Evaluation of the Effect of Mechanical Ventilation on the Pharmacokinetic Parameters of Vancomycin in Intensive Care Patients: A Prospective Cohort Study

Abstract

Objective:

Intensive care unit (ICU) patients on mechanical ventilation, in particular, may experience changes in pharmacokinetics due to alterations in cardiac output and organ blood flow. This study aims to evaluate the effects of mechanical ventilation on the steady-state pharmacokinetic changes of vancomycin.

Methods:

In this prospective cohort study, eligible patients were enrolled into either the ventilated or non-ventilated group. Demographic and clinical data were gathered, and peak and trough blood samples were collected to measure their levels. Additional pharmacokinetic parameters were calculated using appropriate formulas.

Findings:

A total of 39 ICU patients were enrolled in the final analysis. Of these, 32 matched patients were divided into two equal groups: ventilated and nonventilated. None of the pharmacokinetic parameters, including peak plasma concentration (P = 0.878), trough plasma concentration (P = 0.437), volume of distribution (P = 0.468), K (P = 0.234), t_½ (P = 0.266), clearance (P = 0.709), and daily area under the curve (P = 0.418), were significantly different between the two groups. Bivariate correlation analysis showed a significant correlation between estimated glomerular filtration rate, acute physiology and chronic health evaluation II score, and C-reactive protein (CRP) level with vancomycin pharmacokinetic parameters. Furthermore, significant effects of augmented renal clearance (ARC) on vancomycin pharmacokinetics were determined.

Conclusion:

The key finding of this study was that ARC significantly impacted vancomycin serum levels, resulting in subtherapeutic concentrations in 23.1% of patients. High CRP levels also significantly correlated with increased vancomycin distribution volume, emphasizing the role of inflammation. These findings underscore the importance of personalized vancomycin dosing, taking into account factors such as ARC and inflammation, to improve patient outcomes. However, due to the small sample size, further studies with larger cohorts are necessary.

INTRODUCTION

Vancomycin, the oldest glycopeptide antibiotic, is considered the first-line treatment for severe infections caused by methicillin-resistant Staphylococcus aureus (MRSA) in hospitalized patients.[1] Despite the emergence of several new anti-Gram-positive agents, vancomycin remains a cornerstone in the treatment of MRSA infections. However, the challenge lies in the optimal vancomycin dosing in critically ill patients.[2]

Vancomycin pharmacokinetics are notably different in intensive care unit (ICU) patients.[3] Several factors influence the pharmacokinetics of drugs in ICU patients. Three important factors to consider are: (A) these patients receive multiple medications, which increases the risk of drug interactions and side effects, (B) critical illness causes pathophysiological changes that may affect medication exposure, and (C) ICU organ support, such as mechanical ventilation, can cause variability in drug volume of distribution (V_d) and clearance.[4]

In 1947, Drury et al. were the first to demonstrate that mechanical ventilation could impact kidney function and urine output.[5,6]

Over time, research into how mechanical ventilation affects patients’ hemodynamics expanded. It was discovered that mechanical ventilation can influence blood flow to various organs by increasing intrathoracic pressure. The most frequently reported effects were reduced cardiac output and decreased renal blood flow, and the level of positive end-expiratory pressure (PEEP) was directly linked to a reduction in cardiac output. Theoretically, mechanical ventilation could impact the pharmacokinetics of medications administered to patients.[7,8]

Although vancomycin has a long-standing reputation as an effective antibiotic, there is a notable lack of pharmacokinetic data specifically for its use in ICU patients.[9] To date, only one retrospective study has examined the effect of mechanical ventilation on vancomycin pharmacokinetics, revealing a 20% reduction in vancomycin clearance.[10] Therefore, the present study aims to evaluate the effect of mechanical ventilation on the pharmacokinetic parameters of vancomycin in ICU patients.

METHODS

This study was a prospective cohort study from July 2024 to November 2024. The research participants included patients referred to Imam Khomeini Hospital Center in Sari, affiliated with Mazandaran University of Medical Sciences in northern Iran.

The Ethics Committee of Mazandaran University of Medical Sciences approved this study, which has the ethics number: IR.MAZUMS.REC.1403.181. Informed consent was obtained from patients or their authorized substitute decision-makers.

Peak and trough blood samples were collected from ICU patients who were eligible for inclusion in the study. Patients were divided into two groups: a ventilated group and a nonventilated group.[11,12] Nonventilated patients were those who were not mechanically ventilated or were initially in the ventilated group and subsequently transitioned to the nonventilated group after weaning from the device with a 48-h washout period[11] (own-controlled patients). In the ventilated group, the device mode was synchronized intermittent mandatory ventilation, and PEEP was set to 5 cm H₂O in all patients, who remained on these settings for at least 72 h.

All patients were empirically treated with intravenous infusion vancomycin at a daily dose of 15 mg/kg of actual body weight every 12 h[13] without a loading dose. To the extent possible, patients in the ventilated group were matched with those in the nonventilated group based on demographic characteristics and clinical conditions.

