First up, this is an outstanding piece of work which the authors and organizing team should be very proud of. The sections are well set out and easily navigated. The quality of the opinions are excellent and well referenced. I was particularly pleased that the links are included back to PubMed and many of the articles can be retrieved direct on the fly. Discussions of evidence are robust and well informed. The ability to submit evidence is welcome (but may represent a rod for your own backs!).
Where this style excels is for reference before (or possibly even during) an MDM discussion where a particular facet of management is being contemplated. I took it for a test drive whilst filling in my MDM pro forma today and found several (minor) areas where our current practice was not well supported by the evidence. Given that this is a key role for the software it follows that fast navigation will be absolutely critical. Hence I will comment on this first.
The search box doesn't like typos (unsurprising) but if I search for cerebral metastases I get nothing, whilst brain metastases gets many hits but brain mets gets none. Some way of tagging the sections might help the searcher?
If I access via the Lung Cancer page and search for PET/CT I get a number of results for endometrial cancer. Can this be altered so whilst in the lung cancer guidelines search results only pertain to lung cancer? After all, in a lung cancer MDM if I need to look something up I don't want to wade through non-lung results. Conversely if I want to look up a non-lung topic I would specify the organ in that search. Maybe a "Search in Lung" as opposed to "Search all site", but better to have one box with smarter filters.
Another possibility would be some kind of grid for NSCLC where tags could be selected e.g. "stage IIa" "inoperable" and "radiotherapy" could be selected and the relevant article appears. It would have to be neat & tidy though, else would be a distraction from the main body of work. Alternatively, Advice for NSCLC and SCLC could be presented on a different coloured background to allow easy differentiation.
SCLC is still referred to as limited or extensive against 7th TNM staging, which will change with time as the evidence is reappraised I guess. With regards to scope, I assume the SCLC / NSCLC dichotomy has been chosen for pragmatic reasons, as I could find no mention of invasive adenocarcinoma of lepidic type (bronchoalveolar cell carcinoma), or the management of paraneoplastic manifestations, for instance. Not that evidence in these circumstances is thick on the ground.
There are a handful of minor typos and grammatical errors. Should these be commented on in individual sections and if so will the comments remain there to reveal my pedantry forever, or are comments relayed to the author for amendment as needed and then deleted or shunted off into storage? Minor typos etc tend to get kept on wikis but are mainly irrelevant.
Are sections that are recently updated marked out as such? This would be one way of keeping up to date with new evidence. EDIT: I see that can be done after registration, thanks.
Very minor one - when browsing in Safari 5.1.7 on OSX 10.7.3 pressing enter after typing a search query doesn't work, search has to be manually clicked. Works fine on Chrome 20, Firefox 12.0 & Opera 12 though (multiple hits of enter needed). I like the boolean search. I haven't looked at the site in IE, but the formatting is excellent on OSX browsers.
Clearly the guidance relates to treatment only at present - are there plans to extend to investigation etc?
Overall, I think this a fine body of work that should form a standard to which others aspire. Congratulations to all involved. I gather this is a world first, and you are certainly leading the way. Thanks for all your hard work.
I have discussed the guidelines with Ben and essentially endorse all of his comments, with little extra to add.
I agree that the beauty of this format is that it will be very practical for use at MDTs, and has the potential to clearly inform decision making at that meeting in real time. From that point of view it would be great if the few typos and "gremlins" in the search engine etc could be ironed out, in order to make them as easy as possible to utilise. However given the world first status of using of the Wiki format/platform for such guideline, these issues are remarkably few.