Effect of Goal-Directed Intraoperative Fluid Therapy on Duration of Hospital Stay and Postoperative Complications in Patients Undergoing Excision of Large Supratentorial Tumors
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.336329
Source of Support: None, Conflict of Interest: None
Keywords: Brain tumors, goal-directed fluid therapy, neurosurgery, outcome
Optimal fluid management in neurosurgical patients involves a tight balance between overcorrection and undercorrection, which may lead to cerebral edema and hypoperfusion, respectively. Fluid management in patients with brain tumor is tricky and unique as compared to nononcological surgeries due to increased brain bulk, tendency of tumor tissue to swell intraoperatively, use of osmotic diuretics, and blood loss due to increased vascularity. Furthermore, the fluid requirement during surgery for supra- and infratentorial tumors may be different due to positional effects.
Goal-directed fluid therapy (GDFT) is defined as a bundle of care applied to high-risk surgical patients with the goal of improving tissue end-organ perfusion. Studies carried out during past decade indicate that intraoperative GDFT improves postoperative outcome in patients undergoing major non-neurosurgical surgeries., However, the roles of goal-directed fluid strategies in elective neurosurgical patients remain inconclusive and are scarce. Luo and colleagues suggested that goal-directed fluid restriction based on stroke volume variation (SVV) improves outcomes in elective neurosurgical patients, whereas Wu and colleagues suggested that fluid boluses target lower SVV targets of <10%, suggesting a nonrestrictive fluid regimen is beneficial. Various dynamic parameters have emerged as targets in perioperative fluid therapy like systolic pressure variation, pulse pressure variation and SVV. They are good indicators of preload dependency. Modern GDFT management involves optimization of stroke volume (SV) and ensuring preload is replete typically using a crystalloid or colloid-based fluid challenges followed by increasing oxygen delivery by use of inotropic drugs mostly dobutamine and blood transfusion. Previous studies based on similar algorithm have shown to improved postoperative surgical outcome in various noncardiac surgeries.,,,
This study was planned using a simple algorithm based on SVV and cardiac index (CI) for perioperative fluid management. It was hypothesized that the SVV/CI-guided GDFT using the FloTrac/Vigileo™ system with a target to maintain mean arterial pressure (MAP) and hence, peripheral perfusion, rather than using a restrictive or liberal fluid strategy, would improve surgical outcomes in terms of reduced length of hospital stay and incidence of postoperative complications.
The primary objective was to assess the intraoperative use of GDFT on the duration of hospital stay, and the secondary objectives were to compare the volume of fluid used, incidence of perioperative complications, and requirement of inotropes in between the two groups.
This prospective randomized patient and assessor-blinded clinical trial was approved by the Institutional Ethics Committee and was conducted during a period of 2014 to 2016. It was registered with the Clinical Trials Registry India (CTRI/2016/10/007350). The patients were screened for eligibility by a member of the research team, who also obtained written informed consent of the patients, before surgery. Ethics clearance obtained: 25.02.2015.
The sample size was calculated based on the study by Mayer et al. It was observed that, to detect a 4-day difference in duration of hospital stay as statistically significant in two-sided students test with 5% alpha error and 80% power, 17 subjects would be required in each group. Considering a dropout rate of 15%, 20 patients were enrolled in each group for the study.
A total of 40 consecutive patients of American Society of Anaesthesiologists (ASA) physical status I-II, in the age group of 18–65 years, with large supratentorial tumors (tumor size ≥4 cm in at least one dimension) undergoing elective craniotomy and excision, were enrolled. Baseline characteristics like age, sex, weight, height, body mass index, baseline laboratory investigations, and P-POSSUM scores (predict perioperative morbidity and mortality rates) have been validated for elective neurosurgical patients. Patients with coexisting arrhythmias, known cardiac, pulmonary disease, and diabetes, weight ≤50 kg or ≥140 kg, altered sensorium, and skull base tumors were excluded. Enrolled patients were randomly allocated to two groups of 20 each, based on the computer-generated randomization chart [Figure 1].
