In patients with bacterial endocarditis whose symptoms persist despite treatment, it is important to know whether antibiotic therapy really is effective in vivo. Also, some antibiotics, such as gentamicin, have a therapeutic range close to the toxic range. Blood levels of some antibiotics can be measured using various methods. This will provide some assurance that antibiotic blood levels have been reached that ordinarily should be effective. However, this does not guarantee that the organisms are in fact being sufficiently inhibited or killed. To estimate in vivo effectiveness, a Schlichter test may be performed. A patient blood specimen is obtained 15 minutes after IV infusion of the antibiotic (1 hour after an intramuscular dose). This represents the peak antibiotic level. If a trough level is desired, the specimen is drawn just before the next antibiotic dose is to be given. A standard suspension of the organisms previously cultured from the patient is placed in serial dilutions of patient serum and incubated overnight. If the lowest serum dilutions (i.e., those tubes with the least dilution of the antibiotic) do not inhibit growth of the organism, therapy is probably not effective. There is some dispute on which dilution level to use as the peak value cutoff for therapeutic effectiveness; Schlichter accepted 1:2, but currently the majority of investigators require 1:8. In fact, for a test that is widely used and relied on, there is a surprising amount of disagreement about many critical technical aspects, including whether to obtain peak or trough level or both and what their optimal values are; whether to make the patient serum dilutions with saline, Mueller-Hinton broth, nonantibiotic-containing serum, or a mixture of these; what criteria should define the bacteriocidal endpoint; and so forth. For this reason some investigators doubt the value of the test.