Inclusion criteria consisted of the following: EF of at least 41%

1. Age 18 or older

2. Patients receiving vancomycin as an empiric or definitive antibiotic therapy

3. GFR ≥45 mL/min/1.73 m²[14]

4. Acute Physiology and Chronic Health Evaluation II (APACHE II) score ≤22[15]

5. 18.5≤ Baseline body mass index (BMI) <30 kg/m².[16]

Exclusion criteria comprised the following:

1. Pregnant women

2. Baseline renal dysfunction, including patients with acute kidney injury (AKI), chronic kidney disease stage IIIB or greater, end-stage renal disease, and those undergoing hemodialysis

3. Burns

4. Severe hepatic failure (Child-Pugh class C)

5. Serum bilirubin level >2.5 mg/dl

6. Serum albumin level <2 g/dl[14]

7. Breastfeeding

8. Septic shock[17]

9. Hemodynamic instability[12]

10. Metastatic Cancer

11. Hypersensitivity reaction to vancomycin.[18]

12. Septic shock (and any patient receiving vasoactive drugs)

13. Pleural effusion

14. Ascites

To measure serum vancomycin levels at steady state (at least 48 h after the start of vancomycin therapy[1]), 2.5 mL of blood was collected[17] in each of the following conditions:

1. Sample for measuring peak serum levels: 60 min [50-90 min] after the end of the 1-hour infusion[19]

2. Sample for measuring trough serum levels: 30 min before the next dose.[9]

Blood samples were obtained from an arterial catheter, a central venous catheter, or by venipuncture (whichever was possible) from a vein not used for the administration of the vancomycin dose.[20,21] After blood collection, we poured the blood into a test tube (had gel and clot activator), waited a few minutes, and allowed the blood to clot. Then, the sample was centrifuged at 2000 rpm for 10 min (or at 4000 rpm for 5 min), and the obtained serum sample was stored at − 80°C for final analysis.[14,17]

Demographic, clinical, and baseline laboratory findings of eligible patients were recorded, and patients were closely monitored throughout the study. The following information was recorded at the time of blood collection: Mean arterial pressure (MAP), Glasgow Coma Scale (GCS), APACHE II score,[20,22] the number of vancomycin injections, ICU length of stay, mechanical ventilation duration, nutritional support,[17] ventilator settings,[12] and cumulative fluid balance up to the time of sampling.

Fluid balance was calculated as:[18,23,24]

Negative: Cumulative fluid changes <0%

Even: 0% < Cumulative fluid changes <10% body weight

Positive: Cumulative fluid changes ≥10% body weight

We assessed new-onset AKI, defined by the kidney disease improving global outcomes guidelines guideline as having any of the following: a rise in serum creatinine of at least 0.3 mg/dL within 48 h, or an increase in serum creatinine to 1.5 times the baseline within the past 7 days, or urine output below 0.5 ml/kg/h for 6 h.[25] Serum creatinine levels were monitored for 1 week after initiation of dose-adjusted vancomycin to identify new cases of AKI. Additionally, we followed the patients for 4 weeks and recorded mortality on days 7 and 28.

Peak and trough serum levels of vancomycin were measured using a Roche COBAS INTEGRA 400 plus with VANC3 ONLINE TDM Vancomycin Gen. 3-06779336190 assay.

Other pharmacokinetic parameters of vancomycin were calculated with the following formulas,[26] and the effect of ventilation on them was evaluated:

The primary outcomes involved evaluating the pharmacokinetic changes of vancomycin resulting from the use of mechanical ventilation during the steady state in ICU patients.

Secondary outcomes include (a) assessing the relationship between the APACHE II score at the time of sampling and the serum level of vancomycin, (b) evaluating the incidence of infusion-related reactions in patients receiving vancomycin admitted to the ICU, and (c) determining the mean serum level of vancomycin in patients admitted to the ICU.

The sample size was set at 16 patients per group, with an alpha of 5%, 80% power, m1 = 183, m2 = 379, sd1 = 123, and sd2 = 258, based on the study by Polard et al.[27] for the MRTp index and using the relevant formula.

The normality of quantitative datasets was assessed using the one-sample Kolmogorov–Smirnov test. Categorical variables were reported as frequencies and percentages, whereas quantitative variables were expressed as means ± standard deviation or Median (minimum–maximum).

Categorical variables between the two groups were compared using the Chi-square test, and quantitative variables were analyzed using the independent sample t-test or the Mann–Whitney U-test, Paired Samples t-test, or Wilcoxon signed-rank test as appropriate.

Correlation analysis was performed using a scatter plot and the Spearman correlation coefficient.

All statistical analyses were performed using SPSS Software (version 27; IBM SPSS Inc., Chicago, IL) version 27.0 (IBM), with P ≤ 0.05 deemed statistically significant.

RESULTS

Participants and sub-group analyses

Figure 1 shows the enrollment process flowchart. Of the 40 patients included, one was excluded from the analysis due to advanced age (97 years), and 39 patients were included in the final analysis and Marginal GFR [Figure 1].

Figure 1.

Enrollment and subgroup analyses

The mean age of all patients was 45.0 ± 16.1 years, with 3 patients (7.7%) classified as elderly (≥65 years).[28] Demographic information, clinical characteristics, and laboratory data of the included patients are detailed in Appendices 1 and 2. Due to the study design, 16 patients in the ventilated group were matched with 16 patients in the nonventilated group based on baseline conditions and demographic factors to minimize confounding variables (Sub-Group 1). By closely matching age, sex, BMI, and vancomycin dose, the investigators aimed to examine the actual effects of mechanical ventilation on vancomycin pharmacokinetics.

The ventilator settings are presented in Table 1.

Table 1.

Ventilator settings

Mode: SIMV
Parameter	Value all patients (n=16)
PEEP	5 cmH2O
Vt*	5.6–8.46 mL/kg of IBW
	525 (420–600) mL
RR*	14 (10–18) (breaths/min)
FIO2*	50 (45–50) (%)
PS*	12 (10–17) cmH2O

*Data are presented as median (minimum–maximum) for quantitative data with nonnormal distribution. SIMV=Synchronized intermittent mandatory ventilation, PEEP=Positive end-expiratory pressure, Vt=Tidal volume, RR=Respiratory rate, FIO₂=Fraction of inspired oxygen, PS=Pressure support, IBW=Ideal body weight

Consequently, there were no significant differences between the groups in terms of demographic factors, and the two groups were generally similar in age, gender, BMI, and reason for ICU admission [Appendix 3].

