Cardiology developed one the most fantastic medical technologies of the late twentieth century, the percutaneous coronary intervention. The problem that PCI tackled was obvious, patients with coronary artery disease had demonstrable arterial narrowing and we had a technique that could treat this narrowing. How could we not treat it? How could patients not benefit from this? And following these interventions patients had remarkable improvements in angina, the primary symptom of coronary disease. In addition to improving chest pain, this intervention had to prevent heart attacks and save lives. This assumption was accepted fact at the beginning of my career, however some cardiologists were unsatisfied with using intuition to guide therapy and PCI came under the sharp blade of the randomized controlled trial. COURAGE shot down the idea that providing cardiac (bare metal) stents in patients with stable coronary disease provided any survival benefit. This was repeated with ISCHEMIA (now with drug eluting stents). And then ORBITA used sham procedures to question whether PCI even reduced angina, a finding that was at least partly reversed by ORIBITA-2 which removed the use of anti-anginal medications.
This post is not intended to provide a comprehensive review of the use of PCI in coronary disease but to use it as a demonstration how an intuitive therapy that seems to have obvious and unquestioned use, can be questioned and through those qquestions we can reposition the procedure to use it where it is helpful and not waste resources and treat patients with science and not vibes.
This takes us to dialysis. Dialysis unquestionably helps some patients but I don’t think it helps all patients. Nephrology has rarely subjected dialysis to the rigors of a randomized controlled trial and this is to our patients detriment. We have little evidence to guide us as to when to offer dialysis and when not to. One area that has been extensively explored with RCTs is when to initiate dialysis in AKI, and honestly, it wasn’t pretty for dialysis.
Last week Dr. Manjula Tamura published a Target Trial Emulation of an RCT to see how survival differs between elderly patients (age over 65) with a GFR of 12 and immediately starting dialysis (average was 8 days after trial inclusion) versus delaying dialysis at least 30 days (average was 3 years after trial inclusion). The author tracked two outcomes:
- Survival
- Cumulative time at home
The results showed that starting dialysis at 12 ml/min resulted in survival for 770 days. For the cohort who delayed dialysis survival was 761 days. The difference was non-significant.
This image is powerful
The trial was covered in the New York Times!
I am confused by Target Trial Emulation. I am suspicious that people started on dialysis at a GFR of 12 and people that go for 3 years before starting dialysis (in the delayed group) are really interchangeable. I am confused how the late start group can start dialysis an average of 3 years after inclusion in the study while only surviving an average of 761 days (2.1 years) But it is clear to me that nephrology has failed at doing studies to determine who benefits from dialysis. And I believe that old frail people do not get the benefit from dialysis that we dream of. This was shown in an earlier study by Dr. Manjula Tamura which showed dialysis to be a blood bath for nursing home patients. (Tamura, NEJM 2009). I have used this study to guide me in the advice I give patients. It is just an observational study but it showed the folly in believing that the frail patient in the nursing home will turn around as soon you clear the uremia. That’s just a fairy tale we tell ourselves. It almost never happens and the reality is that almost all of these patients are either dead or further debilitated a year later.
As Dr Tamura says in her Tweetorial about the study, it is time for randomized trials to see where dialysis helps and where it falls down. We need to have the COURAGE to test whether what is intuitively helpful actually delivers benefits. The cardiologists have used clinical trials to define the role of PCI in coronary disease, we should do the same for our patients.