Heparin Drips

The pharmacology, and why it needs monitoring

Heparin is not an anticoagulant on its own. It binds antithrombin and changes its shape, accelerating the rate at which antithrombin inactivates thrombin and factor Xa by orders of magnitude. Only the longer heparin chains can bridge antithrombin to thrombin, which is why unfractionated heparin inhibits both IIa and Xa while low molecular weight heparin is weighted toward Xa.

The problem is everything else it binds. Heparin is a heterogeneous mixture of chain lengths that sticks nonspecifically to plasma proteins, endothelial cells, macrophages and von Willebrand factor. That saturable binding gives it dose dependent, nonlinear clearance: the half-life is roughly 60 to 90 minutes at usual therapeutic rates but stretches as the dose climbs. Two patients of the same weight given the same rate will land in different places.

That is the whole argument for monitoring. It is also the argument for using it in the first place. A short half-life and a real antidote make heparin the drug you want in the patient who might go to the operating room, might bleed, or might need a procedure at three in the morning.

Weight based dosing

Two protocols, because the target intensity is not the same. Illustrative examples of the common patterns:

Which weight? Most protocols use actual body weight, with an adjusted body weight in significant obesity to avoid a bolus that produces an immediate and pointless supratherapeutic level. Whatever your protocol says, the weight you dose from should be a weight someone actually measured, not the number the patient gave in triage.

Order a baseline aPTT or anti-Xa, PT/INR, CBC and creatinine before the first dose. You will want the baseline later, and in one specific situation described below you will need it badly.

aPTT versus anti-Xa, and why so many centers switched

The aPTT is a clot based global test. It is sensitive to heparin but it is sensitive to a great many other things as well, which is its weakness. Every laboratory has to establish its own therapeutic aPTT range by correlating it against anti-Xa activity for the specific reagent and instrument in use, so the numbers are not portable between hospitals. A range of roughly 1.5 to 2.5 times the control value is the classic teaching, and locally that might mean 60 to 100 seconds or something quite different.

The anti-Xa assay is a chromogenic test that measures what heparin is actually doing to factor Xa. A therapeutic range of 0.3 to 0.7 IU/mL is the usual target for unfractionated heparin. It is not affected by lupus anticoagulant, by factor deficiencies, or by the acute phase reactants that distort the aPTT. It correlates better with the delivered dose, it reaches a stable target faster, and it produces fewer of the pointless midnight rate changes that come from chasing a noisy number.

That is why many centers moved. The tradeoffs are cost, availability at small hospitals, and a small set of interferences of its own. If you are used to aPTTs and your hospital has switched, the practical adjustment is to stop treating the number as a bleeding risk estimate. It is a drug level.

When to draw the level

• Six hours after starting the infusion, and six hours after every rate change. That is roughly the time needed to approach steady state. Drawing at two hours tells you almost nothing and will make you change a rate that was correct.
• Once two consecutive levels are in range on an unchanged rate, most protocols move to once daily.
• CBC daily for the platelet count. This is not optional and it is how HIT gets caught.
• Recheck after any clinical change: a bleed, a procedure, a transfusion, new renal failure, a new drug.

A worked titration

Nomograms exist so that the nurse can titrate at 2 a.m. without calling you, and so that the titration is consistent. This is an illustrative example of the structure of one, not a protocol to use.

Levels you cannot interpret

Before you change a rate, decide whether the number means anything.

• Lupus anticoagulant. It prolongs the baseline aPTT, sometimes dramatically, and the aPTT then tells you nothing about heparin effect. This is exactly the patient in whom you order the baseline aPTT before starting. If the baseline is already prolonged, monitor with anti-Xa.
• Factor deficiencies and liver disease. A prolonged baseline aPTT from a factor deficiency has the same problem. Anti-Xa again.
• Acute phase response. High factor VIII and fibrinogen in acute illness shorten the aPTT for a given heparin level, so the aPTT reads falsely low and you chase the rate upward into an overdose.
• Sample from the heparin line. Covered above. It happens constantly.
• Interference with the chromogenic assay. Marked hyperbilirubinemia, hypertriglyceridemia and free hemoglobin from hemolysis can affect anti-Xa results. The laboratory will usually comment on a grossly hemolyzed or lipemic sample. Read the comment.
• Concurrent low molecular weight heparin or fondaparinux. Those also produce anti-Xa activity. If the patient got enoxaparin at an outside hospital eight hours ago, the anti-Xa you are measuring is not purely the drip.

