Of course, Scotch is a colloid

In discussing blood pressures, and fluid resuscitation, my current precept asked a simple question, “When would you use crystalloids over colloids, assuming both were available?”.  Further discussion led to a narrowing of the question to, “When would you use crystalloids and pressors, rather than simply switching to colloids?”

Since there are a variety of reasons to use any fluid resuscitation, let’s make this a trauma patient.  Mr. Smith was using his chainsaw to remove a fallen tree on his property when – whoops! – chainsaw slips and he has a deep cut on his anterior thigh.  EMS arrives, bleeding is controlled to an oozing wound.  Mr. Smith is ashen, tachycardic and hypotensive.  The nearest ED is a 20-minute response.

Now, to understand why the precept posed this question, you need to know that it takes far less of a colloid solution to produce the same effect on blood pressure as lots of crystalloid.  Generally, 250ml of colloid has the same BP effect as 4 liters of crystalloid.  That is, 1/16th the amount of colloids does the work of crystalloids.

“What!”  you exclaim.  “DTs, this means that…  hmmm, 4/4 = 1 liter, therefore 250/4 = 62.5… this means that instead of hanging a liter of saline wide open to raise a blood pressure, I can draw up a 50ml syringe of this ‘colloid’ of which you speak and bolus a nice big systolic BP almost immediately!  Why, this revolutionizes EMS!  A guaranteed systolic in my pocket!”

Not so fast!  There is, as you have probably guessed, quite a bit more to it, and as you might also have guessed, we’ll start at the very beginning of the subject.

We first need to discuss pressures.  Any fluid, in any container, exerts hydrostatic pressure.  This is the pressure the fluid (hydro) exerts on the container walls due to gravity, when the fluid is at rest (static).

If we were dumping fluid into a metal bucket, or a Styrofoam cup, we’d pretty much know all we needed to.  Since we’re putting fluid into living things (patients), we need to first explore a couple of concepts.

To get fluids into our patient, we generally introduce the fluids via the vascular bed – which includes the veins, arteries, and capillaries.  That’s usually where we want it to stay, too, if we’re trying to raise the BP.  And we also know that the vascular bed is made up of cells, which have cell walls.  If these walls just allowed anything in to or out of the cell, they’d be pretty worthless.  To work well, they need to be semi-permeable, and to selectively allow the admission or expulsion of fluids or chemicals.

Another important pressure is osmotic pressure.  It’s called “osmotic pressure” because it deals with osmoles, which is the number of osmotically active particles in a kg of solution (there can be non-osmotically active particles in a solution, but we don’t care about those right now).  By osmosis, fluids move from one side of a semi-permeable membrane to the other,  based on which side has the most solutes.   Imagine a solute as acting like a small sponge;  if cell A has a sponge inside, and the blood vessel outside the cell has a fluid which has 20 times the number of solutes (sponges), fluid will flow out of the cell and into the vessel.

Osmotic pressure is used in a lot of different fields, but in the medical field, we’re talking about cell membranes (and no other kind), and usually as it relates to existing blood plasma (your patient wasn’t completely empty, was he?) so we get to have our own term for it – tonicity – which completely ignores a lot of other ugly stuff about osmolality, osmolarity, and other junk that biologists have to worry about.

Our fluids are either hypotonic – containing fewer solutes than surrounding tissue; isotonic – containing the same number of solutes; or hypertonic – containing more solutes.

There are dozens if not hundreds of different IV solutions in existence.  We are ignoring here whole blood, blood plasma, packed red cells, and other mainstays of ED life, and concerned only with crystalloids and colloids.

What is a crystalloid?

A crystalloid is a fluid in which the solutes are dissolved.   If the particles in a fluid do not dissolve, then that fluid is not a crystalloid.  Two very common crystals also happen to make up our two most common crystalloid solutions.  Salt crystals are added to water to make Normal Saline.  Sugar crystals are added to water to make D5W.

Isotonic ( from the Greek isos, meaning “equal”) crystalloids are those fluids which have roughly the same tonicity as blood plasma.  These include:

