From your first paper to your hundredth.

The bigger issue is power, not the test. Four events cannot show a mortality difference between approaches. Expect a wide confidence interval that is compatible with both meaningful benefit and meaningful harm. That is the "underpowered study reported as negative" trap. A non-significant p here does not mean the approaches are equivalent. Say this explicitly.

Confounding. This is retrospective, and minimally invasive patients are usually selected, for example younger, with simpler pathology and better function. With 4 deaths you cannot adjust for anything. The usual rule is about 10 events per covariate, so a multivariable model or propensity-adjusted mortality comparison is not feasible. Show baseline characteristics with standardized differences, so readers can see how comparable the groups were. Use descriptive, non-causal language.

How to report it (STROBE, with SAMPL for the statistics):

• Give the counts and percentages per group, n/N (%), not just percentages.
• Report the risk difference with its 95% CI as the main effect size. Add the risk ratio with its CI. The CI is the headline, not the p-value.
• Give the Fisher's exact p-value, named as two-sided.
• Use wording like: "30-day mortality was X/70 (x%) after minimally invasive repair and Y/70 (y%) after sternotomy (risk difference ..., 95% CI ...; Fisher's exact p = ...)." Fill in the numbers from the tool exactly as given.
• In the limitations, state that the study was not powered for mortality and that selection bias is likely.
• Frame mortality as a safety signal. Do not frame it as evidence of equivalence.

Make the paper stronger. Mortality is too rare to carry the comparison. Prespecify an outcome with more events as the primary endpoint, such as a composite of major morbidity (STS-style), repair durability or residual MR at discharge, length of stay, or transfusion. With more events, an adjusted analysis or propensity score approach becomes realistic. Bring in a biostatistician before you run that.

What to do next

1. Run Fisher's exact test at https://askdimitri.net/tools/stats/#chisquare and record the risk difference and risk ratio with CIs.
2. Build Table 1 with standardized differences between the groups.
3. Choose and prespecify one primary endpoint with enough events. Keep 30-day mortality as a reported safety outcome.
4. Download the current STROBE checklist from the EQUATOR Network and map each item to your manuscript.
5. If you want an adjusted comparison on the new endpoint, book a biostatistician before analysis.

If you paste the tool's output here, I can help you interpret it and tighten the wording.

First, decide what the fellowship is for

A research fellowship is a bridge, so know where it leads. This decision shapes almost every choice below.

• US clinical career: Greek surgical training does not lead to US board certification. You would almost certainly need to repeat residency through the Match. That means ECFMG certification and USMLE alongside the fellowship.
• Academic research career in the US: The fellowship must produce first-author papers and a visible through line toward a postdoc or faculty research role.
• Return to Greece or Europe: The fellowship should give you a specific skill or dataset to bring home, such as outcomes research, trials methodology or a registry.

The 12-month plan

Months 0 to 2: foundation

1. Pick one through line, a single clinical question area you want to be known for. Scattered topics weaken every application.
2. Rewrite your CV in US academic format. Present publications accurately: "submitted" is not "accepted," and first and middle authorship are listed honestly.
3. Close or formally drop open projects in Greece. Unfinished papers cost you attention and credibility.
4. Name one concrete skill you offer. Programs take fellows who reduce their workload. Examples: systematic reviews to PRISMA 2020, R or Stata, database or registry analysis, or trial coordination.

Months 1 to 4: targeting and outreach

1. Build a list of 25 to 40 faculty whose recent papers match your through line. Search PubMed by topic and look at senior authors at US centers.
2. Send short, specific emails, under 150 words. Name one of their recent papers and what you noticed in it. State your concrete skill. Say whether you can bring funding. Attach your CV.
3. Expect a low response rate. Follow up once after two to three weeks, and use warm introductions from Greek mentors with US links.
4. Start a systematic review or meta-analysis in your through-line area now. It needs no visa or institution, it proves your skill, and it carries into the fellowship. This is the compounding engine at work.

Months 3 to 6: offers and funding

1. Expect video interviews. Prepare a two-minute account of your work and the question you want to pursue.
2. Settle funding early. Many research fellowships for IMGs are unpaid or require external support, and visa sponsorship usually requires proof of funds. Look at your home institution, Greek and European foundations, and society grants, and confirm eligibility and deadlines with each funder.
3. Get the offer in writing with salary or funding source, duration, supervisor and expected outputs.