The attending neuroanesthesiologist was not blinded; the neurointensivist posted in the ICU was blinded to the intraoperative fluid protocol and assessed the outcomes such as postoperative complications, durations of mechanical ventilation, ICU, and hospital stay. One of the authors who were blinded to the study protocol and not involved in the patient care assessed all the data independently.
Intraoperative fluid therapy
A baseline maintenance fluid of 3 ml/kg/h was given in all patients in both the groups:
Control Group: Patients of this group received fluids based on the routine hemodynamic monitoring, as below:
Crystalloids were given as maintenance fluid; colloids and vasoactive agents were used as per standard institutional practice and were left at discretion of the attending neuroanesthesiologist. The SVV and cardiac output (CO) data was blinded to the anesthesiologist by placing an opaque cover over the cardiac output monitor; retrospectively, the data were collected, at the end of surgery.
GDFT Group: Patients in this group received fluids based upon a protocolized approach using SVV and CI as per Flotrac Vigileo monitor connected to an arterial pressure line and referenced to the mid-axillary line of the patient as targets.
SVV less than 12%
CI ≥2.5 L/min/m2
In patients who had fall in MAP less than 60 mmHg but with a normal CI and SVV, ephedrine boluses were given.
Anesthesia was induced with fentanyl 2–3 μg/kg and propofol 1–2 mg/kg, and tracheal intubation was facilitated with rocuronium 1 mg/kg. Mechanical ventilation was instituted to maintain an end-tidal CO2 (EtCO2) of 30–35 mmHg, keeping a tidal volume of 4–6 ml/kg. Anesthesia was maintained by using additional boluses of fentanyl and sevoflurane 0.8–1.2% in a mixture of oxygen and N2O to achieve a MAC of 1. Standard monitoring modalities included an electrocardiogram, oxygen saturation (SpO2), EtCO2, invasive blood pressure, central venous pressure, and nasopharyngeal temperature. Mannitol 0.25–0.5 gm/kg was given, for brain relaxation at the time of skin incision, over a period of 20 min.
Parameters such systolic BP (SBP), mean BP (MAP), diastolic BP (DBP), heart rate (HR) were recorded at baseline, 5, 10, 15, and 30 min postinduction and then, every hourly till the end of surgery. Dynamic parameters such as CI, stroke volume index (SVI), SVV were recorded 15 and 30 min after induction and then every hourly till the end of surgery.
The respective protocols were followed in each group till the end of surgery. Blood loss was substituted with fluids, and hemoglobin of ≤8 gm/dl, as per hourly assessment with arterial blood gas (ABG) analysis, was considered as the trigger for transfusion of packed red blood cells. Intraoperatively, total amount of fluids, crystalloids, colloids, rough quantitative estimate of blood loss as per routine visual assessment, and urine output were noted. Intraoperative complications like desaturation, electrolyte abnormalities (on ABG), lactate levels, brain relaxation score (BRS) as assessed by the operating neurosurgeon and hypotension (MAP <60 mmHg; not responsive to fluids and blood transfusion) were noted. Propofol was started, and anesthetic technique was modified to total intravenous anesthesia in the presence of brain bulge.
At the end of surgery, the patients were assessed for tracheal extubation, and if deemed fit, were extubated. Patients whose trachea could not be extubated were shifted to the neuro-intensive care unit for further management by a neurointensivist, who was not part of the study protocol. Any neurological, renal, pulmonary, cardiac, or hematological complications were recorded and managed as per hospital protocol. The discharge criteria were decided by the attending neurosurgeon.
