Clinical benefits of therapeutic drug monitoring of vancomycin therapy in patients with postoperative intracerebral hemorrhage: a retrospective cohort study

Abstract

Objective

The objective of this study was to investigate the clinical efficacy and safety of conducting therapeutic drug monitoring (TDM) of vancomycin in patients with postoperative intracerebral haemorrhage.

Methods

We conducted a retrospective analysis of 435 patients who experienced postoperative cerebral haemorrhage and were treated with vancomycin in the Department of Neurosurgery of Inner Mongolia Autonomous Region People’s Hospital from January 2017 to December 2021. Patients were then matched using the propensity score matching method in a ratio of 1:1. Ninety-two pairs of cases were successfully matched, and the data before and after performing vancomycin TDM were analysed.

Results

After PSM, the baseline data of the two groups were balanced. There were no significant differences in the 14-day mortality and length of hospital stay (p>0.05) between the two groups. Compared with the non-TDM group, the TDM group had a higher proportion of patients with normal white blood cells (83.7% vs 56.5%, p=0.000), neutrophil count (57.6% vs 25.0%, p=0.000) and attaining desirable reductions of 80% in procalcitonin (65.2% vs 10.9%, p=0.000) and C-reactive protein (78.3% vs 41.3%, p=0.000) levels. At US$15.82 per additional TDM, TDM significantly promoted patient outcomes, as seen in improvements in the proportion of patients attaining desirable levels of white blood cells, neutrophil count, procalcitonin and C-reactive protein.

Conclusions

Vancomycin TDM is a safe and effective approach for the treatment of patients with postoperative intracerebral haemorrhage. The empirical use of TDM of vancomycin significantly improved normal values of white blood cells and neutrophil count, achieved desirable reductions of 80% in procalcitonin and C-reactive protein, and reduced nephrotoxicity in patients with postoperative intracerebral haemorrhage.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Vancomycin is the first choice for methicillin-resistant Staphylococcus aureus (MRSA) infections worldwide and is widely used for the treatment of MRSA infections.

• Studies have shown that vancomycin therapeutic drug monitoring (TDM) can significantly improve the efficacy of treatment and reduce nephrotoxicity. Guidelines recommend TDM for the administration of vancomycin in specific populations.

• Many studies have reported that vancomycin TDM is necessary in elderly patients, children, critically ill patients with sepsis and those with renal impairment, but there is no research to support the need for vancomycin TDM for empirical anti-MRSA therapy in patients with postoperative cerebral haemorrhage.

WHAT THIS STUDY ADDS

• In patients with postoperative cerebral haemorrhage, vancomycin TDM can be of clinical benefit in terms of reduced inflammatory markers. At the same time, monitoring blood levels in this population can prevent the development of nephrotoxicity.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• This is a real-world study and the results are true and reliable, providing research-based evidence for the individualised use of vancomycin in patients with postoperative intracerebral haemorrhage. The findings may have a profound impact on optimising treatment planning and the determination of guidelines.

Introduction

Intracerebral haemorrhage is non-traumatic haemorrhage caused by vascular disruption of the brain parenchyma.¹ It is frequently treated surgically, but infection of the central nervous system is a common postoperative complication with infection rates ranging from 1.00% to 19.18%.² Postoperative intracranial infection affects the patient’s prognosis and can even lead directly to death.³ Rapid identification, medical treatment and neurosurgical management are essential to promote recovery in patients with intracerebral haemorrhage. Therapeutic interventions for intracerebral haemorrhage focus on arresting haemorrhage expansion—for example, by improved population blood pressure control, personalised use of antithrombotics and other coagulation complexes and reduction of clot volume in intraventricular and parenchymal haematomas using minimally invasive techniques. Empirical anti-infective therapy is essential in patients with postoperative cerebral haemorrhage. Positive cocci are the main pathogens of postoperative intracranial infection,⁴ and vancomycin can effectively remove such bacteria. Clinicians empirically select vancomycin for the treatment of patients with postoperative intracerebral haemorrhage.⁵ Empirically selected vancomycin for control of infections caused by gram-positive bacteria after neurosurgical procedure is effective.⁶ Vancomycin is one of the most commonly used antibacterial agents for patients with intracerebral haemorrhage after surgery and has an important role in the prevention and treatment of common postoperative infections in patients with postoperative intracerebral haemorrhage.

