Patients that were receiving vancomycin in 8, 12, and 24 h intervals were included for analysis. The primary outcome was to witness the frequency of early vancomycin trough concentrations and to see how many true trough concentrations were considered sub-therapeutic after pharmacokinetic extrapolation. The time of each vancomycin trough concentration was compared to the time of the next scheduled dose and then rounded off to the nearest min interval.
Trough concentrations measured within 0. The min threshold was instilled per institution protocol and also suggested by the report of Ye et al. Creatinine clearance CrCl was estimated using the equation of Cockcroft and Gault [ 10 ]. Trough concentrations measured too early were adjusted pharmacokinetically by predicting how much the serum concentration would have fallen during the time period between when the trough was actually measured and when it should have been measured.
The value of CrCl was estimated from the Cockcroft-Gault equation. Early trough measurement results in an overestimation of the true trough concentration.
Therefore, the potential impact of a timing error for an early trough that is sub-therapeutic is not likely to be significant, because the trough would also have been sub-therapeutic if measured at the correct time.
Then the early troughs must be separated by mild to moderate or complicated infections to determine if the true extrapolated trough concentrations were sub-therapeutic. To be included for sub-therapeutic analysis, the initial measured concentration had to be within the recommended range.
For the secondary outcomes, the relationship between CrCl and the extent of timing error based on the difference between the measured trough and estimated true trough was evaluated. A total of vancomycin trough concentrations, representing patients, were included in the analysis. For each of the four subgroups, the mean and range of the measured trough are reported, along with the estimated true trough and the difference between the measured and estimated true trough.
This observation is further illustrated by the graphical representation of data in Figure 1. The figure displays that when a trough is taken more than 2 h early, there is a possibility that it can become sub-therapeutic regardless of renal function. The most precise trough is the concentration that is measured right before the next dose of vancomycin is administered.
By extrapolating the early trough concentration, the pharmacist or clinician that is monitoring the patient can ensure or adjust the dose so that the regimen is therapeutic due to the linear pharmacokinetics of vancomycin. As reported in Table 3 , the difference between the measured and estimated true trough was at least 2. However, that observation is somewhat misleading, because the results for the category are skewed.
When two extreme values are excluded, the mean difference for the remaining 10 troughs drops from 2. To further illustrate the point, of the 12 troughs in that category, five were measured at least 5 h early and the mean timing error was 4 h.
These findings suggest that the timing of vancomycin trough concentrations is an issue of relevance and importance. Pharmacists should consider evaluating the vancomycin monitoring practices of their respective institutions to determine whether a need exists to implement intervention strategies. Some efforts to improve vancomycin trough monitoring have already been reported [ 12 , 13 , 14 , 15 ].
Cardile et al. The investigators also conducted in-service training for nurses and pharmacists to emphasize the importance of measuring vancomycin trough concentrations at the proper time.
Coleman and Wilson conducted a pre- and post-intervention study on the appropriateness of the timing of vancomycin trough concentrations, for which the intervention involved four voluntary educational sessions that included nurses [ 12 ]. Training sessions focused on the proper use of vancomycin, the importance of blood sample timing, and the principles of serum concentration monitoring of vancomycin. These findings suggest that education alone might not be sufficient to ensure the proper timing of vancomycin trough concentrations.
Traugott et al. After collecting baseline trough data from patients, they implemented a pop-up message containing vancomycin monitoring criteria that was triggered whenever a vancomycin concentration was ordered.
They suggested that prescriber acknowledgement of having reviewed the vancomycin monitoring pop-up message, along with nursing education, could lead to greater compliance rates.
Melanson et al. They proposed that one root cause of early trough measurement was a lack of knowledge about the importance of serum concentration timing. The new intervention involved notifying nurses of the policy for the timing of vancomycin trough concentrations and requiring acknowledgement by having the nurse check a box to indicate that the information had been reviewed.