As shown in Appendix 3, there was a significant difference in APACHE II and GCS scores. Considering the type of patients studied, the level of consciousness in nonintubated patients differed from that of the ventilated group. The GCS score is also a key component of the APACHE II. In laboratory tests, only alkaline phosphatase was higher in the nonventilated group (P = 0.036). In circulatory variables, MAP and cardiac ejection fraction did not differ [Appendix 3].

Appendix 4 presents the clinical conditions of patients at the time of blood collection. There were no significant differences between the ventilated and nonventilated groups in terms of vancomycin number dose, ICU length of stay, cumulative fluid balance, at the time blood was obtained.

The nutritional support status of the patients differed significantly between the two groups, as expected, given the conditions of the ventilated patients. In the correlation analysis between different types of nutritional support and vancomycin pharmacokinetic parameters, no significant relationship was found.

Among the 16 matched pairs (Sub-Group 1), seven patients were their own controls. Because confounding factors were best controlled in these patients, we considered them a subgroup (Sub-Group 2) and analyzed them separately within this group.

Patients were evaluated for augmented renal clearance (ARC) incidence using serum creatinine and the Cockcroft-Gault formula based on adjusted body weight. After excluding 9 ARC patients (Sub-Group 3) and 30 non-ARC patients, they were analyzed in a separate group (Sub-Group 4).

Demographic and clinical characteristics of ARC patients are reported in Appendix 5.

Comparison of pharmacokinetic parameters of vancomycin between the ARC and non-ARC groups is shown in Table 2.

Table 2.

Comparison of vancomycin pharmacokinetic parameters between augmented renal clearance and nonaugmented renal clearance groups (sub-group 3 vs. 4)

Variable	ARC group (n=9)	Non-ARC group (n=30)	P*
Cpeak (mg/L)	20.45±5.22	21 (12–38)	0.604a
Ctrough (mg/L)	6.40±1.21	8.70 (5.00–24.00)	0.002a
Vd (L)	89.50±25.10	107.63 (37.87–200.00)	0.105a
K (h−1)	0.0960±0.0301	0.0639±0.0298	0.008b
t1/2 (h)	7.69±1.79	10.88 (4.10–30.58)	0.004a
Clearance (L/h)	8.48 (5.99–8.95)	6.46±1.88	0.018a
AUC24 (mg/h/L)	287.48±38.49	358.51±97.35	0.003b
Reaching to target trough concentration, n (%)
Yes (10–15)	0	8 (26.7)	-
No less	9 (100)	17 (56.7)
No higher	0	5 (16.7)
Reaching to target AUC24, n (%)
Yes (400–600)	0	7 (23.3)	-
No less	9 (100)	22 (73.3)
No higher	0	1 (3.3)

*Comparison between two groups: ARC versus non-ARC, ^aMann–Whitney U-test, bIndependent samples t-test. A P≤0.05 is considered statistically significant. Data are presented as number (%) for dichotomous variables, mean±SD for quantitative data with normal distribution and Median (minimum–maximum) for non-normal distribution. The sum of percentages may not equal to 100 due to rounding. C_peak=Peak plasma concentration, C_trough=Trough plasma concentration, C_d=Volume of distribution, K=Elimination rate constant, t_1/2=Half-life, AUC₂₄=Daily area under the curve, SD=Standard deviation, ARC=Augmented renal clearance

Except for peak concentration and V_d, significant differences were observed in all other parameters between the two groups.

To create equal groups, two outliers in the nonventilated group were removed, and in Sub-Group 5, 14 patients in each group were analyzed [Figure 1].

Outcomes

Primary outcomes

Our primary outcome was to evaluate the effect of mechanical ventilation on the pharmacokinetic parameters of vancomycin.

The results of our study did not reveal any statistically significant differences in the pharmacokinetic parameters of vancomycin between patients on mechanical ventilation and those not on it in any of the subgroups (Subgroups 1, 2, 4, and 5) [Tables 3 and 4 and Appendices 6 and 7].

Table 3.

Comparison of vancomycin pharmacokinetic parameters between ventilated and nonventilated groups within patient subgroup 1

Variable	Ventilated group (n=16)	Nonventilated group (n=16)	P*
Cpeak (mg/L)	20.68±4.30	20.98±6.25	0.878a
Ctrough (mg/L)	10.48±5.11	8.00 (5.00–18.00)	0.437b
Vd (L)	114.53±32.78	104.38±44.49	0.468a
K (h−1)	0.0631±0.0250	0.0757±0.0329	0.234a
t1/2(h)	10.35 (6.79–30.58)	9.07 (4.10–22.91)	0.266b
Clearance (L/h)	6.65±1.99	6.88±1.55	0.709a
AUC24 (mg/h/L)	357.57±114.13	328.61±83.03	0.418a

*Comparison between two groups: Ventilated versus nonventilated, ^aIndependent-samples t-test, ^bMann–Whitney U-test. A P≤0.05 is considered statistically significant. Data are presented as mean±SD for quantitative data with normal distribution and median (minimum–maximum) for non-normal distribution. C_peak=Peak plasma concentration, C_trough=Trough plasma concentration, V_d=Volume of distribution, K=Elimination rate constant, t_1/2=Half-life, AUC₂₄=Daily area under the curve, SD=Standard deviation

Table 4.

