Removal of common antimicrobial agents by sustained low-efficiency dialysis

ABSTRACT

Adequate dosing of antimicrobials is paramount for treating infections in critically ill patients undergoing kidney replacement therapy; however, little is known about antimicrobial removal by sustained low-efficiency dialysis (SLED). The objective was to quantify the removal of cefepime, daptomycin, meropenem, piperacillin–tazobactam, and vancomycin in patients undergoing SLED. Adult patients ≥18 years with acute kidney injury (AKI) or end-stage kidney disease receiving one of the select antimicrobials and requiring SLED were included. Blood and dialysate flow rates were maintained at 250 and 100 mL/min, respectively. Simultaneous arterial and venous blood samples for the analysis of antibiotic concentrations were collected hourly for 8 hours during SLED (on-SLED). Arterial samples were collected every 2 hours for up to 6 hours while not receiving SLED (off-SLED) for the calculation of SLED clearance, half-life (t_1/2) on-SLED and off-SLED, and the fraction of removal by SLED (f_D). Twenty-one patients completed the study: 52% male, mean age (±SD) 53 ± 13 years, and mean weight of 98 ± 30 kg. Eighty-six percent had AKI, and 4 patients were receiving cefepime, 3 daptomycin, 10 meropenem, 6 piperacillin–tazobactam, and 13 vancomycin. The average SLED time was 7.3 ± 1.1 hours, and the mean ultrafiltration rate was 95 ± 52 mL/hour (range 10–211). The t_1/2 on-SLED was substantially lower than the off-SLED t_1/2 for all antimicrobials, and the SLED f_D varied between 44% and 77%. An 8-hour SLED session led to significant elimination of most antimicrobials evaluated. If SLED is performed, modification of the dosing regimen is warranted to avoid subtherapeutic concentrations.

INTRODUCTION

Sustained low-efficiency dialysis (SLED) is a type of kidney replacement therapy (KRT) combining characteristics of intermittent hemodialysis (IHD) and continuous kidney replacement therapy (CKRT) and is used in hemodynamically unstable patients requiring KRT (1). The SLED procedure differs from standard IHD methods in that lower blood and dialysate flow rates are used to achieve a more gradual removal of fluid and toxins from the patient. SLED, also known as prolonged intermittent kidney replacement therapy or extended daily dialysis, utilizes standard IHD machines for prolonged periods (8–12 hours) with blood flow and dialysate flow rates ranging from 100 to 300 mL/min (1).

While SLED use is increasing, there is a lack of guidance for drug dosing with SLED. In fact, studies that measure drug removal by SLED have been conducted for very few marketed drugs (1, 2). Most patients requiring SLED reside in an intensive care unit (ICU) and require antimicrobial therapy for the treatment of sepsis, the leading cause of acute kidney injury (AKI) in the critically ill population (3). There is a 50% higher mortality risk in septic patients compared to non-septic patients requiring KRT, a problem potentially related to underdosing of antimicrobials in this population (4). Critically ill patients have significant pharmacokinetic (PK) alterations, decreasing the likelihood of achieving optimal PK and pharmacodynamic targets with antimicrobials (5). The addition of SLED introduces even more unpredictability in antimicrobial clearance.

Presently, dosing of many antimicrobials in patients undergoing SLED is based on data for IHD, a different method of KRT compared to SLED. Many commonly used antimicrobials are predominantly eliminated by the kidney; therefore, dose reductions are necessary in patients with kidney failure; however, the contribution of SLED may result in substantial removal, and doses may need to be increased. The problem clinicians face is how to appropriately dose these antimicrobials in patients requiring SLED to account for decreased drug elimination because of kidney failure but increased removal by SLED. In order to address this gap, we quantified clearance of cefepime (C), daptomycin (D), meropenem (M), piperacillin–tazobactam (P/T), and vancomycin (V) in ICU patients undergoing SLED to provide information for dosing these medications in critically ill patients.