Continuous data (demographic data, hemodynamic parameters, duration of surgery, and anesthesia) with normal distribution were tested with paired t-tests; non-normally distributed data like amount of fluids (crystalloids and colloids), urine output, blood loss, duration of postoperative ventilation, duration of hospital, and ICU stay were tested using Mann–Whitney U test. Categorical data (intraoperative complications, blood transfusion, postoperative complications, and Glasgow outcome scale (GOS)) were tested using Chi-square test. Data are presented as mean ± standard deviation when normally distributed and as median [interquartile ranges] in case of abnormal distribution. A P value less than 0.05 was considered as statistically significant.
A total of 44 patients were assessed for eligibility and 4 patients were excluded; two of them for refusal to give consent and two others for not fulfilling the inclusion criteria. A total of 40 patients were included in the study and were divided into two groups, i.e., GDFT and Control groups. None of the patients were excluded from the final analysis [Figure 1].
Baseline patient characteristics, such as age, sex, weight, height, ASA grade, p-POSSUM score, preoperative hemoglobin, biochemical parameters (sodium potassium, urea, and creatinine), and baseline lactate levels; pathological types and size of tumor; and hemodynamic parameters such as HR, SBP, DBP, MAP, SV, CO, SVV, CI, SVI, and CVP were comparable in between the two groups [Table 1].
The amount of crystalloid infused was significantly higher in the control group as compared to GDFT group (P = 0.003); however, the volume of colloids administered was more in the GDFT group (P = 0.03) [Table 2]. The blood loss was significantly higher in the GDFT group (P = 0.048). Number of patients required blood transfusion, duration of anesthesia and surgery were also comparable in between the two groups.
The incidence of intraoperative complications was significantly higher in the control group as compared to GDFT group (P = 0.005); commonest being the occurrence of tight brain and hypotension [Table 2]. There was no episode of oxygen desaturation, electrolyte imbalances, arrhythmias, or decrease in urine output. Two patients in GDFT group required vasoactive support; one required dobutamine infusion, whereas the other was managed with ephedrine (bolus of 3 mg) followed by noradrenaline infusion. Similarly, in the control group one patient required ephedrine of 3 mg boluses.
The hemodynamic parameters at closure were comparable in both the groups except for MAP, which was significantly higher in the GDFT group [Table 3]. The lactate levels at the end of surgery were lower in the GDFT group as compared to the control group (P = 0.018).
The mean duration of mechanical ventilation was higher in the control group as compared to GDFT group (P = 0.07) [Table 4]. Similarly, a trend towards lower duration of ICU stay was seen in GDFT group (P = 0.06). The duration of hospital stay was slightly lower in GDFT group (P = 0.16). The GOS scores were similar (Good: 1-2; Poor: 3–5) in between the two groups.
Postoperative complications were comparable in between the two groups (P = 0.32); major complications are depicted in [Figure 2]. None of the patients had any incidence of hypotension, myocardial infarction, arrhythmias, wound infection, sepsis, and meningitis, renal or hepatic failure. Two patients in control group developed cerebral venous infarct; one patient underwent decompressive craniectomy within 48 h of the first surgery. One patient each in the control group was diagnosed with pulmonary thromboembolism and pneumonia; both were managed conservatively without further complications. In the GDFT group, one patient had generalized seizures and was controlled with medications; another patient developed intracranial hematoma and underwent re-exploration.
This study suggests that perioperative GDFT in elective neurosurgical patients helped in better hemodynamic management signified with optimal MAP, a lower incidence of tight brain, and lower mean lactate levels at the end of surgery. However, it did not show any statistically significant improvement in terms of ICU/hospital stay, duration of mechanical ventilation, or overall GOS. The major strength is it is first study to have a homogeneous elective neurosurgical patient population. The outcome patterns of different types of neurosurgical cases may be different depending upon their nature of urgency, location of tumor, and preoperative neurological status. Luo and colleagues used a fluid protocol based on SVV and found that it reduced incidence of complications and hospital stay. They included a heterogeneous study population and included all types of brain tumors located both in supra- and infratentorial compartments, intracranial aneurysms, and brain abscesses. They also adopted a goal directed fluid restrictive approach using a target SVV >15% as cut-off for fluid boluses and presumed this approach to be beneficial to the patients. In contrast, Wu and colleagues compared high SVV (18%) and low SVV (10%) as targets for fluid boluses and found that the use of a nonrestrictive fluid strategy resulted in better outcome in terms of postoperative complications and ICU stay. In our study, we neither opted for extreme restriction neither a liberal one but rather a balanced fluid regimen. Another study suggested that the SVV value >11.5% can be used to predict positive responsiveness to volume loading in the patients with brain surgery. We opted for SVV target of >12%, as higher targets have been associated with hypovolemia.