However, vancomycin has a narrow therapeutic window and its efficacy is closely related to safety and low blood levels. Therapeutic drug monitoring (TDM) is important for vancomycin in order to ensure the effective and safe use of the drug.⁷ Studies have been conducted to determine the optimal dosing regimen based on patient pharmacokinetic parameters in combination with TDM to optimise exposure to vancomycin, reduce toxicity, and ensure drug efficacy and safety. The need for TDM of vancomycin is controversial,⁸ because some hospitals believe that performing vancomycin blood concentration determinations take a long time and have limited ability to guide clinical treatment, as well as increasing certain medical expenses and imposing a heavy financial burden on patients.⁹ However, some healthcare institutions believe that TDM can improve the clinical efficacy of vancomycin and reduce nephrotoxicity. Vancomycin TDM is also challenging. It requires the involvement of pharmacists, physicians, nurses and laboratory staff as well as complex procedures, making it impractical for clinical use in most hospitals. The consensus review by the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, the Paediatric Infectious Diseases Society and the Society of Infectious Diseases Pharmacists and also the consensus review by the Japanese Society of Chemotherapy and the Japanese Society of Therapeutic Drug Monitoring all recommend monitoring vancomycin serum concentrations to minimise nephrotoxicity and maximise efficacy.^{10 11}

Studies have shown that vancomycin TDM is necessary in elderly patients, children, critically ill patients with sepsis and those with renal impairment.^12–14 In these studies, TDM significantly increased the rate of clinical efficacy and decreased the rate of nephrotoxicity in patients treated with vancomycin. Although the guidelines state that anti-infective treatment with vancomycin is necessary in patients with cerebral haemorrhage, the empirical use of vancomycin using TDM has not been studied. Few data are available regarding the clinical benefits of TDM of vancomycin therapy in patients with postoperative cerebral haemorrhage. We therefore conducted a retrospective study to investigate the clinical benefits of TDM of vancomycin in the treatment of patients with postoperative intracerebral haemorrhage.

Methods

Study design and patients

This single-centre retrospective cohort study collected clinical data from patients with postoperative intracerebral haemorrhage treated with vancomycin between 1 January 2017 and 31 December 2021 at the Department of Neurosurgery of the Inner Mongolia Autonomous Region People’s Hospital (figure 1). For this study, we included a subset of patients with postoperative intracerebral haemorrhage due to small vessel diseases: deep perforator arteriopathy (hypertensive arteriopathy) or cerebral amyloid angiopathy. The patients were diagnosed as having postoperative intracerebral haemorrhage using a baseline severity score recommended by the American Heart Association/American Stroke Association.¹⁵ All patients with intracerebral haemorrhage in our study received surgical treatment. Other medical managements included treatment of coagulopathy, blood pressure, cerebral oedema and empirical anti-infection. The inclusion criteria were: (1) age >18 years; (2) treatment with vancomycin for >3 days in patients with postoperative intracerebral haemorrhage; and (3) patients admitted to the Department of Neurosurgery who underwent neurosurgery. Patients who were: (1) exposed to intravenous vancomycin within the last 7 days, (2) were pregnant, or (3) had incomplete patient records were excluded.

Figure 1.

Flow chart of the study. TDM, therapeutic drug monitoring.

The study was approved by the Ethics Committee of the Inner Mongolia Autonomous Region People’s Hospital and informed consent was not required because it was not an interventional study.

Vancomycin blood level monitoring

We performed vancomycin TDM based on vancomycin troughs. The vancomycin trough target was 15–20 mg/L, and individualised dosing was carried out according to the vancomycin assay. After being given four doses of vancomycin to achieve steady-state blood concentrations (48 hours after dosing or dose change in those with normal renal function and 72 hours after dosing or dose change in those with abnormal renal function), serum was collected from the patients within 30 min before the next dose. Vancomycin blood concentrations were determined using liquid chromatography/mass spectrometry (UPLC-MS/MS; Waters).