RESULTS

A total of 21 adult patients completed this prospective, single-center PK study allowing for the evaluation of one or more of the selected antimicrobials per patient: 4, 3, 10, 6, and 13 measurements for C, D, M, P/T, and V, respectively. The patient and SLED information is shown in Table 1 along with the antimicrobial(s) evaluated for each patient. The mean age and body mass index were 53 ± 13 years and 33.1 ± 9.5 kg/m², respectively, with the majority of patients (86%) receiving SLED for AKI. Antimicrobials were used empirically in 57% of patients. The majority of patients were anuric or oliguic with a urine output of <500 mL over the 24-hour period prior to SLED. The average SLED time was 7.3 ± 1.1 hours with blood and dialysate flow rates maintained at 250 and 100 mL/min and a range of ultrafiltration rates from 10 to 211 mL/hour. The most common site for dialysis access was a femoral location (62%). A summary of the PK data for each antimicrobial is presented in Table 2. SLED CL was substantial for all antimicrobials except D, with the greatest contribution to overall drug elimination for P/T (SLED CL 73 mL/min P, 87 mL/min T; 71%–77% removal by SLED), followed by M (SLED CL 72 mL/min) and C (SLED CL 70 mL/min). There was not a strong correlation between the ultrafiltration rate and SLED CL for the antimicrobials evaluated (r² values 0.52–0.55 for P, T, and V, <0.1 for M, and too few data points to evaluate C and D). A representative example of the concentration versus time profile for M is shown in Fig. 1.

TABLE 1.

Patient and SLED information^a

Patient	Age (yrs)	Weight (kg)	BMI (kg/m2)	Albumin (g/dL)	HCT (%)	SLED time (hr)	Mean UFR (mL/hr)b	Mean PFR (mL/min)	Sex	Race	AKI or ESKD	UOP (mL/day)c	SLED access site	Confirmed infection (Yes/No)	Antimicrobial
1	43	91	36.5	1.5	26.2	8	211	185	Fem	B	ESKD	0	LIJ	Yes	M
2	42	146	50.5	2.5	30.5	8	81	174	Fem	B	AKI	0	L femoral	Yes	M, V
3	55	60	21.3	2.6	36.4	8	74	159	Male	W	AKI	100	R femoral	No	M, V
4	40	55	22.0	4.3	23.4	8	100	192	Fem	W	AKI	60	RIJ	No	M
5	59	100	39.2	3.4	30.5	8	50	174	Fem	B	AKI	225	R femoral	Yes	V
6	67	111	31.4	2	25.1	5	68	187	Male	B	AKI	246	R femoral	No	V
7	21	83	29.4	2.1	26	8	100	185	Fem	B	AKI	415	R femoral	No	M, V
8	64	149	41.3	2.8	34.1	8	162	165	Male	W	AKI	1,185	R femoral	No	C, V
9	56	126	50.8	1.8	24.6	8	122	189	Fem	B	AKI	0	RIJ	Yes	D
10	62	102	28.7	1.5	23.0	6	134	193	Male	B	AKI	0	R femoral	Yes	M, V
11	69	96	29.7	3.1	30.3	8	139	174	Male	W	AKI	15	R femoral	No	M
12	60	97	40.4	2.9	28.6	8	22	179	Fem	B	ESKD	0	R subclavian	No	M, V
13	42	71	28.8	1.3	27.7	8	10	181	Fem	B	ESKD	0	L subclavian	Yes	D
14	53	88	30.1	2.5	21.0	8	100	198	Male	B	AKI	0	L femoral	No	P/T, V
15	51	88	26.5	2.5	25.0	8	10	188	Male	B	AKI	1,300	R femoral	No	C
16	32	53	16.9	2.2	25.9	8	111	185	Fem	W	AKI	0	R femoral	Yes	M, V
17	52	96	28.0	1.5	25.8	8	67	185	Male	B	AKI	125	R subclavian	No	C
18	66	122	35.4	2.3	28.0	5	110	180	Male	B	AKI	575	LIJ	No	C
19	68	53	21.0	3.2	26.5	6	76	184	Fem	W	AKI	0	RIJ	Yes	M, P/T, V
20	51	135	41.8	2.8	21.2	6	50	197	Male	W	AKI	1,710	L femoral	No	P/T, V
21	51	137	44.8	1.5	25.5	6	186	186	Male	B	AKI	740	L femoral	Yes	P/T, V
Mean	53	98	33.1	2.4	26.9	7.3	96	183	52% male	67% B	86% AKI	Range: 0–1,710	L femoral = 4 LIJ = 2 L subclavian = 1 R femoral = 9 RIJ = 3 R subclavian = 2	Yes = 9 (43%) No = 12 (57%)	4 C 3 D 10 M 6 P/T 13 V
SD	13	30	9.5	0.76	3.9	1.1	52	9