Comparison of vancomycin pharmacokinetic parameters between ventilated and nonventilated groups in own-control patients (sub-group 2)

Variable	Ventilated group (n=7)	Nonventilated group (n=7)	P*
Cpeak (mg/L)	19.71±3.35	20.85±3.38	0.244a
Ctrough (mg/L)	7.38±2.02	8.00 (6.00–8.30)	0.599b
Vd (L)	91.23±14.51	84.97±14.83	0.369a
K (h−1)	0.0833±0.0125	0.0851±0.0104	0.758a
t1/2 (h)	8.47±1.30	8.22±1.01	0.688a
Clearance (L/h)	7.57±1.51	7.18±1.15	0.247a
AUC24 (mg/h/L)	300.89±62.19	312.60±43.97	0.440a

*Comparison between two groups: Ventilated versus nonventilated, ^aPaired-samples t-test, ^bWilcoxon signed-rank test. A P≤0.05 is considered statistically significant. Data are presented as mean±SD for quantitative data with normal distribution and median (minimum–maximum) for non-normal distribution. SD=Standard deviation, C_peak=Peak plasma concentration, C_trough=Trough plasma concentration, V_d=Volume of distribution, K=Elimination rate constant, t_1/2=Half-life, AUC₂₄=Daily area under the curve

Secondary outcomes

For a more comprehensive analysis, we examined the correlation between covariates, pharmacokinetic parameters, and clinical conditions. Estimated glomerular filtration rate (eGFR), APACHE II score, and C-reactive protein (CRP) significantly influence the pharmacokinetic parameters of vancomycin [Tables 5-7].

Table 5.

Correlation between estimated glomerular filtration rate and vancomycin pharmacokinetic parameters in all patients (n=39)

Variable	r s	95% CI	P
eGFR, (mL/min/1.73 m2)
Ctrough (mg/L)	−0.612	−0.781 to −0.358	<0.001
Vd (L)	−0.482	−0.698 to −0.188	0.002
K (h−1)	0.713	0.506 to 0.843	<0.001
t1/2 (h)	−0.713	−0.843 to −0.506	<0.001
Clearance (L/h)	0.495	0.204 to 0.706	0.001
AUC24 (mg/h/L)	−0.378	−0.625 to −0.061	0.018

A P≤0.05 is considered statistically significant. 95% CI=Lower to upper. eGFR=Estimated glomerular filtration rate, C_trough=Trough plasma concentration, V_d=Volume of distribution, K=Elimination rate constant, t_1/2=Half-life, AUC₂₄=Daily area under the curve, CI=Confidence interval, r_s=Spearman’s correlation coefficient

Table 7.

Correlation between C-reactive protein and vancomycin pharmacokinetic parameters*

Variable	r s	95% CI	P
CRP (mg/dL)
Vd (L)	0.675	0.582 to 0.751	<0.001

*n=30, remove 9 outliers and use multiple imputation for missing data. A P≤0.05 is considered statistically significant. CRP level (baseline or nearest to sampling day with a 3.3±2.6 day window). r_s=Spearman’s correlation coefficient, V_d=Volume of distribution, CRP=C-reactive protein, CI=Confidence interval

Of the 42 patients who were eligible for inclusion in the study, a vancomycin infusion reaction occurred in 2 patients (4.7%), who were subsequently excluded from the study [Figure 1].

The mean serum vancomycin level in the patients included in the final analysis was 9.4 ± 4.1 mg/L.

Mortality and acute kidney injury

Nephrotoxicity is a major side effect associated with vancomycin treatment. New-onset AKI was observed in three patients; no significant difference was found between the two groups. AKI occurred after our sampling [Appendix 8].

The mortality rate during the 1^st week (7-day mortality) was 12.5% in the ventilated group (2 patients). On the other hand, we monitored patients for mortality status for 28 days, and there were no significant differences between the two groups in 28-day mortality [Appendix 8].

The patient who developed Stage 2 AKI received meropenem along with vancomycin. Among the patients who developed Stage 1 AKI, one received piperacillin/tazobactam, whereas another was treated with meropenem and levofloxacin, along with vancomycin.

Non-steroidal anti-inflammatory drugs (NSAIDs) may influence the pharmacokinetics of vancomycin,[29] but none of our patients were receiving NSAIDs.

DISCUSSION

Several researchers have studied the pharmacokinetics of vancomycin in various populations;[30,31,32] however, little research has been conducted in critically ill patients. This study investigated the pharmacokinetics of vancomycin in patients admitted to the ICU. Despite numerous studies on the pharmacokinetics of this drug, data on patients undergoing mechanical ventilation are limited.

The primary objective was to evaluate the impact of mechanical ventilation on the pharmacokinetic parameters of vancomycin and to identify any additional factors that may contribute to this effect.

This study was conducted on 39 patients admitted to the ICU receiving vancomycin. The patients were divided into two groups: the mechanically ventilated group and the non-mechanically ventilated group. Pharmacokinetic parameters, including peak and trough concentrations, were measured. Furthermore, half-life, V_d, elimination rate, clearance, and daily area under the curve (AUC₂₄), were calculated. In addition, the effects of ARC and inflammation (by measuring CRP) on these parameters were investigated.

In summary, based on the results of our study, mechanical ventilation [with the settings in Table 1] did not significantly alter any of the pharmacokinetic parameters of vancomycin in our study population.

