How to read a master protocol as a referring physician

A patient sits across from you with newly diagnosed advanced non-small cell lung cancer. Her pathology report shows a specific driver mutation. Her family asks about clinical trials. You pull up a trial at a research site 40 minutes away and download the protocol. The PDF is 287 pages. There are eleven appendices. Section 5 splits into "Screening Registration" and "Sub-study Registration" and you are not sure what the difference is. Your next patient is waiting.

This is the moment most referring physicians decide, quietly, that master protocols are not for them. That decision costs patients access. It is also based on a misreading of what the document actually asks of you.

You do not have to read the whole protocol. You do not have to master every appendix. You have to recognize the shape of a fit and hand the patient to the research site team. This piece shows you exactly how, using the same eye path an experienced trialist would use, walked through a realistic scenario section by section.

The two-layer model that unlocks the document

Every master protocol is built in two layers, and once you see the layers, the document becomes readable.

The first layer is the master framework. It sets the shared rules that apply to everyone entering the study: the disease and stage, the performance status, the organ-function thresholds, the shared biomarker screening step, the safety oversight structure, and the schedule backbone. Think of it as the entrance and the hallway.

The second layer is the sub-study appendices. Each appendix is its own mini-protocol, with its own added eligibility criteria, its own investigational product arm or arms, and often its own registry number. Think of these as the rooms off the hallway. A patient can walk through the front door and still not fit any specific room, or fit only one. For a broader view of how modern trial designs are built, see From Design to Discovery: How CROs Power Every Trial Phase.

The section numbering in the actual document usually makes this split literal. Section 5.1 will be labeled something like "Screening Registration" (the master gate). Section 5.2 will be "Sub-study Registration" (the appendix gate). When you see those two headings, you are looking at the two-layer structure in the numbering itself.

Three flavors of master protocol show up in practice. An umbrella trial holds one disease constant and tests multiple investigational products, sorting patients by biomarker into different appendices. A basket trial holds one investigational product constant and tests it across multiple diseases sharing a molecular alteration. A platform trial keeps an ongoing structure open, usually in one disease, and lets investigational arms enter and leave over time against a shared control. You will meet all three inside real protocols; the two-layer model applies to each.

Read the schema first: the one page that saves you an hour

Right after the table of contents, every modern master protocol includes a schema. It is a single-page flow diagram, and it is the highest-value page in the entire document. If you read one page, read this one.

The schema shows the entry point at the top (screening registration), then a shared biomarker profiling step, then a branch point where patients are assigned to sub-studies. Below the branch, each active sub-study appears as its own box, labeled with its target biomarker and its arms inside. A dedicated box usually appears for a non-match cohort, for patients whose profile does not match any of the biomarker-driven appendices.

The footnotes on the schema carry logic you cannot skip. They will tell you what happens when a patient carries more than one qualifying alteration (usually a pre-specified allocation rule or a randomized tie-break), whether patients who progress on one sub-study can enter another, and which appendices are currently open versus closed. Read the footnotes before you read the boxes.

The schema is the map. The full protocol is the territory. You need the map before you enter the territory. For a broader look at how trial designs differ from one another, see Prevention vs Treatment Trials: What's the Difference?.

The 10-minute eye path: what to read, in what order

Reading a master protocol end to end takes days. Reading it well enough to make a good referral takes ten minutes. Here is the ordered path.

First, read the schema. Confirm the entry point, find the shared biomarker step, and identify which sub-study boxes are currently active.

Second, read the synopsis at the front of the document. Two or three pages long, it will confirm the disease setting, the line of therapy, and the overall design. Use it to triage, never to confirm eligibility.

Third, read the master-level eligibility (Section 5.1 in most numbering conventions). This is the shared gate: disease, stage, performance status, organ function, prior therapy, age. If your patient fails these, no appendix matters.

Fourth, read the biomarker routing logic. This usually lives in the schema footnotes and in a biomarker-testing algorithm appendix at the back. It tells you which alteration routes to which appendix, and the tie-break rule when a patient has more than one.

Fifth, open only the one appendix that matches your patient. Read its added eligibility, its arms (including whether one arm is a control or standard of care), and any additional assessments beyond the master schedule.

Sixth, glance at the schedule of assessments. It is usually a dense grid showing assessments as rows and visits as columns. Look at the visit frequency, biopsy requirements, and imaging cadence. This tells you whether the practical burden is realistic for your patient's life. For community physicians thinking about how to widen research access, see How Community Physicians Can Expand Clinical Trial Participant Pools.

Seventh, glance at the safety monitoring section. Confirm that an independent data and safety monitoring board (an independent committee that reviews accumulating data and can pause or stop arms) is in place. You are not auditing the statistics. You are confirming that the trial has real oversight.

When you have completed those seven passes, you have enough to decide whether to refer. You have not confirmed eligibility. The site team will do that. You have confirmed that a referral is worth making.