Months 6 to 10: visa and logistics

1. Research fellows usually come on a J-1 research scholar visa sponsored by the host institution, not through ECFMG. ECFMG sponsors clinical training. Some routes carry a two-year home-country requirement, depending on funding source and other factors. This matters a lot if you later want to stay.
2. Work through the host institution's international office. Before you sign, talk to an immigration attorney about how this visa affects your long-term plan. I am not a lawyer, and these rules change, so confirm current requirements with the institution and the official sources.

Months 6 to 12, in parallel if residency is the goal: Start ECFMG certification steps and USMLE preparation. Confirm current pathway requirements directly with ECFMG, since they have changed in recent years.

A realistic warning

Without your own funding, the search usually takes longer. With even partial funding and a demonstrable skill, your odds improve considerably. Nothing guarantees an offer, but a focused list, specific emails and a review already underway are what move it.

What to do next

1. Write one sentence on what the fellowship is a bridge to.
2. Define your through line and your one offered skill this week.
3. Draft a target list of 25 faculty, with one recent paper each.
4. Begin a systematic review protocol in your area, and consider registering it in PROSPERO.
5. Map funding options and their deadlines.

To make this specific, tell me:

• your surgical specialty
• your publication record, roughly how many papers and how many first-author
• whether your end goal is US residency, a US research career, or returning home

1. Make your record findable and correctly attributed (week 1, about 3 to 4 hours)

Researchers cannot cite what search engines cannot connect to you. Name variants split your record. "D. Smith", "Smith DJ" and a missing middle initial can look like three different people.

• Choose one version of your name and use it everywhere from now on.
• Create an ORCID iD and import all 15 papers. Add it to every future submission.
• Create a Google Scholar profile. Claim every paper, merge duplicates, and set it to public.
• Build a My NCBI bibliography in PubMed so your PubMed record is clean.
• Update your institutional profile with a one-line research focus and links to ORCID and Scholar.

Measure: all 15 papers are correctly attributed on ORCID, Scholar and PubMed.

2. Audit the 15 papers and find your through line (week 2, about 3 hours)

List each paper with its topic, design, journal, your author position, and whether it is open access. Then look for the cluster. Most people with 15 papers have 2 or 3 themes and one that is clearly strongest. That strongest theme is what you want to be known for, and it decides which papers you promote first.

During the audit, check each publisher's self-archiving policy for the paywalled papers. Use the publisher's site or Sherpa Romeo, and confirm the current details there. Where the policy allows, deposit the accepted manuscript in your institutional repository. Free full text gets read and cited more readily than a paywall. Do not upload publisher PDFs to sharing sites unless the licence permits it.

Measure: a one-sentence statement of your position, in the form "I study [topic] using [method] for [audience]", plus the number of paywalled papers you have made legally accessible.

3. Start one sustainable channel and repurpose your best papers (weeks 3 and 4, about 1 to 2 hours a week)

For most clinician-researchers, LinkedIn is the most efficient channel. Write a headline that says what you actually do, using the statement from step 2. Then post once a week, starting with the 3 or 4 papers in your strongest cluster. Each post should cover:

• The question the paper asked and why it mattered clinically.
• What it found, in plain language, with the effect size rather than just "significant".
• What it changes for a reader, and what it does not, including its main limitation.
• Credit to your co-authors, by name.

Teaching what a paper means works better than announcing that it exists. A simple visual abstract roughly doubles the value of each post. Check your institution's social media policy before you start.

Measure: over the next 3 to 6 months, track messages from researchers in your field, collaboration or review invitations, and anyone citing or discussing your work. Likes are not the measure.

What not to do: do not self-cite to inflate your numbers, do not join citation-exchange schemes, and do not respond to unsolicited "high-visibility" journal or award offers that ask for fees. Each of these damages credibility faster than it builds visibility.

What to do next

1. This week: set up ORCID, Google Scholar and My NCBI, using one consistent name.
2. Next week: audit the 15 papers and write your one-sentence position.
3. Deposit accepted manuscripts wherever the publisher policy allows.
4. Update your LinkedIn headline and draft the first post on your strongest paper.
5. Put a recurring one-hour weekly slot in your calendar for the post, and protect it.

If you tell me your specialty, career stage and what the visibility is for, I can help you sharpen the position statement or draft the first post.