Publishing is not that hard. I published 7 during residency, and 2 during med school. I have 2 manuscripts just waiting for boards to be over with to complete. Some were little ditties, others more substantial. Here's my (non-PhD) approach:
1) Find an attending who already has an idea, i.e. "Hey look, I've collected 100 cases of X and always wanted to stain them for Y."
2) Develop a hypothesis. Whatever question you are asking needs to have a yes or no answer essentially, i.e. "Does lesion X stain more diffusely for impox Y than its normal counterpart?" DO NOT pick an ambiguous hypothesis. DO NOT choose a project without a well defined end point. This is very important so that you do not drag on forever with your work.
3) Pick your methods, your (statistical) analytic and grading schema, and your conclusions based on your findings. For instance, if the lesion is more positive for Y than normal, then [insert conclusion A here], or, if the lesion has the same positivity as normal, then [insert conclusion B here].
3) Pick a target journal. Get a copy of the author guidelines and follow them to a tee. Write your manuscript exactly for that journal. Look at work that is similar in concept to yours to see how other authors tackled it.
3) Write the intro / methods as the work (e.g. impox) is being done. The intro in my opinion takes the most work.
4) Analyze your data. Use a statistician (which are often available at large universities) if you are uncomfortable with number crunching.
5) Plot your figures/graphs/charts from the data.
6) Summarize your results from the figures of your data for the manuscript. I think that the entire results section works best if done AFTER you do your figures/tables.
7) Write the discussion /conclusion. Be very careful not to make claims that are not actually supported by your data. Draw specific conclusions--as editors have repeatedly told me, drawing no definite conclusions from your work is bad.
8) Have all of your authors look at your (near-complete) first draft, make changes per their input.
9) Send your manuscript out either online or snail mail per the author's instructions. Wait several months for the revision (I did have 1 paper accepted "as is"), revise right away, and send it out again.
10) Rejected? Who cares, send it to the next bad boy (i.e. journal) in line. Remember, though, similar journals may use the same reviewer who rejected you the first time around. I have NEVER had a paper ultimately be rejected. I have had to journal shop to find a good home though
Other bits of publishing advice: CP labs generate HUGE amounts of data, all for the publishing. Just ask around--someone's bound to have a half-finished project looking for a dedicated first author. Keep in mind, case reports are almost as much work as primary (clinical-oriented, don't want to peeve the PhDs!) research. I have published only one, I think, though I have published small case series. I think in rare entities, case reports can be very important. One paper I wrote specifically dealt with how to manage refractory TTP via a case description. TTP is a fairly rare and very often anectdotally treated condition. In fact, we "cured" this patient by using case report data for a fairly experimental approach as a guide, so felt it was important to add ours to the literature.
Finally, in regard to fellowships and research: I think that the resident who publishes (and is a good pathologist) will be considered a superior candidate to the resident who does not publish. Why? Salaries in academics are based on publication quantity. A fellow who pumps out papers for their attendings is a huge financial plus. A person with a track record for being able to publish and do their work during residency is ideal.
Mindy