top of page

Welcome to the second installment of this metabolic panel series, where we're exploring what happens when various lab values reach their outer limits (or beyond). Each part in the series can be read on its own, but if you want to start at the beginning, go check out the first blog on cations HERE.

This series of blogs and podcasts are meant as a reference for you to come back to. There is a lot of information on each one of the lab values we'll cover, so it might be best to read it in parts. As I mentioned in the last blog, I'm writing these as the reference that I wish I had when I started learning to interpret lab values.

In this portion of the series we'll be discussing anions - the negative changes in the serum. We'll be starting off with chloride, which accounts for the majority of the negative charge in our blood!


Outer Limits: Chloride (Cl-)

Just like sodium (Na+) was our major extracellular positive charge (cation), chloride (Cl-) is our major extracellular negative charge (anion). On a basic metabolic panel, the anions live in the second column.

I'm going to mostly bring up the acid-base implications of chloride (as compared with sodium), and then we'll go over what kind of conditions might bring about a derangement in the chloride level. This general approach to acid-base is derived from the Stewart acid-base model (1,2), with some mixing in of other methodologies as we continue into the bicarbonate section. People often argue over which approach is best, but I just use what makes sense to me. So, what's with this acid-base balance stuff and chloride?

Chloride is involved in two very interesting measurements when it comes to acid-base.

  1. The SID (strong ion difference)

  2. The AG (anion gap)

I'll bet you can guess what the strong ion difference is - it's the difference between two strong ions. Which ones? Sodium and chloride! Remember our reference ranges? Sodium was 135-145, and chloride was 95-105 (both in mEq/L). Since we have the same unit of measure for both, we can easily look at the difference. And, as you can see from their reference ranges, the chloride should always be a good amount lower than the sodium (~38 mEq/L as we're about to find out)(1,2). Let's turn this into a graph so we can see how they stack up against one another.

It looks like the sodium is at about 143, and the chloride is about 105 - what does that make the difference between them? 143 - 105 = 38. That's perfect! 38 is the minimum we want to see in the difference between our two major strong ions. If the difference gets any closer (a number less than 38), we have something called hyperchloremic acidosis, which is when there is too much chloride in relation to the sodium. What I just described is called our SID - our strong ion difference. I keep saying that these are strong ions, but what does that actually mean?

Strong ions are in their final form - they don't go through any changes to become something else. Contrast this with a substance like bicarbonate, which can buffer and become something different. These strong ions influence week ions, such as bicarbonate, and bicarbonate interaction with hydrogen. To put that simply - strong ions influence weak ions. I think of the SID as a magnet system.

More sodium will pull the pH more positive (more alkaline). More chloride will pull the pH more negative (more acidic). This is easy to remember because the negative charge pulls the pH more negative, and the positive charge pulls the pH more positive (1).

This whole SID thing is why people bash on 0.9% NS so much - 0.9% NS causes a SID acidosis by adding more chloride to the system as compared to the amount of sodium it adds. If you look at your saline bag, it will state that it carries 154 mEq/L of sodium and 154 mEq/L of chloride. What does that make the SID of normal saline? Zero. 154 - 154 = 0. When we add volume to the body, the body becomes more like the volume that we add. I'll show you what I mean. Suppose you have 5 liters of blood, and I add 2 liters of normal saline to the body. The blood has a SID of 38, but the fluid has a SID of zero. Let's average that out.

First liter (blood): SID 38

Second liter (blood): SID 38

Third liter (blood): SID 38

Fouth liter (blood): SID 38

Fifth liter (blood): SID 38

Sixth liter (saline): SID Zero

Seventh liter (saline): SID Zero

To find an average, you have to add everything together and divide it by the total number of items you have. (38+38+38+38+38+0+0) / 7 = 27. By adding lower numbers to the mix, I brought the average down substantially. You can think of this as getting straight A's all year, and then bombing your last few classes - this will obviously bring your GPA down. This example above doesn't account for any compensatory mechanisms, and your SID will not likely come down that drastically with 2 liters of saline, but the same general idea holds true - the serum will become like what you put into it (3). But, why does this matter?

Remember how we noted before that the strong ions (like chloride) will influence the weak ions (like bicarbonate)? When chloride takes up more room on the anion side of the stack, there is less room for bicarb. There is a fixed amount of room for them to share, and the strong will overpower the weak because the weak will simply buffer into something else - they crack under the pressure (1,2,3).

Obviously, having low bicarbonate is bad. This is why there has been a push towards using 'balanced solutions' such as Ringer's Lactate, Plasmalyte, and Normosol. The Internet Book of Critical Care has a great series on pH-guided fluid resuscitation HERE if you're looking for more information on this subject (the IBCC podcast is one of the best out there in my opinion). Other than giving 0.9% normal saline, what can cause derangements in the chloride level?

Just like with sodium, excess free water can cause a dilution of chloride.

Certain interventions can cause iatrogenic hypochloremia. medications such as diuretics can cause chloride loss through the urine. Also, anything that we suction from the stomach can cause a loss of chloride. When we use orogastric or nasogastric suction for a prolonged time, the hydrochloric acid can become depleted. This is because the body works to make more of this stomach acid, and ends up using a lot of chloride in the process. For the same reason, vomiting can also cause chloride loss (4).

