Sitting in the bathtub you notice that there is a slow leak around the plug. You adjust the taps to maintain a flow of water that exactly counteracts the loss due to the leak; the water level stays constant. This is called a steady state and the same thing happens with out kidneys and the molecule used to assess their function. Our bodies generate creatinine at a constant rate which finds its way into the blood. Under normal circumstances our kidneys excrete that creatinine into the urine at the same constant
rate. The creatinine concentration in the blood, therefore, stays constant. When our kidneys get injured (as they very often do in hospitalised patients) this is like plugging the leak. Just as the water level in the bathtub would rise slowly – undetectable at first – so too does the creatinine concentration rise slowly. It normally takes a couple of days to be noticed. Most of my work has been about trying to detect this injury to the kidney early. However, if the kidneys start to recover then excess creatinine is only slowly cleared from the blood by the kidney – a process that similarly can take a day or two before it is detected. Just as not knowing if the kidneys have been harmed makes treatment and drug dosing difficult for the nephrologists and intensivists, so too is not knowing if they have recovered. My latest publication (aka a cheesecake file) that has appeared in press presents a simple tool for the physicians to try and determine if kidney function has recovered after having been compromised.
This particular piece of work began when a St Louis Nephrologists (a kidney doc), Dr John Mellas, contacted me to say that although a manuscript of his had been rejected by reviewers, he thought there was merit and could I help him (he found me through a search of the literature). I confessed to being one of the reviewers who had rejected the manuscript! Fortunately, John was forgiving. His problem was that he was called in to the intensive care unit to look at a patient with high blood creatinine concentration. Should he put the patient on dialysis or should he wait? If he knew if the kidney was already recovering, then he would be less likely to put on dialysis. We talked about the issue for a while and eventually settled on a possible tool which we could test by looking at the behaviour of creatinine over time in abut 500 patients in the ICU. The tool is quite simple. It is the ratio of the creatinine that is excreted to the creatinine that is generated. If more creatinine is being generated than excreted then probably the kidney function is still below normal, however, if more is excreted than generated then probably the kidney is recovering. The difficulty is that there is no way to measure in an individual what the creatinine generation is. We ended up using equations based on age, sex, and weight to estimate creatinine generation. This is a bit like using an equation which takes into account pipe diameter, mains water pressure, and how many turns of the screw the tap has had to determine the rate of water flow. Creatinine excretion, though, can be easily measured by recording total urine production over several hours (we suggest 4h) and multiplying this by the concentration of creatinine in the urine.
We discovered that by using the ratio between estimated creatinine generation and creatinine excretion we were able to tell in most patients if the kidney was recovering or not. My hope is that physicians will test this out for themselves. The good thing is that it requires only minimal additional measurements (and costs) beyond what are already made in ICUs, yet may save many from expensive and invasive dialysis.