Do Knowledge Specialists dream of data science?

A personal point-of-view of the PG Certification in Clinical Data Science

It’s been over two years since I wrote about ChatGPT, and a great deal has changed. The number of publications around generative AI has skyrocketed; a clear sign that healthcare professionals are seeing value in utilising these technologies.

I’ve had the extraordinary privilege to chat with like-minded people from across the world; from a GP in the United States, to very lovely librarians in Ireland. I’ve seen many brilliant knowledge and library specialists and clinicians globally putting these tools to good use, critically evaluating their effectiveness and their pitfalls in practice.

Knowledge specialists and librarians up and down the country have been sharing tips, resources, and tricks. It’s what we’re good at. We’ve learned how these tools work and have weighed the benefits of their use. Some of us are showing our users how to use these tools effectively.

Some things have changed since that blog post. Bing’s search-enabled LLM has been transformed; CoPilot is being rolled out and integrated into various Microsoft applications. Google’s Bard is now called Gemini and has been integrated into most Google searches; summarising content, reducing clicks and traffic to websites. Debates and concerns about accuracy and the environment continue.

Different iterations of ChatGPT have popped up too, each improving on the last.

Inevitably, there’s been an explosion of various tools incorporating large language models; tools now exist that can credibly search and summarise information within seconds. There are a few of these about now and knowing which ones are useful, and which are questionable, is a new skill in itself.

Crucially, I’ve changed since writing that blog post too – I’ve learned a lot, and earned a PGCert in Clinical Data Science, thanks to the bursary which is up and running again, closing date 26th May 2025.

The course was equally a lot of fun and a lot of hard work. It was a commitment, but one well worth the effort. I’ve been asked many questions and have detailed some of the common ones below, for those intrepid knowledge and library professionals who are thinking about applying.

“Do I need to know how to code?”

I went in with very minimal understanding of Python and R. While it helps to have a vague understanding of these languages, it’s more important to have a genuine interest in data, and to have the time to learn. The course offers some e-learning too to help with the basics, which I found really useful!

There was a lot of freedom in the course to simply pursue our own interests. Data comes in many forms; from literature search spreadsheets to stuff scribbled down on paper.

You’ll get the opportunity to play with data using very user-friendly software; I tend to learn by doing, so I often spent time simply making my own fictitious (synthetic) datasets to use. This made the course a bit more relevant to me, and more fun.

“Part time or full time?”

I did the course full time – so it was two hours’ worth of online meetings every week or two, and an assignment every couple of months or so (two written essays, two presentation recordings). The meetings were recorded too, so it was accessible.

I found it manageable while working full time, but my commitments outside of work were pretty minimal so I could afford to take the time to study, even if my friends grumbled about it occasionally!

With the course itself, the more you put in, the more you’re likely to get out of it. I took full advantage of the resources on offer and had a lot of fun in the process!

“I’m worried I’ll be rubbish at it and fail!”

Us knowledge and library folk are a strange sort; what we lack in confidence, we often make up for in knowledge and tenacity (and ironically, being blissfully unaware of how clever we can be!).

I see this as a secret burden of our profession. While we might not be proficient in every skill known to humankind, we’ll almost certainly know where to look to learn. And that makes us extremely adaptable and formidable professionals.

I went into the course expecting to fail some things (which thankfully didn’t happen) because it was well outside of my experience and comfort zone. But I was determined to learn and went in with an open mind, without much expectation. Failure can be an important part of learning, after all!

But by studying, being careful with the assignment rubric, and taking my time, my overall average was a humble 67.3%.

I came out of the course feeling more confident in my abilities, and with a renewed appreciation of data. Strangely, my highest scoring module was one I had the least amount of confidence in.   

“What can I do with my new skills?”

I’ve learned how to visualise bibliographic data collected from my literature searches; from the number of papers published in certain countries to trending Medical Subject Headings in a search. These visualisations have proven useful for making my searches look good, presenting information in a more accessible way.

I also made a custom GPT as part of the course; this friendly GPT used the synthetic datasets I made to signpost users to past literature searches, journals, library resources and services, and even fictitious people with similar interests.

I reckon those with an interest in collections management would find this course particularly useful, too.

And naturally, having a PGCert in such a sought-after topic is bound to look great on any CV!

Since completing the course, I’ve built upon the skills I gained. Recently I had the honour of working with some amazing colleagues in NHS England, concerning generative AI detection in education.

“Do you have any tips?”

I’ve had some time to reflect on things, and this is what I’d advise to anyone thinking about applying:

• Being enthusiastic about data or learning about stuff generally can go a long way!
• Be prepared for a learning curve
• There is a time investment involved
• Chat with people who have knowledge or experience of the course

Ultimately, if you can put in the time, I’d recommend this course. That being said, there’s certainly some things I would do differently. Hindsight is 20/20!

What I’d do differently

I’ll be frank; my attendance could have been better! Luckily, I have a supportive boss who had no issue with me attending the weekly meetings, but I opted to watch the recordings in my own time instead. Sometimes I didn’t catch up and likely missed out on some interesting conversations.

I could have spoken with more healthcare professionals on the course. While I did have some lovely chats online and in person, I wish there’d been more opportunities to work more closely with colleagues (slightly related to my first point, ahem!).

I also worked in isolation a lot. Looking back, I could have reached out more to other KLS folk on the course, so we could chat about what we were working on. Piles of work in the day job, studying at weekends, and geographical distances blinkered me a bit, and I became very focused on my own workload. I got absorbed in my own data adventures, and now wish I’d spent more time learning about other people’s adventures too.

So why does this matter?

Knowledge and library professionals have always adapted to the needs of their users. From making books findable with card catalogues, to setting up computer terminals, to offering free WiFi and digital skills training, we’ve taken advantage of new technologies. This is just another opportunity for us to add a few extra tools to our toolboxes.

Knowledge for Healthcare notes that ‘We expect the emergence of new roles and responsibilities for knowledge and library service staff working alongside clinical teams and health informaticians.’

And the Long Term Workforce Plan mentions that  ‘ensuring staff have the right skills to take advantage of new technology that frees up clinicians’ time to care, increases flexibility in deployment, and provides the care patients need more effectively and efficiently’ is one of the priorities for the current and future workforce.

And of course, the upcoming 10 Year Plan has ‘analogue to digital’ as one of the three key strategic shifts planned for the NHS.

If you’ve read this far thank you! For those applying to the course; I wish you all the very best, and I hope you have as much fun as I did.

With new and emerging technologies, the future is bright for knowledge and library professionals. We just have to work for it.