No clamps. Only manual compression.

No clamps. Only manual compression.
Exception : if patient has a high flow access, is unable to hold their own site (hemiparesis, etc) and I am confident that the nurse will check for ongoing flow after clamps placed then I will allow. Especially if patient is on anticoagulant and bleeding time is longer.

Several years ago our dialysis provider insisted on a “no clamps” policy. I argued that there is no evidence that clamps cause access thrombosis and was told that it was “intuitively obvious” and did not need to be studied. I agreed to impose the policy with the expectation that we should see an immediate drop in the number of thrombectomy cases I did. After one year, we saw no such drop so the policy was rescinded. While not the most scientific study, it was more scientific than simple speculation. Furthermore, I could reasonably argue that a patient holding his own sticks might use more pressure than a clamp. In any event, we use clamps freely.

It is very difficult to totally block a strongly flow AV access. When we try to do retrogrades, for any reason (which is a separate issue), it takes a huge amount of direct pressure to occlude an access.
The pressure needed to stop bleeding from a cannulation is less than the pressure required to block the AV access completely.
So one should apply enough pressure to stop bleeding keep the AC access flowing by whatever method or means.
where should the caution be? In grafts that have frequent clotting, I would be weary of prolonged aggressive pressure on the cannulation sites by whatever means.
Listening to the access by stethoscope when clamps are placed or the tough tight dressing is applied would be advised in patients who have frequent AVG thrombosis.

It is very difficult to totally block a strongly flow AV access. When we try to do retrogrades, for any reason (which is a separate issue), it takes a huge amount of direct pressure to occlude an access.
The pressure needed to stop bleeding from a cannulation is less than the pressure required to block the AV access completely.
So one should apply enough pressure to stop bleeding keep the AC access flowing by whatever method or means.
where should the caution be? In grafts that have frequent clotting, I would be weary of prolonged aggressive pressure on the cannulation sites by whatever means.
Listening to the access by stethoscope when clamps are placed or the tough tight dressing is applied would be advised in patients who have frequent AVG thrombosis.

I had to edit my previous note due to typos!!
It is very difficult to totally block a strongly flowing AV access. When we try to do angiographic retrogrades, for any reason (which is a separate issue), it takes a huge amount of direct pressure to occlude an access.
The pressure needed to stop bleeding from a dialysis needle cannulation site is much less than the pressure required to block the AV access completely. However a deranged access with outflow stenosis or soft aneurysm would be problematic and would act differently (but that would be a deranged access that needs fixing).
So one should apply enough pressure (proper pressure) to stop bleeding but not more pressure than necessary, by whatever method or means. More pressure than necessary by whatever means is not good.
Where should the caution be exercised mostly? In grafts that have frequent clotting, I would be weary of prolonged aggressive pressure on the cannulation sites of these graft by whatever means.
Listening to the access by stethoscope when clamps are placed or the tough tight dressing/tape is applied would be advised in patients who have frequent AVG thrombosis. If a short bruit or a high pitched bruit is heard, then they are too tight.

I look at bleeding p dialysis >5 minutes as “prolonged”. Clamps are often used when this happens. First intervention is to adjust the he heparin dose on dialysis. Prolonged bleeding after that should trigger a referral for an angiogram using the indicator venous hypertension and if there is no downstream stenosis and the heparin dose has been adjusted appropriately there should be no problems with prolonged bleeding and no reason to use clamps.

Like others, my advice is also anecdotal. My colleague and I have a saying that we use repetitively at the Access Center: "There isn't any bleeding that I can't stop with one finger" Unfortunately, most tend to over-react when it comes to bleeding from a graft or fistula. Pushing too hard just makes the bleeding worse. Putting a wad of gauze between the finger and the hole only serves to wick up the blood and prevents a good seal/clotting. When clamps are used, both of these issues occur, as the pressure is too great and there is a bunch of gauze under the clamp.

We tell our clinical staff , if your finger is making any type of impression on the skin/access, then you are pressing to hard. We don't like to suture after procedures, because of the risk for infection/ulceration, so light pressure and PATIENCE, are vital. Probably, we've all finished up a thrombectomy procedure, felt a good thrill, but then re-examined while the staff was holding the pressure over the sheath sites only to note pulsatility but no thrill. They relax a bit, and the thrill returns. It's amazing how little pressure it takes to interrupt flow. Certainly, a clamp can easily thrombus our best access: the soft, collapsible radiocephalic fistula.

So, we allow clamps only for emergency situations at the dialysis unit, such as oozing that persists despite 30 mins, and the patient is being transported to the Access Center or Hospital for intervention/stitch placement. Prolonged bleeding (>5 mins) is usually, caused by outflow stricture/thrombosis, coagulopathy, or thin skin/subcutaneous tissue. Therefore, anyone requiring a clamp is immediately referred for a venogram. If the venogram is normal, then the other two issues are addressed.

Clamps IMO are a practical necessity in a busy commercial dialysis unit. Staffing ratios, turnover time, are such that it would be impossible for a tech or RN to manually hold every site that a patient couldn’t hold for themself. There are a subset of accesses for which clamps are probably detrimental (unproven as Max points out). I suspect that a low-flow, low-pressure fistula would be prone to thrombosis with too much or too prolonged pressure, both of which are very possible with a clamp. Admittedly, it is hard to assure a skillful assessment of this in the dialysis unit, so it is tempting to impose a blaket “no clamps” policy. On the other hand, has anyone ever tried to manually induce thrombosis an access when you wanted it to clot? Pretty hard to do. So we use some clamps. Prohibited on new fistulas until cleared for use by NP.

Individualize, especially in a busy clinic with staffing ratios that are not permissible to manual holds for everyone. I would avoid frequent clotters (AVG) as mentioned above or a not so large RC AVF; ultimately every patient is different.

on a number of occasions when I have visited dialysis clinics and seen patients with clamps and have checked and the patients had no pulse in their access

When I was in clinical practice and we were using a large percentage of grafts, I had the nurses check patient's access as there were leaving the dialysis unit after a successful dialysis and found a significant number were leaving with a clotted access - they did not use clamps this was from manual pressure

The end of dialysis is a particularly vulnerable time for access clotting, as blood pressure / fistula flow is often the lowest at this time. It's not unusual for sbp to drop 50 points or > with dialysis treatment and ultrafiltration.

Similar to Dr. Beathard, I have noted how the post-dialysis dressing alone can sometimes occlude flow. You all probably have had the experience of a patient coming to the access center post-dialysis. On occasion, there is no bruit/thrill, but once you remove the dressing and re-examine there is resumption of a good bruit, maybe even a thrill.

I agree with Dr. Saad that its surprising how much it sometimes takes to clot off an access, especially fistulae. I sometimes electively thrombose an access for steal syndrome or ultrafiltration seroma. I leave a balloon inflated in the access for 6-8 hours, while monitoring. There is no flow during this time. A few, 5-10% or so, which demonstrate return of for as soon as the balloon is deflated and removed. These are all grafts, and I anticipate the percentage of non-thrombosis would be markedly higher in fistulae.