Pathology – Harmonisation & Cross-site working
We all know that diagnostic medicine (aka lab and other tests) form the basis of ~70% of clinical pathways/delivery across England and Wales. This is likely the case for most other countries too. There is a reason why the HIMMS Electronic Medical Record Adoption Model (EMRAM) at the lowest levels Stages 1 and 2 (we don’t care about stage 0) talks about 90% of all lab data (and other diagnostic data) being available and processes around them being set in place! So, if we can’t sort labs and diagnostic medicine properly we might as well stop talking about all the other cool semantic interoperability things… These things are all the more relevant now, in the changing world of lab medicine, not just because of an explosion of tests (in the post-genomic world and post-Covid) but also because of other changes.
For those of you who do not work in Lab Medicine (Pathology as it is called in the UK), there are current activities that could determine how healthcare records are used for the next decade or two. The focus of these activities is to enable cross-site working — that is deliberately vague because lab activities (aka performing tests) and sharing data will now need to work across multiple labs/sites. Whether this is the NHS England Pathology Network Consolidation or the SHIELD initiative in the US or other similar initiatives elsewhere, they all want to see a future where lab results across different sites are comparable from the data perspective and an efficiency/costing perspective.
For those of you who are surprised that sharing lab results across `sites` or `organisations` is even an issue, here is a bit of some history/background. Most hospitals (and labs) have traditionally operated as silos. So when your lab result is available, your doctor is able to access and understand what they mean. However, the moment they cross the hospital/lab’s boundaries there is often a loss of meaning, that can increase the patient safety risk. In the UK (and elsewhere), your results could be shared with other systems (e.g. primary care systems) they would go out with PBCL codes which are mostly gross approximations for what was done in a lab. For example, here is what would go out as a result, if the Sodium level in your blood was measured:
For most practical purposes of clinical decision making, this has traditionally been sufficient for the clinicians in primary and secondary outside of labs. However, inside the lab they care about what was actually measured — for example in this case substance/molar concentration. So really this result needs to say:
Notice the difference? It is a molar/substance concentration. Before you generalise and dismiss me, here is almost similar but not the same one… 😬
Note this one says Mass concentration. So why is this relevant?
What makes lab tests comparable or harmonised?
While the clinicians who do not deal with the lab world might not pay attention often enough, every lab result is often accompanied by units and some other bits. It is these other bits that actually determine its meaning and comparability.
So when a lab result’s true meaning is lost in transfer, it becomes less safely comparable/interpretable. This inherently increases the patient safety risks
Let’s take our old friend Serum sodium. Most clinicians worth their salt would remember that its normal ranges are 135–145 mEq/L. So given the chart below, most would immediately posit that something quite bad happened on 05/01/2022 for this patient — they went into hyponatremia (aka low sodium).
However, you’d be remiss of not spotting that they actually went into hypernatremia (aka high sodium) on 03/01/2022. Of course, you’d only spot this, if you were paying attention to the reference range field in the results when you were scanning them (see screenshot).
Question is — do we all do this when we are in a hurry? What if there are 50 rows of such results which is actually quite common in say chronic renal failure patients and we prefer to look at a graph. This is not a made up scenario at all
My Apple watch (or your favourite personal health app) will tell you to enter your lab results. It could produce a technically correct graph, but clinically inaccurate one.
Imagine this from the patient perspective — should they think that they are all good because the app (or graph) didn’t flag up anything? What about the alternative, when a test result actually looks abnormal? So in order to compare lab data, we want to be able to know a lot more than just a test name (aka Serum Sodium).
Comparing lab tests beyond test names
Btw, we have already covered in a previous articl why relying on just the name of a test is not sufficient. If you have not read that article, please read it first here. In fact, this is what I was referring to as the silos — where previously hospitals mostly dealt with results from their own lab or a known lab and they didn’t have to worry about this. But in the changing world of cross-site/regional working or data consolidation (and/or referral labs), we now have to realistically deal with a lot more than just names. Not handling these variations (which might be legitimate) could very easily create a clinical safety risk (or miss one). In fact, I am not the first one saying this — those of you in the UK Pathology world will note that it was Dr Rick Jones who championed the case for `harmonisation` of lab data and making them comparable!
But there are a lot more than you think that could affect comparability of tests. So in fact, to harmonise lab processes and data across sites, the following need to be taken into account
- the property of the thing being measured (molar concentration vs mass concentration)
- the technique that was used to measure the property (e.g. sedimentation, spectrographic analysis, etc)
- the kit/analyser that was used to perform the test (yes, not all test kits are the same)
- the specimen that was used to perform the analysis (e.g. serum vs plasma — since we can’t always assume they are interchangeable)
- … units follow on from here…
These are some of the things that make a lab test comparable and they are some of things that lab specialists care about. They often deal with very precise things and they like everything being very well defined… 😅 So these would be some of the things that determine the sensitivity of a test, not just its comparability and reference ranges. If you need proof, please go look at the number of test kits that were produced to detect SARS-CoV2 and all the things they were measuring (e.g. RNA, Antigen, Antibodies (IgG, IgM)) and all the various techniques they were using. But let us not digress…
Single LIMS Procurement — Changing Landscape
So what does this mean for the changing landscape of lab medicine and cross-site working? Cross-site working is not just being able to login to systems remotely and having calls over Teams/Zoom. As mentioned, various national initiatives like Pathology Network Consolidation, SHIELD etc would like lab tests (and processes) to be comparable. Even if they say interoperable or efficient, they can’t achieve those without making the testing process and the data they produce comparable. What seems to be happening quite a bit now is LIMS procurement — lots of it, with the hope that a Single LIMS will solve all these problems magically.
