← Healthcare & Clinical Practice overview
Healthcare & Clinical Practice — the deeper record
What the coded record (2016–2026) shows for this profile, one population at a time — how each prong clears, the national-interest theories that appear, and the decisions themselves.
How cases like this are judged
Of the 36 decisions in this profile that reached the merits, here is where each Dhanasar prong was found not established. 2 more never reached the merits and are not counted below.
34 of those were the deciding prong.
2 of those were the deciding prong.
What the national-interest theories look like here
Two records, kept apart: the appeal record on the left, and beside it — never merged — what published approvals argued.
The appeal record
Among the 120 Consultants decisions that were deep-read, these are the national-importance theories the petitioners advanced (a case can raise several).
- economic growth generic82 of 120 (68%)
- economic job creation77 of 120 (64%)
- field advancement49 of 120 (41%)
- us competitiveness29 of 120 (24%)
- geographic or shortage area20 of 120 (17%)
- government interest19 of 120 (16%)
- public health14 of 120 (12%)
- critical emerging tech7 of 120 (6%)
- national security7 of 120 (6%)
- stem positive factor6 of 120 (5%)
- other4 of 120 (3%)
What this profile cannot afford to skip
Before the merits even open, 19.5% of decisions in the coded record failed the EB-2 threshold itself (864 of 4,420). For this profile, the record keeps returning to:
- The relevant question is not the importance of the industry or profession in which the individual will work. The endeavor itself must have substantial merit and national importance, distinct from the field's general value.
- Impact must reach beyond a particular geographic area or set of clients. The Policy Manual requires that the benefit be national in scope, not confined to a local or regional area.
- Letters must name your specific endeavor and trace how your work shaped later work. Officers are trained that letters praising skills in general terms do not add value as evidence.
What has shifted recently
- Prong 3 as the deciding prong moved from 0% (2024) to 2.5% (2025).
Read the decisions themselves
- The Petitioner proposed to operate a company providing occupational health and safety training and consultancy services—deep-read
- The Petitioner proposed to develop and implement an AI-enabled multimodal interoperable health data platform for chronicdeep-read
- The Petitioner proposed to provide occupational health and safety training, workshops, consulting, and research servicesdeep-read
- The Petitioner proposed to establish a consulting and construction company focused on sanitation, working with civil engdeep-read
- The Petitioner sought to create an independent consulting service company conducting scientific research on dietary suppdeep-read
- The Petitioner, a former educator with a background in arts and art therapy, proposed to operate a small consulting busideep-read
- The Petitioner, a general and operations manager, initially proposed to provide expert managerial services to U.S. compadeep-read
- The Petitioner, described as a regulatory research consultant, claimed contributions to oncology drug development and redeep-read
Readings of the record
Receipt · S48computed from the record as of 21 July 2026Make the endeavor broad enough to matter nationally, yet concrete enough to picture
An endeavor can fail from either side, and the record carries both failures.
Read the full card →
What the record shows
An endeavor can fail from either side, and the record carries both failures. Too broad reads as vague, coded in 52.0% of prong-1 failures. Too narrow reads as local rather than national, coded in 58.8% (the S48 stack). A strong endeavor sits between the two: large enough to carry national importance, specific enough that a reader knows what the work actually is.
Stronger and weaker, side by side
An illustration of the difference, in a neutral field. Too narrow: tuning one model for one agency, which reads as a task. Too broad: advancing science, which reads as a slogan. The version that clears both bars names a specific line of work with a national application: developing forecasting methods for extreme weather and putting them into public emergency planning.
What to check in your own
Ask both questions at once. Could a reader say why this matters to the country? And could they say, in a sentence, what you actually do? If either answer is no, the endeavor is sitting on one side or the other.
What it does not say
This is a reading of the appeal record, its own population of contested denials, not a sample of filings. It names two opposite failure patterns. It does not predict any outcome, assess any case, or tell you what to write.
Receipt · S48computed from the record as of 21 July 2026Give each recommendation letter one distinct job, so no two make the same point
Letters that all make the same point add up to one point.
Read the full card →
What the record shows
Letters that all make the same point add up to one point. Conclusory letters, which assert importance without carrying a distinct piece of the case, are coded in 18.1% of prong-1 failures (the S48 stack). Coverage of different needs is what makes a set of letters worth more than any single one.
Stronger and weaker, side by side
An illustration of the difference. A weaker set is six letters that each say the person is brilliant and the work matters, so their value collapses into a single claim. A stronger set gives each letter one job: one carries the technical ability, one the field's importance, one the real-world use, one that the person is well positioned to continue, with no two covering the same ground.
What to check in your own
Lay the letters side by side and ask what each one is for. If two are making the same argument, one of them is a wasted slot. Which prong need does each letter close that no other letter covers?
What it does not say
This is a reading of the appeal record, its own population of contested denials, not a sample of filings. It names a failure pattern in how letter sets are built. It does not predict any outcome, assess any case, or tell you what to write, and every letter is prepared for its signer to verify and sign.
Receipt · S48computed from the record as of 21 July 2026Frame your endeavor as a mission you carry out, not the job you happen to hold
The most common way a national-importance argument fails in the appeal record is an endeavor stated as a job: the work of one employer, one role, one place.
Read the full card →
What the record shows
The most common way a national-importance argument fails in the appeal record is an endeavor stated as a job: the work of one employer, one role, one place. Across prong-1 failures, the endeavor is coded as vague in 52.0%, the benefit is tied to a specific employer in 23.9%, and the scope is local rather than national in 58.8% (the S48 stack).