The investigators concluded that educational reminders alone are not sufficient to ensure proper vancomycin serum concentration timing and discovered that many nurses believe that drawing blood for a vancomycin trough concentration too early before 2 h prior to the next dose was not a problem [ 14 ]. Currently available evidence suggests that educational efforts, process changes, systems modifications, and computer strategies produce equivocal improvement in vancomycin trough measurement.
Although such interventions are warranted and are worthy of further study, it appears that a different strategy might be worthwhile. Consideration should be given to establishing a procedure whereby clinicians can project what the true trough would have been for a trough concentration that is measured too early, by applying standard pharmacokinetic principles.
Once the trough concentration is deemed to be drawn early, the pharmacist can use this step wise extrapolation method instead of ordering a repeat level. Then with the difference in time intervals from the actual collection time and scheduled collection time, the clinician needs to convert the time in 0. For example, if the early trough subtraction was an hour and fifteen minutes early, the time interval t would be 1.
When the inevitable mistimed trough is reported, pharmacokinetic extrapolation can render the trough concentration more useful for clinical application and may obviate the need to repeat a concentration. The application can be a clinical guide for a more accurate dose adjustment.
This analysis is based largely on the results from only 48 vancomycin trough concentrations from a single institution. Gaps in the documentation of trough timing complicated the data collection process, and the analysis was dependent on the accuracy of the reported times at which blood samples were drawn.
The pharmacokinetic relationship between CrCl and vancomycin elimination rate constant is based on a linear regression formula derived from a single cohort [ 11 ]. Another limitation is the use of the elimination rate formula that is population based, according to Matzke. The population in this trial could have some inherent variability when applying the pharmacokinetic characteristics of a cohort to a broader population. Switching to continuous infusion instead of a routine daily regimen is another opportunity to prevent mistimed vancomycin trough concentrations [ 16 ].
In these cases, the inappropriate use of levels drawn too early to predict efficacy could have led to underdosing and therapeutic failure. Once the root cause of inaccurate sample timing is established, a more robust infrastructure is required to increase the accuracy of collections for vancomycin levels. Clinicians often adjusted dosing after obtaining an incorrectly timed level Table 2 and Figure 4 , even though these levels did not represent true trough levels and, therefore, should not be used as a basis for clinical decisions.
Moreover, clinicians held, decreased, and discontinued vancomycin at a higher frequency when responding to early levels, particularly when levels were drawn at 8 to 10 hours after the last dose Figure 4A.
These levels, comprising While only a couple of hours off, these samples were still 1. Effect of sample timing on clinical actions. The percentage of vancomycin levels followed by a clinical decision to hold, decrease, or discontinue vancomycin dosing A , increase vancomycin dosing B , or not adjust dosing but obtain a repeat level C is shown for levels obtained at each hour since the last dose. In comparison, levels drawn less than 8 hours after the last dose, which were typically higher Figure 3 , were not as likely to be followed by dosing adjustments and more likely to be repeated Figure 4 , suggesting that clinicians sometimes questioned the accuracy of these results.
The pharmacists at our institution often intervene when levels are high, and they carefully consider pharmacokinetics, which may have contributed to awareness that these levels did not represent true troughs and which may explain the higher rate of repeats.
At institutions with less active clinical pharmacy programs, such recoveries would be expected to be less frequent. However, even if clinicians are realizing the levels are inaccurately timed and thus refraining from inappropriate use of the levels, repeating laboratory tests contributes to delays in patient management and waste in the system. We observed a low percentage of levels within the therapeutic range and a higher than expected percentage below the therapeutic range, even within the group of early levels Table 2 , which may explain the relatively high rates of dosing increases in both groups Figure 4B.
These findings are consistent with previous findings at our institution, and a quality improvement project is currently underway in the pharmacy department to improve vancomycin dosing.