This finding contrasts with some previous research, such as the Medellín-Garibay et al. retrospective study, which reported a 20% reduction in vancomycin clearance.[10]

One of the important findings of this study was the significant impact of a condition called “Augmented Renal Clearance” (ARC), typically defined as a creatinine clearance ≥130 ml/min/1.73 m²,[33] observed in 23.1% of the study patients (9 of 39 patients). These patients showed statistically significantly higher elimination rate and clearance, shorter half-life, and lower trough concentration and AUC₂₄ levels compared with non-ARC patients [Table 2]. None of the ARC patients reached the desired therapeutic target (Trough = 10–15 mg/L) or daily AUC between 400 and 600 mg/h/L, (assuming a minimum inhibitory concentration (MIC) of 1 mg/L),[34] which increases the risk of treatment failure and development of drug resistance [Table 2].

Risk factors associated with ARC included age ≤50 years, male gender, and trauma.[35,36] In our study, all ARC patients were ≤50 years of age, 88.9% were male, and 88.9% were admitted with a diagnosis of trauma medications [Appendix 5].

Two main theories have been proposed to explain the mechanism of ARC: systemic inflammatory response syndrome and renal function reserve.[33] In this regard, in 88.9% of ARC patients, CRP levels were elevated, indicating active systemic inflammation. Furthermore, 44.4% of ARC patients had fluid overload of more than 10% of body weight [Appendix 5], suggesting that fluid therapy may play a key role in the occurrence of ARC.

After excluding ARC patients from the final analysis (to control for confounding effects), analysis was performed on non-ARC patients in Subgroups 4 and 5. Even in these groups, no significant differences in vancomycin pharmacokinetic parameters were observed between patients who were ventilated and those who were not [Appendices 6 and 7].

A possible explanation for the lack of an effect of mechanical ventilation on vancomycin pharmacokinetics could be related to the baseline status of the patients studied. Clearly, the ultimate clinical response to ventilator-induced pressure depends on the patient’s baseline cardiovascular status.[37,38] In patients with decompensated heart failure or pulmonary edema, PEEP can increase cardiac output by reducing left ventricular afterload. In contrast, in healthy subjects, this effect is negligible.[39,40]

There is also some evidence that high levels of PEEP (>10 cm H₂O) are associated with impaired cardiac, renal, and hepatic function compared to low levels.[38,40,41]

Therefore, in our study, due to the relatively good baseline condition of the patients, as well as the use of low PEEP (5 cm H₂O),[22] the ventilator had no significant effect on cardiac output and renal perfusion, and consequently, on the pharmacokinetics of vancomycin.

On the other hand, as long as PEEP and tidal volume are similar, different ventilator modes can have similar effects on patients.[42] Therefore, theoretically, regardless of the ventilator mode, if high-PEEP or higher tidal volumes are studied in the following patients, mechanical ventilation is likely to have a significant effect on the pharmacokinetics of various drugs, including vancomycin: (A) patients with hypervolemic heart failure and (B) patients with pulmonary edema.

Another finding of the study was the high percentage of subtherapeutic vancomycin concentrations, with only 17.9% of patients achieving the target AUC₂₄. In addition to ARC, failure to administer a loading dose and fluid overload could be other important reasons for failure to achieve adequate therapeutic levels.[18,43,44]

Further, using bivariate analysis, a statistically significant inverse correlation was observed between eGFR and parameters such as trough concentration, V_d, half-life, and AUC₂₄, and a significant direct correlation was observed with the elimination rate and clearance of vancomycin, which was to be expected [Table 5]. Furthermore, APACHE-II score was directly correlated with trough concentration, V_d, half-life, and inversely correlated with the elimination rate [Table 6]. The most interesting finding was a strong, statistically significant positive correlation between CRP level and the V_d of vancomycin [Table 7].

Table 6.

Correlation between acute physiology and chronic health evaluation II score at the time of trough sampling and vancomycin pharmacokinetic parameters in all patients (n=39)

Variable	r s	95% CI	P
APACHE II-score
Ctrough (mg/L)	0.402	0.090 to 0.643	0.011
Vd (L)	0.370	0.052 to 0.620	0.020
K (h−1)	−0.471	−0.690 to −0.173	0.002
t1/2 (h)	0.471	0.173 to 0.690	0.002

95% CI=Lower to Upper. A P≤0.05 is considered statistically significant. APACHE II=Acute physiology and chronic health evaluation II, C_trough=Trough plasma concentration, V_d=Volume of distribution, K=Elimination rate constant, t_1/2=Half-life, CI=Confidence intervals, r_s=Spearman’s correlation coefficient

The proposed mechanism for this correlation involves glycocalyx damage and endothelial dysfunction caused by increased inflammation, leading to increased capillary permeability. This causes fluid to shift from the intravascular compartment to the interstitial space, ultimately increasing the V_d of hydrophilic drugs like vancomycin.[44,45,46]

Overall, the incidence of AKI in our patients was low. Among those who experienced AKI, other nephrotoxic antibiotics were also administered, which could have contributed to the occurrence of AKI.

Finally, the present study showed that although mechanical ventilation did not have a significant effect on vancomycin pharmacokinetics in our study population and with the settings used in this study, other factors, such as ARC and systemic inflammation with the biomarker CRP, can have a significant effect on the pharmacokinetic parameters of this drug. These findings emphasize the importance of careful monitoring of renal function and inflammation levels in critically ill patients, as well as dose adjustment based on individual patient characteristics. It is suggested that in future studies, patients with ARC should be excluded from the outset unless ARC itself is the subject of the study.

There were several important limitations in our study:

1. Due to budget limitations and the expense of the vancomycin measurement kit, the study had a small sample size and was conducted at a single center

2. Because of the specific conditions of ICU patients, it was not possible to record exact weight values. In cases where the patient was unable to provide their weight, we used an estimated weight.[43] This is especially important for drugs like vancomycin, which are dosed based on the patient’s weight

3. Estimate GFR using serum creatinine and the Cockcroft-Gault formula. Although this method is widely used in clinical practice, it may lack the necessary accuracy and precision in ICU patients.[47] In addition, currently, estimating GFR through 24-h urine collection remains the best standard for identifying ARC; however, due to its inconvenience and time-consuming nature, it is not routinely used in ICUs.[48,49]

Despite the above limitations, our study’s strengths include a prospective design, proper matching of groups, effective control of confounding factors, and sampling at a clearly defined time point.