Dose evaluation criteria

Vancomycin is commonly used as anti-infective treatment for patients with postoperative cerebral haemorrhage. Due to severe infections in patients with postoperative cerebral haemorrhage, the physician empirically gives a loading dose of 30 mg/kg for the first administration,¹⁶ followed by a standard dose of 15–20 mg/kg. In addition, for obese patients, weight-based administration is used. In this study, vancomycin was not the only antimicrobial agent used and may be combined with other antimicrobials, but vancomycin was the only one used against methicillin-resistant Staphylococcus aureus (MRSA) infections.

Vancomycin was given as a 60 min intravenous injection in patients with normal renal function at a loading dose of 30 mg/kg (1.5 g) and a maintenance dose of 20 mg/kg (1 g) every 12 hours. In patients with renal dysfunction, the dose was adjusted according to recommendations in the Chiang Mai University Hospital (CMUH) manual. The CMUH protocol was initiated by pharmacists and infectious disease physicians, written in the CMUH Antibiotic Manual and recommended for use in hospitals. In patients with renal dysfunction, the vancomycin dosage was adjusted based on their kidney function (creatinine clearance (CLCR) ≥50 mL/min: 1 g every 12 hours; CLCR 40–49 mL/min: 750 mg every 12 hours; CLCR <40 mL/min: 750 g every 48–120 hours).¹⁷

In our study, both groups were tested regularly during the course of treatment. At the time of admission the physician empirically treated the patient with vancomycin according to the patient’s individual circumstances. During the treatment, patients in both groups were monitored for the same clinical indicators including inflammatory factors such as procalcitonin (PCT) and C-reactive protein (CRP), and liver and kidney function such as serum creatinine and urea nitrogen. Patients with severe infection, renal insufficiency or instability, morbid obesity and advanced age were closely monitored during follow-up treatment and TDM was performed if needed to ensure the safety and efficacy of the drug. Renal function tests and complete blood counts were also performed regularly to help monitor the patient’s response to the medication and to assess nephrotoxicity. Clinicians assessed nephrotoxicity by asking patients about their condition in daily check-ups combined with liver and kidney function index tests, and assessed ototoxicity by communicating with them and testing their hearing. Meanwhile, nurses also checked daily to manage the patients. In addition, if the patient had been judged to have developed acute kidney injury related to vancomycin, the drug was promptly discontinued. If renal injury was severe, rapid reduction in vancomycin blood levels and reduced damage to the kidneys can be achieved by continuous renal replacement therapy. In patients with insufficient intravascular volume, appropriate rehydration was given before and during treatment with vancomycin: prevention of hypoproteinaemia in both groups (maintaining serum albumin levels at 25 g/L if possible), avoidance of potentially nephrotoxic drugs when combining vancomycin with other drugs in both groups, and avoidance of continuous dosing for more than 7 days if possible, in addition to the TDM group. Vancomycin overdose was avoided through TDM and dose adjustment (for adult patients with severe MRSA infection, a target vancomycin trough concentration of 10–20 mg/L was recommended, and when the minimum inhibitory concentration of vancomycin was >1 mg/L, a switch to another anti-MRSA agent was considered based on clinical assessment).^{18 19}

Data collection

Data from patient medical records and hospital databases were collected. The data recorded age, sex, comorbidities, cardiac disease, diabetes, hypertension, pneumonia, and severity of disease (Sequential Organ Failure Assessment (SOFA) score) for all included patients. We also collected details of the site of infection, duration of vancomycin treatment, length of hospital stay and laboratory data including leucocyte count, platelet count, neutrophil, PCT, CRP and serum creatinine.

Defining outcome measures

The primary endpoint was 14-day mortality, which was determined as all-cause mortality. The secondary outcomes included the proportion of patients reaching normal values of inflammatory factors, those with nephrotoxicity and length of hospital stay after vancomycin treatment.