Stronger and weaker, side by side
An illustration of the difference, in a neutral field. A weaker endeavor reads as the job itself: building forecasting models for an employer's weather-alert product. A stronger one names the mission the job serves: improving early warning for extreme weather, with the current role as one place that mission is applied, and with room to continue through research, tools, and collaboration across the field.
What to check in your own
Read your endeavor statement and ask: if you changed employers tomorrow, would this endeavor still exist? Does it name a mission, or a job description? Is the benefit national, or bounded to one organization or one place?
What it does not say
This is a reading of the appeal record, which is its own population of contested denials, not a sample of filings. It names a pattern in how endeavors fail. It does not predict any outcome, assess any case, or tell you what to write.
Receipt · S48computed from the record as of 21 July 2026Frame the endeavor so it would continue even if you left your current employer
When the benefit of an endeavor is framed as accruing to one employer, the national-importance argument collapses into a private one.
Read the full card →
What the record shows
When the benefit of an endeavor is framed as accruing to one employer, the national-importance argument collapses into a private one. The record codes this as an employer-specific benefit in 23.9% of prong-1 failures (the S48 stack).
Stronger and weaker, side by side
An illustration of the difference, in a neutral field. A weaker framing rests the importance on the employer: this matters because the company's platform serves many people. A stronger one treats the current role as one place the work is applied, and describes an endeavor that would continue through future roles, independent research, or consulting just as well.
What to check in your own
Ask the survival question directly: if you left your employer next month, does the endeavor still stand on its own? Is your current job cited as evidence that the work is being done, or as the reason the work matters? The first is fine; the second is the trap.
What it does not say
This is a reading of the appeal record, its own population of contested denials, not a sample of filings. It names a failure pattern. It does not predict any outcome, assess any case, or tell you what to write.
Receipt · S48computed from the record as of 21 July 2026Start from the problem you keep working to solve, not from your title or your degrees
A vague endeavor is the single most common prong-1 failure in the appeal record, coded in 52.0% of prong-1 failures (the S48 stack).
Read the full card →
What the record shows
A vague endeavor is the single most common prong-1 failure in the appeal record, coded in 52.0% of prong-1 failures (the S48 stack). A case that opens from the credential often arrives at a list of qualifications with no specific undertaking behind them, which is what "vague" describes.
Stronger and weaker, side by side
An illustration of the difference. A weaker opening leads with the credential: a doctorate and a count of publications offered as the qualification. A stronger one leads with the problem: years of work toward one specific problem, with the degrees and publications standing as the evidence that the expertise to pursue it was built over time.
What to check in your own
Before you name your title or your degrees, ask: what problem have I been moving toward solving? What would I still be working on if the current job ended? Does my evidence show a problem pursued, or a resume assembled?
What it does not say
This is a reading of the appeal record, its own population of contested denials, not a sample of filings. It names a failure pattern. It does not predict any outcome, assess any case, or tell you what to write.
Receipt · S57computed from the record as of 21 July 2026Order your exhibits to follow the story of your expertise, not to group them by file type
On the dismissed side of the record, evidence is credited far less often than filers expect: only 7.5% of the items the decisions discuss are recorded as credited (the S57 evidence-credit reading — these are items the AAO wrote about, not everything filed, and what varies most is the ordering by type).
Read the full card →
What the record shows
On the dismissed side of the record, evidence is credited far less often than filers expect: only 7.5% of the items the decisions discuss are recorded as credited (the S57 evidence-credit reading — these are items the AAO wrote about, not everything filed, and what varies most is the ordering by type). The craft lesson, from the documented preparation record rather than the corpus: an exhibit earns its credit most easily when a reader can tie it to the endeavor without work. A pile sorted by document type leaves that tie for the reader to build, or to miss.
Stronger and weaker, side by side
An illustration of the difference. A weaker arrangement is a filing cabinet: one tab of every diploma, one of every publication, one of every letter. A stronger one follows the arc: how the expertise started, how it specialized, how it was recognized, and how it is applied now, with each exhibit placed where it moves that story forward.
What to check in your own
Take any exhibit and ask what it is doing in your argument. Can a reader see, at a glance, which part of your expertise it supports? If the answer requires them to reconstruct the connection themselves, the order is working against you.
What it does not say
This is a reading of the appeal record, its own population of contested denials, not a sample of filings. It describes how evidence is treated, not how any case was decided. It does not predict any outcome, assess any case, or tell you what to write.
Receipt · S48computed from the record as of 21 July 2026Show how your expertise was built stage by stage, so the endeavor reads as earned
When the arc that led to an endeavor is missing, what remains reads as an assertion of expertise rather than a demonstration of it, and that is the failure the record codes as vague: 52.0% of prong-1 failures (the S48 stack).
Read the full card →
What the record shows
When the arc that led to an endeavor is missing, what remains reads as an assertion of expertise rather than a demonstration of it, and that is the failure the record codes as vague: 52.0% of prong-1 failures (the S48 stack). The officer's own trained standard points the same way: name the specific endeavor and trace its development.
Stronger and weaker, side by side
An illustration of the difference. A weaker statement declares the standing: "I am an expert, and my work is important." A stronger one shows the path that built it: foundations first, then a focus on a specific problem, then application in a real setting, with each stage leading to the next and to the proposed endeavor.
What to check in your own
Read your statement as a stranger would. Can they see how one stage led to the next? Or does it assert expertise and ask the reader to take it on faith? A story that arrives is read differently from a claim that is stated.
What it does not say
This is a reading of the appeal record, its own population of contested denials, not a sample of filings, alongside the published officer standard. It names a failure pattern. It does not predict any outcome, assess any case, or tell you what to write.
More is coming to this page: the profile’s doctrinal signatures once counsel has cleared them. Each lands when its own review does.