CONCLUSION

The final response to the effects of mechanical ventilation on the body depends on the patient’s baseline condition. Since our patients had relatively good cardiopulmonary and renal status and the ventilator settings were set to Low-PEEP, mechanical ventilation did not significantly impact the pharmacokinetic parameters of vancomycin.

Many patients had vancomycin AUC₂₄ below the target. A key reason for this is the high incidence of ARC among ICU patients.

To our knowledge, this study is the first to prospectively assess the impact of mechanical ventilation on vancomycin pharmacokinetic parameters in critically ill adult patients, and it also demonstrates a significant positive correlation between the inflammatory marker CRP and the vancomycin V_d.

This study provides a framework for further research into pharmacokinetics in specific populations of ICU patients and could help optimize dosing regimens and improve treatment outcomes.

Nevertheless, due to the small sample size, further prospective multicenter studies with larger samples should be conducted to inform pharmacokinetic modeling and confirm the role of CRP as a predictive biomarker in vancomycin pharmacokinetics.

AUTHORS’ CONTRIBUTIONS

Sh. A: Conceptualization, methodology, and supervision. F. Gh.: Methodology, investigation, and literature search. Sh. A., FGh: Data Curation, as well as data collection and sample preparation for laboratory tests. F. Gh., Sh. A., F. H., A. Gh., and A. N. contributed to the study recruitment and monitoring. R. A-N. Software and formal analysis. F. Gh. Writing-Original Draft. All authors have critically reviewed and approved the final version and agree that all aspects of the work have been done honestly.

Availability of data and material

The data are available from the corresponding author and can be obtained upon reasonable request.

Conflicts of interest

There are no conflicts of interest.

Appendix 1: Summary of demographic and clinical characteristics for all patients (n=39)

Characteristic	Total (n=39)
Age, years	45.0±16.1
Male sex, n (%)	34 (87.2)
Weight (kg)	75 (5–95)
BMI (kg/m2)	25.3±2.3
EF (%)	55 (4–60)
Ventilated/nonventilated (%)	16/23 (41/59)
Reason for ICU admission, n (%)
M-trauma/trauma	26 (66.7)
Others	13 (33.3)
APACHE II
Baseline	11 (5–22)
Cumulative fluid balance (percentage of body weight)*	8.5±7.7
Cumulative fluid balance, n (%)*
Peak
Negative	4 (10.3)
Even	20 (51.3)
Positive	15 (38.5)
Trough
Negative	5 (12.8)
Even	18 (46.2)
Positive	16 (41.0)
AKI incidence, n (%)	3 (7.7)
AKI Staging-KDIGO, n (%)
Stage 1	2 (5.1)
Stage 2	1 (2.6)
Stage 3	0
Mortality rate, n (%)
7-day ICU mortality	2 (5.1)
28-day hospital mortality	4 (10.3)

*Cumulative fluid balance until trough sampling time. Data are presented as number (%) for dichotomous variables, mean±SD for quantitative data with normal distribution and median (minimum–maximum) for nonnormal distribution. The sum of percentages may not equal to 100 due to rounding. M-Trauma=Multiple trauma, SD=Standard deviation, BMI=Body mass index, EF=Ejection fraction, ICU=Intensive care unit, APACHE II=Acute Physiology and Chronic Health Evaluation II, AKI=Acute kidney injury, KDIGO=Kidney Disease Improving Global Outcomes Guidelines

Appendix 2: Summary of baseline laboratory data for all patients (n=39)

Variable	Total (n=39)
Serum creatinine (mg/dL)	1.0 (0.5–1.6)
eGFR (mL/min/1.73 m2)	103.2±37.7
CRP (mg/dL)	56.8 (0.2–264.7)
Serum albumin (g/dL)	3.5±0.5
WBC ×103/μL	12.6 (4.5–26.5)
Total bilirubin (mg/dL)	0.9 (0.3–2.3)
Direct bilirubin (mg/dL)	0.3 (0.1–1.1)
AST (U/L)	55 (9–282)
ALT (U/L)	32 (11–256)
ALP (U/L)	141 (81–591)
Total protein (g/dL)	6.3 (4.4–8.6)

Data are presented as mean±SD for quantitative data with normal distribution and median (minimum–maximum) for nonnormal distribution. CRP (baseline or nearest to sampling day with a 3.4±2.6 day window). eGFR=Estimated glomerular filtration rate, WBC=White blood cell, AST=Aspartate aminotransferase, ALT=Alanine aminotransferase, ALP=Alkaline phosphatase, CRP=C-reactive protein, SD=Standard deviation

Appendix 3: Summary of demographic, clinical, and baseline laboratory data of the patients within sub-group 1

Characteristic	Total (n=32)	Ventilated (n=16)	Nonventilated (n=16)	P*
Age (years)	43.4±16.5	43.6±16.6	43.1±16.8	0.925a
Male sex, n (%)	30 (93.8)	15 (93.8)	15 (93.8)	1.000b
Weight (kg)	75 (59–93)	76.69±9.3	75 (60–93)	0.789c
BMI (kg/m2)	24.9±2.2	24.4±2.1	25.3±2.4	0.264a
EF (%)	55 (45–60)	53.0±4.8	54.5±3.6	0.445a
Reason for ICU admission, n (%)
M-trauma/trauma	22 (68.8)	12 (75.0)	10 (62.5)	0.446b
Others	10 (31.3)	4 (25.0)	6 (37.5)
Comorbidities, n (%)
No-comorbidity	15 (46.9)	8 (50.0)	7 (43.8)	0.723b
DM	5 (15.6)	3 (18.8)	2 (12.5)	0.626b
Asthma	4 (12.5)	2 (12.5)	2 (12.5)	1.000b
CVD	4 (12.5)	4 (25.0)	0	0.033b
HTN	3 (9.4)	2 (12.5)	1 (6.3)	0.544b
HLP	4 (12.5)	2 (12.5)	2 (12.5)	1.000b
GI	3 (9.4)	1 (6.3)	2 (12.5)	0.544b
Skin cancer	1 (3.1)	0	1 (6.3)	0.310b
APACHE II
Baseline	10.5 (5–22)	13.1±3.9	7.5 (5–22)	0.006c
Peak	9.7±5.6	14.5±2.9	4.9±2.6	<0.001a
Trough	9.5±5.8	14.2±3.9	4.9±3.0	<0.001a
GCS
Baseline	9.2±4.0	6.5 (3–14)	11.1±3.9	0.008c
Peak	10.5 (3–15)	5.5±1.6	15 (12–15)	<0.001c
Trough	10.5 (3–15)	5 (3–9)	15 (12–15)	<0.001c
MAP (mmHg)
Baseline	93.0±12.1	91.4±13.4	91 (77–118)	0.509c
Peak	87.9±9.7	85.9±9.6	90.0±9.6	0.243a
Trough	87.2±9.7	92 (81–102)	84.8±11.9	0.265c
Laboratory findings
Serum creatinine (mg/dL)	0.9±0.2	1.0±0.3	0.9±0.2	0.062a
eGFR (mL/min/1.73 m2)	104.6±38.2	95.2±26.6	114.1±45.9	0.168a
CRP (mg/dL)	58.8 (1.2–264.7)	56.7±25.2	56.9 (1.2–264.7)	0.713c
Serum albumin (g/dL)	3.5 (2.8–4.9)	3.7±0.6	3.2 (2.8–4.3)	0.069c
WBC ×103/μL	12.2 (4.5–26.5)	14.6±6.1	11.6 (4.5–25.1)	0.429c
Total bilirubin (mg/dL)	1.0±0.4	1.0±0.4	0.9 (0.3–2.2)	0.910c
Direct bilirubin (mg/dL)	0.3 (0.1–1.1)	0.4±0.1	0.3 (0.1–1.1)	0.734c
AST (U/L)	61.5 (9–282)	72 (17–282)	59.2±33.5	0.250c
ALT (U/L)	38.5 (11–256)	39 (14–256)	36 (11–104)	0.985c
ALP (U/L)	145 (81–591)	138.8±49.6	174 (81–591)	0.036c
Total protein (g/dL)	6.4±1.0	6.4±1.2	6.4±1.0	0.925a

*Comparison between two groups: Ventilated versus Nonventilated, ^aIndependent-samples t-test, ^bChi-square test, ^cMann–Whitney U-test. A P≤0.05 is considered statistically significant. Data are presented as number (%) for dichotomous variables, mean±SD for quantitative data with normal distribution, and median (minimum–maximum) for nonnormal distribution. The sum of percentages may not equal to 100 due to rounding. CRP level (baseline or nearest to sampling day with a 2.9±2.3 day window). BMI=Body mass index, EF=Ejection fraction, ICU=Intensive care unit, M-trauma=Multiple trauma, DM=Diabetes mellitus, CVD=Cardiovascular disease, HTN=Hypertension, HLP=Hyperlipidemia, GI=Gastrointestinal diseases, APACHE II=Acute physiology and chronic health evaluation II, GCS=Glasgow coma scale, MAP=Mean arterial pressure, eGFR=Estimated glomerular filtration rate, WBC=White blood cell, AST=Aspartate aminotransferase, ALT=Alanine aminotransferase, ALP=Alkaline phosphatase, SD=Standard deviation, CRP=C-reactive protein

Appendix 4: Summary of patients’ clinical characteristics at the time of sampling within sub-group 1

Characteristic	Ventilated group (n=16)	Nonventilated group, (n=16)	P*
The number of vancomycin injections (until sampling), n	5 (5–10)	5 (5–11)	0.265a
ICU length of stay (until sampling), days	6 (3–26)	9±4.4	0.092a
MV duration (until sampling), days	5.5 (3–26)	-	-
Cumulative Fluid balance until trough sampling time (percentage of body weight)	10.7±5.7	8.5±9.2	0.421b
Cumulative fluid balance, n (%)
Peak
Negative	0	2 (12.5)
Even	8 (50.0)	8 (50.0)
Positive	8 (50.0)	6 (37.5)
Trough
Negative	0	3 (18.8)
Even	8 (50.0)	6 (37.5)
Positive	8 (50.0)	7 (43.8)
Nutritional support, n (%)
Peak
PO	0	10 (62.5)	<0.001c
NPO	5 (31.3)	5 (31.3)
Gavage	11 (68.8)	1 (6.3)
Trough
PO	0	12 (75.0)	<0.001c
NPO	4 (25.0)	4 (25.0)
Gavage	12 (75.0)	0

