Nephrology lab flowsheet EHR: trending creatinine and GFR without chart diving
Kidney function is a slope. Most charts store points. The gap between those two facts is where late diagnosis lives, and closing it is a template question rather than a clinical one.
Kidney disease is defined by abnormalities present for at least three months, so a single result cannot establish it. The current quality measure requires both an eGFR and a urine albumin-to-creatinine ratio. A flowsheet that trends both, together, is what turns scattered results into a finding.
Nearly nine in ten American adults with chronic kidney disease do not know they have it.
That figure comes from CDC’s own surveillance data, updated in March 2026, alongside an estimate that around 14 percent of adults have CKD at all. Put those together and the picture is specific: this is a condition that is common, largely silent, and sitting in laboratory values that practices already have.
It is not hiding in a specialist’s office. It is hiding in a results inbox, distributed across three or four ordinary visits a year, where each individual value looks unremarkable and the pattern across them is never assembled.
That is a documentation architecture problem, and it has a cheap fix.
CKD staging in the EHR starts with a rule about time
The definition does the work here. KDIGO’s 2024 guideline defines chronic kidney disease as abnormalities of kidney structure or function, present for a minimum of three months, with implications for health.
Read that as a data requirement rather than a clinical one and the consequence is immediate: a single abnormal result cannot establish the diagnosis. Persistence is part of the definition, which means the record has to be able to demonstrate persistence. A chart that stores each result inside the encounter it arrived in can show you a value. It cannot show you three months.
CDC makes the same point against itself, which is unusual and worth borrowing. Its prevalence estimate rests on a single measure of urine albumin-to-creatinine ratio and serum creatinine, and CDC notes the figure might be overestimated for exactly that reason, because one measurement does not satisfy the three-month persistence requirement. When a federal surveillance programme flags the limits of single-point data in its own methodology, a practice trending the same values in a chart should take the hint.
- Three months of persistence is part of the diagnosis, not a clinical nicety.
- A single abnormal value establishes nothing on its own.
- The record has to be able to show a slope, which means values stored as a series rather than as encounter attachments.
- Two different markers, not one, because function and damage are separate questions.
Race free eGFR is settled, and your calculation may not be
KDIGO’s 2024 guideline moved to the 2021 CKD-EPI creatinine equation, which no longer includes race as a component, and states as a practice point that use of race in the computation of eGFR should be avoided.
That is not framed as an option. A practice whose laboratory interface still returns two values, one with a race coefficient and one without, or whose flowsheet was configured before 2021, is trending a number that current guidance says should not be produced that way. This is worth an explicit check with the laboratory rather than an assumption, because the practice does not usually control which equation the result arrives calculated under.
There is a second point most articles skip. NIDDK notes that using both creatinine and cystatin C to estimate GFR is preferred and more accurate than using serum creatinine alone, and quantifies the residual bias in the creatinine-only equation at roughly four percent in either direction depending on the patient. That is not an argument to stop using creatinine. It is an argument for the flowsheet to record which equation produced each value, because a trend assembled from results calculated different ways is not a trend.
The kidney health evaluation measure requires both markers
The quality side has caught up with the clinical side, and this is where a lot of practices are still operating on an older understanding.
The current kidney health evaluation electronic clinical quality measure covers patients aged 18 to 85 with diabetes who received both an estimated glomerular filtration rate and a urine albumin-to-creatinine ratio within the measurement period. Both. Within the period.
The predecessor measure was looser. It could be satisfied by a nephrology visit or an ACE inhibitor or ARB prescription, neither of which involves testing anything. A practice whose care pathway was built against that older logic is not merely reporting differently; it is doing less testing, and the current measure will say so.
Albuminuria testing is the half that goes missing
KDIGO’s first recommendation on this is unambiguous: test people at risk for and with chronic kidney disease using both urine albumin measurement and assessment of GFR. In practice the creatinine happens, because it rides along on a metabolic panel ordered for other reasons, and the urine albumin-to-creatinine ratio does not, because somebody has to order it deliberately.
That asymmetry is the single most common gap we find in nephrology and primary care flowsheets alike. It is not a knowledge gap. It is an ordering-workflow gap, and it responds to the same fix as any other: a standing order set, a flowsheet row that visibly shows the missing value rather than simply omitting it, and a prompt tied to the diagnosis rather than to a visit type.
Who to prompt for is a question CDC has already answered with numbers.
Nearly half of adults with type 1 diabetes and two in five with type 2. Those are not screening populations that need to be discovered; they are already on the problem list, already coming in, already having blood drawn. The prompt is the only missing part.
Prediabetes at 11 percent is the row worth a second look, because it is the population least likely to have a standing nephrology-adjacent order set, and one in nine is not a small number.
Lab result trending is a template build, not a product purchase
What a working nephrology lab flowsheet EHR build actually contains is unglamorous: a defined row set, a consistent unit, the source equation recorded alongside each eGFR, and both markers visible in one view with their dates.
The thing that makes it useful rather than decorative is that missing values appear as gaps rather than disappearing. A flowsheet that only shows what was collected tells you nothing about what was not, and what was not collected is the entire point. That is a display decision made once, at build time, and it is the difference between a flowsheet somebody uses and one somebody closes.
None of this needs a nephrology module. It needs a row definition and an order set, which is the same class of work as any other template development engagement. Practices running nephrology EHR and IT at scale tend to get there eventually; the ones that get there early do it because somebody treated it as a build rather than waiting for a vendor to ship it.
Can you see the slope, or just the last value?
We will look at how your chart stores kidney function results and show you what it would take to trend them properly.
Where to start
Pull your diabetic patient list and count how many have both an eGFR and a urine albumin-to-creatinine ratio in the last twelve months. That number is your baseline, it takes an afternoon, and it will almost certainly be lower than anyone expects.
Then check which equation your laboratory is returning, and whether the flowsheet records it. Those two facts, together, tell you whether you are looking at a trend or at a collection of numbers that happen to be adjacent. A wider EHR solutions review usually finds the same gap in cardiology and endocrinology for the same reason.
Kidney lab flowsheets: common questions
Why is one abnormal creatinine not enough to act on?
Do we need to change anything if our lab still reports a race-adjusted eGFR?
What does the current quality measure actually require?
Why does the urine albumin test get missed so often?
Is there a HIPAA issue with pulling a patient list to check for missing labs?
Let’s meet
Qventive supports nephrology and primary care practices across New Jersey. Thirty years healthcare-exclusive, from our office in Hackensack.
- Estimated CKD prevalence overall and by risk condition, and the share of adults with CKD who are unaware of it. cdc.gov
- GFR estimating equations, the 2021 race-free creatinine equation and its residual bias. niddk.nih.gov
- Kidney health evaluation measure requiring both an eGFR and a urine albumin-to-creatinine ratio. ecqi.healthit.gov
- KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. kdigo.org