Heparin resistance

The usual working definition is a patient who cannot be brought to a therapeutic level despite very large daily requirements, often quoted at more than about 35,000 units per day. Before you call it resistance, exclude the boring explanations: the drip was off for a CT scan, the sample was drawn wrong, the pump was set to the wrong units.

Real causes include antithrombin deficiency, which is far more often acquired than inherited. Antithrombin is consumed in acute extensive thrombosis, lost in nephrotic syndrome, depleted in disseminated intravascular coagulation, reduced in liver failure, on ECMO and cardiopulmonary bypass, and by L-asparaginase. Also high factor VIII and fibrinogen levels, high heparin binding protein levels in acute inflammation, and increased heparin clearance.

Practically: check an antithrombin activity level, and check an anti-Xa if you have been following aPTTs, because a fair proportion of apparent resistance is an aPTT artifact and the patient is actually therapeutic. If antithrombin is genuinely low, escalating heparin endlessly is not the answer. Discuss antithrombin repletion, or a switch to a direct thrombin inhibitor such as argatroban or bivalirudin, which do not require antithrombin to work. This is a hematology conversation, not a solo decision.

Heparin induced thrombocytopenia

An antibody against the complex of platelet factor 4 and heparin that activates platelets, consumes them, and produces a paradoxical and dangerous prothrombotic state. The patient does not usually bleed. The patient clots.

The 4T score, in outline

Roughly, 0 to 3 is low probability, 4 to 5 intermediate, 6 to 8 high. A low score has a strong negative predictive value and is genuinely useful for not sending the assay on every ICU patient whose platelets drifted down.

Protamine reversal

Protamine sulfate is a positively charged protein that binds the negatively charged heparin molecule and neutralizes it. The dosing logic is what matters, because the number is not fixed.

• 1 mg of protamine neutralizes roughly 100 units of heparin.
• You are only reversing what is still circulating. With a half-life of about an hour, you dose against the heparin given in the last 2 to 3 hours of infusion, not the total the patient has received. A patient running at 1,200 units/hr has had roughly 2,400 units in the last two hours, so on the order of 25 mg of protamine, not 250.
• A single dose above about 50 mg is generally not given.
• Give it slowly, over at least 10 minutes. Rapid administration causes hypotension, bradycardia and anaphylactoid reactions. Risk is higher in patients with prior protamine exposure, prior NPH insulin use, vasectomy, or fish allergy.
• It only partially reverses low molecular weight heparin, roughly 60 percent of the anti-Xa effect, and does essentially nothing for fondaparinux.
• Excess protamine has a weak anticoagulant effect of its own, which is another reason not to overshoot.

Reversal is for serious bleeding or an emergency procedure. For most minor bleeding, stopping the infusion and waiting an hour or two is the entire treatment.

Transitioning to an oral agent

To warfarin. Start warfarin early, usually on day one or two, and overlap with heparin for at least 5 days and until the INR has been in the therapeutic range on two measurements at least 24 hours apart. Both conditions, not either. The INR rises on factor VII long before factor II is depleted, so an INR of 2.4 on day two is a lie about the patient's actual anticoagulation. Stopping the drip at that point is a classic and preventable error.

To a direct oral anticoagulant. Apixaban and rivaroxaban are generally started at the time the infusion is stopped, since their onset is rapid and there is no reason to overlap. Dabigatran and edoxaban are conventionally started after a period of parenteral anticoagulation, typically 5 to 10 days. Check renal function, weight and the indication before choosing, and check for the interactions that matter.

Related: common night calls, and charting to survive a lawsuit for why "held for procedure" needs a reason and a restart time next to it.