A vancomycin case study

"Hindsight, unlike pharmacokinetics, is an exact science"

This is a recent vancomycin case from my files. It is an example of vancomycin's unpredictable pharmacokinetics and also an example of what can go wrong when evaluating vancomycin levels.

This patient was an 86 year old female nursing home patient who was being treated for a staph infection of a knee prosthesis. Her weight was 64.3 kg, height 165 cm, and serum creatinine 0.9 mg%.

Dates Dose given Date levels drawn Peak Trough
11/24 to 11/30 1g Q 12 hrs 11/29 44.1 29.7
11/30 to 12/2 500mg Q 12 hrs 12/2 28.1 22.6
12/2 to 12/10 500mg Q 24 hrs 12/6 17.4 9.8
12/10 15.6 7.8
12/10 to 12/16 1g Q 24 hrs 12/15 13.5 3.1
12/16 to 12/25 1g Q 18 hrs 12/20 28.0 15.6
12/25 33.1 17.7

Dates	Dose given	Date levels drawn	Peak	Trough
11/24 to 11/30	1g Q 12 hrs	11/29	44.1	29.7
11/30 to 12/2	500mg Q 12 hrs	12/2	28.1	22.6
12/2 to 12/10	500mg Q 24 hrs	12/6	17.4	9.8
12/10	15.6	7.8
12/10 to 12/16	1g Q 24 hrs	12/15	13.5	3.1
12/16 to 12/25	1g Q 18 hrs	12/20	28.0	15.6
12/25	33.1	17.7

Here are some lessons we can learn from this case:

Dose it right from the start
With 20/20 hindsight it is apparent that the patient was overdosed from the start. She had a "normal" serum creatinine, so she was given a "normal" dose. But of course, we know that renal function declines with age, therefore an 86 year old is not likely to have "normal" renal function. This patient's calculated creatinine clearance was 43 ml/min. An empiric dose of 1g Q 18 hours would have been a better choice.
Draw serum levels at steady-state
This physician would order a dose change and then a serum level around the "fifth dose". There is nothing magic about the "fifth dose", instead, serum levels should be drawn at steady-state, approximately 3 to 5 times the half-life. Before ordering levels, one should attempt to estimate pk parameters and time to steady-state.
Evaluate serum levels critically
It is apparent that the high trough level on 12/2 was not a true reflection of the dose because it is completely out of context with the empiric model. We must be wary when evaluating serum levels. Remember Murphy's Law, things can and will go wrong. Nursing may administer the dose incorrectly, pharmacy may prepare the dose incorrectly, phlebotomist may draw the level incorrectly or the laboratory may analyze it incorrectly. When serum levels are so out of context, another set of levels should be drawn before making a drastic dose change.
Avoid drastic dose changes
The patient was prescribed 2g/day then 1g/day then 500mg/day then back up to 1g/day. Large dose changes will cause "see-saw" serum levels which are nearly impossible to evaluate, leading to dose chasing. Instead, if the levels are high, hold the next dose long enough to allow the trough level to drop down to normal. Then decrease the maintenance dose, but don't make a 100% change. The high trough level on 12/2 was most likely a holdover from the original dosing regimen, the dose should have been held to allow the trough to fall to WNL, followed by a modest change in dose.
Watch out for accumulation
The last two sets of levels seem to indicate accumulation of vancomycin. There are recent data to suggest that prolonged treatment with vancomycin (>10 days) may result in a decline in the drug's clearance despite stable renal function.

Click here to return to previous page