^a B, Black; ESKD, end-stage kidney disease; F, female; L femoral, left femoral vascath; LIJ, left internal jugular vascath; R femoral, right femoral vascath; PFR, plasma flow rate; RIJ, right internal jugular vascath; UFR, ultrafiltration rate; UOP, urine output; W, white.

^b Range (10-–211 mL/hr).

^c Within 24 hours of study SLED procedure.

TABLE 2.

Summary of pharmacokinetic data by antimicrobial agent^a

Antimicrobial	Number of patients	Extraction ratio (%)b	SLED CL (mL/min)	t1/2 total (hr)	t1/2 off-SLED (hr)	fD (%)
Cefepime	4	42 ± 14	70 ± 21	4.3 ± 0.9	15.4 ± 8.5	66 ± 15
Daptomycin	3	6 ± 2	9 ± 2	17.0 ± 0.6	34.4 ± 4.7c	48 ± 9c
Meropenem	10	43 ± 16	72 ± 12	3.9 ± 1.1	7.2 ± 3.3	45 ± 19
Piperacillin	6	36 ± 14	73 ± 26	4.6 ± 1.5	18.5 ± 10.3	71 ± 13
Tazobactam	6	43 ± 18	87 ± 21	4.1 ± 1.1	20.4 ± 8.3	77 ± 11
Vancomycin	13	23 ± 17	37 ± 13	11.4 ± 6.2	17.1 ± 2.8d	47 ± 12d

^a Shown as mean ± SD.

^b Extraction ratio—weighted mean.

^c Off-SLED data available for two D patients.

^d Off-SLED data available for six V patients.

Fig 1.

Meropenem concentration vs time profile (patient #16). Meropenem concentrations measured during both on-SLED and off-SLED study periods.

DISCUSSION

We found that the removal of the selected antimicrobials by SLED was between that of CKRT and IHD. A comparative summary of antimicrobial clearances with these three dialysis techniques is illustrated in Fig. 2 based on data from the current study for SLED and published literature (for CKRT and IHD) (6–17). Given the obtained results, we propose dosing guidelines for the studied antimicrobials in patients receiving SLED (Fig. 3). We discuss current PK data for each agent in patients receiving SLED compared to published PK data for CKRT and IHD, where available, in the following section.

Fig 2.

Antimicrobial clearance reported for IHD, CKRT, and SLED (6–17). SLED data from this study. Antimicrobials: C= cefepime (6, 7), D=daptomycin (8, 9), M=meropenem (10, 11), P= piperacillin (12, 13), T=tazobactam (13, 14), and V=vancomycin (15, 16). ^*Meropenem clearance by IHD reported for a low flux filter.

Fig 3.

Suggested antimicrobial dosing regimens for SLED performed daily.^{* *}Base dosing on the estimated kidney function of the patient on non-SLED days. ^**Extended infusion method may be used. Higher doses may be warranted depending on severity of infection and source. BFR, blood flow rate; DFR, dialysate flow rate; LD, loading dose; SLED, sustained low efficiency dialysis; TDM, therapeutic drug monitoring.