The recommended treatment duration for vancomycin was 7–14 days, hence the mortality rate at 14 days was analysed. In many clinical trials, treatment efficacy has been assessed by changes in inflammatory indicators in patients. International expert consensus indicates that a fall in PCT by at least 80% from the peak level is a strong indicator for resolution of illness and is used as a stopping threshold.²⁰ A multicentre randomised clinical trial also used a decrease in PCT concentration by 80% of its peak value as the relative stopping threshold.²¹ Therefore, in our study, a reduction of 80% from the peak values of PCT was used as the threshold value. We consulted the guidelines and found that there were no relevant studies on how much CRP was reduced from the peak to basal level, but PCT and CRP have some similarity as inflammatory factors so we referred to the guideline for PCT and used an 80% reduction in CRP as the threshold value. Thus, in this study, normal values of inflammatory factors were defined as the normal human values of white blood cell count and neutrophil ratio, and a reduction by 80% in PCT and CRP compared with the patient’s basal values. Development of nephrotoxicity was defined in line with the Infectious Diseases Society of America Vancomycin Consensus Statement as a raised serum creatinine concentration of ≥0.5 mg/dL or a serum creatinine concentration at least 1.5 times higher than baseline.¹⁰

Sample size estimation

Sample size was measured based on 14-day mortality after vancomycin TDM. According to previous studies in the literature, mortality rates were determined after reviewing the literature: 10% for TDM and 30% for non-TDM. According to the overall mortality sample size estimation formula (Equation 1), the sample size required for the study was calculated while considering the occurrence of data outliers and ensuring the integrity of the sample, and the final number of people required for each group was determined to be 59.

$${\displaystyle n=\frac{10.51[(R+1)-P_2(R^2+1)]}{P_2(1-R)}}$$ (1)

where n is sample size, P represents the exposure rate of the study factors in the population, and R represents the HR, which was the estimated relative risk or ratio of exposure (P₁/P₂) for each study factor, with α value of 0.05.

Statistical analysis

To reduce selection bias between the TDM and non-TDM groups, propensity score matching (PSM) was performed for a total of 290 patients between the two groups. Patients were matched 1:1 in the two groups by a nearest neighbour method with a width of 0.02 of the SD of the logit of the propensity score and without any replacement. The propensity score was calculated based on age, sex, comorbidities such as diabetes, hypertension, pneumonia, SOFA score and duration of vancomycin use.

Continuous variables were presented as mean±SD or median (IQR) and categorical variables were reported as numbers and percentages. Differences between the baseline characteristics of the TDM and non-TDM groups were analysed with t-tests or Mann–Whitney U tests for continuous variables and Pearson’s χ² test or Fisher’s exact test for categorical variables. Multiple logistic regression analysis was performed using 14-day discharge (death/survival) as the dependent variable and age, sex, SOFA score, length of hospital stay and the proportion of patients reaching normal values of white blood cell count and neutrophil ratio and reductions by 80% in PCT and CRP compared with the patient’s base values. The significance level was set at p<0.05. All statistical tests were performed using SPSS 26.0 (SPSS Inc).

Results

Patient characteristics

A total of 458 patients with postoperative intracerebral haemorrhage were included in the study. Twenty-three patients who did not meet the infection criteria and were missing substantial data were excluded, and 435 patients were finally included in the full cohort (138 patients in the TDM group and 297 patients in the non-TDM group) (figure 1). The characteristics of the patients are summarised in table 1. Before matching, a total of 276 of the 435 patients (63.45%) were men and the mean±SD age was 58.12±11.39 years. The metrics at baseline were comparable except for CRP, PCT and vancomycin treatment duration. After PSM, although exact matching could not be obtained in every case, there were 92 pairs of cases with fuzzy matching. All the baseline data of the 92 pairs of successfully matched cases were balanced between the two groups (p>0.05, table 1). In addition, in the TDM group the mean vancomycin blood concentration measured was 9.16±1.43 mg/L. The mean±SD daily dose of vancomycin was 1.43±0.61 g in the TDM group and 1.64±0.63 g in the non-TDM group.

Table 1.