• Normal Saline 0.9%.  This fluid is the most widely used in EMS for volume expansion.  It has no red blood cells, hence no oxygen carrying capability, and includes no electrolytes.  Its administration is purely to increase the hydrostatic pressure in the vessels.  However, it has been noted that about 75% of a saline bolus leaves the vascular bed almost immediately, leaving 25% in circulation.  That 75% can contribute to edema and wet lung sounds if the patient is over-hydrated;
• Lactated ringers.  This fluid contains a bit more dissolved in it – sodium, chloride, lactate, potassium, and calcium.  It is useful in resuscitation because, as the liver metabolizes lactate, the by-products of that metabolism help to counteract acidosis.  For resuscitation the usual dosing is 20-30ml/kg of body weight.  Ringers is not, however, used for long-term drips since the electrolytes sodium (130 mEq/L) and potassium (4 mEq/L) are respectively too high and too low for homeostasis.  That is, while the tonicity of the fluid is the same as the body, the electrolyte balance is not;
• D5W.  This fluid is not used in resuscitation.   Dextrose (the D in D5W) is metabolized by the body and leaves plain water (the W in D5W) behind.  Plain water is hypotonic, containing fewer solutes than blood plasma.  Remembering the “solutes = sponges” concept, if the vascular bed has plain water (fewer solutes) and the surrounding cells have more solutes, fluid will shift OUT of the vascular bed and INTO the cells, resulting in a drop in BP.   Since this is occurring wherever a cell contacts the vascular bed, e.g. everywhere, it happens with all cells.  A common complication is that brain cells may swell, causing headache, weakness, nervousness, vomiting, tremors, convulsions, coma, and dilated pupils.  These are not good things.

Hypertonic (from the Greek prefix hyper-, “over, or excessive”) crystalloids are those whose tonicity exceeds that of plasma.  Again, if the solutes can be thought of as little sponges, this means there are more little sponges going in to the vascular bed than currently exist in the cells.  This results in water being drawn out of the cells and into the vasculature.  The cells shrink, which is called crenation, and this cell-shrinkage is exactly what is sometimes needed:

• 7% hypertonic saline is considered “mucoactive” and is used to hydrate thick secretions to assist in expectoration;
• 7% can be administered via central line for traumatic brain injury;
• 3% hypertonic saline can be used for hemorrhagic shock (drawing water into the vasculature to increase BP), but no other kinds of shock;
• 3% may be used for acute intracranial pressure (this lowers ICP by shrinking the brain cells);
• 3% may be used for severe hyponatremia, but this is controversial

Hypotonic (from the Greek prefix hypo-, “under”) crystalloids are those where tonicity is below that of plasma.  Since surrounding cells will contain more solutes, the fluid is drawn immediately into the cells.  This is, in part, why you get so wrinkly in the bathtub – cells contain more solutes than the surrounding fresh water.   We’ve mentioned what happens with over-zealous administration of D5W.  When cells swell to bursting (which they can), the process is known as osmotic lysis or  cytolysis.  There is currently no out-of-hospital use that I know of for hypotonic fluids.

What is a colloid?

A colloid is a fluid which has something in it which is not dissolved.   The particles in colloids are larger, and do not fit through the vascular pores, and so they tend to stay in the vascular bed.  None of the particles in a colloid are osmotically active, and so we don’t have “hypotonic colloids” or “isotonic colloids”.

Colloids are used mainly for fluid expansion, and since it doesn’t leak as readily from the vascular bed smaller amounts (1/16 by volume) can be used to achieve the same results as crystalloids.  However, as hydrostatic pressure increases, the vascular pores “stretch” and allow the larger colloid particles to migrate out of the vascular bed, into cells and interstitial spaces.  When the hydrostatic pressure lowers, those pores “snap shut” and the colloidal particles are trapped outside of the vascular bed.  Therefore, edema caused by colloid administration takes much longer to resolve than edema caused by over-hydrating with crystalloids.

Examples of colloid fluids include:

• Human albumin, used for trauma, burns, surgeries, and liver disease with ascites;
• Hetastarch, a synthetic starch used for hemorrhage, burns, surgery, sepsis, and trauma.  Hetastarch has no O2 carrying capabilities or plasma proteins, and a couple of important contraindications

So there we have it, crystalloids and colloids, and completely ignoring Hartmann’s solution, blood plasma, PRBC, and about 99 other IV fluids that a patient can receive.  And we can probably, at this point, answer the initial question:  “What is best for our Mr. Smith?”

Of course, Answer #1 is, “always follow your local protocols”.  But presuming we had, say, normal saline 0.9%; D5W; Lactated Ringer’s solution, and for some reason human albumin on hand, which would be better?

Mr. Smith exhibits ashen skin and tachycardia, and a wound that bled heavily prior to EMS arrival.  He clearly needs fluids.

