A Dedicated Metrics Window for Longitudinal Patient Data
H
Henning Van Den Heever
We run an ophthalmology practice and need to capture, track and compare a large number of clinical data points for each patient across multiple consultations. These include visual acuity (VA), intraocular pressure (IOP), central retinal thickness (CRT), and many other clinical observations. The number of metrics that may need to be monitored can be quite extensive.
Until now, we have entered this information either in the Context section or via the Speak function, relying on a customised template to retrieve and organise data from both the current and previous consultations into tables. Unfortunately, this approach has proved unpredictable and unreliable, particularly when it comes to accurately extracting and comparing historical values.
I believe this is a common need across many medical disciplines. For example, physicians may want to monitor blood pressure, HbA1c, BMI, cholesterol and other clinical parameters over time.
My suggestion is to introduce a dedicated window alongside the existing "Context" window, perhaps called "Metrics".
This would function as a structured, customisable repository for patient-specific clinical measurements. Each practice could configure its own template according to its specialty and requirements.
The basic concept would be a longitudinal reference table for each patient, structured as follows:
- First column: The name of the metric (e.g. VA, IOP, HbA1c).
- Subsequent columns: Dates of consultations or measurements, with the corresponding value recorded under each date.
This would allow the clinician to quickly review a patient's historical measurements, compare values across visits and identify trends at a glance.
Ideally, the system could also automatically generate trend graphs from the recorded data. Furthermore, customised consultation templates could be instructed to reference selected metrics, or combinations of metrics, and incorporate the relevant tables and/or trend graphs into clinical notes and reports.
The benefit would be substantial: clinicians could reliably monitor how a patient's blood pressure, HbA1c, visual acuity, retinal thickness or other parameters change over time, without having to depend on AI to extract and interpret historical information from unstructured consultation notes.
An important consideration is data accuracy. Information entered into the Metrics window would not necessarily need to be captured through speech recognition. In fact, I would prefer the option of manually entering values. Given that these are clinical measurements, absolute data integrity and reliability are essential. A simple manual-entry interface would be perfectly adequate if it ensured that the recorded values were accurate and consistently available for future consultations.
I believe this would be a valuable addition, not only for ophthalmology practices but for virtually any medical discipline that requires longitudinal monitoring of patient-specific clinical data.
Anya Sharma
Thank you for taking the time to share such a thoughtful and detailed suggestion.
It sounds like the underlying need is to maintain a reliable, structured record of patient measurements over time, which can then be easily reviewed, compared and incorporated into future clinical documentation.
I have a couple of questions to help us better understand your ideal workflow:
- Where do these measurements currently live? Are they already stored in your practice's EHR, or would Heidi need to become the place where you maintain them?
- Would you primarily want to manually enter and update these values, or would it be valuable for Heidi to suggest measurements extracted from consultations for you to review and confirm?
- Of the capabilities you've outlined, which would make the biggest immediate difference: reliable access to historical values, automatic comparisons and trend visualisation, or incorporating these measurements into generated notes?
H
Henning Van Den Heever
- We put the in the context window. For example "IOP R = 10mmHg". We don't use an EMR. Heidi could maintain the labelled data points (BP, HR, IOP, VA etc) in a simple table as xplained in my first post, and add and retreive as needed.
- It is actually better to enter it manually by default as most of the data points are awkward and impractical to speak especially in front of the patients. The assistant adds it. Yes, Heidi could also suggest measurements to be added BUT it must be double checked then. For an ophthalmologist the difference between 550microns and 515microns is NOT insignificant. A value picked up by Heidi can be highlighted in yellow and when clicked on it (to confirm validity) turn into a green highlight and then after few seconds no highlight (accepted).
- Reliable access to historical values which much more valuable if it can be visualized easily. Trend graphs are great for this. Requires minimal reading. Gives instantaneous info.
- There should be a prompt that you can add to your templates to list the trend in a table and/or show it as a graph. Let's say the doctor chose to follow BP and BMI trends, the prompt can be [table:BP (for past 12 months)] or [line graph:BMI for past 6 years)]