Without first getting the tests themselves harmonised, any shiny new LIMS has no chance of actually making things comparable and delivering promised benefits, leave alone solving all the other problems.
For a start, before any LIMS can be rolled out/shared across multiple sites, all previous silos need to understand:
- What makes their tests comparable — look at more than just test names
- Where their tests and processes are different (and legitimately so).
- What do we need to know, if we are dealing with lab processes or results from one site as opposed to another?
Without this, there might be multiple instances of a single LIMS product (often the case), but each of these is configured slightly differently with just the test names being comparable. So organisations would have gone through a whole lot of heartache for over a year or two, with the end-result that there are silos being recreated at every site under the name of a single LIMS.
The biggest challenge we see here is the ability of leadership to balance the future vision of safe cross-site LIMS from both process and data perspective, with the demands of the past/now.
At least for the UK, this could be a `once in a generation opportunity` to truly harmonise lab testing, but we hear arguments like but in our lab, we have always done it this way (in Excel) or we are too busy, so we’ll just focus on test names. But remember, harmonising lab test names alone is not sufficient to enable cross-site working and ensure patient safety! Yes, your past practice (from 2 decades ago) is great and you currently have pressing demands on your time, but Winston Churchill famously said:
If we open a quarrel between the past and present, then we shall find that we have lost the future!
So yes you might have always done things this way in your lab (the past) and yes you might be very busy now (the present), but is this how we always want things to be?
Principles of Harmonisation
While this might be entirely wrong, there is a reason why we choose the word harmonisation over standardisation. We think standardisation is a route to achieving harmonisation which itself is a process of evolving to clinical excellence. We tend to think of it as — going from the silos of labs to a state where cross-site lab processes and data actually work as expected. So here are some quick principles that stand out:
- Harmonisation aims to enable cross-site working and cross-site data interoperability in a network/region/hub/cluster.
- Harmonisation of Pathology data covers more than just codes and includes specimens, units, reference ranges, methods, analysers, etc.
- Harmonisation aims to ensure that legitimate variations in lab test processes and data are accurately recorded, so they can be addressed if legitimate clinical safety concerns can be addressed.
- Harmonisation is an on-going process of data capture → data analysis → process improvement → clinical excellence with the eventual goal of continuous clinical improvement while allowing phased improvements across sites.
What is not Harmonisation
- If you aim to only create a single test name for your new LIMS, then you are only addressing a single aspect of harmonisation. You risk patient safety involving units, reference ranges, and test comparability.
- If you only aim to maintain an Excel spreadsheet of all your test names like you always have, then you risk the loss of meaning of what a test is across sites. You have already lost the future where new tests are introduced (with different units, reference ranges, methods, test kits/analysers) across different sites. Instead, you want to move to a single readily accessible web reference of tests (processes, units, reference ranges, etc) across all your sites.
- If you aim to handle/explain on paper somehow why two test results from two labs with the same test name are the same (or not), then you risk every downstream system having access to that `paper`. You have already lost the future where clinical safety is a concern for regional/network-wide consolidation/aggregation of data. Instead, you want the ability to use a reference master list to track legitimate differences in tests across sites/labs.
- If you aim to allow new tests to be created/added to your Excel spreadsheet of tests by different sites, without overarching governance, then you have not removed the risk of `tests with the same name` meaning slightly different things. In a few years, you will likely succeed in recreating the current complicated mess of local codes per site. Instead, you want to establish a clinical governance process across sites (per speciality) that continuously reviews and signs off changes backed by a tool that audits/tracks these decisions transparently.
- If you aim to only standardise the test names that are sent out to downstream systems (aka primary care, etc), then you lose the ability to improve clinically when downstream systems can inform/influence lab practice. You have already lost the future where diagnostic medicine (and variability of test sensitivity/results) is taken into account by clinical colleagues in patient care. Instead, you want all clinical colleagues (even outside labs) to understand what your lab test means (whether it is mass concentration or something else, what units, etc).
But, all this sounds very complicated… 🙈😱 Yes, no one said trying to break down silos to truly share lab results was easy. But remember, without sorting lab tests that underpin ~70% of clinical encounters/pathways what we are doing other than just repeating history? However, enabling cross-site working and data sharing with or without single LIMS does not have to be a challenge.
Pathnexus is our flagship data harmonisation platform for lab medicine.
Pathnexus allows you not only to harmonise your tests, but also also add your own content like techniques, properties and more — all designed to ensure your lab tests and content are ready for cross-site working!
Whether you are thinking of starting down the single LIMS journey or have started, Pathnexus can enable the vision of cross-site working and take you to a harmonised single test list across multiple disciplines, beyond just test names!
Come talk to us about how our clients use Pathnexus to achieve cross-site working and single test lists beyond just names…
Want to discuss lab data harmonisation, single test lists or enabling cross-site working? We’re always open to chat
We at Termlex believe that all clinicians and patients would like to have the ability to access and compare diagnostic results from anywhere irrespective of their origin. If this is something you believe in and would like to chat, please get in touch with us.