• Hypotonic fluids are right out – they would, as we’ve seen, speed in through our IV and straight into cells and just make everything worse;
• Hypertonic fluids *might* make some sense – they would draw fluid from the cells and interstitial spaces and into the vasculature.  But Mr. Smith has an overall deficit of fluid and needs more added, not just what he has shifted around

Isotonic or colloid it is.  Of the isotonic, D5W is right out – the dextrose will be metabolized and the water will enter cellular space, not stay in vascular space to help with BP.  Of the two remaining, Lactated Ringer’s solution might do well for a bolus, and may correct some of the acidosis we might expect from his initial trauma, but Ringers might not work well as an ongoing drip (due to electrolyte imbalance)

So, we’re down to Normal Saline, and a colloid (we’re pretending we have albumin).  And there, sorry to wuss out on you, is where the jury is still out.  Studies are being done all the time, coming to one conclusion (“Yay Saline!”) or another (“Yay Colloids!”), and the next study claims to shoot that idea down.  We just don’t know which is better.  We do know that colloids are much more expensive than crystalloids.  It would seem that Mr. Smith is getting saline today.

But at least we know why, right?

Science!

Blinded Enlightened by Science!

In the post, Reading a Map, we found our hero DTs attempting to yet again simplify EMS math, as he has done so successfully in the past.  And lo! it seemed he had again succeeded!

Indeed.  Behold the hideous formula for mean arterial pressure,

((Diastolic BP * 2) + Systolic BP) / 3

Too horrible for words!  Too grotesque for thought!  And too much damned work for 2 am.  With parry and jab, the plucky DTs vanquished the offending formula with a simple,

(Systolic / 10) + Diastolic

Yes, as flash bulbs popped, our hero stood proudly upon the podium and explained his conquest, with concrete examples – and even a table!  Yet even as he spoke, the silhouette of the beastly equation (quite undead) rose stealthily in the background to the horrified gasps of the press…

In other words, it seems that simplistic equation don’t work so well.

The two methods agree completely when (Systolic / Diastolic) = 1.43.  For instance, 120/84 results in MAP=96 using either formula.  80/56 results in MAP=64, again using either formula.

The examples in the original post, plucked randomly from mine own head, all just happened to work out to within a few mmHg, making it an attractive theory.  Without peer review, my team published (I count my hands as two separate co-workers, while typing, to help spread the blame).

Further field research blew the thing apart.  A simple 120/61 provides traditional MAP=81, DTsMAP=73 – too much error to ignore.  As did 137/76, MAP=96 and DTsMAP=90.

A random number generator was quickly pressed into service – with rules (eg Systolic must always be greater than diastolic, etc.)  The results did not bear out the usefulness of the formula.

And THIS, folks, is why we have to relearn CPR every couple of years, always with new rules; and why ET tubes in the field are losing support, and a host of other data-driven changes we see all the time in the field.

Cuz it’s Science!

Reading a MAP

I’ve figured a 2-am cheat for the MAP, which doesn’t work half-bad.  Not the linesy-roadsy MAP, the other kind.

Blood pressure is one of the more important measurements we can take, we all know that and I won’t belabor the point.  If we’ve been in the business long enough, we get a feel for a blood pressure that’s “not right”, in the overall picture of patient age, habitus, etc.

The real golden nugget of the BP is, of course, the mean arterial pressure or MAP.  This is the number which some studies suggest must be maintained over 60 (other sources state 65), and failure to do so results in poor organ perfusion or even organ ischemia.  We’re talking kidney failure, liver problems, the works.

So, what is the mean arterial pressure or MAP?

Wikipedia defines it as “… a term used in medicine to describe an average blood pressure in an individual.  It is defined as the average arterial pressure during a single cardiac cycle.”  Great.  Okay.

The article proceeds to inform us that to find the MAP, all we need to do is multiply the cardiac output by the systemic vascular resistance, and add the central venous pressure.  Wiki tells us that the CVP “is usually small enough to be neglected in this formula”.

So the MAP is (CO x SVR).

And cardiac output is…?  Along with systemic vascular resistance, it is hard to measure in the field, that’s what it is mes amis.

Wiki goes on to state that there are several ways to estimate the map, using the systolic blood pressure (SBP) and diastolic blood pressure (DBP).  This is more my speed – I got those numbers.  There are a few ways to use them to figure out MAP, to whit:

MAP = DBP + (0.33 x (SBP – DBP))

(English translation:  Subtract diastolic from systolic, multiply that number by 0.33, then add diastolic back in.)

-OR-

MAP = 2/3 DBP + 1/3 SBP

(English: multiply diastolic by 0.66, multiply systolic by 0.33, add those products)

-OR-

MAP = ((2 * DBP) + SBP ) / 3

(multiply diastolic by 2, add in systolic, divide this number by three)

Yeah, right.  This is just uno poquito mas math than I like doing.

