Many patients with CKD have a slowly progressive decline in function. Often these patients get admitted with a higher creatinine than their previous documented creatinine. We then often quickly label them as having AKI. But we should also consider the possibility that the patient’s new creatinine represents continued deterioration of their CKD.
While not perfect (few if any formulas used for renal disease or fluid and electrolyte disorders are), we can graph 1/creatinine versus time to get a reasonable estimate of expected progression. Here is the idea. For a majority of patients the endless progression to end stage progresses in a straight line of 1/creatinine versus time. One can find this concept used in the trials that showed that ACE-I or ARB slowed the progression of diabetic nephropathy. The investigators calculated the curves slope before and after intervention. A slower slope gave evidence that the drugs delayed the inevitable progression to end stage.
If we have access to records (as we do at the VA system), we can draw a line representing the previous serum creatinines. Extrapolating that line allows us to estimate what the patient’s creatinine would most likely be on admission. A significantly higher creatinine suggests AKI on CKD.
As an example, let’s assume a patient has diabetes mellitus and proteinuria. We have creatinine measurements every 6 months:
| Time in months | Creatinine | 1/Creatinine |
| 0 | 1.2 | 0.83 |
| 6 | 1.3 | 0.77 |
| 12 | 1.4 | 0.71 |
| 18 | 1.6 | 0.63 |
| 24 | 1.8 | 0.56 |
| Expected 30 | 2 | 0.5 |
| Actual 30 | 2.5 | 0.4 |
Using this progression you can graph a reasonably straight line with 1/creatinine vs. time. With the patient presenting at 30 months and most recent creatinine measurement at 24 months, we would expect an increase from 1.8 to 2. But if the patient actually presents at 2.5 we should strongly consider the possibility of acute kidney injury.
I would not state that the patient as a baseline creatinine of 1.8 since the creatinine has steadily increased for the past 2 years. To be complete I would point out that the most creatinine was 1.8 6 months ago, with an expected increase to approximately 2. The increase to 2.5 is thus unexpected and should trigger a careful evaluation (r/o obstruction, check volume status, look for events that might have triggered kidney injury such as medications).
I hope this clarifies my concern about the term baseline creatinine. If the creatinine stays steady for a period of time, then we could use the term, but often the creatinine inevitably increases.