Clinical features before and after matching

Features	All patients			Propensity-matched patients
	TDM (n=138)	Non-TDM (n=297)	P value	TDM (n=92)	Non-TDM (n=92)	P value
Age, years, median (IQR)	58 (50,63)	57 (51,67)	0.874	59 (49.25,63)	57 (49.25,63)	0.947
Gender, n (%)
Men	93 (67.4)	183 (61.6)	0.244	61 (66.3)	56 (60.9)	0.444
Women	45 (32.6)	114 (38.4)		31 (33.7)	36 (39.1)
SOFA, points, median (IQR)	6 (5,7)	6 (5,8)	0.699	6 (5,7)	6 (5,7)	0.397
Hypertension, n (%)	84 (60.9)	222 (74.7)	0.003	63 (68.5)	69 (75.0)	0.326
Diabetes, n (%)	18 (13.0)	18 (6.1)	0.014	7 (7.6)	13 (14.1)	0.155
Pneumonia, n (%)	21 (15.2)	69 (23.2)	0.055	15 (16.3)	19 (20.7)	0.447
Infection type, n (%)
Pulmonary	65 (47.1)	150 (50.5)	0.262	46 (50.0)	46 (50.00)	0.248
Intracranial	31 (22.5)	78 (26.3)		18 (19.6)	26 (28.3)
Pulmonary and intracranial	42 (30.4)	69 (23.2)		28 (30.4)	20 (21.7)
WBC, ×109 cells/L, median (IQR)	13.41 (10.72,17.11)	13.62 (10.15,20.17)	0.495	13.26 (10.72,17.09)	11.85 (9.17,17.26)	0.744
PLT, ×109 cells/L, median (IQR)	252.50 (185.75,303)	234.00 (185,307)	0.799	263 (192,303)	205 (173,285)	0.331
NEUT, %, median (IQR)	83.75 (78.5,89.1)	84.6 (80.1,87.9)	0.441	83.30±7.06	83.98±7.83	0.745
CRP, mg/L, median (IQR)	80.96 (53.74,168.73)	37.35 (15.05,87.03)	0.000	97.72 (44.92,192.83)	49.08 (12.48,98.37)	0.060
PCT, ng/L, median (IQR)	0.41 (0.28,1.33)	0.30 (0.11,0.61)	0.000	0.46 (0.29,1.32)	0.34 (0.099,0.62)	0.090
Albumin, g/L, median (IQR)	34.1 (31.15,37.3)	33.30 (29.80,37.3)	0.321	34.2 (31.6,37.6)	34 (31.3,36.9)	0.730
Serum creatinine, μmol/L, median (IQR)	55.25 (40.88,68.43)	52.00 (43.00,62.2)	0.424	58.4 (43.6,69.1)	56.1 (47.2,60.1)	0.554
CKD stage, n (%)			0.582	–		0.236
CKD1	90 (65.22)	200 (67.34)	–	68 (73.91)	66 (71.74)	–
CKD2	33 (23.91)	71 (23.91)	–	18 (19.57)	24 (26.09)	–
CKD3	12 (8.70)	24 (8.08)	–	6 (6.52)	2 (2.17)	–
CKD4	3 (2.17)	2 (0.67)	–	0	0	–
CKD5	0	0	–	0	0	–
Vancomycin days, median (IQR)	12.5 (10,18.25)	9 (7.00,12.00)	0.000	12 (8,14)	11.5 (7,15)	0.997

Statistically significant values are shown in bold.

CKD, chronic kidney disease; CRP, C-reactive protein; NEUT, neutrophil; PCT, procalcitonin; PLT, platelet; SOFA, Sequential Organ Failure Assessment; TDM, therapeutic drug monitoring; WBC, white blood cell.

Comparison of clinical outcomes in the two groups

In patients with empirically selected vancomycin for postoperative intracerebral haemorrhage, the rates achieved for inflammatory factors and for nephrotoxicity were meaningful in both groups (table 2). Compared with the non-TDM group, patients in the TDM group had higher percentages of reduction to normal values for WBC (83.7% vs 56.5%, p=0.000), NEUT (57.6% vs 25.0%, p=0.000) and attained reductions of 80% in values of PCT (65.2% vs 10.9%, p=0.000) and CRP (78.3% vs 41.3%, p=0.000). Likewise, the incidence of vancomycin-related nephrotoxicity was lower in the TDM group (6.5% vs 17.4%, p=0.023). However, there was no significant difference in 14-day mortality (p=0.0.351) and length of hospital stay (p=0.135) between the two groups. Multifactorial logistic regression analysis showed the proportion of patients attaining an 80% reduction in CRP levels compared with baseline (OR 0.165, 95% CI 0.030 to 0.892, p=0.036). Likewise, length of hospital stay also affected 14-day mortality (OR 6.904, 95% CI 1.074 to 44.401, p=0.042). It was concluded that CRP and length of hospital stay were factors influencing 14-day mortality in patients with postoperative intracerebral haemorrhage treated with vancomycin (table 3).

Table 2.

Comparison of clinical outcomes between the two groups after propensity score matching analysis

Variables	TDM (n=92)	Non-TDM (n=92)	OR	95% CI	P value
WBC attainment rate, n (%)	77 (83.7)	52 (56.5)	0.253	0.127 to 0.553	0.000
NEUT attainment rate, n (%)	53 (57.6)	23 (25.0)	0.245	0.131 to 0.459	0.000
PCT 80% reduction, n (%)	60 (65.2)	10 (10.9)	0.065	0.030 to 0.142	0.000
CRP 80% reduction, n (%)	72 (78.3)	38 (41.3)	0.195	0.102 to 0.373	0.000
PLT, ×109 cells/L, median (IQR)	276 (243,313)	344 (223,441)	–	–	0.830
Albumin, g/L, median (IQR)	31.45 (30.6,35.95)	33.15 (29.83,36.60)	–	–	0.376
Nephrotoxicity, n (%)	6 (6.5)	16 (17.4)	3.018	1.124 to 8.102	0.023
14-day mortality, n (%)	4 (4.34)	7 (7.6)	1.812	0.512 to 6.414	0.351
Length of hospital stay, days, median (IQR)	30 (28,40)	27 (20,36)	–	–	0.135

*Statistically significant values are shown in bold.

CRP, C-reactive protein; NEUT, neutrophils; PCT, procalcitonin; PLT, platelets; TDM, therapeutic drug monitoring; WBC, white blood cells.

Table 3.

Logistic regression analysis for 14-day mortality

Covariate	OR (95% CI)	P value
Age	0.980 (0.916 to 1.050)	0.569
Sex	1.431 (0.211 to 9.724)	0.714
WBC	1.501 (0.256 to 8.814)	0.653
NEUT	0.186 (0.026 to 1.327)	0.093
PCT	0.640 (0.088 to 4.636)	0.658
CRP	0.165 (0.030 to 0.892)	0.036
SOFA	5.042 (0.378 to 67.206)	0.221
Length of hospital stay	6.904 (1.074 to 44.401)	0.042

Statistically significant values are shown in bold.

CRP, C-reactive protein; NEUT, neutrophils; PCT, procalcitonin; SOFA, Sequential Organ Failure Assessment; WBC, white blood cells.

Cost-effectiveness

TDM for vancomycin may be an economically choice for patients with postoperative intracerebral haemorrhage treated with vancomycin, and the cost-effectiveness analysis is shown in table 4. The average total direct medical costs in the TDM group were lower than in the non-TDM group (US$19 803.20±8713.93 vs US$20 321.99±13 643.11). Although the cost of each additional TDM was US$15.82, vancomycin TDM significantly promoted patient outcomes, as seen in improvements in the proportion of patients attaining desirable outcomes on WBC, NEUT, PCT and CRP. The incremental cost effective ratios (ICERs) of WBC, NEUT, PCT and CRP were 3.46, 3.46, 1.95 and 3.13, respectively. Thus, it was more economical to perform vancomycin TDM in patients with postoperative intracerebral haemorrhage.

Table 4.