Cefepime

As a small-molecular-weight drug with low protein binding and small volume of distribution, cefepime is removed by IHD and CKRT (6, 7, 18–21). While cefepime is expected to be removed by SLED, the extent of removal compared to other modalities has not previously been reported in a formal PK study. Jang et al. used Monte Carlo simulations (MCSs) to assess the likelihood of attaining target cefepime concentrations using previously published PK data and clearance data from CKRT procedures to simulate SLED conditions (22). Hemodialysis clearance values entered into this model ranged from ~46 to 55 mL/min depending on the effluent flow rate incorporated for SLED conditions (4–5 L/hour). The dosing strategy with ≥90% probability of achieving the target minimum inhibitory concentration (MIC) was a 2-g loading dose followed by 1 g every 6 hours or 2 g pre-SLED with 3 g administered post-SLED and dosing based on the estimated kidney function on non-SLED days.

The cefepime clearance determined from our study of 70 ± 21 mL/min was higher than the values used in the aforementioned study. SLED conditions differed in that our dialysate flow rate of 6 L/hour (100 mL/min) was higher. Based on our data showing that cefepime clearance is intermediate between that of IHD and CKRT and considering that the duration of SLED is longer than IHD yet shorter than CRRT, we suggest a higher dose and increased frequency compared to IHD: a 2-g loading dose followed by 2 g every 12 hours on days when SLED is performed with one dose to be administered after SLED considering that typical SLED durations are 8–12 hours. An alternative is to administer 1 g every 6 hours. These options provide a total daily dose of 4 g, which is lower than the total amount suggested for CKRT (6 g administered as 2 g every 8 hours) and higher than the total daily dose of 1–2 g for IHD (23–25). Extended infusion administration (e.g., over 4 hours) is also an option, although this was not a dosing regimen we used at the time of this study. There is a risk of neurotoxicity with cefepime in patients with reduced kidney function; therefore, regular monitoring and reassessment of the need for cefepime and the clinical condition of the patient (e.g., recovery of kidney function for a patient with AKI and changes in the SLED regimen) are necessary (26).

Daptomycin

While the low volume of distribution (0.1 L/kg) of daptomycin makes removal by KRT likely, the high degree of protein binding (~90%) and relatively large molecular weight (1,620 Daltons) are factors unfavorable for removal (27). As shown in Fig. 2, clearance by IHD and CKRT is below that of the other antimicrobials evaluated: range of 13–21 mL/min for IHD and 4.5–8.3 mL/min for CKRT (8, 9, 28–32). SLED CL was previously reported for 10 critically ill patients with AKI receiving SLED for ~7.6 hours at mean blood and dialysate flow rates of 166 mL/min (33). Total SLED CL was 1.03 ± 0.29 L/hour (16.7 ± 4.8 mL/min). The dosing strategy proposed was 6 mg/kg given daily 8 hours prior to SLED (as opposed to every 48 hours in IHD) (33). This is the same dosing strategy recommended for CKRT (34). A similar daptomycin SLED CL was also reported in one critically ill patient receiving 12 hours of SLED at blood and dialysate flow rates of 200 and 100 mL/min, respectively. Total clearance was 0.92 L (15 mL/min) (35). Daptomycin clearance in our study for our three patients was lower than previous studies at 9 ± 2 mL/min but still between that of IHD and CKRT. Based on our findings, we support the dosing regimen of 6 mg/kg daily with administration following SLED. Higher doses (e.g., 8–12 mg/kg) may be warranted for more severe infections such as endocarditis and vancomycin-resistant enterococcus.