A worked example: reading a lung cancer umbrella protocol

Bring back the patient from the opening scene. Newly diagnosed advanced non-small cell lung cancer. A specific driver mutation on the molecular report. A large adaptive umbrella trial open at a nearby research site. Here is how you actually read the protocol.

Step 1: The schema (2 minutes)

You open the PDF and skip past the title page, participants list, and table of contents. On page 8, you find the schema. It shows an entry funnel labeled "Pre-Screening/Screening Registration" at the top, feeding into a box labeled "Common Broad Platform Biomarker Profiling." Below that, a branch point splits patients into eight lettered sub-studies, each named for a specific driver alteration and its matched investigational product, plus a non-match sub-study for patients whose profile does not fit any of the eight.

The footnote beneath the branch tells you two things you need. Patients eligible for more than one biomarker-driven sub-study are randomized to sub-study assignment, weighted by the rarity of each biomarker. Patients who progress on one sub-study may be eligible to enter another. That second footnote is a small mercy. It means a referral now is not a one-shot bet.

Step 2: The synopsis (2 minutes)

You flip forward to the synopsis on page 12. It confirms the trial is a phase II/III adaptive umbrella in advanced or metastatic non-small cell lung cancer, biomarker-driven, with the option for accelerated registration pathways in select sub-studies. Your patient is newly diagnosed with advanced disease. The setting fits.

Step 3: Master-level eligibility (3 minutes)

You go to Section 5.1, the screening registration criteria. This is the shared gate. You scan the inclusion criteria in order.

Histologically confirmed non-small cell lung cancer. Your patient has it. Stage IV or recurrent. She does. Measurable disease by RECIST criteria (the standardized rules for measuring tumor response on imaging). She has a measurable lung lesion and a liver metastasis. Performance status 0 or 1. She is ECOG 1. Adequate organ function as defined in a linked table. Her labs meet the thresholds. Age 18 or older. She is 62. No prior systemic therapy for advanced disease (the trial is looking for frontline patients). She has not been treated. She clears the master gate.

You scan the exclusion criteria. Active untreated central nervous system metastases. Her brain MRI is clean. Recent major surgery. None. A specific list of prior therapies that would exclude. None apply. She passes.

You have spent seven minutes so far. You know the patient can enter the front door.

Step 4: Biomarker routing (1 minute)

You flip to the biomarker-testing algorithm appendix at the back of the document. It shows a table with each qualifying alteration on the left and the corresponding sub-study letter on the right. Your patient's specific driver mutation routes her to Sub-Study C. The footnote adds that if a patient carries both the Sub-Study C alteration and a Sub-Study F alteration, the randomization tie-break weights toward the rarer of the two. Your patient carries only the Sub-Study C alteration. She routes cleanly.

The appendix also lists alterations that EXCLUDE from certain sub-studies. Your patient does not carry any of the exclusionary alterations for Sub-Study C. She stays routed.

Step 5: The matching appendix (2 minutes)

You open Sub-Study C. Its added eligibility criteria are short. The patient must have the specific driver mutation confirmed by a certified laboratory assay (her report meets this). The patient must not have received prior therapy targeting the same pathway (she has not). A washout period applies for certain prior medications (none of her current medications trigger it).

Sub-Study C has two arms. Arm 1 is the investigational product. Arm 2 is standard-of-care chemotherapy. Patients are randomized 2:1 in favor of the investigational product. You note this. It means your patient could be randomized to standard chemotherapy and never receive the investigational product. The informed-consent conversation, which the site team will hold with her, must cover both arms honestly.

For a broader view of how technology is reshaping the referral pathway between physicians and research sites, see AI and Doctors Unite: New Era of Clinical Trial Referrals.

Step 6: The schedule (30 seconds)

You look at the schedule of assessments. Visits every three weeks for the first six months, then every six weeks. Imaging every nine weeks. A required tumor biopsy at progression. You know your patient's life circumstances. She lives 45 minutes from the site. She has family who can drive her. The schedule is doable.

Step 7: Safety oversight (30 seconds)

You scan Section 11. An independent data and safety monitoring board reviews accumulating data at pre-specified intervals and can pause or close any arm. Pre-specified futility rules govern whether any arm is dropped early. You do not need to read the statistics in detail. Independent oversight is present.

Total reading time: about ten minutes. You now know your patient plausibly fits the master gate, routes to Sub-Study C, and could realistically complete the schedule. You call the site's research coordinator. The site team will run formal eligibility, hold the informed-consent conversation, and, if she qualifies, enroll her.

Two variations on the same reading path

The seven-step path scales across master protocol types. Two brief variations.

A neurologist reading an ALS platform protocol

A patient with early ALS. A perpetual adaptive platform trial runs several concurrent investigational-product arms against a shared placebo. The schema shows one shared master screening step, then random assignment to whichever arms are currently open, then within-arm randomization weighted toward active product. The master eligibility is diagnosis by accepted criteria, a time-since-onset window, and the ability to complete study procedures. The trap specific to platforms is that active arms change by amendment. The regimen you remember from a conference may have closed. Check the current version of the protocol and the current registry records, not a downloaded PDF from six months ago. Consent happens in two stages: a master-protocol consent at screening, then a regimen-specific consent after assignment.