If the renal system chooses to reabsorb bicarbonate instead of chloride, alkalosis will occur, along with hypochloremia. As we rid the body of chloride, bicarbonate is free to take up the room on the 'stack' that chloride was previously occupying (4).

Cystic fibrosis can cause hypochloremia due to genetic mutations that cause them to lose salts in their sweat (sodium and chloride) (5).

Hyperchloremia, on the other hand, is usually the result of either dehydration or saline administration. The lack of free water in dehydration will cause a concentration of ions, resulting in a relative increase of chloride. We already went over why 0.9% NS would cause hyperchloremia - it has more chloride than the body per liter (154 mEq vs. ~100 mEq) (4).

Just like you can increase the chloride concentration from 0.9% normal saline administration, you can also increase it by taking in high amounts of salt as well. Just like saline, table salt is also equal parts sodium and chloride as well (NaCl).

Metabolic acidosis that is unrelated to saline administration can cause the kidneys to retain chloride to maintain electrical neutrality (if an anion like bicarbonate is missing, something has to take its place) (4).

Other electrolyte disorders that raise the cation side of the 'stack' can also cause the body to retain chloride to balance out the charges (2).

Wrapping up this chloride section, we're mostly worried about acid-base disturbances when we look at our chloride. When we take the chloride into context with sodium, we can draw some valuable information out of these values. Also, we're now seeing how one lab can start to influence others as the body attempts to maintain electrical neutrality (keeping the same charge on both the positive and negative side).

"But Saaaam..... the BaSICS Trial just came out and they said there's no difference in mortality between balanced fluids and normal saline at 90 days in ICU patients - doesn't even seem like it matters."

Here's what you should do:

  1. Read this article on trial design. Note the points on trial end-points, patient selection, and confounding factors in these types of patient populations.

  2. Now read this entire study and see if you can see any similarities in the trial design to what was noted in the article above by Dr. Farkus. Don't just read the abstract. Get your hands on the whole paper and read through the methods, limitations, and exclusions, and see if you think the conclusion they reached is representative of a practice-changing study.

  3. Evaluate other human and animal studies on the subject and see if this trial was able to re-create their results. If they were not, why do you think that is?

The only thing I will say is that this trial did not change the way I wrote this portion of the blog at all.


Outer Limits: Bicarbonate (HCO3-)

Even though a lab sheet or diagram will label the bicarbonate as "CO2", always remember this is actually a bicarbonate level when obtained through a chemistry panel. We learn about bicarbonate usually from learning to interpret an ABG. For example:

pH = 7.15

pCO2 = 40

HCO3 = 16

I'm not sure how your brain works when you analyze this, but I always look at the pH, and then ask which of the other two matches this issue. Since the pH above is acidic, and the bicarbonate is acidic, it's metabolic acidosis. Then, I'll see if the other value is compensating or not. In the example above, the pCO2 is not compensating. The pCO2 should be about 32. If you're wondering how to do that formula, it's called a Winter's Formula(6):

(Bicarb x 1.5) + 8 = compensated pCO2 (plus or minus two).

(16 x 1.5) + 8 = 32 (range from 30-34).

I also find that just adding 16 to the bicarb usually lands you just about where you want to be. In the above example, adding 16 would have landed us at 32 as well. Keep in mind that ETCO2 is always lower than pCO2, so you'll want to keep that gradient in mind, especially if it's a large gap due to a V/Q mismatch.

Aright, so we see that bicarbonate and CO2 have this relationship whereas bicarbonate goes down, the CO2 should go down to compensate for it. What else can we use this bicarbonate for? Our Anion Gap (AG)!

Remember how our strong ion difference (SID) was the difference between our sodium and our chloride? For the anion gap, we're simply adding the bicarbonate to the anion side. The easiest way to evaluate this is to first figure out your SID (sodium - chloride), and then subtract the bicarbonate from the SID. Example:

Sodium = 138

Chloride = 100

Bicarb = 25

The SID is 38, and then we would simply take the bicarb off of that to equal an anion gap of 13. As a rough estimate, 12 is the average anion gap.

In the previous images, there are very small letters above the bicarbonate that say "GOLDMARK". This row can grow as unmeasured anions are added to the system. I'm not going to cover what each of these items is, because I wrote an entire blog on this topic HERE. Essentially what happens in this process is that as that row of unmeasured ions grows, it leaves less room for bicarbonate, again causing bicarbonate to buffer out of the anion stack.

These GOLDMARK items are strong ions that will not buffer unless their main cause is treated. A very common example is DKA (diabetic ketoacidosis). When ketones are present in large amounts, they crush the bicarbonate and cause it to buffer (resulting in an elevated anion gap acidosis). Giving bicarbonate will not help this patient because their anion stack is already full - we would need to make room for bicarbonate by ridding the serum of ketones (by giving insulin). Interestingly, the bicarbonate usually increases by itself because the body must fill that negative electrical charge with something as the ketones leave. Buffering CO2 and H2O back into bicarbonate just happen to be the easiest way to fill this new gap. Pretty cool!