The root cause for the high percentage of inappropriately timed levels is currently under investigation, but we suspect the cause is multifactorial. While clinicians are prompted at the time of placing an electronic order for a vancomycin level with ordering instructions and recommendations, the menu option does not default to a trough level. After a clinician places an order, the nurse must schedule for the level to be drawn, sometimes requiring coordination with the phlebotomy team, which may further complicate the process of getting a correctly timed sample.
Finally, neither the sample collection time nor any dosing information is displayed with the vancomycin result in our LIS, making it difficult for clinicians to be cognizant of sample timing relative to dose administration and providing 1 explanation why inappropriate clinical actions were observed for early levels. The increased adoption of clinical information systems presents new opportunities to address the issue of correct timing of monitoring for vancomycin and other therapeutic drugs through real-time display of dose administration, specimen collection, and test result data, as well as automated guidance to help clinicians time samples correctly.
Of note, the method we used to gather data and evaluate the timing of specimen collection was automated and used data recorded by our LIS and eMAR. Our ongoing efforts are aimed at linking the LIS, positive patient identification system, and eMAR such that we can display the time relative to last administration along with each drug level.
It is possible that some of the levels we evaluated were intended as peaks or random levels, although peaks are rarely clinically indicated, and we carefully designed criteria to exclude random levels. We also did not evaluate whether trough levels were clinically appropriate eg, had the patient reached steady state. Furthermore, it is possible that clinical factors beyond timing of vancomycin levels contributed to some of our findings, although we did not observe any significant baseline differences in age, sex, renal function, or previous vancomycin levels between the 2 groups.
Samples for vancomycin trough levels were frequently drawn too early, resulting in higher vancomycin concentrations that may have contributed to a high rate of repeat vancomycin levels and possibly inadvertent underdosing. Further effort is needed to identify the root causes of incorrect sample timing and to implement solutions to improve the accuracy of vancomycin monitoring. Physicians should claim only the credit commensurate with the extent of their participation in the activity.
The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Questions appear on p Exam is located at www. Moellering RC Jr. Vancomycin: a year reassessment. Clin Infect Dis. Google Scholar. Rybak MJ. The pharmacokinetic and pharmacodynamic properties of vancomycin. Griffith RS.
Introduction to vancomycin. Rev Infect Dis. November—December ; 3 suppl : S — S Saunders NJ. Why monitor peak vancomycin concentrations? Levine DP. Vancomycin: a history. Therapeutic monitoring of vancomycin in adults summary of consensus recommendations from the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists.
Appropriateness of sampling times for therapeutic drug monitoring. Am J Hosp Pharm. Appropriateness of antiepileptic drug level monitoring. Nighttime dosing assures postdistribution sampling for therapeutic drug monitoring of digoxin. Clin Chem. Jones H Lindsay Ballard T. Improving outcomes in therapeutic drug monitoring: a case history. Clin Lab Manage Rev. Strategies for physician education in therapeutic drug monitoring.
Bates DW. Improving the use of therapeutic drug monitoring. Drug monitoring levels in the NICU. Antibiotics should be withheld until the level is known, and an order to that effect should be written. If the trough level is greater than acceptable, the drug should not be given and another level checked 6 hours later, and this should be repeated as needed until a safe level is obtained.
Peak levels are not necessary for patients being treated with a course of antibiotics without an identified organism. If a blood culture is positive, and an organism and sensitivities are identified, both peak and trough levels should be obtained to ensure adequate dosing.
If 10 days of antibiotics are planned, and the first trough level obtained is acceptable, consider repeating the trough levels at 4 to 6 days into therapy to ensure non-toxic levels, especially if there is evidence or concern of impaired renal function.
Gentamicin A gentamicin trough level should be obtained within 1 hour of the dose: Obtain level prior to the administration of the 2nd dose. If impaired renal function is a concern, a level should be obtained before the 2nd dose.
Theophylline Either a trough or peak level will provide adequate information regarding dosing.
0コメント