Meropenem

Meropenem characteristics (low molecular weight, 383 Daltons; low protein binding, 2%) make removal by KRT likely (10, 11, 36). Meropenem removal by SLED has been studied in prospective PK studies and using simulations to predict optimal dosing based on PK parameters (37–41). A PK study performed on 19 patients with AKI receiving SLED with mean blood and dialysate flow rates of 250 mL/min for 5.3 hours resulted in a meropenem SLED clearance of 7.9 L/h (131 mL/min) (40). Meropenem clearance during SLED in 10 critically ill patients with blood and dialysate flow rates of 160 ± 3 mL/min and SLED duration of 8 hours was 2.3 L/h (38 mL/min) when estimated using the amount of drug recovered in dialysate and 5.1 L/h (85 mL/min) when using drug concentrations before and after dialysis (37). The meropenem regimen recommended was 0.5–1 g every 8 hours. Meropenem clearance was also evaluated in a study of 10 ICU patients receiving 8 hours of SLED at mean blood flow rates and dialysate flow rates of 160 and 170 mL/min, respectively. The dosing strategy of 1 g every 12 hours was recommended because plasma meropenem levels remained above the MIC during the entirety of SLED (38).

The meropenem SLED CL in our study was 72 ± 12 mL/min for the 10 patients evaluated, which can contribute to substantial drug removal. We suggest a more agressive dosing regimen similar to that recommended for CKRT of 2 g as a loading dose followed by 1 g every 12 hours with adjustments made based on the offending organism and residual kidney function of the patient. An extended infusion method may also be used for this antimicrobial (e.g. 1 gram over 4 hours).

Piperacillin–tazobactam

Piperacillin–tazobactam is a beta-lactam/beta-lactamase inhibitor combination with characteristics that make removal by KRT likely due to its low molecular weight (P 516 Daltons; T 300 Daltons), small volume of distribution (15–18 L), and low protein binding (~20%) (42). SLED CL was reported as 6 L/hour (100 mL/min) in a PK study performed on six patients receiving a 6-hour SLED session with blood and dialysate flow rates at 200 mL/min (43). The recommended dosing strategy of 4 g of piperacillin every 8–12 hours with a 2 g replacement dose post-SLED was proposed. In 34 ICU patients receiving 8 hours of SLED at blood and dialysate flow rates of 200 and 300 mL/min, respectively, SLED CL was 4.8 L/hour (80 mL/min) (44). Simulations were used to determine an optimal dosing strategy of 3.375 g P/T every 8 hours. MCS data were used to determine the dosing regimen with ≥90% probability of achieving the target MIC during SLED conditions (hemodialysis clearance values entered into this model ranged from ~40 to 50 mL/min for P and ~50 to 70 for T) depending on the effluent flow rate incorporated for SLED conditions (4–5 L/hour) (22). The recommended P/T dose was 4.5 g every 6 hours initiated at the start of SLED or immediately post-SLED (22). Our SLED clearance for P of 73 ± 26 mL/min was in the range of clearance values entered in these simulation studies (22, 44). Given the relatively wide therapeutic index of P/T and the need to achieve sufficient antimicrobial concentrations, we recommend a P/T dose of 4.5 g every 8 hours for SLED performed under similar conditions.

Vancomycin

Vancomycin removal by KRT is possible with high flux dialyzers capable of removing this large-molecular-weight antimicrobial (1,450 Daltons) (15, 16). Data available on vancomycin disposition during SLED are from both prospective and retrospective evaluations and MCS (37, 45–50). In 11 patients with AKI (blood and dialysate flow rates of 200 and 100 mL/min over a 24-hour period) (45), the mean systemic vancomycin clearance was 24.3 ± 8.9 mL/min. An initial dose of 15 mg/kg was recommended followed by TDM to guide further dosing. In 10 ICU patients (blood and dialysate flow rates of 100–200 mL/min, 8-hour duration), a 36% reduction in vancomycin concentrations was reported (46). Additionally, in 10 patients (blood and dialysate flow rates of 160 ± 3 mL/min and SLED duration of 8 hours), vancomycin SLED CL was 2.1 L/h (35 mL/min) when estimated using the amount of drug recovered in dialysate. An initial dose of 20–25 mg/kg followed by TDM was recommended (37). Similarly, in 11 patients (blood and dialysate flow rates of 160–300 mL/min for 6–8 hours), vancomycin SLED CL was 3.5 ± 2.0 L/hour (58 ± 33 mL/min) (48). Finally, in 10 patients receiving SLED for 8 hours with blood and dialysate flow rates of 200 and 300 mL/min, SLED CL was 3.84 ± 1.78 L/hour (64 ± 30 mL/min) with 59% removal during the procedure (47). The recommended vancomycin loading dose was 20–25 mg/kg prior to SLED, followed by 15 mg/kg post-SLED with TDM for further adjustments.