Now, I’ve noticed a lot of ambulance folk are equipped with PDAs and the like, which is wonderful if you don’t mind whipping it out to calculate all this – with blood or vomit or worse on your gloves.  Better and easier to do it in your head, if you need it.

Here’s how:

For comparison purposes we’re going to use the third MAP equivalency formula, 2 times the Diastolic, plus Systolic, then divide the whole shebang by three.  That’s the formula I’ve most seen touted in books and such for us field grunts.  Using that formula, we see that for a patient with a BP of 80/40, this equals ((2 x 40) + 80 / 3), or (80 + 80)/3, or a MAP of 53.33.

Again, this is too much work.

The DTs 2-am MAP formula is:  Systolic / 10 + Diastolic.  Easy-peasy.  This yields, from a BP of 80/40:  80/10 = 8, plus 40 is 48.

Like any good 2am rule, this is fast, easy, and wrong.  Notice we’re a full five mmHg off the “official” estimated MAP.

Notice also that you wouldn’t probably bother figuring this out in this example, anyway – 80/40 is Not a Good BP, and you already know that.   But if you’re wondering about the mean arterial pressure for a patient with a better-sounding BP, the formula works very nicely:

BP	MAP	DTs MAP
100/65	77	75
108/75	86	86
144/100	115	114
136/90	105	104
192/160	171	179

… and so on.  Again, not many systems ask “What is the patient’s mean arterial pressure?”  If you want to ballpark it, though, Systolic/10 + Diastolic is probably an easier way to go.

So, there it is.

The Cause of, and Solution to, All Life’s Problems

I wandered into the TV room and sat as my family watched the old, 1978 version of Battlestar Galactica.  The scene on the bridge was tense, as some poofy-haired guy wandered up to the admiral and reported, “Sir!  An incredible number of Cylons are approaching!”, at which line we all burst into laughter.  “Incredible number?”  Of what possible help could that report be?  “Perhaps we should formulate an unbreachable defense!”  Tactically, it would be better to have an actual count, right?

And so it is for EMS.  We don’t say, “Doc, the patient BP is high!” – it’s 180/110, or whatever.  Pulse isn’t “Racing!”, it’s 120.  We use actual numbers because they suggest what our treatment should be, and by comparing them afterward allow us to know if our treatment is working.

Now, we’ve been saddled with some useless numbers – GCS for instance, of which I’ve written previously.  But there’s always been a number I’ve wanted, something I think an ED, ICU, or floor could really use – and despite looking everywhere I couldn’t find it, until I got a Christmas present.

In school one of the instructors told us that one of the first things a doc will do, on entering a patient room, is glance at the Foley bag (if there is one).  For those unfamiliar, a Foley is a bag used to drain urine.  A catheter is inserted into the patient’s urethra and threaded into the bladder.  A small balloon on the catheter tip is inflated with saline, making it too big to slide back out again.  The distal end of the tube connects to a clear bag, which is hung on the side of the bed.  The bag has a provision for emptying it without removing the catheter from the patient’s bladder, and the whole setup can be left in place for days at a time.  This is usually not a field procedure in our area, but other parts of the country, where transports are long, might do so.

The three things a doc is usually looking at are:  Volume (amount of urine), Color, and Flocculence (stuff floating it).

So, Volume.  When the catheter is first inserted, there should be some urine output.  Over time, of course, the bag fills – the bags are of different capacities, but in general one or two liters.  If the patient has had a Foley for a while, and the bag isn’t filling, that’s a Bad Thing – kidneys might not be working.  If that’s what you see, ask the RN when the bag was last emptied.  If it’s not been emptied by the nursing staff, and it’s only a few milliliters full, you may be seeing kidney failure.

Flocculence is usually Bad.  It’s appearance varies – I’ve seen what look like soggy cornflakes floating in the urine – but it can be sediment-like or a simple cloudiness.  Flocculence implies bacteria, usually – UTI, bacterial infection of the kidneys, that sort of thing, but can be other material I’m sure.

And Color.  This is where I wanted a number. Generally, urine should be clear or pale if the patient is adequately hydrated.  Urine is of less volume, but a darker color, if the patient is underhydrated – the urine is more concentrated because the body is trying to conserve what little water it has.  Conversely, if you see a patient whose Foley is filled to the brim with clear liquid, chances are they’re way overhydrated, although this can be on purpose if they’re flushing his kidneys.

Rhabdomyolysis is the breakdown of muscle tissue, from trauma or burns, sometimes stroke.  It releases a red-pigmented chemical which can overwhelm the kidneys and turns the urine copper- or red-colored, or dark.

