Monday, April 20, 2009

Keeping up with the literature

I’ve never been good at keeping up with Literature.

In times past, the department librarian would circulate journals with a list of names stapled to the cover. Researchers were asked to scan the table of contents, read the articles of interest, and pass it on within a day.  Within a week, the publication would get back to the library.  In reality, it would hang up in a pile on someone’s desk, out of reach.

Scheduling a half day to go to the library, browsing the shelves or pulling a stack of journals, is similarly impractical: time at work is too fragmented and scheduled, and time at home requires taking away the journals overnight (our librarian was always checking me for filched issues).

The Internet has changed a lot of that: we have broad on-line archives stretching back many years and fully indexed by topic, keyword.

About once a week I need to search a clinical topic in depth: I like using Stanford’s Highwire interface over Pubmed.  Their links to cited and citing literature, and to related literature and reviews, are marvelous. cover_imageI can complete a substantial research project in an afternoon, building a .pdf archive for the project and printing a few key papers to read on the exercise bike.

It’s still hard to browse an e-journal, though.  Circulation comes by email, JAMA and the New England Journal offer podcast updates, but neither has access to full-content. To get an article, I have to log into the company server, route through many security screens, then drill down to specific issue and article.

I do like using RSS feeds to aggregate content across sources. InCirculation  feeds about 30 summaries each day, and Medworm also looks promising.  I collect new literature into Netvibes, along with blog updates and news feeds, and dispatch it all over coffee in about 30 minutes.

Unfortunately, I’ve found nothing in the technical or business literature that offers the same broad, deep, cross-linked access capability.  I’ve tried subscribing to Wired (too much) or IEEExplore (no detail) for technology, or EBSCO (no access) for management information, but it’s really unsatisfactory.

Worse, I’ve found that I really have to play with technology to learn about it.  Today I saw a reference to Curl in a tweet from Adam Curry; it’s a new language for creating rich internet applications developed at MIT.  Wikipedia gave me a bit of background and a couple of key links, but I feel like I’ll need to invest days to really understand what it’s all about.

All of which begs a recurring question: What matters to me as a manager?  My role is

  1. Support management with informed strategic insights and accurate project estimates,
  2. Supply cross-functional teams with technology details and clinical data for planning and resourcing, and
  3. Lead and evaluate the work of scientists and engineers in my group, evaluating their proposals and intermediate results and coaching them on alternative approaches and career development.

imageSo, in my role (circa 1985, right) I need to be top-line familiar with the latest medical news and research findings, and can increasingly do that from feeds and searches for clinical literature.

But it’s the details of technologic innovations where I worry: I can apply a process to assessing the quality of results, and an intuition to whether an engineer is getting lost.  But I can’t help redirect them unless I’m prepared to invest a lot more time in learning, or trust outside experts with the task.

No comments: