Which factors are associated with good screening and recognition of chronic kidney disease (CKD) in the Veterans Affairs (VA) health care system?
Our providers, I do not know, they’re doing a good job, at least in the chart, they’re good. About 90% of the patients with hypertension, diabetes, more than 90%, they have serum creatinine in the chart. But the detection or ordering for urine protein and albumin is low. It’s specifically low if you have only hypertension. If you have diabetes, or both hypertension and diabetes, then screening for albuminuria is decent. But if you have only hypertension, then more than 65% of patients do not get albumin detected. That rate is about 20%, if you have hypertension and diabetes or diabetes alone. The screening for proteinuria in general is low.
Then recognization, in our study we identified that only 43% of the patients who had lab evidence of CKD were identified as having CKD. So 57% of the patients who had CKD, they were not identified as having CKD. Again you look at, because CKD can be identified because of the low estimated glomerular filtration rate (eGFR) or high urine albumin/creatinine ratio. Most of the time the unidentified CKD was when they had higher albuminuria ratio. The primary care [doctors were] kind of not thinking about albuminuria as CKD. That gave us idea that we need to do more education with the primary care regarding the role of albuminuria. Albuminuria in itself is an independent prognostic factor, as well as mediator of worse outcome and we have clearly seen patients who have higher albuminuria they have higher chances of uncontrolled hypertension too.
In terms of factors, you have univariate analyses as well as multivariate. In the univariate analyses, we saw the patients with race other than white, so the minority races, they had lower chances of screening. Then we saw that patients who have comorbidities, they have heart disease, stroke, tumor, diabetes, hypertension, or both, then they have higher chances of getting screened. Then patients who have a higher number of primary care visits per year, or they have a higher number of specialty care visits, and more elective procedures, either radiological or endovascular or cardiac, they had higher chances of getting screening tests in the chart. Then patients are on angiotensin-converting enzyme (ACE) inhibitors, diuretics, non-steroidal anti-inflammatory drugs (NSAIDs), they have higher chances of getting a screen test. These are basically the univariate analyses. But when you do the multivariate where you are adjusting for all these other factors for each other, comorbidities still stay positive. All these factors kind of still stood independent of each other and they were associated with higher screening rates.
Same thing on the recognization side also. One thing which was striking was that we saw that the minority races had less probability of getting screen testing, but they had higher probability of getting recognized for CKD. If they had CKD they’re a lot higher chances of them having CKD recognized. It’s kind of difficult to analyze from this work what could be the cause. But our hypothesis, or we postulate the reasons are 2: that other providers are familiar that minority races are more prone for CKD development, or these patients have advanced stages of CKD. Because in our analysis, we found that people with more advanced CKD, more stages 4 and 5 and higher albuminuria, they had higher chances of recognization. There’s a possibility that these minority races had higher advanced stages CKD or albuminuria. They were recognized higher although, they had lower chances of getting screened for CKD.
Another thing which will be found for the screening and recognization, is that [if you have a] high number of primary care visits or high number of specialty care visits or high number of elective procedures, you have a higher probability of having screen tests done. So you think maybe, ‘Oh, primary care are ordering these tests because there’s something wrong.’ However, that does not translate into higher recognization. They have more eGFR and serum creatinine and urine albuminuria in their chart, but they are not identified as having CKD despite having the lab evidence of CKD. Clearly, these tests were ordered, just for those procedures, or because patients have comorbidities, and they’re going to specialty clinics or they’re needing the procedure, and that’s why they are getting them. But nobody’s looking at them, or even if they’re looking at it, they’re not identifying that patient has CKD. That was kind of a striking feature. Again, when patients had both diabetes and hypertension, they were more likely to have screening as well as recognization. That makes sense, when you have both diabetes and hypertension, you kind of really think about it.
In terms of medications, patients who were on ACE and angiotensin II receptor blockers (ARB) and who were on diuretics, they were more prone for screening, as well as recognization even in our regression analysis. So that tells that providers who know that patients who need diuretics or patients who need ACE and ARB, they’re aware of the CLD existence. However, one thing, which was not very intuitive was the NSAIDs. We saw that patients were on NSAIDs were less likely to be recognized for CKD, which was kind of a little counterintuitive. NSAIDs is one of the risk factors for kidney injury, right? Then patients who were on NSAIDs, they were less likely to be screened or less likely to be recognized. We were just wondering about the cause. It could be that most of the time, what kind of patients need NSAIDs? Those who have pain. The providers were spending more time talking about their pain, rather than looking at their other comorbidities and addressing those comorbidities. Their providers’ time gets consumed by management of the pain, and they have less time or they do not have time to spend on other chronic care. Because CKD screening and CLD recognition is part of your routine care. Those kind of get jeopardized. That was kind of counter intuitive we found. These were the factors. But we clearly saw the patients who were recognized or who were screened, they had lower rates of NSAIDs. That’s a good thing. Basically, if they get recognized, they are prescribed NSAIDs less.
The gist of our work, I would say that, yes, screening for CKD, especially the proteinuria is low. Once they have CKD, the identification or recognization of CKD is on the lower side than what we expect patients to have. The factors which we came across, many times, clearly these tests were ordered not for the screening purposes, but for other purposes, because a high number of screenings is not translating into a high number of recognization. Because albuminuria is one of the big reasons, more education of the primary care about the importance of albuminuria for diagnosis as well as prognosis is warranted. We have to have more work at the primary care level. Once we identify CKD, then we can do more things to take care of the process of care and improve the prognosis of CKD patients.
There’s a lot of data coming out, that show if you take care of these few things early on in CKD, like aggressive control of blood pressure, aggressive control of blood sugars, and the statins, and ACE and ARB, then you’re going to improve their CKD progression rate and reduce the risk of end stage renal disease. But for that to happen, you have to implement all these interventions earlier. And for that you need to recognize CKD first, right? That’s what this whole thing is about, that we need to do a better job of screening and recognizing the CKD for these interventions to implement and improve the prognosis of our patients.