Comparison of cost-effectiveness between the two groups

Group	Total direct medical costs (US$) Mean±SD	C (US$)	ICER-WBC (△C/△E)	ICER-NEUT (△C/△E)	ICER-PCT (△C/△E)	ICER-CRP (△C/△E)
TDM	19803.20±8713.93	15.82	3.46	3.46	1.95	3.13
Non-TDM	20321.99±13 643.11	0	–	–	–	–

CRP, C-reactive protein; ICER, incremental cost effective ratio; NEUT, neutrophil; PCT, procalcitonin; TDM, therapeutic drug monitoring; WBC, white blood cell.

Discussion

Patients with postoperative intracerebral haemorrhage often have the risk of central infection by MRSA. In the USA, 53% of the Staphylococcus aureus isolated in the clinical setting was MRSA.²² Empirical vancomycin has generally been chosen for the anti-infective treatment of patients with postoperative intracerebral haemorrhage. Vancomycin has an important role in the prevention and treatment of common postoperative infections in neurosurgery. Meta-analysis has shown that TDM significantly improves the clinical outcome of patients with Gram-positive infections treated with vancomycin (OR 2.62, 95% CI 1.34 to 5.11, p=0.005).⁸ Guidelines recommend that TDM of vancomycin is used clinically in neonates and in patients with obesity, renal dysfunction and those receiving prolonged courses or continuous infusion with vancomycin.¹⁰ However, our study suggests that vancomycin TDM is beneficial in the routine treatment of patients with postoperative intracerebral haemorrhage.

In our study, after vancomycin treatment the proportion of patients in whom WBC and NEUT levels were reduced to normal was higher in the TDM group than in the non-TDM group (P _WBC = 0.000, P _NEUT = 0.000). This was consistent with the results of Maulina et al,²³ who found that hospitals implementing TDM protocols achieved better treatment outcomes than hospitals without TDM. Our results showed that more patients in the TDM group had a fall in PCT from the peak level of at least 80% than in the non-TDM group (65.2% vs 10.9%, p=0.000). A retrospective case–control study showed that, in patients with severe infections, vancomycin TDM had a higher overall efficacy rate than non-TDM (95.16% vs 91.94%), with statistically significant PCT in the TDM group (p<0.05).²³ We also conducted an analysis of the reduction in CRP levels by at least >80% from the peak, amd found that more patients reached this threshold in the TDM group than in the non-TDM group (78.3% vs 41.3%, p=0.000). Although there are no guidelines recommending a fall in CRP levels by at least >80% from the peak as an indicator of improvement in treatment outcome, this value can be used as a reference for future studies for further exploration.

Since 90% of vancomycin is excreted by the kidneys, an overdose could lead to accumulation of the drug, which in turn could lead to kidney injury.²⁴ In adult patients with severe infection by MRSA, a target vancomycin serum trough concentration of 15.0–20.0 mg/L should be maintained.¹⁰ The incidence of nephrotoxicity reached 29.6% at vancomycin trough concentrations >15 mg/L.²⁵ The risk of nephrotoxicity does not become significant until the trough concentration is >20 mg/L.¹⁰ Vancomycin TDM was performed to help determine the optimal dose and guide rational clinical use to avoid the occurrence of vancomycin nephrotoxicity. In our study, the nephrotoxicity results in both groups showed that the incidence of nephrotoxicity was significantly higher in the non-TDM group than in the TDM group (17.4% vs 6.5%, p=0.023). Iwamoto et al ²⁶ also reported a lower incidence of nephrotoxicity in the TDM group than in the non-TDM group (8.2% vs 18%). Meta-analyses showed a significantly lower incidence of nephrotoxicity in the TDM group, and subgroup analyses indicated that TDM was associated with a lower incidence of nephrotoxicity.²⁷ Thus, vancomycin TDM is used to monitor the trough plasma concentration to control the drug dose in order to maximise clinical efficacy while minimising the risk of nephrotoxicity, and to select the most appropriate dose for individualised drug therapy based on the patient’s individual circumstances to improve the safety, rationality and feasibility of treatment.