The GDFT was useful to achieve a better degree of brain relaxation. This could be due to less requirement of total fluid in the GDFT approach. Two other studies evaluated the effect of GDFT on BRS., Xia et al. found similar brain relaxation, but they used GDFT in both groups and not compared it to the conventional fluid therapy. Hasanin et al., did not find better BRS as compared to usual care group possibly due to use of crystalloid-based GDFT and use excessive amount of intraoperative fluids. We used colloids in our GDFT group that decreased the requirement of total amount of fluids, which is in accordance with observation made by a recent trial that suggests colloid use decreases the total requirement of fluids. The total amount of fluid administered in the GDFT group was much less, which was in contrast to the finding of Wu and colleagues; they observed fluid administration to be higher with GDFT. The possible explanation for this difference could be because they adapted a restrictive fluid strategy in their control group as evidenced by zero balance at the end of the surgery using 50 ml colloid boluses for correction of CVP and MAP. Additionally, they have preloaded all patients with colloid infusions in the preinduction and preoperative period that might have impacted the fluid requirements.
The blood loss was significantly higher in the GDFT group possibly due to a higher MAP achieved, or due to higher amount of colloid administration. However, the total amount of transfusion was similar in both groups. The reason for this finding could be hemodilution as a result of infusion of significant amount of fluid (mainly crystalloids), and hence, more amount of blood was transfused in the control group. Therefore, GDFT helped us to guide intraoperative blood transfusion and to maintain a better microcirculation and tissue perfusion as indicated by lactate levels at the end of surgery. Bacchin et al. demonstrated reduced requirement in blood transfusion in spine surgery; however, no robust literature is available in this regard.
Perioperative fluid replacement is a challenging issue during surgical care, especially in a procedure-specific model. Multiple studies investigated the utility of GDFT in patients undergoing major abdominal, vascular, and orthopedic surgeries,,,, suggested an overall reduction in postoperative morbidity (25–50%). However, in this study, we did not find any effect of the GDFT on postoperative mechanical ventilation, ICU stay, or length of hospital stay. The most plausible explanation could be that in neurosurgery the spectrum of complications is quite different from abdominal and cardiothoracic surgeries where medical complications are higher with increased morbidity. The specific complications in neurosurgery like venous infarct, seizures, intracranial hematomas depend mostly on surgical techniques involved rather than fluid therapy. Future studies may be carried out along with specific neuromonitoring techniques like intracranial pressure monitoring and trans-cranial Doppler to observe the effects of GDFT on cerebral perfusion indices.
This study was limited by a single center observation and small sample size. It was carried out in low-risk patients undergoing high-risk surgeries. The goal-directed fluid protocol was done only in the intraoperative period, and the postoperative period was not monitored in terms of fluid balance, blood transfusion, and serum creatinine levels that might have affected the overall postoperative outcomes. Lastly, long-term follow up was not done in these patients.
This study did not show beneficial effects of GDFT over conventional intraoperative fluid therapy in terms of incidence of postoperative complications, hospital and ICU stay, GOS at-discharge in patients undergoing craniotomy for excision of large supratentorial tumors. Further studies are required to monitor the effects of GDFT on intracranial pressure indices and perioperative blood transfusion practices and their effects on long-term outcome of patients.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]