*Comparison between two groups: Ventilated versus nonventilated, ^aMann–Whitney U-test, ^bIndependent-samples t-test, ^cPearson’s Chi-square test. A P≤ 0.05 is considered statistically significant. Data are presented as number (%) for dichotomous variables, mean±SD for quantitative data with normal distribution, and median (minimum–maximum) for nonnormal distribution. The sum of percentages may not equal to 100 due to rounding. SD=Standard deviation, ICU=Intensive care unit, MV=Mechanical ventilation, PO=Per OS (Latin term): “Through the mouth” or “by mouth,” NPO=Nil per os (Latin term): Nothing by mouth

Appendix 5: Summary of demographic, clinical, and laboratory data of the patients within sub-Group 3

Characteristic	ARC group (n=9)
Age (years)	36.5±7.1 37 (29–48)
Male sex, n (%)	8 (88.9)
eGFR (mL/min/1.73 m2)	155.7±26.5
CRP (mg/dL)	62.1±32.3
CRP, elevated (>10 mg/dL)	8 (88.9) Missing: 1 (11.1)
Reason for ICU admission, n (%)
Multiple-trauma/trauma	8 (88.9)
Others	1 (11.1)
Cumulative fluid balance until trough sampling time (percentage of body weight)	9.4±8.6
Cumulative fluid balance, n (%)
Peak/trough
Negative	1 (11.1)
Even	4 (44.4)
Positive	4 (44.4)

Data are presented as number (%) for dichotomous variables, mean±SD for quantitative data with normal distribution, and median (minimum–maximum) for nonnormal distribution. The sum of percentages may not equal to 100 due to rounding. ARC=Augmented renal clearance, eGFR=Estimated glomerular filtration rate (baseline), CRP=C-reactive protein, ICU=intensive care unit, SD=Standard deviation

Appendix 6: Comparison of vancomycin pharmacokinetic parameters between ventilated and nonventilated groups within patient sub-Group 4

Variable	Ventilated group (n=14)	Nonventilated group (n=16)	P*
Cpeak (mg/L)	21.21±4.28	21.16±6.10	0.982a
Ctrough (mg/L)	11.05±5.23	8.25 (5.00–18.00)	0.631b
Vd (L)	115.52±35.06	112.82±45.98	0.859a
K (h−1)	0.0608±0.0259	0.0666±0.0335	0.601a
t1/2 (h)	10.88 (6.79–30.58)	12.37±4.93	0.819b
Clearance (L/h)	6.37±1.98	6.53±1.85	0.822a
AUC24 (mg/h/L)	371.22±115.41	347.40±80.60	0.513a

*Comparison between two groups: Ventilated versus nonventilated, ^aIndependent-samples t-test, ^bMann–Whitney U-test. A P≤0.05 is considered statistically significant. Data are presented as mean±SD for quantitative data with normal distribution and median (minimum–maximum) for nonnormal distribution. C_peak=Peak plasma concentration, C_trough=Trough plasma concentration, V_d=Volume of distribution, K=Elimination rate constant, t_½=Half-life, AUC₂₄=Daily area under the curve, SD=Standard deviation

Appendix 7: Comparison of vancomycin pharmacokinetic parameters between ventilated and nonventilated groups within patient sub-Group 5

Variable	Ventilated group (n=14)	Nonventilated group (n=14)	P*
Cpeak (mg/L)	21.21±4.28	19.76±4.30	0.380a
Ctrough (mg/L)	11.05±5.23	8.35 (7.00–18.00)	0.890b
Vd (L)	115.52±35.06	121.77±41.55	0.671a
K (h−1)	0.0608±0.0259	0.0576±0.0188	0.710a
t1/2 (h)	10.88 (6.79–30.58)	13.31±4.47	0.765b
Clearance (L/h)	6.37±1.98	6.60±1.98	0.765a
AUC24 (mg/h/L)	371.22±115.41	344.17±85.69	0.488a

*Comparison between two groups: Ventilated versus nonventilated, ^aIndependent-samples t-test, ^bMann–Whitney U-test. A P≤0.05 is considered statistically significant. Data are presented as mean±SD for quantitative data with normal distribution and median (minimum–maximum) for nonnormal distribution. C_peak=Peak plasma concentration, C_trough=Trough plasma concentration, V_d=Volume of distribution, K=Elimination rate constant, t_1/2=Half-life, AUC₂₄=Daily area under the curve, SD=Standard deviation

Appendix 8: Comparison of acute kidney injury incidence and mortality between ventilated and nonventilated groups within patient sub-group 1

Characteristic	Total (n=32), n (%)	Ventilated (n=16), n (%)	Nonventilated (n=16), n (%)	P*
AKI incidence, n (%)	3 (9.4)	2 (12.5)	1 (6.3)	0.544a
AKI staging-KDIGO, n (%)
Stage 1	2 (6.3)	1 (6.3)	1 (6.3)
Stage 2	1 (3.1)	1 (6.3)	0
Stage 3	0	0	0
Mortality rate, n (%)
7-day ICU mortality	2 (6.3)	2 (12.5)	0	0.144a
28-day hospital mortality	3 (9.4)	2 (12.5)	1 (6.3)	0.544a

*Comparison between two groups: Ventilated versus nonventilated, ^aChi-square test. A P≤0.05 is considered statistically significant. Data are presented as numbers (%) for dichotomous variables. The sum of percentages may not equal to 100 due to rounding. AKI=Acute kidney injury, KDIGO=Kidney Disease Improving Global Outcomes Guidelines, ICU=Intensive care unit

Funding Statement

The results are from the clinical pharmacy specialist thesis of Farnia Ghanbarveisi. This study was financially supported by a grant from the Research Council of Mazandaran University of Medical Sciences. The authors thank the Vice Chancellor for Research and Technology at Mazandaran University of Medical Sciences for their financial support. This study is an independent research project, and the funding source had no influence on data collection, analysis, or publication.