Decision analysis? – where are the details?

November 15, 2015


In the Milan conference (1st EFLM Strategic Conference Defining analytical performance goals) one of the papers (1) suggests that analytical performance specifications should be prepared from indirect outcome studies using decision analysis. The only example presented is a simulation, which is not decision analysis. Decision analysis is also discussed in this section but on an abstract level.

I have performed decision analysis and discuss it in my book (2). Decision analysis requires a quantitative variable that is either maximized or minimized. In my case, we performed financial decision analysis and the parameter to be maximized was net present value (NPV) of future cash flows. The Milan paper never identifies a quantitative parameter to be optimized.

I don’t understand how decision analysis can be recommended without any known examples or details about how one would go about it.


  1. Horvath AR, Bossuyt PMM, Sandberg S, Setting analytical performance specifications based on outcome studies – is it possible? Clin Chem Lab Med 2015; 53(6): 841–848.
  2. Assay Development and Evaluation: A Manufacturer’s Perspective. Jan S. Krouwer, AACC Press, Washington DC, 2002, see Chapter 3.

How to insult clinicians

November 5, 2015


In my blog post Total Error and Milan, I mentioned how clinician surveys were in the draft consensus statement but dropped from the final and published consensus statement. (The draft consensus statement is no longer available on the EFLM site).

I had occasion to read a paper from the Milan conference (1) where it is clear why clinician surveys were dropped.

“RCVs from vignettes should probably not be used on their own as a basis for setting analytical performance specifications, since clinicians seem “uninformed” regarding important principles.”

RCV = reference change values.

For an example of how clinician surveys were used to set analytical performance specifications, see reference 2.


  1. Thue G and Sandberg S: Analytical performance specifications based on how clinicians use laboratory tests Clin Chem Lab Med 2015; 53(6): 857–862.
  2. Klonoff DC, Lias C, Vigersky R, et al The surveillance error grid. J Diabetes Sci Technol. 2014;8:658-672.

Fake Reviews

October 22, 2015


Amazon has been in the news regarding their suing people who write fake reviews of their products. A recent article in the New England Journal of Medicine describes cases where authors supply fake names and emails when a journal to which they have submitted a publication asks for suggested reviewers. These emails belong to the author, who then writes fake reviews so the article gets accepted.

I don’t there if there is a connection but in the past year or so, I get an email about once a week from different journals asking me to submit an article. I assume that many people gets these emails.

And don’t forget this site, which allows one to generate a fake paper.

The most cited method evaluation paper

October 10, 2015


Last year, Nature compiled the 100 most cited scientific papers. The top paper that deals with method evaluation is Bland and Altman’s paper from the Lancet in 1986. Number 29 on the all-time list, this paper is the source for the Bland-Altman (or difference) plot and associated statistics. One of the reasons that this paper is so popular is that is makes a complex topic simple. What could be easier than to plot differences. A paper in Clinical Chemistry showed that in spite of the popularity of the Bland-Altman method, it was not used as much as one might expect (only about a third of papers used it).

Here’s a thought about total error

October 5, 2015


If total error – calculated by the usual models – is set to equal 95% of the results, then how can 95% instead of 100% represent total error? Shouldn’t “total” imply all results? Putting things another way, what do you call the additional 5%?

Word doesn’t recognize measurand

October 1, 2015


I was working on a paper and decided to comply with the nomenclature expected by the journal and used the word “measurand.” The word was underlined as unknown by the dictionary used by “Word.” I went to an online version of the Merriam Webster dictionary and no match for found for measurand. So much for ISO nomenclature.

Triangulating Dynamic? – You’ve got to be kidding

September 21, 2015


Clinical Chemistry has published one of its Q&A pieces (subscription required) with the title “Triangulating Dynamic of Clinical Laboratory Testing.” With a title like that it’s not surprising that 8 pages contained no useful information. Given that some of these people work for laboratory medicine consulting companies, it would have been nice to provide information such as test volume by analyte, company revenues and profits, revenue by analyte, etc. But don’t expect to see any Letters about this piece because Clinical Chemistry won’t accept Letters about anything other than articles.


Get every new post delivered to your Inbox.