In a February 10 article for The Denver Post, Katy Human wrote, “Children with health insurance, studies have shown, are less likely than uninsured kids to end up in emergency rooms, more likely to get key vaccinations and less likely to be absent from school.”
The article promoted tax-funded health programs and included not a word from critics, but I was first interested in Human’s claim about the studies. Which studies did she have in mind? I asked Human via e-mail, “Is the lack of insurance causing the problems mentioned, or is the lack of insurance itself a symptom of having poorer and less educated parents (on average)?”
Human responded on February 12:
Oh, I know you know the answer to this, Ari. There are many many many studies on this — and all, of course, control for factors such as income and education of parents. There are also studies showing before-and-after for same kids and same families, once the families make a change (adding or dropping insurance.)
But what Human did not do is provide me with a single citation regarding these “many many studies.” I suppose that at least some of the studies that she had in mind do contain the sorts of controls that she mentioned. However, I wanted to look for myself, not take Human’s word on faith. Moreover, not only did I want to see for myself whether the statistical controls are adequate, but I wanted to learn what is the magnitude of difference. How much difference is there between the insured and uninsured?
I asked Human on February 12 and again on February 13 for her citations. I was hardly being overly demanding in my request; I wrote, “You mentioned that there are many such studies; citations for the two or three that you find most persuasive would suffice.” This would have taken only a minute or two of Human’s time, as obviously she is already familiar with the studies in question.
I have yet to hear back from her.
I suggest that The Denver Post adopt the following policy. If reporters, for lack of space, mention but do not specify studies or other sources, the reporters should be required to provide the names of those studies or sources to interested readers. Otherwise, readers have no way to verify the reliability of the studies or sources.
February 21 Update: In response to this post, Human send me a list of citations:
Here you go. I won’t be doing this for you again – you can do it yourself, and I don’t have time to repeat these types of searches for everyone who asks.
http://www.ncbi.nlm.nih.gov/pubmed/15121980?ordinalpos=1&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_RVAbstractPlusDrugs1
http://www.ncbi.nlm.nih.gov/pubmed/18219242?ordinalpos=2&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_RVDocSum
http://www.ncbi.nlm.nih.gov/pubmed/18045482?ordinalpos=10&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_RVDocSum
http://www.ncbi.nlm.nih.gov/pubmed/17805222?ordinalpos=9&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_RVDocSum
ADULTS: http://www.ncbi.nlm.nih.gov/pubmed/18096863?ordinalpos=5&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_RVDocSum
I replied, “Thanks! However, you are incorrect that I can correctly guess the studies that you have in mind on my own.” After all, I am not a mind reader. Moreover, I do not regard my request as an imposition, given that Human already knew which studies she had in mind. I could have spent hours trying to guess the studies to which Human was referring and still not guessed correctly, while it took Human perhaps a minute or two to send me the links.
If Human does not wish to respond to readers about her citations, then she is free to include them briefly within her articles. In this case, all of the links point to the National Center for Biotechnology Information. Including that information would have added seven words to Human’s article, including the worlds “by” and “the.” Even that brief citation would have pointed readers in the right direction.
I will evaluate the studies within the next couple of days and then discuss the political implications of them. February 22 Update: I’ve started to work on this, but it will take me some days to write up the results, which I may release with a co-author and perhaps first through another outlet.