A specialist reading a tumor-agnostic basket protocol

A patient whose solid tumor carries a specific gene fusion. A basket trial tests one investigational product across any solid tumor with that alteration. The synopsis confirms the trial is histology-agnostic: eligibility keys on the alteration, not the organ. The master eligibility requires an advanced or metastatic solid tumor, the fusion confirmed by a certified laboratory assay, progression on standard options or absence of satisfactory standard options, and measurable disease. Baskets are often single-arm rather than randomized, but confirm this in the appendix rather than assuming. The assay requirement is worth flagging: a real-world report may not automatically qualify if the trial requires central confirmation.

The pitfalls that trip up first-time readers

Six mistakes come up over and over when physicians new to master protocols try to read them. Each one is easy to avoid once named. For a discussion of how adaptive trial elements interact with operational timelines, see The Evolving Role of CROs in a Patient-Centric World.

Confusing master-level eligibility with appendix-level eligibility. A patient can clear the master gate and fail an appendix criterion, or the reverse. Always read both layers.

Missing that arms open and close by amendment. The active list changes. What was open last quarter may be closed now, and a new arm may have opened last month. Check the current version, not a cached copy.

Confusing screening with randomization. Master protocol screening registers the patient for biomarker profiling and sub-study assignment. Randomization is the separate step that puts the patient onto an arm inside the sub-study. A patient can be screened and assigned but never randomized (progression, deterioration, or failing the appendix gate in between). Passing screening is not enrollment.

Missing the shared control implication. Many sub-studies randomize between the investigational product and a shared control or placebo. A referred patient may be randomized to control and not receive the investigational product. Set this honestly with the patient before referral; the site team's informed-consent conversation will cover it in full.

Reading the synopsis and stopping there. The synopsis is a triage tool. Eligibility nuance, biomarker routing, and the schedule of assessments live in the full document.

Reading an outdated version. Master protocols amend frequently. The version date on the title page matters. If the site coordinator quotes a criterion you cannot find in your PDF, you probably have an older version.

Where your role ends and the research site team's begins

DecenTrialz pre-screens only. The research site team handles walk-through, eligibility, consent, and enrollment.

That sentence describes the whole architecture of the referral relationship, and it defines the limit of your reading task. You are not adjudicating final eligibility. You are not walking the patient through the protocol. You are not obtaining informed consent. You are recognizing a plausible fit and connecting the patient to the people who do those things. The site's principal investigator and study coordinators are trained to close every gap in the criteria, cover every arm including control in the consent conversation, and enroll the patient if she qualifies.

What if you are not sure your patient fully qualifies?

Refer anyway if the disease and general profile fit. Master protocols are designed to screen against multiple sub-studies at once, and the site team will run the formal eligibility check. A referral based on plausible fit is more useful than a referral held back for perfect certainty. Time-to-referral is one of the biggest drivers of whether a patient ever enrolls, and the patients most hurt by delay are the ones who cannot afford to wait.

Nothing about a master protocol changes your clinical relationship with the patient. You remain the medical home. The research site team communicates study-related events. Your care continues in parallel.

How pre-screening compresses the referral into a phone call

Pre-screening platforms sit between the referring physician and the research site team. Their job is to identify patients who may match one or more sub-studies within a master protocol before the site's formal eligibility determination. When pre-screening works, the site coordinator's first conversation about your patient is with a warm, filtered candidate rather than a chart pulled from a stack.

DecenTrialz uses AI-assisted matching and registered nurse-led pre-screening to identify patients who may fit a specific sub-study within a master protocol. The nurse pre-screen adds a clinical layer to the algorithmic match, catching nuance in the patient's history that a matching model alone might miss. The platform then routes the referral to the research site team, which owns the eligibility decision and every step that follows. For you, the practical effect is that a single phone call or referral form starts a process that would otherwise require you to read the whole appendix and pre-adjudicate the criteria yourself. Explore how DecenTrialz supports referring physicians and research sites.

Pre-screening does not replace the site team. It sorts the pipeline so the site team spends its time on plausible matches rather than on chains that end in a mismatch. It also does not replace your ten-minute read. Reading the schema and the master gate yourself is still what tells you a patient is worth pre-screening in the first place.

Refer with confidence

Master protocols look intimidating because the document is long. They read quickly once you know the shape. Two layers. One schema. Seven steps. Ten minutes. A referral to a research site team that owns everything that comes next.

The patient in the opening scene is one of a growing number who will only get access to research if her referring physician recognizes the shape of a fit in the first visit after her diagnosis. Master protocols are the reason that recognition is now possible for many more patients than it was even five years ago, because one protocol is quietly screening against many opportunities at once. Reading one well is a small skill with a large downstream effect.

Learn how DecenTrialz can help you refer patients to the right master protocol sub-study.