Google has changed
Retrieving useful search results from a search engine is not always reliable. At this time, I look to Google as the standard-bearer of how well a search engine should perform. The original page-rank algorithm was impressive, and showed how poorly other search engines performed. However, Google's business model has changed, and it is clear that its algorithm now seems to take other factors into account. They got into the business of filter-bubbling, and preferentially showing you results you are likely to click on. The end result is that your search results are often different from other people's results. Also, the more you click on certain links, the more this is taking into account by the algorithm, and this affects what you will see the next time you enter the search terms. And now, Google has taken it upon themselves to decide if a website contains "disinformation" or information that might be harmful to their preferred political candidate. All this decreases the usefulness of Google, but it is still the default search engine used by most people. Gonna be a tough habit to break.
It can't parse metonyms or understand the nuance behind the query
A blogger named Sandy Maguire <a href="https://reasonablypolymorphic.com/blog/monotonous-web/index.html">recently posted</a> on his frustrations with Google. First, he decried that Google was not smart enough to understand metonymy. He gives the example of really enjoying the vibe at Koh Lanta, Thailand, and so he asked Google to find the "Koh Lanta of Croatia". He expected Google to figure out that he wanted Google to understand that he was searching for corresponding place in Croatia that would give him the same "vibe" as Koh Lanta. Not surprisingly Google failed. I mean, c'mon.
But having worked for Google, he knows that Google also rewards a website for how long you stay on the search result. (Did you know that?) Websites designed with SEO in mind will rank higher, even though their content is not what you wanted.
Then he describes that Google, being the most-used search engine, feels that because it is under constant government scrutiny, that it modifies its search results accordingly. He yearns for the Google of 2006, when you got good search results, as you expected. Instead of what you get now. Sandy feels that Google is the victim of its own success, and would like to see another search engine that ignores the tactics of "aggressive SEO", and boosts personal blogs, where people often have more interesting things to say. A search engine "by the people and for the people".
But we don't even have anything approaching a Google-type search engine for the medical field
So how shall one design a search engine that will pull up relevant and quality results for a physician who needs this information to treat patients? One tactic would be to avoid casting a net over the entire Web, but instead focus on select sites likely to contain the specific and relevant information. This would be medical journals and published guidelines, but also meeting summaries from relevant symposia as well as smaller meetings. Of course, the content from these meetings are the property of certain groups and academic centers, and they will not likely yield up their intellectual property without remuneration and possibly some editorial control. This limitation will be difficult to overcome without financial backing and some political clout.
Only after it were made possible to have medical journals and meeting organizers allow their content to be indexed and be made retrievable by some entity other than PubMed, would it be possible to study the benefit of more sophisticated search strategies to retrieve relevant documents for a query. Google already does index some content in medical journals, so I do pay attention to what Google is doing. So far, however, they are not much use as a clinical query tool. At least not for physicians, nurses and other healthcare providers who want content not intended for the lay audience.