Most of the time, low bicarbonate is going to be a result of buffering. However, there are times when we can lose bicarbonate, or not create enough.

Diarrhea can result in hypobicarbonatemia through the stool. It's interesting to note that a patient who solely has diarrhea may develop metabolic acidosis due to loss of bicarbonate (7). On the other hand, a patient who is only vomiting will likely develop a metabolic alkalosis due to hydrogen and chloride loss (loss of hydrochloric acid). If a patient has both vomiting and diarrhea, the disorders will probably even each other out and not result in a severe acid-base disturbance in either direction.

The renal system is also a source of an absolute loss of bicarbonate. If the renal system is diseased or injured, the body may not produce or reabsorb adequate amounts of bicarbonate (8). In these two cases - a loss of bicarbonate through diarrhea or a failure to produce or reabsorb it by the renal system, it may be appropriate to replace it with exogenous bicarbonate (9). What about hyperbicarbonatemia?

Winter's formula (mentioned above) talked about how the pCO2 should compensate if there is metabolic acidosis. However, what about if things go the other way around? If the patient has a chronic CO2 issue (usually retaining it such as in COPD), the bicarbonate will elevate. This can happen in acute respiratory acidosis, as well as in the chronic CO2 retainer. The general thought is that in acute respiratory acidosis, the bicarbonate will raise by 1 mEq/L for every 10 mmHg rise in CO2. If the process is chronic, the bicarbonate may rise up to 3 mEq/L for the same 10 mmHg rise in CO2. In an acute or chronic respiratory alkalosis, we would see the opposite changes in the bicarbonate (10).

To wrap up this bicarbonate section, it's important to realize that bicarbonate is the buffer on the anion side - it will change its level based on what's happening around it. This is why it is said that "strong ions influence weak ions" - the concentration of ions that cannot buffer will raise or lower the concentration of ions that can buffer (like bicarbonate and/or hydrogen). The bicarb is useful for evaluating the presence of unexpected and unmeasured anions such as those found in the GOLDMARK memory aid. As a shameless plug, this lab value reference includes SID, AG, and GOLDMARK. How the card displays GOLDMARK:

The bicarbonate level can also lead us towards suspicion of an absolute loss or failure to produce bicarbonate - a bowel or kidney issue.

Alright, let's now move on to a less familiar anion - albumin!


Outer Limits: Albumin

When we think about albumin, we generally reference it in relation to plasma protein binding. Plasma protein binding is a process in which albumin soaks up a lot of some of the drugs that we administer. For example, if we administer fentanyl and versed together, this would come into play. We actually just had a blog and podcast on warfarin, which is highly plasma protein-bound.

Warfarin: ~97% plasma protein-bound.

This means that 97% of warfarin is bound to plasma protein (mostly albumin), and only 3% of that drug is actually free for use in the body, causing its clinical effects. What would happen if we administered another high plasma protein-bound medication? It would knock some of the warfarin off the plasma protein, increasing its levels in the body. In addition to causing more warfarin to be released into the body, we would also have more free-drug of whatever we're giving. Since spots on the plasma protein were already vastly occupied by warfarin, our propofol (for example) will have a higher concentration of free-drug in the serum. Who would have thought that these two seemingly unrelated drugs would have such an important thing in common? Jake Good wrong a great blog on this topic HERE.

The classic example of a plasma protein binding usually follows this story:

An elderly patient is scheduled for surgery and is told to discontinue the warfarin (97% PPB) that they take for their a-fib 2 weeks prior to the surgery to reduce their risk for bleeding. The patient also takes phenytoin (90% PPB). When the patient stops their warfarin, that opens up spots on albumin, and the albumin soaks up a bunch of the free phenytoin floating around the blood. The result? The patient has seizures because the amount of free phenytoin they were used to was soaked up by the new open units on the albumin. In emergency and critical care medicine, a fair amount of the drugs we give are pretty highly plasma protein-bound, and knowing about this process can help us to anticipate how a patient will respond to our medications.

Here's a picture from Jake's blog illustrating active and inactive drug (you'll have to read it to understand the Finding Nemo reference).

Albumin is made by the liver and has another important function other than soaking up drugs in our system. In cirrhosis of the liver, the liver does not produce adequate amounts of albumin. Albumin, while it is not part of our traditional serum osmolality formula (sodium/BUN/glucose), still plays an important role in colloid oncotic pressure in our vasculature (albumin accounts for about 80%) (17). In addition to conditions such as abdominal hypertension, portal hypertension, and liver scarring, lack of albumin contributes to ascites in the belly. This is usually exacerbated by sodium and water balance issues seen in hepatic and renal disease.

Besides binding drugs, helping us hold fluid in our vessels, and possibly indicating fluid and liver status, what else can we take away from the albumin level?

The more thing we count on our anion side, the less we have to wonder about what's in the 'unmeasured' areas. When we have an anion gap, part of that gap is albumin. Once we account for the albumin, there are fewer things that we're wondering about. However, a derangement in albumin might cause us to have a false negative in our anion gap (11,12). How come? Check out this picture: