Delayed Dialysis Disasters

Working in the Kimberley is a constant challenge.  The medicine is fascinating, the logistics can be difficult.  Fortunately, our patients tend to be quite robust.  There is one particular scenario that epitomises this dynamic perfectly – the “missed dialysis” patient.

What am I talking about?  There are a few things to you need to know to understand how we end up in this situation regularly.

Firstly, end-stage renal disease requiring chronic dialysis is very common in Northern Australia.  There is a huge demand for “HD chairs” to maintain a growing number of dialysis-dependent folk in the Kimberley.

Secondly, there are no “acute dialysis” options available in the Kimberley.  Most of our renal patients receive thrice-weekly HD in one of several units spread over the Kimberely.  These dialysis units are generally not based in hospitals, but in Aboriginal Health centres or in some places at the patients home.  As such we can only really offer dialysis to folk who are well, stable and needing pretty standard filtration goals.  The closest tertiary dialysis unit that can offer acute dialysis or HD for sick patients is 2000 km south in Perth.  About an 8 – 12 hour transfer by plane.

Lastly, there are the challenges that our patients face in going about life on dialysis.  Many Aboriginal people in the Kimberely travel long distances on a regular basis to seek healthcare, visit family or attend important cultural events such as land meetings or funerals.  As such they may be away from their HD-base for prolonged periods.

So that is the setup that lands us in the “missed dialysis” scenario.  I thought I would share my basic approach to this scenario as it can be quite scary and may quickly spiral out of control if not managed early and aggressively.  There is a lot of good medicine to do… but crucially to really get it right we need to communicate and manage the logistics carefully.  Let’s use a case to illustrate:

Aloysius is a 48-year-old man who has been receiving haemodialysis for the last 3 years.  He is from a remote community but has been living in Broome to receive dialysis.  He has a background of:

  • type-2 diabetes and hypertension. (the presumed causes of his renal failure)
  • hypertensive cardiomyopathy and has had at least one “non-STEMI” about 5 years ago that required a stent in his right coronary.
  • He has also had rheumatic heart disease as a child
    • subsequently got mitral valve disease,
    • his most recent ECHO shows mitral stenosis a dilated LA, hypertrophic LV with normal systolic function.
  • Gouty arthritis
  • He is a non-drinker and quit smoking when he commenced dialysis.

Medications:  aspirin 100, atorvastatin 40, ramipril 10, amlodipine 10, CaCO3, gliclazide 30, sitagliptin 25.

Aloysius’s uncle passed away 2 weeks ago and he travelled to his home community ( 10 hours drive away) to spend time with family and attend the funeral.  Initially, he has planned to use a “chair” in the local clinic about 200 km from the community.  Unfortunately, after he arrived the renal clinic nurse became unwell and had to leave.  Aloysius stayed for the funeral and then a few more days.  His car had some engine trouble and that extended his stay.  It is now day 7 since his last dialysis session.  He is becoming unwell.

The ED phone rings.  It is the remote area nurse (RAN) at Aloysius’s small community.  She has just met Aloysius and is very concerned.  He presented with dyspnoea.

His observations: RR 36/min,  BP 240/140,  HR 110/min,  Spo2 83% on air but came up to 91% with a Hudson mask.  He is afebrile.  He reports coughing up a lot of sputum which is pink.  He currently weighs 102 kg and tells the nurse his “dry weight” is 93 kg.  He has a mild bitemporal headache. He denies any chest pain.

The clinic has no way of checking any blood test other than Hb and BSL.  Aloysius states that he has not had his usual medication for the last few days as his blister pack was empty – he was hoping to be back in town by now!  The RFDS retrieval has been deployed, but it is going to be a few hours until they can get Aloysius back to Broome…  what do we do in the meantime?