A common theme in terms of vancomycin dosing recommendations is to provide standard loading doses (range of 15–25 mg/kg recommended) followed by TDM. A key concept to consider is the goal of achieving target drug exposure to treat the offending organism, commonly methicillin-resistant Staphylococcus aureus (e.g., an area under the curve/minimum inhibitory ratio or AUC/MIC of 400–600 mg*hour/L assuming an MIC of 1 mg/L) (51). Based on our findings, approximately 47% of vancomycin was removed during the 7-hour procedure. We recommend a loading dose of 20–25 mg/kg followed by approximately half of this dose (10–15 mg/kg) at the end or post-SLED for an 8–12-hour procedure. Higher doses may be considered for longer SLED duration. If doses are administered just prior to or during SLED, there is greater potential for removal, and higher doses may be warranted. The use of TDM should guide dosing on non-SLED days.

Advantages and limitations

This is one of the few prospective studies to evaluate antimicrobial removal during SLED. Most antimicrobials evaluated (C, D, M, and P/T) do not have routine TDM available; therefore, information about elimination with SLED is essential for adequate dosing in critically ill patients. Given that PK studies are very small, even confirmatory findings for vancomycin are important and serve as internal controls for our findings. Limitations include the fact that there were some agents for which we recruited a relatively small number (e.g., three patients for D). The dosing regimens proposed for each antimicrobial agent were based on the consideration that for many patients the risk of negative outcomes from severe sepsis often outweighs the risk of adverse drug reactions for these antimicrobials with a relatively wide therapeutic index (e.g., piperacillin/tazobactam) or for which therapeutic drug monitoring may be performed to guide therapy (e.g., vancomycin). We also made recommendations for daily SLED, which is a common scenario for critically ill patients with sepsis in the ICU setting. Dosing decisions on non-SLED days should consider the estimated kidney function for the patient (e.g., based on a glomerular filtration rate <15 mL/min for an ESKD patient or based on an assessment of changes in kidney function for a patient with AKI). In addition, we did not evaluate a particular dosing regimen to determine the achievement of a target MIC. This was outside of the scope of this study, however, and our intent was to quantify drug removal under common SLED conditions.

Conclusions

Per unit of time, drug removal by SLED is intermediate between CKRT (less aggressive removal) and IHD (most aggressive removal). While one needs to account for the more aggressive removal of these antimicrobials with SLED compared to CKRT, the duration and frequency of the procedure and the severity and source of the infection are factors to consider when designing a dosing regimen. Elimination was especially significant for cefepime, meropenem, and piperacillin–tazobactam, suggesting a need for increased doses. While dosing recommendations have been proposed, individualization of drug regimens is necessary to account for variability in pharmacokinetics in the critically ill and variation in the KRT prescription.

