Recently, due to budget cuts at AHRQ https://guidelines.gov and https://www.qualitymeasures.ahrq.gov/ have been scheduled for shutdown. This blog post documents our efforts to mirror those resources and our reasons for doing so. We have published the raw files that we mirrored at
Updates:
- Multiple different efforts to backup the websites have been successful. Take a look at this tweet for others besides the one featured here. Several different approaches were taken, which should hopefully ensure that we can resurrect most of the sites on shutdown day. We will see, and I need to sleep.
- Twitter user @randyprine got a quick reply from the AHRQ folks that he posted. Indicated that the folks at AHRQ are trapped by lack of funds.
- We are coordinating further progress on the twitter hashtag #SaveTheGuidelines
- If you do not ALREADY HAVE a mirror bot running, please do not start now, we have multiple people using multiple approaches to mirror the site, and it is obvious that we are causing something of an organic DDOS..
- If you want to help, take a look at https://github.com/CareSet/AHRQ_search_clone It is both sourcecode for cloning and the resulting mirrored html files.. Help me figure out what I am missing. The assumption should be that if way back machine does not have it, or we do not have it by midnight.. it could be lost forever.
- Also, although the decision to take down the web resource lies with the Trump administration, the decision to substantially defund AHRQ was an Obama-era change. So not sure exactly who the “credit” for this should stay with.
We have known for some time that AHRQ resources might be targeted for deletion. Hard to say whether this started in the Trump administration, since apparently funding was originally cutoff during the Obama administration, but Trump certainly put the nail in the coffin for this web resource. Thankfully, we had taken some steps to study the problem in advance. Sadly, that preparation did not keep us from being surprised that by the press that several important sites will be taken down tomorrow. Thanks to the dailybeast and sunlight foundation for cluing us in. Specifically, the sites and resources that were targeted include:
- https://www.guidelines.gov – a website provides meta-information for clinical guidelines, including evaluation mechanisms, comparison tools and of course, links.
- https://www.qualitymeasures.ahrq.gov/ – the quality measures clearing house, is searchable from this main page. This is a webapp that provides meta information for clinical quality measures.
- https://www.qualitymeasures.ahrq.gov/hhs/index.aspx – The clinical measures inventory. At this point I do not understand the difference between the “clearinghouse” and the “inventory” but they have different search functions and results pages, and the number of results is not identical… so I am backing them both up distinctly.
- https://www.qualitymeasures.ahrq.gov/expert – Expert commentary, I think these are all backed up in wayback, but I am making sure.
(Note: Are there other top level domain names that are threatened? Let me know)
For those who have no idea what this means, these two domains provide critical information on evidence-based medicine. Frequently, clinicians who provide services can become blind to the effectiveness of their treatments. While not necessarily malicious this has happened multiple times over the years. Specifically in the controversy around the effectiveness of spinal surgery, high-dose chemotherapy and the early aggressive treatment of prostate cancer. In all of these cases, physicians recommended expensive and traumatic treatment in the face of evidence that those treatments were ineffective or unnecessary. If this topic interests you I recommend the book “How we do harm” by Otis Webb M.D.
The movement to prioritize science over profit or instinct when determining which medical treatments are recommended is called “Evidence Based Medicine”. AHRQ as a government agency, along with IHI, Academy Health, The Cochrane Collaboration and even Wikipedia have been central figures in this ongoing movement, to ensure that patients are treated ethically, from science-based principles.
I am not sure which lobby was tired of having AHRQ resources recommend against their clinically ineffective services or products, but we suspect that a “union” of Spinal Surgeons was involved.
It is not actually clear, however, just how important these websites are in the larger scheme of promoting evidence-based medicine. AHRQ as an organization has clearly been a great benefit. But these sites, for the most part, simply provide links to the Guidelines and Standards that other organizations maintain. It is not clear that they contribute substantially to the understanding of the average clinician has on evidenced based medicine, as opposed to the work of the underlying standards organizations.
There have been criticisms that the AHRQ website has been out-of-date, for instance, and did not actually do that great of a job at ensuring that the latest and greatest information was available. Especially as we move forward it is critically important that people check with the original source of the guideline or measure, to ensure that they are following the latest evidence. This will become a bigger and bigger problem as the “last day” snapshot that we are taking ages, and eventually we will likely take down this cache, once a suitable replacement arrives. (Thanks to Jenniffer Hinkle @Oncotastic for the insights here)
However, there must be some legitimizing effect that having this information on a “dot gov” website must have.
Now, we are in a desperate situation, we need to backup the sites that are being taken down. Thankfully, archive.org is very good about mirroring static websites. The concern has been that search functionality and the contents of the underlying databases would go missing. A report from the Sunlight Foundations Web Integrity Project says as much.
So far, we have been able to mirror the contents of the “search all” and the individual guidelines and measures pages. That work, both the code and the resulting archives, are available at the following url:
https://github.com/CareSet/AHRQ_search_clone
Here are two videos, that I will likely delete from the final version of this article to catch you up on what the cloning repo has done, and what might still need to be done.
How you can help!!
- Create github issues for functionality, sites or data that we are not yet mirroring.
- If that is too geeky for you, contact me at http://twitter.com/fredtrotter to let me know if I am missing something.
- As you are browsing the websites, if you see “data full” pages, use the chrome wayback machine extension to convince archive.org to make a quick backup of the page.
Regards,
-FT