There are a few ways to think about this situation.  The standard problem list for renal failure is  a fair place to start:

  1. Volume overload / APO
  2. Acidosis
  3. Hyperkalemia
  4. Uremia

Patients in Aloysius’s situation are also at risk for other lurking diagnoses that may be hidden in all the deranged physiology eg. acute coronary syndromes, sepsis, pneumonia, drug toxicity… all are not improbable.  However, these can be tough to diagnose in the immediate setting and should be considered again if he does not respond to interventions as we might expect.  Our priorities for now are:

  1. Get him to somewhere that he can receive dialysis safely
  2. Deal with the physiology that is about to overwhelm his ability to compensate.

The immediate threat is the APO and hypertensive crisis which is really very similar to “sympathetic, crashing, acute pulmonary oedema” we see in other heart failure patients… the exception being that we have no way to pee out any excess volume.  The principles remain the same – control the preload and afterload.  In a remote setting there are not a lot of options available.  But here are a few:

  • Give the patient’s usual antihypertensives.  This step is often forgotten. It may not work for some time but it is useful to get the oral meds on board early.  Luckily missing dialysis usually goes along with not having taken meds – so it is a cheap play to give them ASAP and can be done almost anywhere.
  • GTN is your friend.  GTN comes in many forms and can be dosed aggressively in this situation.  For me this is a “Stand by the bed” therapy.  I am giving GTN and watching closely, not waiting for more than 5 minutes to adjust the dose.  If IV GTN is available that is ideal. However, in smaller places, we usually have sublingual and topical preparations.
    • My starting dose if using IV (50mg/100ml = 500 micrograms/ml) is around 60 ml/hr ( = 500mcg/min) for a few minutes then slow it down with regular BP checks or an arterial line.  Target a BP of 140-150 systolic. I find most patients end up requiring an infusion of ~ 20 ml/hr which is 166 mcg/minute.  If you start low and titrate slow you end up taking hours to get control of the haemodynamics and risk spiralling fast into a hypoxic arrest.
    • Now, this may seem like a lot, but it isn’t really.  We have sublingual GTN tablets that are 600 mcg that get well absorbed over a few minutes.  So if you don’t have IV capability then giving SL GTN is pretty good as a start.  It can be easily repeated and spat out if you overshoot.
    • I do see the application of topical GTN patches in this situation.  This has a place once the dust has settled but rarely works in the true crashing SCAPE patient. The 15 mg/24 patch, in theory, would give you at best about 10 micrograms/minute.  So you would need to place quite a few to get to the doses required for acute BP / afterload control.
  • OXYGEN is always needed.  Our dialysis patients have bad vessels and Aloysius could collapse at any moment – so just give oxygen up front.  Oxygen requirements will hopefully decrease as our other interventions kick in.  Particularly the GTN and PEEP.  Titrating down the FiO2 is a good way to monitor if we are winning.  See logistics section below – ongoing oxygen requirements will be one of the key decision points for where the patient needs to go to next.  In reality we often get these patients down to a few litres by nasal prongs or nothing in order to facilitate movement to a chronic dialysis chair.
  • PEEP is even better than oxygen.  Using non-invasive ventilation does a lot of good things in this situation. Unfortunately, it is not often available in small sites and remote communities.  Most transport ventilators can be used to give CPAP with a well-fitted mask.  Even the humble Air-Viva BVM set up can have a cheap, low-tech PEEP valve attached to deliver some PEEP.  What we are wanting to do is increase the intrathoracic pressure.  This:
    • Decreases preload by impeding venous return to the right heart.
    • Decreases the work of breathing which is usually high in these tachypnoeic folk.
    • increases the transmural pressure on the LV wall which increases contractility and hence the cardiac output.
    • Avoids intubation in most cases.  Usually, a decent trial of CPAP at a decent dose (> 10 cm at least) can turn things around quickly enough to make transfer safe by air UNINTUABTED.



If you have no kidney function, and you miss dialysis you will become acidotic.  You body can compensate for a while by exhaling CO2 and shifting ions around in the plasma to buffer… but eventually, the pH will drop.  There is not much you can do about this in a remote setting.  Crucially you should avoid interventions that will make it worse.

  • Do not sedate these folk, anything that drops their intrinsic respiratory rate is bad.
  • If you think they need intubation for transfer then you need an aggressive ventilation strategy with a high rate and supernormal minute volume… and careful monitoring after the fact.
  • Bicarb drips were popular in the past.  It makes sense eh? Give more HCO3… not really.  The evidence is scanty, and it involves mixing up a solution that will not give too much volume.  Tough to balance in a remote clinic.  The BICAR-ICU trial used “avoiding dialysis” as a nifty end=point… but this is clearly not an option in our patient. It may help a bit with hyperkalemia though.


    • If you can measure K levels with a VBG machine then great.  However, most small clinics do not have these.
    • ECG changes of hyperkalemia are notoriously inaccurate and insensitive.  Many dialysis patients have wildly abnormal ECGs as a result of comorbid heart disease.  So do not place too much weight on them.  If you see peaked T waves or widening QRS then it is likely true.
    • DO NOT EXCLUDE high K on the basis of an ECG!
    • Don’t’ PANIC about potassium levels around 6 or even 7.  Many of the HD patients seem to cope with this.  I am not sure if their myocytes are conditioned or what the mechanism is, but a mid-range high K seems less toxic in these folk.
  • Empirical treatment is usually a good idea if you are far from a dialysis site either in place or time.
  • GIving Ca-gluconate is a relatively safe option and can buy time.  There are few downsides to doing this off the bat.
  • Most patients are also diabetic and may have really high BSL if they are off meds and on an adrenaline-fuelled hypertensive surge.  So giving insulin will help…. how much?  well 5 units seems to work as well as 10 units with less hypoglycemia (STUDY: LaRue et al).  You can always give another 5 in an hour if the K and BSL remain too high.
  • Salbutamol is probably the last line in a small clinic.  It does work, it may help with breathing a little too…  but it can also cause a lot of tachycardia and potentially lactic acidosis.  These patients are already on the brink of acidotic plummets, so think a bit about this before giving a neb.
  • Frusemide is usually pointless in this scenario.  Some of the HD patients will still make urine and may be on massive doses of frusemide.  Give it if they are, but it is unlikely to help anytime soon.
  • Fluid boluses are usually given for non-dialysis patients.  BEWARE: a few litres of normal saline is a really bad idea in this case.  It will make their APO worse, and push them into more acidosis.  ZERO fluid is ideal.


The real key to treating the patient who desperately needs dialysis is…. DIALYSIS.  No-brainer!  However, to make this happen the patient needs to get to a place that can dialyse them SAFELY.

Most small rural dialysis units cannot (and should not) deal with patients on infusions of vasoactive agents or non-invasive ventilation.  They are just not set u that way.  So although the temptation is to get the patient “home” to dialyse… this may not be the best move.  Sometimes they need to fly over “home” and go to a bigger centre where they can have supportive intervention as they get treated and dialysed.

This is a real judgement call, a tough one at that.  Knowing the trajectory and watching their response to initial therapy can sometimes mean we can get them well enough to dialyse locally.  This is cheaper and easier than a cross-country retrieval.

But… if you get this wrong and the retrieval team bring them to a centre that cannot deal with their needs – they then need a secondary transfer and you have wasted precious time and possibly risked them needing intubation etc.  So you really need to know what your local chronic dialysis team can and cannot do.  Then you need to aim to get your patient to that level of support as soon as possible.

This is a wonderful example of how we need to act as a cohesive team in remote areas with good early communication from the remote sites, retrieval team and the receiving hospitals.


As mentioned above, dialysis patients are sick folk.  They are at high risk for a lot of other complications and it can be easy to miss these if you treat them all as simple ‘fluid overload”.

  • Think about sepsis ( they will not go hypotensive like your average bacteremic punter).
  • Consider ACS as a cause of decompensation… troponin measurement in this context is tiger territory, but a good history and attention to ECG details might catch it.  Trending troponin is useful still.
  • Look for other lung problems eg. pneumonia or effusions.  They will not get better with HD… you may be able to intervene and buy some breathing space

Add a Comment

Your email address will not be published. Required fields are marked *