MATERIALS AND METHODS

Study design, patients, and setting

A prospective, single-center PK study was performed in adult (age 18–80 years) male and female patients admitted to Methodist University Hospital, Memphis, TN, who required SLED. Patients with either ESKD or AKI who had received or were scheduled to receive at least one dose of one or more of the selected antimicrobials prior to SLED were eligible. Selected antimicrobials included C, D, M, P/T, and V. Patients with severe anemia (hemoglobin <7 g/dL) or an interval >72 hours between the antimicrobial dose and the start of SLED or unable to give informed consent directly or via their legal representative were not eligible. The study was approved by the Institutional Review Board of the University of Tennessee Health Science Center, Memphis, TN (UT-IRB #14-02951). All procedures were followed in accordance with the ethical standards of the UT-IRB and the Helsinki Declaration of 1975.

Study procedures

Antimicrobials were administered according to the prescriber. Blood sampling occurred during the on-SLED and off-SLED periods to determine residual clearance of the antimicrobial by the patient and clearance by SLED.

The on-SLED sample collection occurred during a SLED session that followed antimicrobial administration (termed the study SLED session). The attending nephrologist ordered the standardized blood and dialysate flow rates of 250 and 100 mL/min, respectively, during the study SLED session. If a patient required a change in flow rates during the SLED session as determined by the attending nephrologist, this information was documented. Either a Fresenius Optiflux 50 (F50) or 160 (F160) polysulfone dialyzer (Fresenius Medical Care North America, Waltham, MA) was used for all SLED sessions. The off-SLED sampling occurred during a 6-hour period when the patient was not receiving SLED. No antimicrobials evaluated for this study were administered during the sampling periods.

Data collection and sampling

For on-SLED sampling, blood samples (~3.0 mL per sample) were collected immediately prior to the start of SLED and then simultaneously from the arterial (blood line prior to the dialyzer) and venous (blood line exiting the dialyzer) ports at 0.25, 1, 2, 3, 4, 5, 6, 7, and 8 hours.

For off-SLED sampling, arterial blood samples (~3.0 mL per sample) were collected every 2 hours for up to 6 hours during a period when the patient was not receiving SLED. Sampling occurred either the day before or after the study SLED procedure.

For sample processing, all samples were collected into BD Vacutainer tubes containing ethylenediaminetetraacetic acid and centrifuged at 1,300 × g for 10 min. Plasma samples were kept frozen at −70°C until analysis. Measurement of antimicrobial concentrations was performed using liquid chromatography–mass spectrometry methods as was previously described (52).

Data collected included patient demographic information (age, gender, race, total body weight, and height), classification of kidney function (AKI or ESKD), previous antimicrobial doses received during hospitalization, time of antimicrobial administration, total time on SLED, flow rates during sample collection, total ultrafiltration volume removed during SLED, vascular access location for SLED, urine output (mL/day in the 24 hours prior to SLED sampling), serum albumin, and hemoglobin and hematocrit on the day of SLED. A central venous catheter was used as the SLED access in all patients.

Pharmacokinetic analysis

SLED CL, k (elimination rate constant), half-life (t_1/2) on-SLED and off-SLED, and the fraction of removal by SLED (f_D) for each antimicrobial were calculated:

SLED CL = (Ca − Cv)/Ca × [Qb × (1 − HCT)], where Ca and Cv represent the arterial and venous antimicrobial concentrations in milligrams per deciliter, respectively. Qb is blood flow during SLED in milliliters per minute, and HCT is the hematocrit. The correction of blood flow rate based on hematocrit provides the plasma flow rate, which is more appropriate when determining the clearance of drugs that distribute into plasma. Mean SLED CL in milliliters per minute was the average of the individual clearances determined at each sampling point.

The elimination rate constant during SLED (k_total), which includes the patient’s own elimination and the contribution of SLED, and the off-SLED elimination rate constant (k_off-SLED) were calculated from the slope of the logarithmic arterial plasma concentration–time profile during SLED. The elimination half-life (t_{1/2 total}) was determined as 0.693/k_total. The k_SLED was calculated as k_SLED = k_total – k_off-SLED. The fraction of drug removed by SLED (f_D) was calculated as f_D = k_SLED/k_total. This information was used to make dosing suggestions for patients receiving daily SLED.