In addition, in our study the survival rate was higher in the TDM group than in the non-TDM group (95.66% vs 92.40%). Meta-analysis showed no significant difference in mortality risk between the two vancomycin minimum inhibitory concentration doses managed by TDM (p=0.43).²⁷ This is consistent with the results of our study. In our study, multifactorial logistic regression analysis showed that the proportion of patients attaining an 80% reduction in CRP levels compared with baseline (OR 0.165, 95% CI 0.030 to 0.892, p=0.036) and length of hospital stay (OR 6.904, 95% CI 1.074 to 44.401, p=0.042) were factors influencing 14-day mortality. Meanwhile, the results of our study showed no statistically significant differences in length of hospital stay and duration of vancomycin treatment between the two groups (p=0.135, p=0.997), which is consistent with the results of a previous meta-analysis.⁸ This might be caused by our retrospective single-centre study and the small number of cases.

From the economic perspective, our findings suggest that the TDM group cost US$15.82 more per additional TDM than the non-TDM group. In 2021, the per capita disposable income of urban permanent residents in Hohhot, Inner Mongolia Autonomous Region was US$7928.98, and the increase of US$15.82 per TDM performed was equivalent to 0.20% of the per capita disposable income. Although patients in the TDM group incurred additional blood monitoring costs compared with those in the non-TDM group, the average cost of the total direct medical care in the TDM group was lower than in the non-TDM group; the attainment rate of the WBC count was 84.4%, the attainment rate of CRP was 68.8%, the attainment rate of PCT was 62.50% and the attainment rate of NEUT was 59.4%. A study in elderly patients showed that total medical costs (US$841.40) and drug costs (US$16.70) per patient were significantly lower in the TDM group than in the non-TDM group, and concluded that vancomycin TDM was cost-effective in elderly patients.²⁸ Again, studies in critically affected patients have shown that TDM with vancomycin was cost-effective, reduces the incidence of nephrotoxicity and shortens the stay in hospital.²⁹ This finding is consistent with the results of a meta-analysis.⁸

In summary, vancomycin TDM has been applied to elderly patients, paediatric patients and seriously ill patient groups, but there are few data regarding the efficacy and safety of vancomycin TDM in the treatment of patients with postoperative intracerebral haemorrhage. Our study suggests that vancomycin TDM has an important role in the anti-infective treatment of patients with postoperative intracerebral haemorrhage. Vancomycin TDM significantly improved the outcomes of patients with postoperative intracerebral haemorrhage, with an increase in the proportion of patients attaining desirable outcomes in WBC, NEUT, PCT and CRP and reduced nephrotoxicity. Although there was no significant difference in 14-day mortality and length of hospital stay, an economic evaluation showed cost-effectiveness in the proportion of patients reaching normal values of inflammatory factors. The ICERs of WBC, NEUT, PCT and CRP were >1. Thus, vancomycin TDM could be clinically beneficial in patients with postoperative intracerebral haemorrhage and it is shown to be a cost-effective procedure.

This study has several limitations. First, the study method was a single-centre retrospective cohort study and no data were available for calculating the damage caused by vancomycin-induced nephrotoxicity, ototoxicity or its treatment costs. Therefore, the cost-effectiveness evaluated in this study is limited to improvement in inflammatory markers. Second, Bayesian modelling for TDM was not performed in this study, which is the new standard of care. Due to the limitations of our hospital, there is currently no Bayesian-based modelling for TDM. However, this is a future direction for our hospitals, combined with Bayesian modelling for TDM to optimise vancomycin delivery. Third, although the PSM compensated for confounding factors and increased the credibility of the evidence level of the study, there were still selection biases. Finally, after PSM 1:1 matching, only 92 pairs of the 435 samples were included in the study. The sample size was small, which affected the credibility of the research results, and the statistical power may not have been high enough to analyse the results. Therefore, multicentre randomised trials are need to further verify the benefits of TDM.

Conclusions

The results of this study show that vancomycin TDM is important in the treatment of patients with postoperative intracerebral haemorrhage. It can improve the compliance rate of inflammatory factors and can also reduce vancomycin-related nephrotoxicity. In addition, although additional costs were incurred per TDM, clinical benefits can be improved. In patients with postoperative intracerebral haemorrhage, TDM has been shown to be a cost-effective procedure.

Data availability statement

No data are available. no data are available.

Ethics statements

Patient consent for publication

Not applicable.