We can go on, but basically the point is that the color of the urine can be indicative of underlying processes, and this would be good information to have.  And rather than running to Commander Adama on the bridge of the Galactica and shouting like a goof, “Sir, the patient’s urine is really messed up!”, it would be nice to quantify it somehow.

Enter Christmas, for which I received a startup Home Brewing kit.  What fun!  And in learning all I could about this fun and rewarding hobby, I came across something called the SRM.

The SRM is the Standard Research Method, a scale of color.  In Europe they use something called the European Brewing Convention or EBC – same colors, different numbers.

There are hundreds if not thousands of types of beer, and interestingly they range from Clear, through straw colors (pale yellow), to amber, all the way to very, very Black.  Just like urine.  And there’s a number for each of these.  The SRM ranges from 0 through 40;  here’s an example, shamelessly pick-pocketed from the Web:

I believe that with some minor tweaking, or perhaps simply expanding the scale, this might serve the purpose of quantifying urine color.  Is this a desirable thing?  I think so.  Charting over time that the patient’s urine changed from SRM 12/S (with sediment) to SRM 6/C (clear of sediment) is showing progress, that kidney function is moving in the right direction.  The other way, not so much, the patient’s condition is deteriorating.

Anyway, there it is.

The Numbers Game

Is anybody using anything other than the GCS for field assessment of head injuries?

Way, way back in 2003 DTs pointed out in a paper that there were only two GCS scores that were reliable:  GCS 3 and GCS 15.  And a 3 can be obtained by the CPR dummy, a chair, a rock…

The main problem is the number of ways a patient can score a GCS.  Different values for Eye, Verbal, and Motor can change and still give an overall GCS that remains the same.

A pre-hospital provider reports to medical control that his patient has a GCS  score of 9.  There are eighteen combinations of the three sub-scores which will result in a GCS of 9:

E4V4M1, E4V3M2, E4V2M3,  E4V1M4,  E3V5M1, E3V4M2, E3V3M3, E3V2M4, E3V1M5, E2V5M2, E2V4M3, E2V3M4, E2V2M5, E2V1M6, E1V5M3, E1V4M4, E1V3M5, and E1V2M6.  Each of these combinations is attainable; that is, it is not impossible for a patient to be E4V4M1.

Overall Glasgow Coma Score	3	4	5	6	7	8	9	10	11	12	13	14	15
Number of Sub-score Combinations To Total This Score	1	3	6	10	14	17	18	17	14	10	6	3	1
Patient’s Sub-scores Inferred Accuracy	100%	33%	17%	10%	7%	6%	6%	6%	7%	10%	17%	33%	100%

Glasgow Coma Score Distributions

As the apex of the Bell curve is approached, the individual sub-scores comprising the GCS total score become less predictable.  There are, for example, seventeen (17) possible combinations each to account for a GCS total of eight (8) or ten (10); if one were to guess the individual sub-scores, one would have a 3 in 50 chance of pinning down the appropriate values.

The inherent problem with this scoring method is easily illustrated.  A patient who at the scene scores an E3V4M3 on initial examination receives a GCS of 10.  En route to the hospital, after interventions have been applied (e.g. O2, bleeding control, etc.) another GCS of 10 is derived – this time, however, from sub-scores E2V3M5.  The patient’s overall condition, according to the GCS scale, has neither degraded nor improved, as both are GCS 10.  However, the individual scores have changed significantly either because of or in spite of prehospital interventions. In this case, the Best Eye response has degraded from “Opens on command” to “Opens on pain”; the Best Verbal response has degraded from “Disoriented speech” to “Inappropriate words”, while the Best Motor response has changed from “Flexion withdrawal” to “Localizes pain.”

To the receiving ER physician or Medical Control the changes in these individual performance criteria may provide significant insights to the patient’s condition or underlying problem, but reporting only the GCS total, which remains constant (10 in the example) will impart none of this information.

A workaround might be to report “Eye, Verbal, and Motor” scores separately rather than their sum.  Care would be needed in reporting over the radio, as “E” can sound the same as “V” if one is reporting “E 2 V 3 M 5” for instance.  Or I suppose we could just say “Eye”, “Verbal”, and “Motor”, but this seems unwieldy.

But somehow I think we can come up with something just as quick but more useful.  We in EMS are used to scrapping stuff all the time when something better comes along – MAST, paper bags for hyperventilation, tourniquets, then bring back the tourniquets – we’re flexible.  The GCS itself is a replacement for a previous system.

Of course, we’d need to overhaul the Trauma Score (which uses GCS as one of its inputs), but, hey.

Anyone?