clinical team still needs to know who will continue the work.
A strong letter preserves the underlying clinical reasoning
A well-structured document can make supporting information easier to review. Its quality depends on the relationship between the claims, the record and the applicable requirement.
An evaluation should examine that relationship sentence by sentence. Does the document state something the clinician knows, something supported by the cited source or something the system has inferred? Are those categories distinguishable?
A letter can be fluent and still require substantial correction. It may omit a necessary qualification, assume a fact absent from the record or provide general evidence that does not address the actual criterion.
The useful comparison is the work required to reach an accurate, relevant final document. Generation time is only one part of that process. Review, correction, missing information and follow-up all consume staff time.
The objective is a better-supported submission, not simply a longer or more assertive letter. A clinician should remain able to explain the argument after the prose has been simplified or reformatted.
Follow a fictional case through the whole process
Imagine a specialist has discussed a treatment with a patient, who then raises a concern about obtaining it. This is a fictional administrative example; it does not recommend a treatment or predict what OpenEvidence would output.
Start by recording the unresolved question precisely. Is it uncertainty about coverage, a financial barrier, missing paperwork or a combination? Without that distinction, the team may solve the easiest visible task while leaving the main obstacle unchanged.
Next, identify the information needed for the chosen route. Compare the relevant programme or policy criteria with the information the team is authorised to use. Ask the patient or responsible service for what is missing rather than filling the gap with an assumption.
If a document is prepared, review its factual statements and supporting rationale. Record what was changed and which source settles the important criterion. Then establish whether the next step is internal review, an external submission or a conversation with the patient.
Finally, follow the process to a defined endpoint. A saved draft is one endpoint for the drafting task. An acknowledged submission is another. The patient receiving the intended support is a later outcome that should be measured separately.
This sequence allows a team to see where software helped and where human coordination remained necessary. It also makes a disappointing result more informative: the problem may have been irrelevant programme information, incomplete documents or an external decision, and each needs a different response.
Evaluate outcomes with the right denominator
A document count is a measure of activity. To understand operational value, a service needs to know what happened to a defined set of cases.
Begin with the eligible cases and the baseline process. Record the work completed, the staff time required and the cases that could not proceed. Keep corrections and unsuccessful attempts in the assessment.
For an application workflow, distinguish cases requiring more information, submitted cases and completed decisions. For affordability, establish what the patient actually paid or received where the service has an appropriate way to measure it.
The denominator should remain consistent. Comparing all generated documents with only successful historical submissions would produce a misleading impression. Likewise, excluding the most difficult cases after the evaluation starts could make the software look more effective than it was.
A modest evaluation with clear definitions can answer a useful local question. A large number without a known pathway from activity to outcome cannot do the same work.
Free access still has product and regional conditions
OpenEvidence’s published standard terms exclude service access from the UK, EU and EEA. Separate properties and applicable institutional agreements can carry their own conditions. Its acceptable-use policy separately restricts inputs concerning residents of those regions. Network terms, acceptable-use policy
For an eligible user elsewhere, the next check is the actual feature and intended workflow. The phrase “free for clinicians” should not be expanded into a promise that every country, profession, programme or organisational deployment is supported.
For UK readers, the launch is therefore useful market intelligence rather than evidence that an equivalent NHS access workflow is now available. A US coverage process and a local NHS pathway can involve different decisions, responsibilities and sources.
Where iatroX makes a positive contribution
Ask iatroX provides free clinical reference. Selected question banks are also available without a subscription; all-access supports wider preparation, continued Tutor use and reviewed professional learning in one connected routine. Current iatroX access
That is valuable when an administrative difficulty exposes a clinical knowledge gap. The clinician may need to understand the recommendation more clearly, distinguish two plausible approaches or revisit why a particular patient group was excluded from the evidence.
Use Ask iatroX to investigate the reference question, inspect the source and identify the point worth learning. For supported exam preparation, an attempted question gives the subsequent discussion a concrete focus.
The advantage is purposeful continuity. Instead of leaving the uncertainty as another saved answer, the clinician can turn it into a specific activity and a later check of understanding. An access tool and a learning platform can each earn a place by addressing the work that remains.
This does not imply an integration or automatic exchange of patient information between the services. It is a choice about their distinct professional purposes.
The launch changes what should be compared
OpenEvidence should now be assessed beyond the quality of a clinical answer alone. The relevant questions include the accuracy of administrative information, the completeness of prepared documents, the handling of actions and the outcome of the access process.
For clinicians, that broader scope is a reason to inspect the exact workflow. For institutions, it creates a more demanding evaluation than counting generated letters or timing an initial response.
iatroX retains a clear, affirmative proposition within this changing market: source-linked reference connected to learning the clinician can deliberately pursue. A clinician can value help completing administrative work and still choose a specialist environment for understanding the decisions behind it.
The most useful arrangement makes both forms of progress visible: the patient’s next access step and the clinician’s next learning step. Access Intelligence makes the first question more prominent. iatroX provides a focused way to work on the second.